build: generated presets and a check entry point

tools/make_presets.py emits the uniform pipeline the hand-grown file had
drifted from — generated configure/build/test presets and workflows for
all 37 chips, reflect configure/build for libavr's 12-chip spot set —
and tools/check.sh runs every chip's workflow (--full adds the reflect
spot) as the port's gate.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_019553N6pLmaKhxEwkiCs3RX

.gitmodules

@@ -1,3 +0,0 @@
-[submodule "libavr"]
-	path = libavr
-	url = ../libavr.git

CMakeLists.txt

@@ -2,19 +2,18 @@ cmake_minimum_required(VERSION 3.28)
 
 project(tsb_libavr LANGUAGES CXX)
 
-# libavr rides as the pinned submodule; LIBAVR_ROOT (cache or environment)
+# libavr from a local checkout (LIBAVR_ROOT) or the forge; the toolchain file
-# overrides it for tandem development against a working tree. The toolchain
+# comes from the same checkout via CMakePresets.json.
-# file comes from the submodule via CMakePresets.json either way.
+include(FetchContent)
 if(NOT LIBAVR_ROOT AND DEFINED ENV{LIBAVR_ROOT})
 	set(LIBAVR_ROOT $ENV{LIBAVR_ROOT})
 endif()
-if(NOT LIBAVR_ROOT)
+if(LIBAVR_ROOT)
-	set(LIBAVR_ROOT ${CMAKE_CURRENT_SOURCE_DIR}/libavr)
+	FetchContent_Declare(libavr SOURCE_DIR ${LIBAVR_ROOT})
+else()
+	FetchContent_Declare(libavr GIT_REPOSITORY git@git.blackmark.me:avr/libavr.git GIT_TAG main)
 endif()
-if(NOT EXISTS ${LIBAVR_ROOT}/CMakeLists.txt)
+FetchContent_MakeAvailable(libavr)
-	message(FATAL_ERROR "libavr not found at ${LIBAVR_ROOT} — run: git submodule update --init libavr")
-endif()
-add_subdirectory(${LIBAVR_ROOT} libavr-build)
 
 if(PROJECT_IS_TOP_LEVEL)
 	add_compile_options(-Werror) # warnings are errors for the port's own code
@@ -22,17 +21,15 @@ if(PROJECT_IS_TOP_LEVEL)
 
 	# The behavioral tests drive the real wire protocols over a simavr pty
 	# (as the host tools do) and actually flash the device. The runners are
-	# host programs built at configure time against libsimavr (C++23 — what
+	# host programs built at configure time against libsimavr; if they or
-	# the distribution's compiler speaks in full); if they or Python are
+	# Python are missing, only the size tests run.
-	# missing, only the size tests run.
+	find_program(_host_cc NAMES cc gcc)
-	find_program(_host_cxx NAMES c++ g++)
 	find_package(Python3 COMPONENTS Interpreter)
-	if(_host_cxx AND Python3_FOUND)
+	if(_host_cc AND Python3_FOUND)
 		set(PB_DEVICE ${CMAKE_BINARY_DIR}/pureboot_device)
 		execute_process(
-		    COMMAND ${_host_cxx} -std=c++23 -Wall -Wextra -O2
+		    COMMAND ${_host_cc} -O2 -I/usr/include/simavr -I/usr/include/simavr/parts
-		            -I/usr/include/simavr -I/usr/include/simavr/parts
+		            -o ${PB_DEVICE} ${CMAKE_CURRENT_SOURCE_DIR}/test/pureboot_device.c
-		            -o ${PB_DEVICE} ${CMAKE_CURRENT_SOURCE_DIR}/test/pureboot_device.cpp
 		            -lsimavr -lsimavrparts -lelf -lutil
 		    RESULT_VARIABLE _pbdev_res ERROR_VARIABLE _pbdev_err)
 		if(NOT _pbdev_res EQUAL 0)
@@ -42,9 +39,8 @@ if(PROJECT_IS_TOP_LEVEL)
 		if(LIBAVR_MCU STREQUAL "atmega328p")
 			set(TSB_DEVICE ${CMAKE_BINARY_DIR}/tsb_device)
 			execute_process(
-			    COMMAND ${_host_cxx} -std=c++23 -Wall -Wextra -O2
+			    COMMAND ${_host_cc} -O2 -I/usr/include/simavr -I/usr/include/simavr/parts
-			            -I/usr/include/simavr -I/usr/include/simavr/parts
+			            -o ${TSB_DEVICE} ${CMAKE_CURRENT_SOURCE_DIR}/test/device.c
-			            -o ${TSB_DEVICE} ${CMAKE_CURRENT_SOURCE_DIR}/test/device.cpp
 			            -lsimavr -lsimavrparts -lelf
 			    RESULT_VARIABLE _dev_res ERROR_VARIABLE _dev_err)
 			if(NOT _dev_res EQUAL 0)
@@ -69,18 +65,16 @@ function(add_image_outputs name)
 	                           $<TARGET_FILE:${name}> $<TARGET_FILE:${name}>.bin)
 endfunction()
 
-# The TinySafeBoot protocol reimplemented on libavr in variants that trade
+# The TinySafeBoot protocol reimplemented on libavr in three variants that trade
 # clarity for size. Each links into the ATmega328P boot section (BOOTSZ selects
 # its size; BOOTRST vectors a reset to its base) with -nostartfiles — a polled
-# loader has no use for the crt or the vector table. The entry sits in
+# loader has no use for the crt or the vector table. The naked entry sits in
-# .vectors, laid first, and runs — avr::startup::entry on the policy tier,
+# .vectors, laid first, and runs. The boot base is FLASHEND+1 minus the section
-# the experiment tiers' own naked stubs elsewhere, each documented in its
+# size; the linker section-start and the source's boot_bytes agree. tsb_app is
-# source. The boot base is FLASHEND+1 minus the section size; the linker
-# section-start and the source's boot_bytes agree. tsb_app is
 # the application's reset vector, pinned to 0 here so the loaders jump to a
 # named function; --pmem-wrap-around lets relaxation turn that absolute jump
 # into the wrapped rjmp AVR's modulo-flash PC actually executes.
-# All four implement the full oracle feature set (see oracle/README.md):
+# All three implement the full oracle feature set (see oracle/README.md):
 # watchdog bail, one-wire half-duplex, config-page activation timeout, password
 # gate, emergency erase, config/flash/EEPROM read-write. They differ only in how,
 # and the size gradient is the cost of that "how" — see dev/lessons.md.
@@ -100,10 +94,6 @@ endfunction()
 #   tsb_pure  — pure idiomatic libavr, one function per command, TU-local
 #                 (internal linkage), streaming (no SRAM page buffer): 836 B in
 #                 the 1 KB section.
-#   tsb_policy — the policy floor: pureboot's rules (no asm, no register
-#                 variables) with every pureboot lesson applied. 638 B in the
-#                 1 KB section — the measured evidence that the 512 B fit is a
-#                 property of the mechanisms philosophy #5 bans.
 #
 # add_tsb_variant(<name> <boot-section-bytes>)
 function(add_tsb_variant name bytes)
@@ -131,13 +121,8 @@ endfunction()
 # chips build pureboot alone.
 if(LIBAVR_MCU STREQUAL "atmega328p")
 	add_tsb_variant(tsb_asm 512)
-	add_tsb_variant(tsb_policy 1024)
 	add_tsb_variant(tsb_pure 1024)
 	add_tsb_variant(tsb_tricks 1024)
-	# The policy tier's floor is measured with the loop flags pureboot's size
-	# work found (a loader's loop bodies all contain calls); the other tiers
-	# keep the flag set their recorded floors were measured with — none.
-	target_compile_options(tsb_policy PRIVATE -fno-move-loop-invariants -fno-tree-ter)
 endif()
 
 # pureboot — the pure-constraint port (see pureboot/README.md): one source,
@@ -165,25 +150,10 @@ if(PROJECT_IS_TOP_LEVEL)
 	if(Python3_FOUND)
 		add_test(NAME pureboot.pi
 		         COMMAND ${Python3_EXECUTABLE} ${CMAKE_CURRENT_SOURCE_DIR}/test/check_pi.py
-		                 ${CMAKE_OBJDUMP} ${CMAKE_OBJCOPY} ${CMAKE_CXX_COMPILER} ${LIBAVR_MCU}
+		                 ${CMAKE_OBJDUMP} ${CMAKE_NM} $<TARGET_FILE:pureboot> ${PUREBOOT_BASE_HEX})
-		                 $<TARGET_FILE:pureboot>
-		                 ${CMAKE_BINARY_DIR}/CMakeFiles/pureboot.dir/pureboot/pureboot.cpp.obj
-		                 ${PUREBOOT_BASE_HEX})
 		add_test(NAME pureboot.planner
 		         COMMAND ${Python3_EXECUTABLE} ${CMAKE_CURRENT_SOURCE_DIR}/test/test_planner.py
 		                 ${CMAKE_CURRENT_SOURCE_DIR}/pureboot/pureboot.py)
-		add_test(NAME pureboot.scan
-		         COMMAND ${Python3_EXECUTABLE} ${CMAKE_CURRENT_SOURCE_DIR}/test/test_scan.py
-		                 ${CMAKE_CURRENT_SOURCE_DIR}/pureboot/pureboot.py)
-		# CMakePresets.json is generated; hand edits drift the moment the
-		# generator reruns, so the gate holds the pair together.
-		add_test(NAME presets.generated
-		         COMMAND ${Python3_EXECUTABLE} ${CMAKE_CURRENT_SOURCE_DIR}/tools/make_presets.py
-		                 --check)
-		add_test(NAME pureboot.handshake
-		         COMMAND ${Python3_EXECUTABLE} ${CMAKE_CURRENT_SOURCE_DIR}/test/test_handshake.py)
-		add_test(NAME pureboot.updatelink
-		         COMMAND ${Python3_EXECUTABLE} ${CMAKE_CURRENT_SOURCE_DIR}/test/test_update_link.py)
 	endif()
 
 	# The protocol test flashes this fixture through the loader with the real
@@ -202,38 +172,6 @@ if(PROJECT_IS_TOP_LEVEL)
 		                 ${CMAKE_BINARY_DIR}/pbtest-work)
 		set_tests_properties(pureboot.protocol PROPERTIES TIMEOUT 180)
 
-		# The activation window as a measured duration: application installed,
-		# line idle, the first transmit is the application's banner — its
-		# cycle is the window the source declares, held to ±2 % (one
-		# mis-counted cycle per poll is a 10 % shift).
-		add_test(NAME pureboot.window
-		         COMMAND ${Python3_EXECUTABLE} ${CMAKE_CURRENT_SOURCE_DIR}/test/pbwindow.py
-		                 --device ${PB_DEVICE} --loader $<TARGET_FILE:pureboot>
-		                 --mcu ${PUREBOOT_SIM_MCU} --hz ${_pb_stock_hz}
-		                 --base ${PUREBOOT_BASE_HEX} --page ${PUREBOOT_PAGE}
-		                 --baud ${_pb_stock_baud} --app $<TARGET_FILE:pbapp>.bin
-		                 --seconds ${PUREBOOT_TIMEOUT}
-		                 --tool ${CMAKE_CURRENT_SOURCE_DIR}/pureboot/pureboot.py
-		                 --workdir ${CMAKE_BINARY_DIR}/pbwindow-work)
-		set_tests_properties(pureboot.window PROPERTIES TIMEOUT 300)
-
-		# The half-duplex loader's window, same gate: its poll runs through
-		# rx_ready()'s release-line test, whose outlined call re-shapes the
-		# whole loop — a per-class cycle count (poll_cost() in pureboot.cpp)
-		# that only the built image can prove, chip by chip.
-		if(PUREBOOT_HAS_USART)
-			add_test(NAME pureboot.window.halfduplex
-			         COMMAND ${Python3_EXECUTABLE} ${CMAKE_CURRENT_SOURCE_DIR}/test/pbwindow.py
-			                 --device ${PB_DEVICE} --loader $<TARGET_FILE:pureboot_hd>
-			                 --mcu ${PUREBOOT_SIM_MCU} --hz ${_pb_stock_hz}
-			                 --base ${PUREBOOT_BASE_HEX} --page ${PUREBOOT_PAGE}
-			                 --baud ${_pb_stock_baud} --app $<TARGET_FILE:pbapp>.bin
-			                 --seconds ${PUREBOOT_TIMEOUT}
-			                 --tool ${CMAKE_CURRENT_SOURCE_DIR}/pureboot/pureboot.py
-			                 --workdir ${CMAKE_BINARY_DIR}/pbwindow-hd-work)
-			set_tests_properties(pureboot.window.halfduplex PROPERTIES TIMEOUT 300)
-		endif()
-
 		# The position-independence acceptance test: the identical image,
 		# installed one slot lower, must serve the full command set.
 		add_test(NAME pureboot.reloc
@@ -245,22 +183,6 @@ if(PROJECT_IS_TOP_LEVEL)
 		set_tests_properties(pureboot.reloc PROPERTIES TIMEOUT 180
 		                     ENVIRONMENT "PB_OBJCOPY=${CMAKE_OBJCOPY}")
 
-		# Entering the loader from a running application with no reset
-		# between, over a page buffer the application dirtied — the case the
-		# loader declines to guard and the host repairs. Hardware forbids the
-		# state here (SPM runs only from the boot section); simavr does not,
-		# which is what makes it constructible.
-		if(LIBAVR_MCU STREQUAL "atmega328p")
-			add_test(NAME pureboot.dirty
-			         COMMAND ${Python3_EXECUTABLE} ${CMAKE_CURRENT_SOURCE_DIR}/test/pbdirty.py
-			                 ${PB_DEVICE} $<TARGET_FILE:pureboot> ${PUREBOOT_SIM_MCU} ${_pb_stock_hz}
-			                 ${PUREBOOT_BASE_HEX} ${PUREBOOT_PAGE} ${_pb_stock_baud}
-			                 $<TARGET_FILE:pbapp>.bin
-			                 ${CMAKE_CURRENT_SOURCE_DIR}/pureboot/pureboot.py
-			                 ${CMAKE_BINARY_DIR}/pbdirty-work)
-			set_tests_properties(pureboot.dirty PROPERTIES TIMEOUT 180)
-		endif()
-
 		# Re-homing: a loader mistakenly programmed at address 0 (a raw .bin
 		# handed to a programmer) or sitting in the staging slot must heal
 		# into the canonical slot through the ordinary --update-loader flow.
@@ -293,13 +215,12 @@ if(PROJECT_IS_TOP_LEVEL)
 	endif()
 
 	# The size matrix: every configuration axis that could move the image
-	# size — the serial backend (different code), the USART instance
+	# size — the serial backend (different code), the clock and its ladder
-	# (different registers), the clock (different constants), the baud
+	# baud (different constants and divisor shapes), the USART instance
-	# through the shapes its bit timing takes, and the pins through the one
+	# (different register class) — each combination must still fit the
-	# thing they decide (whether a bit-banged link has to release the USART
+	# chip's slot budget. Pins are size-neutral (port and bit are immediate
-	# that owns them) — each combination must still fit the chip's slot
+	# operands) and the timeout is a constant, so neither adds an axis. The
-	# budget. The timeout is a constant and adds no axis. The stock build is
+	# stock build is one point of this matrix and already has its test.
-	# one point of this matrix and already has its test.
 	function(pureboot_size_variant name)
 		pureboot_add_loader(${name} ${ARGN})
 		add_test(NAME ${name}.size
@@ -307,218 +228,27 @@ if(PROJECT_IS_TOP_LEVEL)
 		                 -DLIMIT=${PUREBOOT_LIMIT} -P ${CMAKE_CURRENT_SOURCE_DIR}/test/check_size.cmake)
 	endfunction()
 
-	# The autobaud loader: one clock-agnostic image per chip, so it has no
-	# clock x baud axis of its own — the matrix below sweeps those for the
-	# fixed-baud builds, and this one binary has to serve all of them at run
-	# time. Size-tested against the same per-chip budget as every other variant.
-	pureboot_add_loader(pureboot_autobaud SERIAL autobaud)
-	add_test(NAME pureboot_autobaud.size
-	         COMMAND ${CMAKE_COMMAND} -DSIZE_TOOL=${CMAKE_SIZE} -DELF=$<TARGET_FILE:pureboot_autobaud>
-	                 -DLIMIT=${PUREBOOT_LIMIT} -P ${CMAKE_CURRENT_SOURCE_DIR}/test/check_size.cmake)
-	# The measured unit's home is wire contract, not layout accident: the
-	# host reads the bit period from it (--info's measured clock). In the
-	# GPIOR home the image must carry no RAM copy at all; in the RAM home it
-	# is the loader's only RAM object, at the very start of SRAM.
-	add_test(NAME pureboot_autobaud.unit
-	         COMMAND ${CMAKE_COMMAND} -DOBJDUMP=${CMAKE_OBJDUMP} -DELF=$<TARGET_FILE:pureboot_autobaud>
-	                 -DRAM_START=${PUREBOOT_RAM_START} -DGPIOR=${PUREBOOT_UNIT_GPIOR}
-	                 -P ${CMAKE_CURRENT_SOURCE_DIR}/test/check_unit.cmake)
-
-	# One point of the exhaustive matrix, named from its resolved parameters
-	# so the enumeration cannot collide with itself. `pins` is empty for the
-	# default pair, or the index of the USART whose own pins a bit-banged
-	# link sits on. Unreachable rates drop out here rather than aborting the
-	# configure.
-	# The optional trailing argument is the one-wire shape of the same link:
-	# ONE_WIRE folds a software point onto its RX pin (the default, or the
-	# named USART's RXD), HALF_DUPLEX is the hardware USART's turn-around.
-	function(pureboot_matrix_point hz baud link pins)
-		set(_name pbm_${hz}_${baud}_${link})
-		if(link STREQUAL "software")
-			pureboot_baud_feasible(${hz} ${baud} 1 _ok)
-			set(_args SERIAL software)
-			if(NOT pins STREQUAL "")
-				list(APPEND _args RX ${PUREBOOT_USART${pins}_RX} TX ${PUREBOOT_USART${pins}_TX})
-				set(_name ${_name}_on${pins})
-			endif()
-			if(ARGC GREATER 4 AND ARGV4 STREQUAL "ONE_WIRE")
-				if(NOT pins STREQUAL "")
-					set(_args SERIAL software RX ${PUREBOOT_USART${pins}_RX} TX ${PUREBOOT_USART${pins}_RX})
-				else()
-					list(APPEND _args RX pb0 TX pb0)
-				endif()
-				set(_name ${_name}_1w)
-			endif()
-		else()
-			pureboot_baud_feasible(${hz} ${baud} 0 _ok)
-			set(_args USART ${link})
-			if(ARGC GREATER 4 AND ARGV4 STREQUAL "HALF_DUPLEX")
-				list(APPEND _args HALF_DUPLEX)
-				set(_name ${_name}_hd)
-			endif()
-		endif()
-		if(_ok)
-			pureboot_size_variant(${_name} CLOCK ${hz} BAUD ${baud} ${_args})
-		endif()
-	endfunction()
-
 	# Clock points: the shipped-fuse floor (CKDIV8), the calibrated RC, and
 	# the crystal the stock build assumes (the tiny13's ladder is its own RC
 	# menu — it has no crystal option).
 	if(LIBAVR_MCU MATCHES "^attiny13")
 		set(_matrix_clocks 1200000 4800000 9600000)
-		set(_full_clocks 128000 600000 1200000 4800000 9600000)
 	else()
 		set(_matrix_clocks 1000000 8000000 16000000)
-		set(_full_clocks 128000 1000000 1843200 2000000 3686400 4000000 7372800 8000000
-		                 11059200 12000000 14745600 16000000 18432000 20000000)
 	endif()
-
+	foreach(_matrix_hz IN LISTS _matrix_clocks)
-	# The exhaustive cross product: every clock a deployment plausibly runs
+		math(EXPR _matrix_khz "${_matrix_hz} / 1000")
-	# — the internal oscillators, the shipped CKDIV8 floor, the plain
+		if(PUREBOOT_HAS_USART OR NOT _matrix_hz EQUAL _pb_stock_hz)
-	# crystals and the UART crystals — against every rate, against every
+			pureboot_size_variant(pureboot_sw_${_matrix_khz}k CLOCK ${_matrix_hz} SERIAL software)
-	# backend. Beyond the ladder the list carries the slow rates a
-	# sub-megahertz oscillator is left with, which no ladder rate reaches
-	# (16000 Bd is the only rate the 128 kHz oscillator holds exactly); at
-	# the fast clocks those same rates also select the software UART's
-	# 16-bit _delay_loop_2 bit spin (two words more setup at each of its five
-	# sites), the largest image the space produces and a shape the ladder
-	# default — always the *fastest* rate a clock reaches — never picks.
-	#
-	# Every chip runs the full cross product: the size-bearing classes (flash
-	# addressing, hand-over shape, page size, USART inventory) are what make
-	# the image differ, and a chip outside them is expected to match its class
-	# — but "expected" is what a matrix is for, and the whole sweep is cheap
-	# enough to run rather than reason about. PUREBOOT_FULL_MATRIX is what
-	# selects it; the compact matrix below is the per-commit default.
-	get_property(_full_bauds GLOBAL PROPERTY PUREBOOT_BAUD_LADDER)
-	list(APPEND _full_bauds 16000 4800 2400 1200)
-	if(DEFINED ENV{PUREBOOT_FULL_MATRIX})
-		foreach(_matrix_hz IN LISTS _full_clocks)
-			foreach(_matrix_baud IN LISTS _full_bauds)
-				pureboot_matrix_point(${_matrix_hz} ${_matrix_baud} software "")
-				pureboot_matrix_point(${_matrix_hz} ${_matrix_baud} software "" ONE_WIRE)
-				if(PUREBOOT_HAS_USART)
-					pureboot_matrix_point(${_matrix_hz} ${_matrix_baud} software 0)
-					pureboot_matrix_point(${_matrix_hz} ${_matrix_baud} software 0 ONE_WIRE)
-					pureboot_matrix_point(${_matrix_hz} ${_matrix_baud} 0 "")
-					pureboot_matrix_point(${_matrix_hz} ${_matrix_baud} 0 "" HALF_DUPLEX)
-				endif()
-				if(PUREBOOT_HAS_USART1)
-					pureboot_matrix_point(${_matrix_hz} ${_matrix_baud} software 1)
-					pureboot_matrix_point(${_matrix_hz} ${_matrix_baud} software 1 ONE_WIRE)
-					pureboot_matrix_point(${_matrix_hz} ${_matrix_baud} 1 "")
-					pureboot_matrix_point(${_matrix_hz} ${_matrix_baud} 1 "" HALF_DUPLEX)
-				endif()
-			endforeach()
-		endforeach()
-	else()
-		foreach(_matrix_hz IN LISTS _matrix_clocks)
-			math(EXPR _matrix_khz "${_matrix_hz} / 1000")
-			if(PUREBOOT_HAS_USART OR NOT _matrix_hz EQUAL _pb_stock_hz)
-				pureboot_size_variant(pureboot_sw_${_matrix_khz}k CLOCK ${_matrix_hz} SERIAL software)
-			endif()
-			if(PUREBOOT_HAS_USART AND NOT _matrix_hz EQUAL _pb_stock_hz)
-				pureboot_size_variant(pureboot_hw_${_matrix_khz}k CLOCK ${_matrix_hz} SERIAL hardware)
-			endif()
-			if(PUREBOOT_HAS_USART1 AND NOT _matrix_hz EQUAL _pb_stock_hz)
-				pureboot_size_variant(pureboot_usart1_${_matrix_khz}k CLOCK ${_matrix_hz} USART 1)
-			endif()
-		endforeach()
-		list(GET _matrix_clocks -1 _matrix_top_hz)
-		pureboot_size_variant(pureboot_sw_wide CLOCK ${_matrix_top_hz} BAUD 9600 SERIAL software)
-		# The pin axis at the widest software image — the slowest ladder rate
-		# against the fastest clock, whose bit spin needs the 16-bit delay
-		# loop — with the USART release on top of it. The exhaustive sweep
-		# above carries the same axis across its whole cross product.
-		if(PUREBOOT_HAS_USART)
-			pureboot_size_variant(pureboot_sw_wide_on_usart0 CLOCK ${_matrix_top_hz} BAUD 9600
-			                      SERIAL software RX ${PUREBOOT_USART0_RX} TX ${PUREBOOT_USART0_TX})
 		endif()
-		if(PUREBOOT_HAS_USART1)
+		if(PUREBOOT_HAS_USART AND NOT _matrix_hz EQUAL _pb_stock_hz)
-			pureboot_size_variant(pureboot_sw_wide_on_usart1 CLOCK ${_matrix_top_hz} BAUD 9600
+			pureboot_size_variant(pureboot_hw_${_matrix_khz}k CLOCK ${_matrix_hz} SERIAL hardware)
-			                      SERIAL software RX ${PUREBOOT_USART1_RX} TX ${PUREBOOT_USART1_TX})
 		endif()
-	endif()
+	endforeach()
 	if(PUREBOOT_HAS_USART1)
 		pureboot_size_variant(pureboot_usart1 USART 1)
 	endif()
 
-	# The pin axis at its fixed points, in both matrix modes. The autobaud
-	# loader carries no clock and no baud, so the sweep has nothing to vary
-	# for it — yet it is the tightest image in the space, and on a USART's
-	# own pins it pays the release too: that combination is the one that
-	# overflowed the 1284's slot. The software build on those pins is the
-	# same deployment the mute test drives.
-	if(PUREBOOT_HAS_USART)
-		pureboot_size_variant(pureboot_sw_on_usart0 SERIAL software
-		                      RX ${PUREBOOT_USART0_RX} TX ${PUREBOOT_USART0_TX})
-		pureboot_size_variant(pureboot_autobaud_on_usart0 SERIAL autobaud
-		                      RX ${PUREBOOT_USART0_RX} TX ${PUREBOOT_USART0_TX})
-	endif()
-	if(PUREBOOT_HAS_USART1)
-		pureboot_size_variant(pureboot_sw_on_usart1 SERIAL software
-		                      RX ${PUREBOOT_USART1_RX} TX ${PUREBOOT_USART1_TX})
-		pureboot_size_variant(pureboot_autobaud_on_usart1 SERIAL autobaud
-		                      RX ${PUREBOOT_USART1_RX} TX ${PUREBOOT_USART1_TX})
-	endif()
-
-	# The OSCCAL axis at its fixed points: the stock shape, and the tightest
-	# image in the space with the trim on top — the axis adds one register
-	# write, and these points hold both of its addressing encodings to every
-	# chip's budget.
-	pureboot_size_variant(pureboot_osccal OSCCAL 0x9c)
-	pureboot_size_variant(pureboot_autobaud_osccal SERIAL autobaud OSCCAL 0x9c)
-	if(PUREBOOT_HAS_USART)
-		pureboot_size_variant(pureboot_autobaud_osccal_on_usart0 SERIAL autobaud OSCCAL 0x9c
-		                      RX ${PUREBOOT_USART0_RX} TX ${PUREBOOT_USART0_TX})
-	endif()
-
-	# The one-wire axis at its fixed points, in both matrix modes (the
-	# exhaustive sweep carries the same shapes across its cross product):
-	# the software link folded onto one pin, the tightest autobaud image
-	# likewise — on the default pin and on the USART's own RXD, whose
-	# release the now-driven shared pin needs where a receive-only link
-	# would not — and the hardware USART's half-duplex turn-around, stock
-	# and at the widest fixed-baud shape.
-	# The two spellings deliberately split across the two points: HALF_DUPLEX
-	# folds TX onto RX, RX == TX states the same thing directly.
-	pureboot_size_variant(pureboot_1w SERIAL software RX pb0 HALF_DUPLEX)
-	pureboot_size_variant(pureboot_1w_autobaud_osccal SERIAL autobaud OSCCAL 0x9c RX pb0 TX pb0)
-	if(PUREBOOT_HAS_USART)
-		pureboot_size_variant(pureboot_1w_on_usart0 SERIAL software
-		                      RX ${PUREBOOT_USART0_RX} TX ${PUREBOOT_USART0_RX})
-		pureboot_size_variant(pureboot_1w_autobaud_osccal_on_usart0 SERIAL autobaud OSCCAL 0x9c
-		                      RX ${PUREBOOT_USART0_RX} TX ${PUREBOOT_USART0_RX})
-		pureboot_size_variant(pureboot_hd HALF_DUPLEX)
-		list(GET _matrix_clocks -1 _hd_top_hz)
-		pureboot_size_variant(pureboot_hd_wide CLOCK ${_hd_top_hz} BAUD 9600 HALF_DUPLEX)
-	endif()
-	if(PUREBOOT_HAS_USART1)
-		pureboot_size_variant(pureboot_usart1_hd USART 1 HALF_DUPLEX)
-	endif()
-
-	# The trim byte, observed through the wire from the first prompt — one
-	# chip per OSCCAL addressing class: extended I/O on the 328P (data 0x66,
-	# an sts — DS40002061B §36), plain I/O on the 85 (data 0x51, an out —
-	# Atmel-2586 §21).
-	if(LIBAVR_MCU MATCHES "^(atmega328p|attiny85)$" AND DEFINED PB_DEVICE)
-		if(LIBAVR_MCU STREQUAL "atmega328p")
-			set(_osccal_addr 0x66)
-		else()
-			set(_osccal_addr 0x51)
-		endif()
-		get_target_property(_osccal_hz pureboot_osccal PUREBOOT_HZ)
-		get_target_property(_osccal_baud pureboot_osccal PUREBOOT_BAUD)
-		add_test(NAME pureboot.osccal
-		         COMMAND ${Python3_EXECUTABLE} ${CMAKE_CURRENT_SOURCE_DIR}/test/pbosccal.py
-		                 ${PB_DEVICE} $<TARGET_FILE:pureboot_osccal> ${PUREBOOT_SIM_MCU}
-		                 ${_osccal_hz} ${PUREBOOT_BASE_HEX} ${PUREBOOT_PAGE} ${_osccal_baud}
-		                 ${_osccal_addr} 0x9c ${CMAKE_CURRENT_SOURCE_DIR}/pureboot/pureboot.py
-		                 ${CMAKE_BINARY_DIR}/pbosccal-work)
-		set_tests_properties(pureboot.osccal PROPERTIES TIMEOUT 120)
-	endif()
-
 	# One configured deployment end to end — a real board's shape rather
 	# than the stock assumption: the ATmega328P on its shipped 1 MHz fuses,
 	# the software UART on hand-picked pins (TX = PB1, RX = PB5), the ladder
@@ -547,85 +277,6 @@ if(PROJECT_IS_TOP_LEVEL)
 		set_tests_properties(pureboot.custom PROPERTIES TIMEOUT 180)
 	endif()
 
-	# Hand-over with the USART that owns the loader's pins left enabled — the
-	# state an application reaches by jumping in without a reset, and the one
-	# that made a bit-banged loader on PD0/PD1 (where the Uno's USB bridge
-	# lands) receive and obey while answering nothing. Run where it was found
-	# on silicon; the runner supplies the pin ownership simavr has no model
-	# for, which is what lets this fail when the release is gone.
-	if(LIBAVR_MCU STREQUAL "atmega328p" AND DEFINED PB_DEVICE)
-		get_target_property(_mute_hz pureboot_sw_on_usart0 PUREBOOT_HZ)
-		get_target_property(_mute_baud pureboot_sw_on_usart0 PUREBOOT_BAUD)
-		get_target_property(_mute_link pureboot_sw_on_usart0 PUREBOOT_LINK)
-		add_executable(pbapp_handover test/pbapp.cpp)
-		target_link_libraries(pbapp_handover PRIVATE libavr)
-		target_compile_definitions(pbapp_handover PRIVATE PUREBOOT_CLOCK_HZ=${_mute_hz}
-		                           PUREBOOT_BAUD=${_mute_baud} PUREBOOT_HANDOVER)
-		add_custom_command(TARGET pbapp_handover POST_BUILD
-		                   COMMAND ${CMAKE_OBJCOPY} -O binary
-		                           $<TARGET_FILE:pbapp_handover> $<TARGET_FILE:pbapp_handover>.bin)
-		add_test(NAME pureboot.mute
-		         COMMAND ${Python3_EXECUTABLE} ${CMAKE_CURRENT_SOURCE_DIR}/test/pbmute.py
-		                 ${PB_DEVICE} $<TARGET_FILE:pureboot_sw_on_usart0> ${PUREBOOT_SIM_MCU} ${_mute_hz}
-		                 ${PUREBOOT_BASE_HEX} ${PUREBOOT_PAGE} ${_mute_baud}
-		                 $<TARGET_FILE:pbapp_handover>.bin
-		                 ${CMAKE_CURRENT_SOURCE_DIR}/pureboot/pureboot.py
-		                 ${CMAKE_BINARY_DIR}/pbmute-work ${_mute_link})
-		set_tests_properties(pureboot.mute PROPERTIES TIMEOUT 180)
-
-		# The same hand-over against the one-wire deployment on that USART's
-		# RXD: RXEN forces the shared pin's direction, so a loader that only
-		# released the transmit-side hold would read the wire and answer into
-		# a pin it cannot drive. The host runs with the --one-wire echo
-		# discard, which the bridge's shared-line model feeds for real.
-		get_target_property(_mute1w_link pureboot_1w_on_usart0 PUREBOOT_LINK)
-		add_test(NAME pureboot.mute.onewire
-		         COMMAND ${Python3_EXECUTABLE} ${CMAKE_CURRENT_SOURCE_DIR}/test/pbmute.py
-		                 ${PB_DEVICE} $<TARGET_FILE:pureboot_1w_on_usart0> ${PUREBOOT_SIM_MCU} ${_mute_hz}
-		                 ${PUREBOOT_BASE_HEX} ${PUREBOOT_PAGE} ${_mute_baud}
-		                 $<TARGET_FILE:pbapp_handover>.bin
-		                 ${CMAKE_CURRENT_SOURCE_DIR}/pureboot/pureboot.py
-		                 ${CMAKE_BINARY_DIR}/pbmute-1w-work ${_mute1w_link})
-		set_tests_properties(pureboot.mute.onewire PROPERTIES TIMEOUT 180)
-
-		# The full protocol suite over one shared pin: the loader folded onto
-		# PB0, the bridge following the pin's direction, the fixture
-		# bannering as a guest on the same line, and the host discarding its
-		# own echo throughout.
-		get_target_property(_1w_hz pureboot_1w PUREBOOT_HZ)
-		get_target_property(_1w_baud pureboot_1w PUREBOOT_BAUD)
-		get_target_property(_1w_link pureboot_1w PUREBOOT_LINK)
-		add_executable(pbapp_1w test/pbapp.cpp)
-		target_link_libraries(pbapp_1w PRIVATE libavr)
-		target_compile_definitions(pbapp_1w PRIVATE PUREBOOT_CLOCK_HZ=${_1w_hz}
-		                           PUREBOOT_BAUD=${_1w_baud} PUREBOOT_SOFT_SERIAL
-		                           PUREBOOT_RX=pb0 PUREBOOT_TX=pb0)
-		add_custom_command(TARGET pbapp_1w POST_BUILD
-		                   COMMAND ${CMAKE_OBJCOPY} -O binary
-		                           $<TARGET_FILE:pbapp_1w> $<TARGET_FILE:pbapp_1w>.bin)
-		add_test(NAME pureboot.onewire
-		         COMMAND ${Python3_EXECUTABLE} ${CMAKE_CURRENT_SOURCE_DIR}/test/pbtest.py
-		                 ${PB_DEVICE} $<TARGET_FILE:pureboot_1w> ${PUREBOOT_SIM_MCU} ${_1w_hz}
-		                 ${PUREBOOT_BASE_HEX} ${PUREBOOT_PAGE} ${_1w_baud} ${PUREBOOT_EEPROM}
-		                 $<TARGET_FILE:pbapp_1w>.bin ${CMAKE_CURRENT_SOURCE_DIR}/pureboot/pureboot.py
-		                 ${CMAKE_BINARY_DIR}/pb1w-work ${_1w_link})
-		set_tests_properties(pureboot.onewire PROPERTIES TIMEOUT 180)
-
-		# The hardware USART's half-duplex turn-around, end to end: every
-		# reply byte runs drive-line, TXC-hold, release — against simavr's
-		# RXEN-gated receiver, which drops input to a disabled receiver the
-		# way silicon does. The pty is a two-wire transport, so the host
-		# needs no echo discard here; the off-chip tie itself is the
-		# hardware bench's item.
-		add_test(NAME pureboot.halfduplex
-		         COMMAND ${Python3_EXECUTABLE} ${CMAKE_CURRENT_SOURCE_DIR}/test/pbtest.py
-		                 ${PB_DEVICE} $<TARGET_FILE:pureboot_hd> ${PUREBOOT_SIM_MCU} ${_pb_stock_hz}
-		                 ${PUREBOOT_BASE_HEX} ${PUREBOOT_PAGE} ${_pb_stock_baud} ${PUREBOOT_EEPROM}
-		                 $<TARGET_FILE:pbapp>.bin ${CMAKE_CURRENT_SOURCE_DIR}/pureboot/pureboot.py
-		                 ${CMAKE_BINARY_DIR}/pbhd-work)
-		set_tests_properties(pureboot.halfduplex PROPERTIES TIMEOUT 180)
-	endif()
-
 	# The second USART, driven for real on one chip: instance selection is
 	# compile-checked everywhere, but only a live session proves the loader
 	# initialized and polls the USART it claims to. The fixture application
@@ -648,67 +299,4 @@ if(PROJECT_IS_TOP_LEVEL)
 		                 ${CMAKE_BINARY_DIR}/pbusart1-work usart1)
 		set_tests_properties(pureboot.usart1 PROPERTIES TIMEOUT 180)
 	endif()
-
-	# The autobaud loader driven end to end over the software-UART bridge:
-	# the host sends the 0xC0 calibration pulse, the loader times it, locks,
-	# and programs. Run on the near-flash 328P
-	# and the word-addressed 1284P — the two flash-addressing classes — and each
-	# at two clocks with the one binary, which is the clock-agnostic property
-	# autobaud exists for (test/pbautobaud.py). The fixture application banners
-	# over the same software link at the first clock's rate.
-	if(LIBAVR_MCU MATCHES "^atmega(328p|1284p)$" AND DEFINED PB_DEVICE)
-		add_executable(pbapp_autobaud test/pbapp.cpp)
-		target_link_libraries(pbapp_autobaud PRIVATE libavr)
-		target_compile_definitions(pbapp_autobaud PRIVATE PUREBOOT_CLOCK_HZ=1000000
-		                           PUREBOOT_BAUD=9600 PUREBOOT_SOFT_SERIAL PUREBOOT_TX=pb1)
-		add_custom_command(TARGET pbapp_autobaud POST_BUILD
-		                   COMMAND ${CMAKE_OBJCOPY} -O binary
-		                           $<TARGET_FILE:pbapp_autobaud> $<TARGET_FILE:pbapp_autobaud>.bin)
-		add_test(NAME pureboot.autobaud
-		         COMMAND ${Python3_EXECUTABLE} ${CMAKE_CURRENT_SOURCE_DIR}/test/pbautobaud.py
-		                 ${PB_DEVICE} $<TARGET_FILE:pureboot_autobaud> ${PUREBOOT_SIM_MCU}
-		                 ${PUREBOOT_BASE_HEX} ${PUREBOOT_PAGE} $<TARGET_FILE:pbapp_autobaud>.bin
-		                 1000000 9600 ${CMAKE_CURRENT_SOURCE_DIR}/pureboot/pureboot.py
-		                 ${CMAKE_BINARY_DIR}/pbautobaud-work)
-		set_tests_properties(pureboot.autobaud PROPERTIES TIMEOUT 240)
-
-		# The tightest deployment in the space, end to end: the autobaud
-		# loader folded onto the USART's own RXD with the OSCCAL trim baked
-		# — one-wire calibration, the receive-side release, and the host's
-		# echo discard, over the same two-clock sweep. One chip carries it;
-		# the shape is chip-independent.
-		if(LIBAVR_MCU STREQUAL "atmega328p")
-			get_target_property(_ab1w_link pureboot_1w_autobaud_osccal_on_usart0 PUREBOOT_LINK)
-			add_executable(pbapp_autobaud_1w test/pbapp.cpp)
-			target_link_libraries(pbapp_autobaud_1w PRIVATE libavr)
-			target_compile_definitions(pbapp_autobaud_1w PRIVATE PUREBOOT_CLOCK_HZ=1000000
-			                           PUREBOOT_BAUD=9600 PUREBOOT_SOFT_SERIAL
-			                           PUREBOOT_RX=pd0 PUREBOOT_TX=pd0)
-			add_custom_command(TARGET pbapp_autobaud_1w POST_BUILD
-			                   COMMAND ${CMAKE_OBJCOPY} -O binary
-			                           $<TARGET_FILE:pbapp_autobaud_1w> $<TARGET_FILE:pbapp_autobaud_1w>.bin)
-			add_test(NAME pureboot.autobaud.onewire
-			         COMMAND ${Python3_EXECUTABLE} ${CMAKE_CURRENT_SOURCE_DIR}/test/pbautobaud.py
-			                 ${PB_DEVICE} $<TARGET_FILE:pureboot_1w_autobaud_osccal_on_usart0>
-			                 ${PUREBOOT_SIM_MCU} ${PUREBOOT_BASE_HEX} ${PUREBOOT_PAGE}
-			                 $<TARGET_FILE:pbapp_autobaud_1w>.bin
-			                 1000000 9600 ${CMAKE_CURRENT_SOURCE_DIR}/pureboot/pureboot.py
-			                 ${CMAKE_BINARY_DIR}/pbautobaud-1w-work ${_ab1w_link})
-			set_tests_properties(pureboot.autobaud.onewire PROPERTIES TIMEOUT 240)
-		endif()
-
-		# The autobaud window: the calibration poll budget, at the measured
-		# 10 cycles a poll (pbwindow.py pins the constant the README's
-		# seconds arithmetic uses; the budget itself is the clock-free knob).
-		add_test(NAME pureboot.window.autobaud
-		         COMMAND ${Python3_EXECUTABLE} ${CMAKE_CURRENT_SOURCE_DIR}/test/pbwindow.py
-		                 --device ${PB_DEVICE} --loader $<TARGET_FILE:pureboot_autobaud>
-		                 --mcu ${PUREBOOT_SIM_MCU} --hz 1000000
-		                 --base ${PUREBOOT_BASE_HEX} --page ${PUREBOOT_PAGE}
-		                 --baud 9600 --app $<TARGET_FILE:pbapp_autobaud>.bin
-		                 --autobaud-polls 4000000 --link sw
-		                 --tool ${CMAKE_CURRENT_SOURCE_DIR}/pureboot/pureboot.py
-		                 --workdir ${CMAKE_BINARY_DIR}/pbwindow-autobaud-work)
-		set_tests_properties(pureboot.window.autobaud PROPERTIES TIMEOUT 300)
-	endif()
 endif()

CMakePresets.json

@@ -6,7 +6,7 @@
    "hidden": true,
    "generator": "Ninja",
    "binaryDir": "${sourceDir}/build/${presetName}",
-   "toolchainFile": "${sourceDir}/libavr/cmake/avr-toolchain.cmake",
+   "toolchainFile": "$env{LIBAVR_ROOT}/cmake/avr-toolchain.cmake",
    "cacheVariables": {
     "CMAKE_BUILD_TYPE": "Release",
     "CMAKE_EXPORT_COMPILE_COMMANDS": "ON",

ide/README.md

@@ -1,77 +0,0 @@
-# Atmel Studio
-
-`master` carries `bootloader.atsln`, so this branch does too: `ide/bootloader.atsln`
-builds the loaders from the same sources Ninja does, to a **byte-identical
-`.text`** — 390 B for the 328P pureboot loader, 510 B for the `tsb_asm` tier in
-its 512-byte section. CMake remains the build system; the solution is here so the
-port opens in Studio as its predecessor did.
-
-## The two projects, and why two
-
-pureboot is a chip × backend × clock × baud matrix — `pureboot_add_loader()`
-resolves a deployment into compile definitions — and a `.cppproj` is one binary
-at one set of flags, so a project can only ever be one point of it. `pureboot`
-is that point: the stock 328P deployment, USART0 at 115200 on a 16 MHz crystal,
-an 8-second activation window. `tsb_asm` is the TinySafeBoot tier that occupies
-the same 512-byte section `master`'s `tsb` project targeted.
-
-The other three tsb tiers (`tsb_pure`, `tsb_tricks`, `tsb_policy`) are not here.
-They differ from `tsb_asm` in their source file, their section size, and — for
-`tsb_policy` — two loop flags; nothing about that is a Studio concern, and what
-they exist to demonstrate is a size gradient only the CMake size tests measure.
-Adding one is a copy of `tsb_asm/tsb_asm.cppproj` in its own directory, with its
-name, its GUID, its source path and its `--section-start` changed (`0x7c00` for
-the 1 KiB tiers), plus four lines in the solution.
-
-`avrdevice` is a project property, so each project gets its own directory:
-Studio builds into `<project dir>/<Configuration>` whatever `OutputDirectory`
-says, and two projects sharing a directory would share one object file.
-
-## Debug keeps `-Os`
-
-Both configurations compile at `-Os`; Debug adds only `-gdwarf-4`. The `.text`
-is therefore identical in both, which is the point — a loader's section is a
-**correctness** bound and not a budget. `-Og` builds this same source to 590 B,
-and linking it at `--section-start=.text=0x7e00` on a 32 KiB part puts 78 bytes
-past flash end **without a diagnostic**: `rcall`/`rjmp` targets there wrap
-modulo flash size, so the image dies right after activation. A debug
-configuration that silently produces that is worse than none, and DWARF costs no
-flash, so the optimisation level stays where correctness needs it.
-
-## What Studio needs from the machine
-
-libavr from the **submodule**, found at
-`$(MSBuildProjectDirectory)\..\..\libavr\include` — correct by construction, and
-anchored to the project because a plain relative path resolves against the
-generated makefile's directory (the configuration's output directory), not the
-project's. There is no `LIBAVR_ROOT` escape hatch: a variable exported in a
-shell is invisible to Studio launched from the Start menu, and the failure reads
-as a missing `libavr/libavr.hpp` — which is what the submodule answers.
-
-A GCC 16.1 toolchain registered as flavour `avr-g++-16.1.0`, nothing older
-reaching `-std=c++26`.
-
-## Generating and gating
-
-One generated file is required before a project will load at all, and one command
-per project checks the flags have not drifted (both from libavr's
-`tools/atmelstudio/`):
-
-```sh
-for name in pureboot tsb_asm; do
-    python libavr/tools/atmelstudio/componentinfo.py \
-        "ide/$name/$name.componentinfo.xml" --device ATmega328P
-    python libavr/tools/atmelstudio/check-flags.py \
-        --solution ide/bootloader.atsln --project "$name" --target "$name" \
-        --compile-commands build/atmega328p-generated/compile_commands.json \
-        --log "build/as-$name.log"
-done
-```
-
-`--project` because one reference describes one binary; `--target` because
-`pureboot.cpp` is compiled by every point of the size matrix and the flags
-differ per point, so the basename alone does not name a reference. Release is
-what the gate compares — the presets define no debug build, and Debug differs
-from Release only in `-gdwarf-4`.
-
-Legacy (the yazoalfa-era submodules) stays on `master`.

ide/bootloader.atsln

@@ -1,28 +0,0 @@
-
-Microsoft Visual Studio Solution File, Format Version 12.00
-# Atmel Studio Solution File, Format Version 11.00
-VisualStudioVersion = 14.0.23107.0
-MinimumVisualStudioVersion = 10.0.40219.1
-Project("{E66E83B9-2572-4076-B26E-6BE79FF3018A}") = "pureboot", "pureboot\pureboot.cppproj", "{99067222-32D5-49E3-B4F8-5ABA0F7722B7}"
-EndProject
-Project("{E66E83B9-2572-4076-B26E-6BE79FF3018A}") = "tsb_asm", "tsb_asm\tsb_asm.cppproj", "{6618D3BE-7EB3-49A2-9113-F128E396FF06}"
-EndProject
-Global
-	GlobalSection(SolutionConfigurationPlatforms) = preSolution
-		Debug|AVR = Debug|AVR
-		Release|AVR = Release|AVR
-	EndGlobalSection
-	GlobalSection(ProjectConfigurationPlatforms) = postSolution
-		{99067222-32D5-49E3-B4F8-5ABA0F7722B7}.Debug|AVR.ActiveCfg = Debug|AVR
-		{99067222-32D5-49E3-B4F8-5ABA0F7722B7}.Debug|AVR.Build.0 = Debug|AVR
-		{99067222-32D5-49E3-B4F8-5ABA0F7722B7}.Release|AVR.ActiveCfg = Release|AVR
-		{99067222-32D5-49E3-B4F8-5ABA0F7722B7}.Release|AVR.Build.0 = Release|AVR
-		{6618D3BE-7EB3-49A2-9113-F128E396FF06}.Debug|AVR.ActiveCfg = Debug|AVR
-		{6618D3BE-7EB3-49A2-9113-F128E396FF06}.Debug|AVR.Build.0 = Debug|AVR
-		{6618D3BE-7EB3-49A2-9113-F128E396FF06}.Release|AVR.ActiveCfg = Release|AVR
-		{6618D3BE-7EB3-49A2-9113-F128E396FF06}.Release|AVR.Build.0 = Release|AVR
-	EndGlobalSection
-	GlobalSection(SolutionProperties) = preSolution
-		HideSolutionNode = FALSE
-	EndGlobalSection
-EndGlobal

ide/pureboot/pureboot.cppproj

@@ -1,118 +0,0 @@
-<?xml version="1.0" encoding="utf-8"?>
-<Project DefaultTargets="Build" xmlns="http://schemas.microsoft.com/developer/msbuild/2003" ToolsVersion="14.0">
-  <PropertyGroup>
-    <SchemaVersion>2.0</SchemaVersion>
-    <ProjectVersion>7.0</ProjectVersion>
-    <ToolchainName>com.Atmel.AVRGCC8.CPP</ToolchainName>
-    <ProjectGuid>99067222-32d5-49e3-b4f8-5aba0f7722b7</ProjectGuid>
-    <avrdevice>ATmega328P</avrdevice>
-    <avrdeviceseries>none</avrdeviceseries>
-    <OutputType>Executable</OutputType>
-    <Language>CPP</Language>
-    <OutputFileName>$(MSBuildProjectName)</OutputFileName>
-    <OutputFileExtension>.elf</OutputFileExtension>
-    <OutputDirectory>$(MSBuildProjectDirectory)\$(Configuration)</OutputDirectory>
-    <AssemblyName>pureboot</AssemblyName>
-    <Name>pureboot</Name>
-    <RootNamespace>pureboot</RootNamespace>
-    <ToolchainFlavour>avr-g++-16.1.0</ToolchainFlavour>
-    <KeepTimersRunning>true</KeepTimersRunning>
-    <OverrideVtor>false</OverrideVtor>
-    <CacheFlash>true</CacheFlash>
-    <ProgFlashFromRam>true</ProgFlashFromRam>
-    <RamSnippetAddress>0x20000000</RamSnippetAddress>
-    <UncachedRange />
-    <preserveEEPROM>true</preserveEEPROM>
-    <OverrideVtorValue>exception_table</OverrideVtorValue>
-    <BootSegment>2</BootSegment>
-    <ResetRule>0</ResetRule>
-    <eraseonlaunchrule>0</eraseonlaunchrule>
-    <EraseKey />
-    <AsfFrameworkConfig>
-      <framework-data xmlns="">
-  <options />
-  <configurations />
-  <files />
-  <documentation help="" />
-  <offline-documentation help="" />
-  <dependencies>
-    <content-extension eid="atmel.asf" uuidref="Atmel.ASF" version="3.52.0" />
-  </dependencies>
-</framework-data>
-    </AsfFrameworkConfig>
-  </PropertyGroup>
-  <PropertyGroup Condition=" '$(Configuration)' == 'Release' ">
-    <ToolchainSettings>
-      <AvrGccCpp>
-        <avrgcc.common.Device>-mmcu=atmega328p</avrgcc.common.Device>
-        <avrgcc.common.outputfiles.hex>True</avrgcc.common.outputfiles.hex>
-        <avrgcc.common.outputfiles.lss>True</avrgcc.common.outputfiles.lss>
-        <avrgcc.common.outputfiles.eep>True</avrgcc.common.outputfiles.eep>
-        <avrgcc.common.outputfiles.srec>True</avrgcc.common.outputfiles.srec>
-        <avrgcc.common.outputfiles.usersignatures>False</avrgcc.common.outputfiles.usersignatures>
-        <avrgcccpp.compiler.symbols.DefSymbols>
-          <ListValues>
-            <Value>NDEBUG</Value>
-            <Value>PUREBOOT_CLOCK_HZ=16000000</Value>
-            <Value>PUREBOOT_BAUD=115200</Value>
-            <Value>PUREBOOT_TIMEOUT=8</Value>
-          </ListValues>
-        </avrgcccpp.compiler.symbols.DefSymbols>
-        <avrgcccpp.compiler.directories.IncludePaths>
-          <ListValues>
-            <Value>$(MSBuildProjectDirectory)\..\..\libavr\include</Value>
-          </ListValues>
-        </avrgcccpp.compiler.directories.IncludePaths>
-        <avrgcccpp.compiler.optimization.level>Optimize for size (-Os)</avrgcccpp.compiler.optimization.level>
-        <avrgcccpp.compiler.optimization.PrepareFunctionsForGarbageCollection>True</avrgcccpp.compiler.optimization.PrepareFunctionsForGarbageCollection>
-        <avrgcccpp.compiler.optimization.PrepareDataForGarbageCollection>True</avrgcccpp.compiler.optimization.PrepareDataForGarbageCollection>
-        <avrgcccpp.compiler.warnings.AllWarnings>True</avrgcccpp.compiler.warnings.AllWarnings>
-        <avrgcccpp.compiler.miscellaneous.OtherFlags>-std=c++26 -Wextra -Werror -mrelax -fno-exceptions -fno-rtti -fno-threadsafe-statics -fno-ivopts -fira-algorithm=priority -fno-tree-ter -fno-split-wide-types</avrgcccpp.compiler.miscellaneous.OtherFlags>
-        <avrgcccpp.linker.optimization.GarbageCollectUnusedSections>True</avrgcccpp.linker.optimization.GarbageCollectUnusedSections>
-        <avrgcccpp.linker.miscellaneous.LinkerFlags>-mrelax -nostartfiles -Wl,--section-start=.text=0x7e00 -Wl,--defsym=pureboot_app=0 -Wl,--pmem-wrap-around=32k</avrgcccpp.linker.miscellaneous.LinkerFlags>
-      </AvrGccCpp>
-    </ToolchainSettings>
-  </PropertyGroup>
-  <PropertyGroup Condition=" '$(Configuration)' == 'Debug' ">
-    <ToolchainSettings>
-      <AvrGccCpp>
-        <avrgcc.common.Device>-mmcu=atmega328p</avrgcc.common.Device>
-        <avrgcc.common.outputfiles.hex>True</avrgcc.common.outputfiles.hex>
-        <avrgcc.common.outputfiles.lss>True</avrgcc.common.outputfiles.lss>
-        <avrgcc.common.outputfiles.eep>True</avrgcc.common.outputfiles.eep>
-        <avrgcc.common.outputfiles.srec>True</avrgcc.common.outputfiles.srec>
-        <avrgcc.common.outputfiles.usersignatures>False</avrgcc.common.outputfiles.usersignatures>
-        <avrgcccpp.compiler.symbols.DefSymbols>
-          <ListValues>
-            <Value>DEBUG</Value>
-            <Value>PUREBOOT_CLOCK_HZ=16000000</Value>
-            <Value>PUREBOOT_BAUD=115200</Value>
-            <Value>PUREBOOT_TIMEOUT=8</Value>
-          </ListValues>
-        </avrgcccpp.compiler.symbols.DefSymbols>
-        <avrgcccpp.compiler.directories.IncludePaths>
-          <ListValues>
-            <Value>$(MSBuildProjectDirectory)\..\..\libavr\include</Value>
-          </ListValues>
-        </avrgcccpp.compiler.directories.IncludePaths>
-        <avrgcccpp.compiler.optimization.level>Optimize for size (-Os)</avrgcccpp.compiler.optimization.level>
-        <avrgcccpp.compiler.optimization.PrepareFunctionsForGarbageCollection>True</avrgcccpp.compiler.optimization.PrepareFunctionsForGarbageCollection>
-        <avrgcccpp.compiler.optimization.PrepareDataForGarbageCollection>True</avrgcccpp.compiler.optimization.PrepareDataForGarbageCollection>
-        <avrgcccpp.compiler.warnings.AllWarnings>True</avrgcccpp.compiler.warnings.AllWarnings>
-        <avrgcccpp.compiler.miscellaneous.OtherFlags>-std=c++26 -Wextra -Werror -mrelax -fno-exceptions -fno-rtti -fno-threadsafe-statics -fno-ivopts -fira-algorithm=priority -fno-tree-ter -fno-split-wide-types -gdwarf-4</avrgcccpp.compiler.miscellaneous.OtherFlags>
-        <avrgcccpp.linker.optimization.GarbageCollectUnusedSections>True</avrgcccpp.linker.optimization.GarbageCollectUnusedSections>
-        <avrgcccpp.linker.miscellaneous.LinkerFlags>-mrelax -nostartfiles -Wl,--section-start=.text=0x7e00 -Wl,--defsym=pureboot_app=0 -Wl,--pmem-wrap-around=32k</avrgcccpp.linker.miscellaneous.LinkerFlags>
-      </AvrGccCpp>
-    </ToolchainSettings>
-  </PropertyGroup>
-  <ItemGroup>
-    <Compile Include="..\..\pureboot\pureboot.cpp">
-      <SubType>compile</SubType>
-      <Link>pureboot\pureboot.cpp</Link>
-    </Compile>
-  </ItemGroup>
-  <ItemGroup>
-    <Folder Include="pureboot" />
-  </ItemGroup>
-  <Import Project="$(AVRSTUDIO_EXE_PATH)\Vs\Compiler.targets" />
-</Project>

ide/tsb_asm/tsb_asm.cppproj

@@ -1,112 +0,0 @@
-<?xml version="1.0" encoding="utf-8"?>
-<Project DefaultTargets="Build" xmlns="http://schemas.microsoft.com/developer/msbuild/2003" ToolsVersion="14.0">
-  <PropertyGroup>
-    <SchemaVersion>2.0</SchemaVersion>
-    <ProjectVersion>7.0</ProjectVersion>
-    <ToolchainName>com.Atmel.AVRGCC8.CPP</ToolchainName>
-    <ProjectGuid>6618d3be-7eb3-49a2-9113-f128e396ff06</ProjectGuid>
-    <avrdevice>ATmega328P</avrdevice>
-    <avrdeviceseries>none</avrdeviceseries>
-    <OutputType>Executable</OutputType>
-    <Language>CPP</Language>
-    <OutputFileName>$(MSBuildProjectName)</OutputFileName>
-    <OutputFileExtension>.elf</OutputFileExtension>
-    <OutputDirectory>$(MSBuildProjectDirectory)\$(Configuration)</OutputDirectory>
-    <AssemblyName>tsb_asm</AssemblyName>
-    <Name>tsb_asm</Name>
-    <RootNamespace>tsb_asm</RootNamespace>
-    <ToolchainFlavour>avr-g++-16.1.0</ToolchainFlavour>
-    <KeepTimersRunning>true</KeepTimersRunning>
-    <OverrideVtor>false</OverrideVtor>
-    <CacheFlash>true</CacheFlash>
-    <ProgFlashFromRam>true</ProgFlashFromRam>
-    <RamSnippetAddress>0x20000000</RamSnippetAddress>
-    <UncachedRange />
-    <preserveEEPROM>true</preserveEEPROM>
-    <OverrideVtorValue>exception_table</OverrideVtorValue>
-    <BootSegment>2</BootSegment>
-    <ResetRule>0</ResetRule>
-    <eraseonlaunchrule>0</eraseonlaunchrule>
-    <EraseKey />
-    <AsfFrameworkConfig>
-      <framework-data xmlns="">
-  <options />
-  <configurations />
-  <files />
-  <documentation help="" />
-  <offline-documentation help="" />
-  <dependencies>
-    <content-extension eid="atmel.asf" uuidref="Atmel.ASF" version="3.52.0" />
-  </dependencies>
-</framework-data>
-    </AsfFrameworkConfig>
-  </PropertyGroup>
-  <PropertyGroup Condition=" '$(Configuration)' == 'Release' ">
-    <ToolchainSettings>
-      <AvrGccCpp>
-        <avrgcc.common.Device>-mmcu=atmega328p</avrgcc.common.Device>
-        <avrgcc.common.outputfiles.hex>True</avrgcc.common.outputfiles.hex>
-        <avrgcc.common.outputfiles.lss>True</avrgcc.common.outputfiles.lss>
-        <avrgcc.common.outputfiles.eep>True</avrgcc.common.outputfiles.eep>
-        <avrgcc.common.outputfiles.srec>True</avrgcc.common.outputfiles.srec>
-        <avrgcc.common.outputfiles.usersignatures>False</avrgcc.common.outputfiles.usersignatures>
-        <avrgcccpp.compiler.symbols.DefSymbols>
-          <ListValues>
-            <Value>NDEBUG</Value>
-          </ListValues>
-        </avrgcccpp.compiler.symbols.DefSymbols>
-        <avrgcccpp.compiler.directories.IncludePaths>
-          <ListValues>
-            <Value>$(MSBuildProjectDirectory)\..\..\libavr\include</Value>
-          </ListValues>
-        </avrgcccpp.compiler.directories.IncludePaths>
-        <avrgcccpp.compiler.optimization.level>Optimize for size (-Os)</avrgcccpp.compiler.optimization.level>
-        <avrgcccpp.compiler.optimization.PrepareFunctionsForGarbageCollection>True</avrgcccpp.compiler.optimization.PrepareFunctionsForGarbageCollection>
-        <avrgcccpp.compiler.optimization.PrepareDataForGarbageCollection>True</avrgcccpp.compiler.optimization.PrepareDataForGarbageCollection>
-        <avrgcccpp.compiler.warnings.AllWarnings>True</avrgcccpp.compiler.warnings.AllWarnings>
-        <avrgcccpp.compiler.miscellaneous.OtherFlags>-std=c++26 -Wextra -Werror -mrelax -fno-exceptions -fno-rtti -fno-threadsafe-statics</avrgcccpp.compiler.miscellaneous.OtherFlags>
-        <avrgcccpp.linker.optimization.GarbageCollectUnusedSections>True</avrgcccpp.linker.optimization.GarbageCollectUnusedSections>
-        <avrgcccpp.linker.miscellaneous.LinkerFlags>-mrelax -nostartfiles -Wl,--section-start=.text=0x7e00 -Wl,--defsym=tsb_app=0 -Wl,--pmem-wrap-around=32k</avrgcccpp.linker.miscellaneous.LinkerFlags>
-      </AvrGccCpp>
-    </ToolchainSettings>
-  </PropertyGroup>
-  <PropertyGroup Condition=" '$(Configuration)' == 'Debug' ">
-    <ToolchainSettings>
-      <AvrGccCpp>
-        <avrgcc.common.Device>-mmcu=atmega328p</avrgcc.common.Device>
-        <avrgcc.common.outputfiles.hex>True</avrgcc.common.outputfiles.hex>
-        <avrgcc.common.outputfiles.lss>True</avrgcc.common.outputfiles.lss>
-        <avrgcc.common.outputfiles.eep>True</avrgcc.common.outputfiles.eep>
-        <avrgcc.common.outputfiles.srec>True</avrgcc.common.outputfiles.srec>
-        <avrgcc.common.outputfiles.usersignatures>False</avrgcc.common.outputfiles.usersignatures>
-        <avrgcccpp.compiler.symbols.DefSymbols>
-          <ListValues>
-            <Value>DEBUG</Value>
-          </ListValues>
-        </avrgcccpp.compiler.symbols.DefSymbols>
-        <avrgcccpp.compiler.directories.IncludePaths>
-          <ListValues>
-            <Value>$(MSBuildProjectDirectory)\..\..\libavr\include</Value>
-          </ListValues>
-        </avrgcccpp.compiler.directories.IncludePaths>
-        <avrgcccpp.compiler.optimization.level>Optimize for size (-Os)</avrgcccpp.compiler.optimization.level>
-        <avrgcccpp.compiler.optimization.PrepareFunctionsForGarbageCollection>True</avrgcccpp.compiler.optimization.PrepareFunctionsForGarbageCollection>
-        <avrgcccpp.compiler.optimization.PrepareDataForGarbageCollection>True</avrgcccpp.compiler.optimization.PrepareDataForGarbageCollection>
-        <avrgcccpp.compiler.warnings.AllWarnings>True</avrgcccpp.compiler.warnings.AllWarnings>
-        <avrgcccpp.compiler.miscellaneous.OtherFlags>-std=c++26 -Wextra -Werror -mrelax -fno-exceptions -fno-rtti -fno-threadsafe-statics -gdwarf-4</avrgcccpp.compiler.miscellaneous.OtherFlags>
-        <avrgcccpp.linker.optimization.GarbageCollectUnusedSections>True</avrgcccpp.linker.optimization.GarbageCollectUnusedSections>
-        <avrgcccpp.linker.miscellaneous.LinkerFlags>-mrelax -nostartfiles -Wl,--section-start=.text=0x7e00 -Wl,--defsym=tsb_app=0 -Wl,--pmem-wrap-around=32k</avrgcccpp.linker.miscellaneous.LinkerFlags>
-      </AvrGccCpp>
-    </ToolchainSettings>
-  </PropertyGroup>
-  <ItemGroup>
-    <Compile Include="..\..\tsb\tsb_asm.cpp">
-      <SubType>compile</SubType>
-      <Link>tsb\tsb_asm.cpp</Link>
-    </Compile>
-  </ItemGroup>
-  <ItemGroup>
-    <Folder Include="tsb" />
-  </ItemGroup>
-  <Import Project="$(AVRSTUDIO_EXE_PATH)\Vs\Compiler.targets" />
-</Project>

pureboot/CMakeLists.txt

@@ -1,16 +1,27 @@
-# pureboot as a consumable CMake unit: the per-chip geometry, the default baud
+# pureboot as a consumable CMake unit: the per-chip geometry, the default
-# ladder, and pureboot_add_loader() — the one way a loader target is created.
+# baud ladder, and pureboot_add_loader() — the one way a loader target is
-# A downstream project brings its usual libavr setup (the `libavr` target and
+# created, both by this port's own build and by a downstream project. A
+# downstream project brings its usual libavr setup (the `libavr` target and
 # the LIBAVR_MCU toolchain preset), adds this directory, and states its
-# deployment; every argument is optional (README.md):
+# deployment:
 #
 #   add_subdirectory(bootloader/pureboot)
 #   pureboot_add_loader(myboot CLOCK 1000000 SERIAL software TX pb1 RX pb5)
+#
+# Every argument is optional — CLOCK defaults to the family assumption
+# below, BAUD to the fastest standard rate the clock reaches within 2.5 %
+# (the ladder), SERIAL to the chip's hardware USART where it has one
+# (`hardware`/`software` force a backend, USART 1 picks the second
+# instance), RX/TX to pb0/pb1 for the software UART, TIMEOUT to 8 s.
+# Infeasible picks fail the build by name: libavr's baud-error and
+# software-UART cycle-floor static asserts re-check whatever is passed.
 
-# Per-family geometry, deployment defaults, and the linker wrap the PC modulo
+# Per-family geometry: flash/page/EEPROM sizes and the linker wrap the PC
-# needs. The slot is 512 bytes on every chip. The USART flags mirror the
+# modulo needs, the loader slot (each chip's smallest boot sector — 1 KiB on
-# hardware inventory the loader's own static asserts check — the plain 644 is
+# the word-addressed 1284s), and the deployment defaults (crystal assumption
-# the x4 family's one single-USART die (Atmel-2593).
+# on the megas, calibrated RC on the tinies). The USART flags mirror the
+# hardware inventory the loader's own static asserts check (the plain 644 is
+# the x4 family's one single-USART die, Atmel-2593).
 set(_pb_has_usart 1)
 set(_pb_has_usart1 0)
 if(LIBAVR_MCU MATCHES "^attiny13a?$")
@@ -80,9 +91,10 @@ elseif(LIBAVR_MCU MATCHES "^atmega324(a|p|pa)$")
 	set(_pb_eeprom 1024)
 	set(_pb_has_usart1 1)
 elseif(LIBAVR_MCU MATCHES "^atmega644(a|p|pa)?$")
-	# 64 KiB is exactly the 16-bit byte space, so plain LPM still reaches
+	# 64 KiB is exactly the 16-bit byte space: plain LPM reaches everything,
-	# everything and the wire stays byte-addressed. The plain 644 is the
+	# and the smallest boot section (1 KiB) holds the loader and its staging
-	# family's one single-USART die.
+	# slot together (see README.md). The plain 644 is the family's one
+	# single-USART die.
 	set(_pb_flash 65536)
 	set(_pb_wrap -Wl,--pmem-wrap-around=64k)
 	set(_pb_page 256)
@@ -92,41 +104,38 @@ elseif(LIBAVR_MCU MATCHES "^atmega644(a|p|pa)?$")
 		set(_pb_has_usart1 1)
 	endif()
 elseif(LIBAVR_MCU MATCHES "^atmega1284p?$")
-	# 128 KiB: wire addresses are words, reads go through ELPM, and the PC's
+	# 128 KiB: wire flash addresses are word addresses, reads go through
-	# modulo wrap exceeds what --pmem-wrap-around models.
+	# ELPM, and the PC's modulo wrap exceeds what --pmem-wrap-around models.
+	# The slot is 1 KiB — this chip's own smallest boot sector; the far
+	# machinery cannot fit 512 B (see README.md).
 	set(_pb_flash 131072)
 	set(_pb_wrap "")
 	set(_pb_page 256)
 	set(_pb_hz 16000000)
 	set(_pb_eeprom 4096)
+	set(_pb_slot 1024)
+	set(_pb_limit 1024)
 	set(_pb_has_usart1 1)
 else()
 	message(FATAL_ERROR "pureboot: no geometry for ${LIBAVR_MCU}")
 endif()
-set(_pb_slot 512)
+if(NOT DEFINED _pb_slot)
+	set(_pb_slot 512)
+endif()
 math(EXPR _pb_base "${_pb_flash} - ${_pb_slot}")
 math(EXPR _pb_base_hex "${_pb_base}" OUTPUT_FORMAT HEXADECIMAL)
 # Patched-vector chips hand over through the trampoline word below the slot,
 # which is also the slot's own last word — their budget is slot − 2.
 if(LIBAVR_MCU MATCHES "^atmega" AND NOT LIBAVR_MCU MATCHES "^atmega48")
 	set(_pb_app 0)
-	set(_pb_limit ${_pb_slot})
+	if(NOT DEFINED _pb_limit)
+		set(_pb_limit ${_pb_slot})
+	endif()
 else()
 	math(EXPR _pb_app "${_pb_base} - 2")
 	math(EXPR _pb_limit "${_pb_slot} - 2")
 endif()
 
-# The pins each USART owns. A bit-banged link deployed on them has to release
-# that USART before it can drive the line, and those instructions are the one
-# way the choice of pins moves the image — so a size matrix needs them as an
-# axis even though pins are otherwise immediate operands. Uniform across every
-# mega libavr covers: USART0 (the classics' un-numbered USART included) on
-# PD0/PD1, USART1 on PD2/PD3.
-set(_pb_usart0_rx pd0)
-set(_pb_usart0_tx pd1)
-set(_pb_usart1_rx pd2)
-set(_pb_usart1_tx pd3)
-
 # simavr names its cores after the base dies; the A revisions run on them
 # (the 644PA on the 644P core).
 set(_pb_sim_mcu ${LIBAVR_MCU})
@@ -136,28 +145,6 @@ elseif(LIBAVR_MCU STREQUAL "atmega644pa")
 	set(_pb_sim_mcu atmega644p)
 endif()
 
-# Where SRAM begins: the classic megas keep it right after the plain I/O
-# registers, the x8/x4 generations push it past their extended I/O file, and
-# the tinies match the classics. An autobaud loader keeps its measured unit
-# in GPIOR2:GPIOR1 wherever the chip has the pair (data 0x32 on the
-# t25/45/85, 0x4A from the x8 generation on) and as the first RAM object at
-# SRAM start where it does not (the t13s and classic megas). The host reads
-# whichever home applies (pureboot.py's geometry), and the unit-position
-# test holds the image to the same split.
-if(LIBAVR_MCU MATCHES "^atmega(8|16|32)a?$")
-	set(_pb_ram 0x60)
-	set(_pb_unit_gpior "")
-elseif(LIBAVR_MCU MATCHES "^atmega")
-	set(_pb_ram 0x100)
-	set(_pb_unit_gpior 0x4A)
-elseif(LIBAVR_MCU MATCHES "^attiny13")
-	set(_pb_ram 0x60)
-	set(_pb_unit_gpior "")
-else()
-	set(_pb_ram 0x60)
-	set(_pb_unit_gpior 0x32)
-endif()
-
 # The function runs in its caller's scope, so everything it needs crosses
 # scopes as global properties.
 set_property(GLOBAL PROPERTY PUREBOOT_BASE_HEX ${_pb_base_hex})
@@ -166,10 +153,6 @@ set_property(GLOBAL PROPERTY PUREBOOT_WRAP "${_pb_wrap}")
 set_property(GLOBAL PROPERTY PUREBOOT_DEFAULT_HZ ${_pb_hz})
 set_property(GLOBAL PROPERTY PUREBOOT_HAS_USART ${_pb_has_usart})
 set_property(GLOBAL PROPERTY PUREBOOT_HAS_USART1 ${_pb_has_usart1})
-set_property(GLOBAL PROPERTY PUREBOOT_USART0_RX ${_pb_usart0_rx})
-set_property(GLOBAL PROPERTY PUREBOOT_USART0_TX ${_pb_usart0_tx})
-set_property(GLOBAL PROPERTY PUREBOOT_USART1_RX ${_pb_usart1_rx})
-set_property(GLOBAL PROPERTY PUREBOOT_USART1_TX ${_pb_usart1_tx})
 
 # The port's own build (tests, the size matrix) reads the geometry from the
 # parent scope; a downstream consumer gets the same variables for free.
@@ -179,93 +162,54 @@ set(PUREBOOT_SLOT ${_pb_slot} PARENT_SCOPE)
 set(PUREBOOT_LIMIT ${_pb_limit} PARENT_SCOPE)
 set(PUREBOOT_EEPROM ${_pb_eeprom} PARENT_SCOPE)
 set(PUREBOOT_DEFAULT_HZ ${_pb_hz} PARENT_SCOPE)
-set(PUREBOOT_RAM_START ${_pb_ram} PARENT_SCOPE)
-set(PUREBOOT_UNIT_GPIOR "${_pb_unit_gpior}" PARENT_SCOPE)
 set(PUREBOOT_HAS_USART ${_pb_has_usart} PARENT_SCOPE)
 set(PUREBOOT_HAS_USART1 ${_pb_has_usart1} PARENT_SCOPE)
 set(PUREBOOT_SIM_MCU ${_pb_sim_mcu} PARENT_SCOPE)
-set(PUREBOOT_USART0_RX ${_pb_usart0_rx} PARENT_SCOPE)
-set(PUREBOOT_USART0_TX ${_pb_usart0_tx} PARENT_SCOPE)
-set(PUREBOOT_USART1_RX ${_pb_usart1_rx} PARENT_SCOPE)
-set(PUREBOOT_USART1_TX ${_pb_usart1_tx} PARENT_SCOPE)
 
-# The rates a default may pick, fastest first.
+# The fastest standard rate the clock reaches within 2.5 % — the same
-set_property(GLOBAL PROPERTY PUREBOOT_BAUD_LADDER 115200 57600 38400 19200 9600)
+# best-of-U2X-and-plain divisor search libavr's solve_baud runs, so a
-
+# default never trips the compile-time error it is checked against. A
-# Whether <baud> is reachable from <clock> within 2.5 %, by the same
+# software build additionally requires the polled receiver's 100-cycles-a-bit
-# best-of-U2X-and-plain divisor search libavr's solve_baud runs, so a build
+# floor (its own static assert): at low clocks the U2X divisor still reaches
-# never trips the compile-time error it is checked against. A software build
+# rates the bit-banged sampler cannot, so the backend gates the ladder.
-# also needs the polled receiver's 100-cycles-a-bit floor: at low clocks the
+function(pureboot_default_baud clock software outvar)
-# U2X divisor reaches rates the bit-banged sampler cannot.
+	foreach(baud 115200 57600 38400 19200 9600)
-function(pureboot_baud_feasible clock baud software outvar)
+		math(EXPR _cycles "${clock} / ${baud}")
-	set(${outvar} 0 PARENT_SCOPE)
+		if(software AND _cycles LESS 100)
-	math(EXPR _cycles "${clock} / ${baud}")
-	if(software AND _cycles LESS 100)
-		return()
-	endif()
-	foreach(divisor 8 16)
-		math(EXPR _step "${divisor} * ${baud}")
-		math(EXPR _n "(${clock} + ${_step} / 2) / ${_step}")
-		if(_n LESS 1 OR _n GREATER 4096)
 			continue()
 		endif()
-		math(EXPR _actual "${clock} / (${divisor} * ${_n})")
+		foreach(divisor 8 16)
-		math(EXPR _delta "${_actual} - ${baud}")
+			math(EXPR _step "${divisor} * ${baud}")
-		if(_delta LESS 0)
+			math(EXPR _n "(${clock} + ${_step} / 2) / ${_step}")
-			math(EXPR _delta "-(${_delta})")
+			if(_n LESS 1 OR _n GREATER 4096)
-		endif()
+				continue()
-		math(EXPR _error_bp "${_delta} * 10000 / ${baud}")
+			endif()
-		if(_error_bp LESS_EQUAL 250)
+			math(EXPR _actual "${clock} / (${divisor} * ${_n})")
-			set(${outvar} 1 PARENT_SCOPE)
+			math(EXPR _delta "${_actual} - ${baud}")
-			return()
+			if(_delta LESS 0)
-		endif()
+				math(EXPR _delta "-(${_delta})")
+			endif()
+			math(EXPR _error_bp "${_delta} * 10000 / ${baud}")
+			if(_error_bp LESS_EQUAL 250)
+				set(${outvar} ${baud} PARENT_SCOPE)
+				return()
+			endif()
+		endforeach()
 	endforeach()
-endfunction()
+	message(FATAL_ERROR "pureboot: no standard baud rate fits a ${clock} Hz clock within 2.5 %")
-
-# The fastest ladder rate the clock reaches.
-function(pureboot_default_baud clock software outvar)
-	get_property(_ladder GLOBAL PROPERTY PUREBOOT_BAUD_LADDER)
-	foreach(baud ${_ladder})
-		pureboot_baud_feasible(${clock} ${baud} ${software} _ok)
-		if(_ok)
-			set(${outvar} ${baud} PARENT_SCOPE)
-			return()
-		endif()
-	endforeach()
-	message(FATAL_ERROR "pureboot: no standard baud rate fits a ${clock} Hz clock within 2.5 % "
-	        "— pass BAUD <rate> to deploy a non-standard one")
 endfunction()
 
 # pureboot_add_loader(<name> [CLOCK <hz>] [BAUD <bd>]
-#                     [SERIAL auto|hardware|software|autobaud] [USART <n>]
+#                     [SERIAL auto|hardware|software] [USART <n>]
-#                     [RX <pin>] [TX <pin>] [TIMEOUT <s>] [OSCCAL <byte>]
+#                     [RX <pin>] [TX <pin>] [TIMEOUT <s>])
-#                     [HALF_DUPLEX])
 #
-# The loader target plus its flashable images (<name>.hex for a programmer,
+# Creates the loader target plus its flashable images (<name>.hex for a
-# <name>.bin for --update-loader). The resolved deployment is stamped on the
+# programmer, <name>.bin for --update-loader) and stamps the resolved
-# target as PUREBOOT_HZ / PUREBOOT_BAUD / PUREBOOT_LINK (the link spelled
+# deployment on the target: the PUREBOOT_HZ, PUREBOOT_BAUD and PUREBOOT_LINK
-# usart0, usart1, or sw:<RX>,<TX> with a trailing @<n> where those pins are a
+# properties (the link as usart0/usart1/sw:<RX>,<TX> — what a test harness
-# USART's own) — what a test harness speaks to it with.
+# needs to speak to the build).
-#
-# HALF_DUPLEX is the one-wire deployment, per backend: on the hardware USART
-# it enables the library's .half_duplex turn-around (RXD and TXD tied
-# together off-chip); on a software or autobaud link it puts both directions
-# on the RX pin — the same thing RX == TX spells directly.
-#
-# SERIAL autobaud measures the host's bit timing at run time, so the image
-# carries no clock and no baud: CLOCK and BAUD are not build parameters there,
-# and one binary per chip serves every F_CPU and every rate. The stamped
-# PUREBOOT_HZ/PUREBOOT_BAUD then record what a harness should *drive* it at,
-# not what it was built for.
-#
-# OSCCAL bakes a measured oscillator trim into the loader (README.md: the
-# RC-oscillator deployment answer): the byte is written at the top of run(),
-# so every reset path — the watchdog hand-over included — runs on the
-# corrected clock. Orthogonal to the backend: an autobaud build may carry it
-# purely for the application's benefit, its own link being clock-free. No
-# value, no code.
 function(pureboot_add_loader name)
-	cmake_parse_arguments(PB "HALF_DUPLEX" "CLOCK;BAUD;SERIAL;USART;RX;TX;TIMEOUT;OSCCAL" "" ${ARGN})
+	cmake_parse_arguments(PB "" "CLOCK;BAUD;SERIAL;USART;RX;TX;TIMEOUT" "" ${ARGN})
 	if(PB_UNPARSED_ARGUMENTS)
 		message(FATAL_ERROR "pureboot_add_loader(${name}): unknown arguments ${PB_UNPARSED_ARGUMENTS}")
 	endif()
@@ -285,8 +229,8 @@ function(pureboot_add_loader name)
 	if(NOT PB_SERIAL)
 		set(PB_SERIAL auto)
 	endif()
-	if(DEFINED PB_USART AND NOT PB_SERIAL MATCHES "^(auto|hardware)$")
+	if(DEFINED PB_USART AND PB_SERIAL STREQUAL "software")
-		message(FATAL_ERROR "pureboot_add_loader(${name}): USART ${PB_USART} contradicts SERIAL ${PB_SERIAL}")
+		message(FATAL_ERROR "pureboot_add_loader(${name}): USART ${PB_USART} contradicts SERIAL software")
 	endif()
 	if(DEFINED PB_USART)
 		set(PB_SERIAL hardware)
@@ -302,9 +246,6 @@ function(pureboot_add_loader name)
 			message(FATAL_ERROR "pureboot_add_loader(${name}): ${LIBAVR_MCU} has no hardware USART")
 		endif()
 		set(_serial_defines PUREBOOT_USART=${PB_USART})
-		if(PB_HALF_DUPLEX)
-			list(APPEND _serial_defines PUREBOOT_HALF_DUPLEX)
-		endif()
 		set(_link usart${PB_USART})
 	else()
 		if(PB_SERIAL STREQUAL "auto")
@@ -314,26 +255,14 @@ function(pureboot_add_loader name)
 			endif()
 			if(_usart)
 				set(_link usart0)
-				if(PB_HALF_DUPLEX)
-					set(_serial_defines PUREBOOT_HALF_DUPLEX)
-				endif()
 			else()
 				set(PB_SERIAL software)
 			endif()
 		endif()
-		if(PB_SERIAL MATCHES "^(software|autobaud)$")
+		if(PB_SERIAL STREQUAL "software")
 			if(NOT PB_RX)
 				set(PB_RX pb0)
 			endif()
-			if(PB_HALF_DUPLEX)
-				# One-wire: both directions on the RX pin. RX == TX spells
-				# the same deployment directly.
-				if(PB_TX AND NOT PB_TX STREQUAL PB_RX)
-					message(FATAL_ERROR "pureboot_add_loader(${name}): HALF_DUPLEX puts both "
-					        "directions on RX (${PB_RX}); TX ${PB_TX} contradicts it")
-				endif()
-				set(PB_TX ${PB_RX})
-			endif()
 			if(NOT PB_TX)
 				set(PB_TX pb1)
 			endif()
@@ -342,35 +271,13 @@ function(pureboot_add_loader name)
 					message(FATAL_ERROR "pureboot_add_loader(${name}): pin '${_pin}' is not of the form pb1")
 				endif()
 			endforeach()
-			if(PB_SERIAL STREQUAL "autobaud")
+			set(_serial_defines PUREBOOT_SOFT_SERIAL PUREBOOT_RX=${PB_RX} PUREBOOT_TX=${PB_TX})
-				set(_serial_defines PUREBOOT_AUTOBAUD PUREBOOT_RX=${PB_RX} PUREBOOT_TX=${PB_TX})
+			# The link spec a test harness drives a GPIO bridge with: sw:<RX>,<TX>
-			else()
+			# as the port letter and bit, the loader's own pin naming upcased.
-				set(_serial_defines PUREBOOT_SOFT_SERIAL PUREBOOT_RX=${PB_RX} PUREBOOT_TX=${PB_TX})
-			endif()
-			# sw:<RX>,<TX> as port letter and bit, upcased — with @<n> where
-			# the TX pin is a USART's own TXD, since a harness driving that
-			# link has to know the USART owns the pin until the loader
-			# releases it.
 			string(SUBSTRING ${PB_RX} 1 2 _rx_pin)
 			string(SUBSTRING ${PB_TX} 1 2 _tx_pin)
 			string(TOUPPER "sw:${_rx_pin},${_tx_pin}" _link)
 			string(REPLACE "SW" "sw" _link ${_link})
-			get_property(_tx0 GLOBAL PROPERTY PUREBOOT_USART0_TX)
-			get_property(_tx1 GLOBAL PROPERTY PUREBOOT_USART1_TX)
-			get_property(_rx0 GLOBAL PROPERTY PUREBOOT_USART0_RX)
-			get_property(_rx1 GLOBAL PROPERTY PUREBOOT_USART1_RX)
-			if(_usart AND PB_TX STREQUAL _tx0)
-				set(_link "${_link}@0")
-			elseif(_usart1 AND PB_TX STREQUAL _tx1)
-				set(_link "${_link}@1")
-			elseif(PB_TX STREQUAL PB_RX AND _usart AND PB_RX STREQUAL _rx0)
-				# One-wire on a USART's RXD: RXEN forces that pin's direction
-				# (§20.7.3), so the driven shared pin is held exactly like a
-				# TXD — the harness models the hold either way.
-				set(_link "${_link}@0")
-			elseif(PB_TX STREQUAL PB_RX AND _usart1 AND PB_RX STREQUAL _rx1)
-				set(_link "${_link}@1")
-			endif()
 		endif()
 	endif()
 	if(NOT PB_BAUD)
@@ -381,50 +288,18 @@ function(pureboot_add_loader name)
 		endif()
 	endif()
 
-	if(PB_SERIAL STREQUAL "autobaud")
+	set(_defines PUREBOOT_CLOCK_HZ=${PB_CLOCK} PUREBOOT_BAUD=${PB_BAUD} PUREBOOT_TIMEOUT=${PB_TIMEOUT}
-		# No clock and no baud reach the image; the window is a poll budget.
+	    ${_serial_defines})
-		set(_defines ${_serial_defines})
-	else()
-		set(_defines PUREBOOT_CLOCK_HZ=${PB_CLOCK} PUREBOOT_BAUD=${PB_BAUD} PUREBOOT_TIMEOUT=${PB_TIMEOUT}
-		    ${_serial_defines})
-	endif()
-	if(DEFINED PB_OSCCAL)
-		math(EXPR _osccal "${PB_OSCCAL}" OUTPUT_FORMAT DECIMAL)
-		if(_osccal LESS 0 OR _osccal GREATER 255)
-			message(FATAL_ERROR "pureboot_add_loader(${name}): OSCCAL ${PB_OSCCAL} is not one byte")
-		endif()
-		list(APPEND _defines PUREBOOT_OSCCAL=${_osccal})
-	endif()
 
 	add_executable(${name} ${CMAKE_CURRENT_FUNCTION_LIST_DIR}/pureboot.cpp)
 	target_link_libraries(${name} PRIVATE libavr)
 	target_compile_definitions(${name} PRIVATE ${_defines})
-	# Codegen shaping for the loader TU only. At -Os GCC otherwise rewrites the
-	# byte-stream loops' counters into end-pointer forms that cost registers
-	# (-fno-ivopts, -fno-split-wide-types), leaves register pressure on the
-	# table with the default allocator (-fira-algorithm=priority), and keeps
-	# expression temporaries in registers (-fno-tree-ter) — but every loop body
-	# here contains a call, so a register held across it costs more than the
-	# load-immediate it saves. The set is fitted to the loader's body and has to
-	# be re-measured when that body changes: -fno-move-loop-invariants belonged
-	# here while the command loop carried four transfer bodies and costs bytes
-	# now that it carries one, and -fno-ivopts is fitted per backend — an
-	# autobaud body needs ivopts to keep the calibration countdown a single
-	# induction variable (without it the counter is duplicated and the
-	# measurement loop runs 9 cycles instead of its contracted 7), while the
-	# fixed-baud bodies still measure smaller with it off.
-	if(PB_SERIAL STREQUAL "autobaud")
-		target_compile_options(${name} PRIVATE
-		    -fira-algorithm=priority -fno-tree-ter -fno-split-wide-types)
-	else()
-		target_compile_options(${name} PRIVATE
-		    -fno-ivopts -fira-algorithm=priority -fno-tree-ter -fno-split-wide-types)
-	endif()
 	target_link_options(${name} PRIVATE -nostartfiles -Wl,--section-start=.text=${_base_hex}
 	                    -Wl,--defsym=pureboot_app=${_app} ${_wrap})
 	add_custom_command(TARGET ${name} POST_BUILD COMMAND ${CMAKE_SIZE} $<TARGET_FILE:${name}>)
-	# The ELF is a container, never flashed: .hex for a programmer, .bin (the
+	# The ELF is a container (symbols, section headers), never flashed; the
-	# slot's bare bytes) for --update-loader.
+	# flashable forms sit beside it: .hex for a programmer, .bin (the slot's
+	# bare bytes) for the host tool's raw path and --update-loader.
 	add_custom_command(TARGET ${name} POST_BUILD
 	                   COMMAND ${CMAKE_OBJCOPY} -O ihex -R .eeprom
 	                           $<TARGET_FILE:${name}> $<TARGET_FILE:${name}>.hex
@@ -433,4 +308,3 @@ function(pureboot_add_loader name)
 	set_target_properties(${name} PROPERTIES PUREBOOT_HZ ${PB_CLOCK} PUREBOOT_BAUD ${PB_BAUD}
 	                      PUREBOOT_LINK ${_link})
 endfunction()
-

pureboot/README.md

@@ -2,388 +2,200 @@
 
 A serial bootloader on [libavr](https://git.blackmark.me/avr/libavr), pure by
 constraint: one C++ source, no inline assembly, no global register variables
-(attributes and compiler flags allowed), **a 512-byte slot on every chip
+(attributes allowed), built for **every chip libavr targets — all 37 —
-libavr targets — all 37**. The device speaks primitives; every composite — verify,
+fitting each chip's smallest boot sector**: 512 bytes everywhere — 470 B on
-erase, reset-vector surgery, updating the loader itself — lives in the host
+the tiny13s, ~484 B on the tiny25/45/85, 458–506 B across the megas and
-tool (`pureboot.py`).
+every configured variant of them (the m8's software-serial build is the
+fattest) — except the ATmega1284/1284P, whose smallest boot sector is 1 KiB
+and whose far-flash machinery (ELPM reads, RAMPZ page commands,
+word-addressed wire) lands at 556–562 B in a 1 KiB slot: the 512-byte
+figure is a hardware boundary those chips simply do not have, and no
+implementation of this feature set fits it there. Clock, baud, serial
+backend and pins are per-build configuration (below); the size matrix in
+the test suite holds every combination inside its slot. The device speaks
+primitives; every composite — verify, erase, reset-vector surgery, updating
+the loader itself — lives in the host tool (`pureboot.py`).
 
 The image is **position-independent**: control flow is PC-relative, the
-transfer paths take wire addresses, the write guard protects the slot the code
+read/write paths take wire addresses, the write guard protects the slot the
-is *running* in (from the runtime return address), nothing else is
+code is *running* in (from the runtime return address), the info block is
-flash-resident to address at all, and the application jump is an indirect call
+addressed from that same anchor, and the application jump is an indirect
-to an absolute entry. The identical binary therefore runs from any slot with
+call to an absolute entry. The identical binary therefore runs from any
-every command intact, which makes pureboot **its own staging loader**: the host
+slot with every command intact — which makes pureboot **its own staging
-installs the same binary one slot below the resident, jumps into it, and lets
+loader**: the host installs the same binary one slot below the resident,
-it rewrite the resident. The lint holds it to that literally — the image must
+jumps into it, and lets it rewrite the resident. The slot is 512 bytes
-come out byte-identical linked at a different base.
+(1 KiB on the word-addressed large chips, matching their boot-sector
-
+minimum); on the tinies the budget is 510, not 512: a slot's last word
-## Chips
+belongs to the host-managed trampoline (below).
-
-The Stock column is the default configuration: the hardware USART0 at 115200
-8N1 on a 16 MHz crystal, or the software UART on RX = PB0 / TX = PB1 at
-57600 8N1 on the tinies' RC oscillator (9.6 MHz on the t13s, 8 MHz above).
-Every axis moves per build — see *Configuration*. The Autobaud column is the
-worst configuration the space produces for the chip: the clock-free build —
-it alone carries the calibration machinery — with the `OSCCAL` trim baked
-and, where the chip has a USART, the link deployed on that USART's own pins,
-which the loader then has to release (*Pin ownership*). Folding the same
-build onto a single pin (*One-wire*) measures identically on every chip, so
-the column covers that twin too. On default pins without the trim the same
-loaders run 4–10 B smaller.
-
-| Chip | Flash | Loader at | Link | Stock | Autobaud |
-|---|---|---|---|---|---|
-| ATtiny13, ATtiny13A † | 1 KiB | 0x0200 | software | 384 B | 474 B |
-| ATtiny25 † | 2 KiB | 0x0600 | software | 388 B | 466 B |
-| ATtiny45 † | 4 KiB | 0x0e00 | software | 388 B | 466 B |
-| ATtiny85 † | 8 KiB | 0x1e00 | software | 388 B | 466 B |
-| ATmega8, 8A | 8 KiB | 0x1e00 | USART0 | 362 B | 494 B |
-| ATmega16, 16A | 16 KiB | 0x3e00 | USART0 | 364 B | 496 B |
-| ATmega32, 32A | 32 KiB | 0x7e00 | USART0 | 364 B | 496 B |
-| ATmega48, 48A, 48P, 48PA † | 4 KiB | 0x0e00 | USART0 | 378 B | 468 B |
-| ATmega88, 88A, 88P, 88PA | 8 KiB | 0x1e00 | USART0 | 388 B | 478 B |
-| ATmega168, 168A, 168P, 168PA | 16 KiB | 0x3e00 | USART0 | 390 B | 480 B |
-| ATmega328, 328P | 32 KiB | 0x7e00 | USART0 | 390 B | 480 B |
-| ATmega164A, 164P, 164PA | 16 KiB | 0x3e00 | USART0 | 390 B | 480 B |
-| ATmega324A, 324P, 324PA | 32 KiB | 0x7e00 | USART0 | 390 B | 480 B |
-| ATmega644, 644A, 644P, 644PA | 64 KiB | 0xfe00 | USART0 | 384 B | 474 B |
-| ATmega1284, 1284P | 128 KiB | 0x1fe00 | USART0 | 410 B | 502 B |
-
-† No hardware boot section: the host patches the reset vector, and the budget
-is 510 bytes, since the slot's last word is the trampoline.
-
-The tightest fit in the whole space is therefore the 1284s' 502 of their
-512: they alone carry the far-flash machinery (ELPM reads, RAMPZ page
-commands) on top of everything the column already stacks. The flash bank
-riding in a transfer's selector byte keeps even those chips' addressing the
-same 16-bit form every other chip uses, which is why they are no longer the
-outlier they were.
-
-The software UART enables the RX pull-up; TX idles high. All multi-byte wire
-quantities are little-endian.
 
 ## Configuration
 
-Every deployment axis is a build parameter of `pureboot_add_loader()` (in
+Every deployment axis is a build parameter, resolved by the CMake function
-`pureboot/CMakeLists.txt`) — the one way a loader target is created, by this
+`pureboot_add_loader()` (in `pureboot/CMakeLists.txt`) — the one way a
-repo's build and by a downstream project alike:
+loader target is created, by this repo's own build and by a downstream
+project alike:
 
 | Argument | Meaning | Default |
 |---|---|---|
 | `CLOCK <hz>` | the clock the board runs | 16 MHz megas, 8 MHz t25/45/85, 9.6 MHz t13s |
 | `BAUD <bd>` | the wire rate | the ladder below |
-| `SERIAL auto\|hardware\|software\|autobaud` | the link backend | `auto`: the hardware USART where the chip has one |
+| `SERIAL auto\|hardware\|software` | the link backend | `auto`: the hardware USART where the chip has one |
 | `USART <n>` | the USART instance (x4 megas carry two) | 0 |
 | `RX <pin>`, `TX <pin>` | software-UART pins | `pb0`, `pb1` |
 | `TIMEOUT <s>` | the activation window | 8 |
-| `OSCCAL <byte>` | a measured oscillator trim, applied before anything runs | none — no value, no code |
-| `HALF_DUPLEX` | one-wire: both directions on one line (*One-wire* below) | off |
 
 The default baud is the fastest of 115200/57600/38400/19200/9600 the clock
 reaches within 2.5 % — the same U2X-included divisor search libavr's baud
 solver runs — and on a software build additionally within the polled
-receiver's 100-cycles-a-bit floor. Whatever is picked or overridden is
+receiver's 100-cycles-a-bit floor. 16 MHz lands 115200, 8 MHz 57600,
-re-checked in the compile: an infeasible combination, or a USART the chip does
+1 MHz 9600. Whatever is picked or overridden is re-checked in the compile:
-not have, fails with a named static assert.
+an infeasible clock/baud/backend combination, or a USART the chip does not
+have, fails with a named static assert.
 
-Putting a bit-banged link on a USART's own pins is a supported deployment, and
+A downstream project brings its usual libavr setup (the `libavr` target,
-the usual one where a board's USB bridge is wired to RXD/TXD: the link's `init`
+the chip via the `LIBAVR_MCU` toolchain preset), consumes this directory,
-clears that USART's `UCSRnB` first, because while its `TXEN` is set the USART —
+and states its deployment — for example an ATmega328P on its shipped
-not the port register — owns the TX pin, and a loader entered from an
+1 MHz fuses with the software UART on hand-picked pins:
-application that left it enabled would receive and obey while answering nothing
-(§20.6.3). It costs one store — four bytes on the extended-I/O chips, two on
-the classic megas — and only on those pins.
-
-`SERIAL autobaud` takes neither: the loader **measures** the host's bit timing
-at run time, so `CLOCK` and `BAUD` are not build parameters there and one
-binary per chip serves every clock and every rate. It is for the deployments
-whose clock is not known at build time and does not hold still — the internal
-RC oscillator, ±10 % from the factory and moving with supply and temperature —
-where a fixed-baud software build has to be rebuilt per clock and still drifts
-out of tolerance. The cost is that it is software-serial only (a hardware USART
-needs its divisor programmed) and that activation counts poll iterations rather
-than seconds, since there is no clock to convert them against
-(`PUREBOOT_AUTOBAUD_POLLS`, default 4,000,000). The wait spends nine cycles a
-poll (measured, and held by the `pureboot.window.autobaud` gate), so the
-default window is 36 M cycles: 4.5 s at 8 MHz, 3.75 s at 9.6 MHz, 36 s at
-1 MHz.
-
-**Pick the rate by cycles a bit, and leave the oscillator room.** What the
-calibration can measure is bounded by how many clock cycles one bit lasts, so a
-rate is only ever sensible relative to the clock. Two different floors matter:
-
-| | cycles a bit |
-|---|---|
-| the logic's floor — exact clock, simulated | solid to ~36, fails outright by ~31 (`pureboot.autobaud` gates a point here) |
-| a factory-trimmed internal RC, measured on an ATtiny13A | reliable at ~118; already locking 1 attempt in 5 by ~59 |
-
-The gap is the oscillator's own jitter, and no exact-clock simulation shows it.
-So on an RC part, **budget about 100 cycles a bit** — the same order as the
-fixed-baud software receiver's floor — rather than the logic's ~36. Measured
-envelope on that ATtiny13A, with an application resident: 9.6 and 4.8 MHz reach
-115200, 1.2 MHz reaches 9600, 600 kHz reaches 4800, 128 kHz reaches 2400.
-
-One trap in testing this: on a patched-vector chip an **erased** application
-region walks straight back up into the loader, so every expired window opens
-another one and the host's retries eventually catch the pulse. That reads as far
-more reliable than the same part with an application resident, which gets one
-window per reset. Measure with an application in place.
-
-## One-wire
-
-`HALF_DUPLEX` puts both directions on one line — the deployment for a board
-with a single spare pin, or a native-UART bootloader's shared-line wiring.
-Each backend has its shape:
-
-- **Software and autobaud links** fold onto the RX pin (`RX == TX` spells
-  the same deployment directly). The pin idles as the receiver's pull-up
-  input; each transmitted frame takes the pin's direction and hands it back
-  with the stop bit's level already on the pull-up, so neither flip makes
-  an edge. This costs nothing: the frame's direction wrap is exactly what
-  the dropped second-pin init paid, and the tightest image in the space —
-  the 1284s' autobaud + `OSCCAL` on their USART's RXD — measures the same
-  502 bytes one-wire as two-wire. On a USART's own pin the release applies
-  as ever, RXD included: `RXEN` forces that pin's direction (§20.7.3),
-  which a receive-only link could live with and a driven shared pin cannot.
-- **The hardware USART** (`SERIAL hardware`/`auto` + `HALF_DUPLEX`) uses
-  libavr's `.half_duplex` turn-around — exactly one direction enabled at a
-  time, each written byte held to transmit-complete before the line can be
-  released — and needs RXD and TXD tied together off-chip. It costs
-  +42…50 B over the stock loader (m8 404, m328P 440, 1284P 460 — all far
-  inside the slot); the activation window is unchanged, its poll merely
-  runs through the release-line test (18 cycles a poll in bit-addressable
-  I/O, 22 in extended — measured, and held per chip by
-  `pureboot.window.halfduplex`).
-
-Host wiring, for an FTDI-style adapter: **adapter TX through ~1 kΩ to the
-line, adapter RX and the MCU pin directly on it.** The resistor lets the MCU
-win the line while it answers; the price is that the adapter reads back every
-byte it transmits. `pureboot.py --one-wire` consumes that echo byte for byte
-— a missing echo is reported as the wiring fault it is, and a device reply
-that lands between the echoes of the knock (a loader already in session
-re-prompts mid-knock) is held for the reader. The knock is the protocol's
-one blind multi-byte write, so on real wiring its second byte can be lost to
-that collision outright; the tool's knock retries absorb it. Everything else
-is ack-paced and cannot collide.
-
-A downstream project brings its usual libavr setup (the `libavr` target, the
-chip via the `LIBAVR_MCU` toolchain preset), consumes this directory, and
-states its deployment — an ATmega328P on its shipped 1 MHz fuses with the
-software UART on hand-picked pins, say. A submodule pins the loader version
-(the tags name them; this repo pins its own libavr the same way), where
-FetchContent tracks whatever `main` is:
 
 ```cmake
-# git submodule add <forge>/avr/bootloader.git bootloader   — or FetchContent
+FetchContent_Declare(bootloader GIT_REPOSITORY git@git.blackmark.me:avr/bootloader.git GIT_TAG main)
-add_subdirectory(bootloader/pureboot pureboot)
+FetchContent_MakeAvailable(bootloader)
+add_subdirectory(${bootloader_SOURCE_DIR}/pureboot pureboot)
 
 pureboot_add_loader(myboot CLOCK 1000000 SERIAL software TX pb1 RX pb5)
 ```
 
 The function emits the ELF plus `myboot.hex` (the programmer artifact) and
-`myboot.bin` (the self-update image), prints the size, and stamps the resolved
+`myboot.bin` (the self-update image), prints the size, and stamps the
-deployment on the target as the `PUREBOOT_HZ`, `PUREBOOT_BAUD` and
+resolved deployment on the target as the `PUREBOOT_HZ`, `PUREBOOT_BAUD`
-`PUREBOOT_LINK` properties — what a flashing script or test harness needs to
+and `PUREBOOT_LINK` properties — what a flashing script or test harness
-speak to the build. This exact deployment runs the full protocol suite in CI
+needs to speak to the build. This exact example deployment runs the full
-(`pureboot.custom`).
+protocol suite in CI (`pureboot.custom`).
+
+## Link
+
+The stock builds assume the family's natural deployment; any axis moves
+per build (above).
+
+| Chip | Serial | Baud | Clock assumed |
+|---|---|---|---|
+| every ATmega | the hardware USART (USART0), RXD/TXD per pinout | 115200 8N1 | 16 MHz crystal |
+| ATtiny25/45/85 | software UART, RX = PB0, TX = PB1 | 57600 8N1 | 8 MHz internal RC |
+| ATtiny13/13A | software UART, RX = PB0, TX = PB1 | 57600 8N1 | 9.6 MHz internal RC |
+
+The software-UART RX pin has its pull-up enabled; TX idles high. All
+multi-byte quantities on the wire are little-endian.
 
 ## Activation
 
-Reset enters the loader (BOOTRST on the boot-sectioned megas, the patched
+Reset enters the loader (BOOTRST on the boot-sectioned megas; the patched
-reset vector elsewhere) — except a watchdog reset, which hands straight to the
+reset vector on the tinies and the boot-section-less m48s) — except a
-application with no activation window, since the application owns its watchdog.
+watchdog reset, which hands straight to the application (the application
-This is deliberate: it lets an application reboot itself instantly rather than
+owns its watchdog; it must clear WDRF itself, which also releases the
-sit through the window. The application must clear WDRF itself (libavr's
+WDRF-forced WDE).
-`watchdog::disable()` does). **Gotcha:** WDRF is sticky (cleared only by
-software, not by a later reset), so an application that watchdog-resets and
-never clears it diverts *every* subsequent reset — external ones included —
-past the window too, and the loader becomes reachable only through an external
-programmer until the flag is cleared. A serial recovery path therefore assumes
-the application clears WDRF on its own reset path.
 
-The host then knocks `p` then `b`, each awaited byte under a fresh activation
+The host then has one activation window per awaited byte to knock: `p` then
-window; any other byte is discarded and awaited again, so line noise can delay
+`b`. Each awaited byte gets a fresh window; any other byte is discarded and
-the loader but never lock it. A window expiring on an idle line boots the
+awaited again (line noise cannot lock the loader, only delay it). A window
-application.
+expiring with an idle line boots the application.
 
-An autobaud build opens differently, because it has to learn the rate before it
+The window length is a compile-time constant — 8 s by default, another
-can read a byte at all: the host sends the **calibration byte 0xC0** — a start
+value via `pureboot_add_loader(... TIMEOUT <s>)` (the stock target keeps
-bit plus six zero data bits form one low pulse of seven bit-times — and the
+the `PUREBOOT_TIMEOUT` cache variable) — so the whole EEPROM belongs to
-loader times that pulse into its bit period. A single `p` then activates; the
+the application; pureboot never uses it for its own state. Re-timing a
-pulse has already proven a host is present, which the two-byte knock exists to
+deployed loader is a self-update with a re-timed build (below).
-establish elsewhere. Both waits are bounded, so a stray low pulse with no host
-behind it costs one window and then boots the application rather than holding
-the loader.
-
-The window is a compile-time constant (`TIMEOUT`, 8 s by default), so the whole
-EEPROM belongs to the application — pureboot keeps no state of its own.
-Re-timing a deployed loader is a self-update with a re-timed build. An autobaud
-build counts poll iterations instead (`PUREBOOT_AUTOBAUD_POLLS`), there being
-no clock to turn into seconds.
 
 ## Session
 
 After the knock the loader stays in its command loop until `J` jumps away or
 the chip resets. Before reading each command it waits for any pending EEPROM
-write and sends the prompt `+` (0x2b), which is therefore also the previous
+write to finish and sends the prompt `+` (0x2b) — the prompt is therefore
-command's completion ack. A session is: await `+`, send a command, read its
+also the completion ack of the previous command. A session is: await `+`,
-reply, repeat.
+send a command, read its reply, repeat.
 
-Addresses are **byte addresses within a 64 KiB bank**, and the bank rides in
+On chips whose flash exceeds 64 KiB (the 1284s — info-block flag bit 1) the
-the command's selector byte, so no command has to speak word addresses. `J` is
+`R`/`W` flash addresses are **word** addresses; everywhere else they are byte
-the exception: its address is a word address, because that is what the
+addresses (the 644s' 64 KiB is exactly the 16-bit byte space and stays
-hardware's own jump takes — it still carries a selector byte (reserved,
+byte-addressed). EEPROM addresses are always bytes, counts always bytes.
-ignored) so its decode is the same three reads as every other command's.
-EEPROM and data-space addresses and all counts are bytes.
-
-The loader trusts the host to keep addresses in range: it does not bound them
-against the chip. **Gotcha:** a write (or read) that runs past `E2END` wraps —
-EEAR is only as wide as the array, so an address past the end truncates onto
-low EEPROM and the write silently overwrites it. Keeping transfers within the
-real sizes is the host's job (the shipped tool does); the flash budget is
-better spent on features than on re-checking a bound the host already holds.
 
 | Cmd | Arguments | Reply |
 |---|---|---|
-| `b` | — | 4 bytes: the pureboot version, then the three signature bytes |
+| `b` | — | the 12-byte info block |
-| `G` | sel8, addr16, n8 | n bytes from the selected space (n = 0 means 256) |
+| `R` | addr16, n8 | n flash bytes (n = 0 means 256) |
-| `g` | sel8, addr16, n8, then n data bytes | `+` per byte, sent once its write has begun |
+| `W` | addr16, then one page of data | — (completion = next prompt) |
-| `W` | sel8, addr16, then one page of data | — (completion = next prompt) |
+| `r` | addr16, n8 | n EEPROM bytes (n = 0 means 256) |
-| `J` | sel8 (reserved), word address (16-bit) | `+`, then execution continues there |
+| `w` | addr16, n8, then n data bytes | `+` per byte, sent once its write has begun |
+| `F` | — | 4 bytes: low fuse, lock, extended fuse, high fuse |
+| `J` | word address (16-bit) | `+`, then execution continues there |
 | other | — | ignored; the loop re-prompts (send a junk byte, await `+`, to resync) |
 
-`G` and `g` are one letter in two cases, which is the whole command set for
+`W` streams exactly one SPM page (size from the info block) into the buffer,
-every memory: the **selector** byte's low nibble names the space and its high
+then erases and programs; the address must be page-aligned. Pages inside the
-nibble carries the flash bank.
+512-byte slot the loader is *running* in are drained but never programmed — a
+broken host cannot brick the running copy, and a staged copy may rewrite the
+resident slot. `w` is host-paced: send the next byte only after the previous
+byte's `+`. `F` returns the bytes in the hardware's Z order; on a chip
+without an extended fuse byte (the ATtiny13A) that slot carries no meaning.
+Fuse *writing* does not exist: SPM reaches flash (and, on the mega, lock
+bits) only — fuse bytes are external-programming territory by hardware.
 
-| Space | | |
+`J` is the one control-transfer primitive: the host uses it to run the
-|---|---|---|
+application (word 0 on the mega, the trampoline word on the tinies — both
-| 0 | flash | read-only here; it is written through `W` and the SPM space |
+known from the info block) and to move between loader copies during a
-| 1 | EEPROM | |
+self-update. A jump to a loader slot's base re-enters that copy's own
-| 2 | data | SRAM — and with it the register file and every I/O register, which share the data address space on AVR |
+startup; it must then be knocked afresh.
-| 3 | fuse and lock | index 0..3 in the hardware's own Z order: low, lock, extended, high |
-| 4 | SPM | write-only: the byte goes to SPMCSR and fires the instruction at the address |
 
-The data space is worth more than it looks. pureboot keeps **zero static RAM**
+The info block (`b`):
-and pushes no register, so at loader entry an application's SRAM is still
-whatever the application left there, bar the handful of bytes of return-address
-stack — which makes `G` over space 2 a post-mortem of a running application,
-not just a poke hole. The same address space carries the register file and the
-I/O registers, so peripheral state is readable too; reading some of those has
-side effects (reading UDR clears its flags), which is the host's business to
-know.
 
-Programming a page is therefore `W` to fill the buffer, then a `g` to the SPM
+| Offset | Content |
-space for the erase, another for the write, and on a boot-sectioned chip a
+|---|---|
-third to re-enable the RWW section — `0x03`, `0x05` and `0x11`, the SPMCSR
+| 0–2 | `'P'`, `'B'`, protocol version (1) |
-encodings every part pureboot targets shares. The loader carries no page-commit
+| 3–5 | device signature |
-logic of its own, and the same primitive reaches every other SPM operation,
+| 6 | SPM page size in bytes (0 means 256) |
-lock bits included.
+| 7–8 | loader base — application flash ends here (a word address when bit 1 is set) |
+| 9–10 | EEPROM size |
+| 11 | bit 0: host must patch the reset vector (no hardware boot section); bit 1: flash wire addresses are word addresses |
 
-The SPM store and the SPM instruction must issue within four cycles of each
+Composites are the host's job: verify = read back and compare, erase =
-other (§26.2), which no host can hit across a serial link — so this one
+write `0xff` (per page for flash, per byte for EEPROM).
-primitive is *fused* rather than being a poke of SPMCSR followed by a poke of
-something else. That four-cycle window is the floor on how low-level a
-bootloader's primitives can go; it is not a byte-count decision.
-
-An SPM command aimed at the 512-byte slot the loader is **running in** is
-dropped, so a broken host cannot brick the running copy, while a staged copy
-one slot lower may rewrite the resident — which is what a self-update is.
-
-The loader never clears the SPM buffer before a fill, so **one `W` may program
-the wrong bytes, and the host is what fixes it**. The buffer is write-once per
-word until cleared, and two things leave words in it: a refused page, and —
-where SPM runs from anywhere, the tinies and the m48s — an application that
-self-programmed before entering. The next page write takes those stale words
-and clears them, since a page write auto-erases the buffer (§26.2.1; §19.2 on
-the tinies), so repeating it programs correctly. The host therefore verifies
-every page it writes and rewrites what comes back wrong (three retries, then it
-stops).
-
-`g` is host-paced: send the next byte only after the previous byte's `+`. Fuse
-*writing* does not exist — SPM reaches flash and boot lock bits only.
-
-`J` is the one control-transfer primitive: it runs the application (word 0 or
-the trampoline word, both derived from the chip) and moves between loader
-copies during a self-update. A jump to a slot's base re-enters that copy's own
-startup, which must then be knocked afresh.
-
-`b` answers with the loader's identity — its version and the chip's signature —
-and nothing else. Everything else the host needs (page size, loader base,
-EEPROM size, whether the reset vector must be patched, how many flash banks)
-follows from the signature, and the host holds that table; the loader derived
-the same facts from its own chip database at build time, so nothing is guessed,
-it is simply not sent twice.
-
-An update image, though, is a bare 512-byte slot with no device to ask, and
-installing one built for another chip bricks the target. Every loader image
-therefore carries a six-byte **stamp** — `'P'`, `'B'`, the version, the three
-signature bytes — which the loader itself never reads and the host tool refuses
-to install a mismatch against.
-
-## Version
-
-`b`'s first byte is the **pureboot version** — the loader's one identity
-number, and the only way to tell what a deployed loader is. Nothing else is
-numbered: the wire protocol has no version, a pureboot version implies it, and
-the host tool holds that map. The tool states the window of loader versions it
-speaks (`OLDEST_LOADER`/`NEWEST_LOADER` in `pureboot.py`), and a version that
-changes the protocol becomes the new floor there. A loader newer than the tool
-is refused by name rather than decoded on the assumption that nothing moved.
-
-Two generations exist. **1 through 4** speak one session — a 12-byte info block
-from `b`, and a command per memory (`R`/`W` flash, `r`/`w` EEPROM, `F` fuses).
-**5** replaced those with the single `G`/`g` pair over selector-named spaces
-above; the shipped tool speaks both, choosing on the version it reads, so a
-deployed pureboot 4 stays drivable and self-updatable to 5. **6** changes
-nothing on the wire: it marks the builds that may carry a baked `OSCCAL` trim
-(Configuration), so a tool driving an update knows such images exist. **7**
-moves `J` onto the unified decode — it gains the selector byte the table
-shows, which older loaders do not read, so the tool sends each form to the
-version that speaks it — and re-homes the autobaud unit into the GPIOR pair
-on the chips that have one (Session: what must not be written), which is
-where `--info`'s measured clock now reads it on those parts. **8** changes
-nothing on the wire either: it marks the builds whose deployment may be
-one-wire (*One-wire* above) — the hardware USART's half-duplex turn-around,
-or a software link folded onto a single pin. The host-side trace is
-`--one-wire`, the echo discard a shared line requires of any tool driving
-it.
-
-Every closed generation is tagged in this repo at its era's last commit — the
-commit just before the next version bump, so a tag holds everything its
-version ever gained — and each tag carries the `libavr/` submodule pinned to
-the libavr that loader was built against, as the whole libavr era does commit
-by commit. `git checkout v3 && git submodule update --init libavr` followed by
-the usual preset build therefore reproduces the v3 loader exactly; the open
-generation is `main`.
-
-Collapsing four command bodies into one transfer loop is what paid for the
-version: the data space, the host-issued SPM operations and the fuses now share
-the loop, the cursor and the argument decode that `R`/`r`/`w` each carried a
-copy of. The loader shrank while gaining all three.
-
-The tool carries its own version, free to drift; `--version` prints it and the
-window.
 
 ## Deployment
 
 The build leaves three artifacts per chip. The ELF is a container for the
-tests and objcopy, never flashed. The **.hex is the programmer artifact**: it
+tests and objcopy — never flashed. The **.hex is the programmer artifact**:
-carries its own addresses and lands the loader in its top slot, touching
+it carries its own addresses and lands the loader in its top slot,
-nothing else. The **.bin is the self-update image** — the slot's bare bytes.
+touching nothing else. The **.bin is the self-update image** — the slot's
+bare bytes with no addressing, which a programmer would put at address 0.
+On a boot-sectioned mega a copy at 0 is dead weight (SPM only executes
+from the boot section, so it cannot even heal itself — reflash the .hex);
+on the patched-vector chips it *runs* (the image is position-independent
+and reset enters word 0), reports its canonical geometry, and the ordinary
+`--update-loader` flow re-homes a build into the top slot from any
+position — the staging install and the word-0 redirect execute from
+copies outside page 0's slot, and a copy sitting in the staging slot
+itself is recognized as the installed staging copy and left in place (it
+streams the new resident like any staged copy, so an older build installs
+a newer one). `pureboot.rehome` is the acceptance test for both
+positions. Flashing the application afterwards overwrites the stale copy,
+vector surgery included.
 
-**Boot-sectioned megas**: program the loader at `flash − 512` with an external
+**Boot-sectioned megas**: program the loader at `flash − slot` with an
-programmer. Every such mega has a BOOTSZ step whose boot section is exactly
+external programmer. Every such mega has a BOOTSZ step whose boot section
-the 512-byte slot — the second-smallest step on the 8 KiB and 16 KiB chips,
+is exactly the loader slot — 512 B, the second-smallest step on the 8 KiB
-the smallest on the 32 KiB ones — so the ATmega328P profiles below apply to
+and 16 KiB chips (m8, m88, m16, m168, m164), the smallest on the 32 KiB
-every one of them with its own addresses; the per-chip BOOTSZ ladders live in
+ones (m32, m328, m324); on the 1284s that step is the smallest, 512 words,
-the host tool (`BOOT_FUSE`).
+which is why their slot is 1 KiB — so the ATmega328P profiles below apply
+to every one of them with its own addresses and slot size; the per-chip
+BOOTSZ ladders live in the host tool (`BOOT_FUSE`). The 1284s' numbers:
+standalone = BOOTSZ 512 words (reset at the loader base 0x1fc00);
+self-update = 1024 words, covering both 1 KiB slots, the loader-first
+reset landing at 0x1f800 — the staging slot, walked across when erased.
 
-The **644s and 1284s** are the geometry's sweet spot: their smallest boot
+The **644s** are the geometry's sweet spot: their smallest boot section
-section (512 words = 1 KiB) is exactly *two* slots, so the resident and its
+(512 words = 1 KiB) is exactly *two* 512-byte slots, so the resident and
-staging slot both live inside the minimum section. Self-update needs no fuse
+its staging slot both live inside the minimum section — self-update needs
-step up, and the standalone profile does not exist — reset lands one erased
+no fuse step up, and the standalone profile does not exist (reset lands at
-slot below the loader (0xfc00 / 0x1fc00) and walks up into it.
+0xfc00, one erased slot below the loader: the loader-first walk built in).
 
 ATmega328P profiles (addresses for its 32 KiB):
 
@@ -393,312 +205,141 @@ ATmega328P profiles (addresses for its 32 KiB):
 | 512 words (1 KB) | unprogrammed | *Self-update, app-first*: reset always boots the application, which owns all 31.5 KB and must offer its own jump to 0x7e00 to reach the loader (a virgin chip reaches it by reset across erased flash). Updates are power-fail-safe except mid-rewrite of the resident slot itself (no reset path leads to the staging copy then). |
 | 512 words (1 KB) | programmed | *Self-update, loader-first*: reset lands at 0x7c00 — the staging slot, normally erased, so execution walks up into the loader; during an update it is the staging copy itself, so a mid-rewrite power loss recovers by reset. The loss windows move to the staging install/retire page writes instead (page-write scale). The host keeps `[0x7c00, 0x7e00)` clear of application data (`--force` overrides). |
 
-Applications are flashed unmodified here — word 0 stays the application's own
+Applications are flashed unmodified — word 0 stays the application's own
 reset vector, and the hand-over jumps to 0.
 
-**Patched-vector chips — the tinies and the m48s** (no boot section; the m48s'
+**Patched-vector chips — the tinies and the m48s** (no boot section; the
-SPM runs from the entire flash, Atmel-8271 §26): program the loader at
+m48s' SPM runs from the entire flash, Atmel-8271 §26): program the loader
-`flash − 512`; erased flash below it walks up into the loader, so a virgin
+at `flash − 512`; erased flash below it walks up into the loader, so a
-chip activates. Flashing an application then takes reset-vector surgery: word
+virgin chip activates. When flashing an application the host performs
-0 becomes an `rjmp` to the loader base, and the application's own entry is
+reset-vector surgery: word 0 is rewritten to `rjmp` to the loader base, and
-re-encoded as a trampoline `rjmp` in the word just below the loader
+the application's own entry is re-encoded as a trampoline `rjmp` in the
-(`base − 2`, where the hand-over jumps). Every other vector stays the
+word just below the loader (`base − 2`, where the hand-over jumps). Every
-application's. The patched page 0 and the trampoline page are written *first*
+other vector stays the application's. The patched page 0 and the trampoline
-and an erase runs top-down, so from the first write on an interruption still
+page are written *first*, so from the first write on an interrupted flash
-resets into the loader.
+still resets into the loader; an erase runs top-down for the same reason.
-
+The m48s speak this profile over their hardware USART — no fuse preflight,
-A .bin programmed at address 0 by mistake is dead weight on a boot-sectioned
+BOOTRST does not exist there.
-mega (SPM only executes from the boot section — reflash the .hex), but *runs*
-on a patched-vector chip, and the ordinary `--update-loader` flow re-homes it
-into the top slot from there (`pureboot.rehome`).
-
-**Fixed-baud on an internal RC oscillator is a deployment risk the build
-cannot see.** The factory trim is ±10 % where an 8N1 frame survives about
-±4: a part at the edge answers nothing at the built rate, and the symptom —
-silence — reads as a wiring fault (a real ATtiny13A measured −5.5 %, outside
-every standard rate at its own documented default). The **autobaud build is
-the deployment-proof backend**: it has no rate to miss. Where fixed-baud on
-RC is wanted anyway, measure first and bake the trim: an autobaud session's
-`--info` prints the part's true clock from the loader's own measured bit
-period, OSCCAL moves the oscillator about 1 % per step, and `OSCCAL <byte>`
-builds the correction in — one build–measure iteration converges. A loader
-already deployed and silent is diagnosed with `--scan` (Host tool).
 
 ## Updating the loader
 
 `pureboot.py --update-loader new_pureboot.bin` replaces the resident loader
-with any pureboot build — a re-timed window, a newer version — using the
+with any pureboot build — a re-timed window, a newer protocol — using the
 loader itself as its own staging loader. The image is the loader's own 512
-bytes as a raw binary, or the Intel HEX the build emits beside it.
+bytes as a raw binary, or the Intel HEX the build emits beside it, which
+links the loader at its base inside an otherwise blank flash image:
 
-One thing the image cannot tell the host: **which link it speaks.** The update
+The preflight refuses an image built for another chip: the info block
-works by entering copies of the *new* image (steps 3 and 4 below), so a build
+embedded in every pureboot binary (signature, page size, loader base,
-made for another baud or another backend answers on that one and not on the
+EEPROM size, flags) must match the device's own, and the error names both.
-session's — and 512 bytes of position-independent code carry no header to read
+Die revisions share their base signature and geometry, so their images are
-it from. Where the new image's link differs, name it:
+interchangeable — as the silicon is. `loader_image()` also accepts a
+padded image (a raw .bin padded from 0, or a whole-flash read-back with
+the loader resident) and peels it to the slot content by the embedded base.
 
-```sh
+1. The staging slot `[base−slot, base)` is saved to a host-side state file
-# a 57600 fixed-baud resident, replaced by an autobaud build
+   (on the 1 KB tiny13s that is the whole application, vectors included).
-pureboot.py --port … --baud 57600 --update-loader ab.bin --staged-autobaud
+2. The resident installs the identical update image there. On the
-# …or by a 38400 build of the same backend
+   patched-vector chips the host composes the slot's last word — the same
-pureboot.py --port … --baud 57600 --update-loader sw38400.bin --staged-baud 38400
+   address as the resident's trampoline — as a jump to the resident base,
-```
+   so even an abandoned staging copy times out into a loader, never into
-
+   garbage. A loader already sitting whole in the staging slot (its info
-The host retunes on the open port, so no DTR pulse resets the copy it is talking
+   block in place, the slot unchanged since the update began) is left as
-to. Omit them against a changed link and the update stops after installing the
+   the staging copy instead — rewriting it would only meet its own
-staging copy, saying so and naming this as the cause.
+   running-slot guard.
-
+3. `J` enters the staging copy, which rewrites the resident slot. On the
-An `OSCCAL`-baked image is a link change in effect even at an unchanged rate
+   patched-vector chips whose staging slot sits away from page 0 the host
-on paper: the staging copy shifts the physical clock the moment its `run()`
+   first re-aims word 0 at the staging copy, so a power loss mid-rewrite
-starts, and from then on speaks exactly what it was built for. Declare it
+   still resets into a loader; on the tiny13s the staging slot carries the
-like any other link change — `--staged-baud` with the new build's rate.
+   reset vector itself.
-
-The preflight refuses an image built for another chip: the stamp every pureboot
-binary carries must resolve to the device's own geometry, and the error names
-both. Die revisions share their base signature and geometry, so their images
-are interchangeable — as the silicon is.
-
-1. The staging slot `[base−512, base)` is saved to a host-side state file (on
-   the 1 KB tiny13s that is the whole application, vectors included).
-2. The resident installs the update image there. On the patched-vector chips
-   the host composes the slot's last word as a jump to the resident base, so
-   even an abandoned staging copy times out into a loader. A loader already
-   sitting whole in the staging slot is left as the staging copy instead —
-   rewriting it would only meet its own running-slot guard.
-3. `J` enters the staging copy, which rewrites the resident slot. Where a
-   patched reset vector routes through the resident, the host first re-aims
-   word 0 at the staging copy, so a power loss mid-rewrite still resets into a
-   loader; on the tiny13s the staging slot carries the reset vector itself.
 4. `J` enters the new resident, which restores the staging slot's saved
-   content, and the state file is discarded.
+   content (word 0 and the trampoline with it) and the state file is
+   discarded.
 
-Every phase is idempotent and keyed off the actual flash state, so re-running
+Every phase is idempotent and keyed off the actual flash state: re-running
-the same command after any interruption resumes and completes — with one
+the same command after any interruption resumes and completes. The state
-qualification, which is the link again: from step 2 on, the copy the re-run has
+file carries the only bytes not recoverable from the device; if it is lost
-to reach is the *new* image, so a resumed run needs the same `--staged-*` as the
+mid-update the update still completes, and the staging region is restored by
-first one. On a patched-vector part step 3 also re-aims word 0 at the staging
+reflashing the application. A boot-sectioned mega needs its fuses for the
-copy, so after that point a reset reaches the new image's link and **only** that
+preflight (BOOTSZ gate, profile notes) — read from the device, or supplied
-one; a re-run on the resident's link finds nothing at all. The state file carries
+with `--assume-fuses` where reading is impossible (simulators); the
-the only bytes not recoverable from the device; losing it mid-update still
+patched-vector chips need none.
-completes the update, and the staging region comes back by reflashing the
-application. A boot-sectioned mega needs its fuses for the preflight — read
-from the device, or supplied with `--assume-fuses` where reading is impossible
-(simulators).
 
 ## Host tool
 
 `pureboot.py` — Python 3, standard library only. The port layer is the one
-platform-specific part: termios drives any tty on POSIX (a USB adapter as well
+platform-specific part: termios drives any tty on POSIX (a USB adapter as
-as a simavr pty), the Win32 serial API through `ctypes` drives a COM port on
+well as a simavr pty), the Win32 serial API through `ctypes` drives a COM
-Windows (`--port COM6`; the `\\.\` form for two-digit ports is supplied by the
+port on Windows (`--port COM6`; the `\\.\` form for two-digit ports is
-tool). Opening the port asserts DTR and RTS on both, so a board that wires DTR
+supplied by the tool). Opening the port asserts DTR and RTS on both, so a
-to reset gets its reset pulse and opens the activation window by itself.
+board that wires DTR to reset gets its reset pulse and opens the activation
+window by itself.
 
     pureboot.py --port /dev/ttyUSB0 --baud 57600 \
         --info --fuses --flash app.hex
 
 Operations run in a fixed order within one session: info, fuses, loader
-update, flash (erase / program / read / verify), EEPROM (the same), then
+update, flash (erase / program / read / verify), EEPROM (erase / program /
-`--peek`/`--poke` — then the loader hands over to the application. `--stay` keeps the session alive
+read / verify) — then the loader hands over to the application; `--stay`
-instead, and a later invocation reconnects into it. `--flash` and `--eeprom`
+keeps the session alive instead, and a later invocation reconnects into it
-verify by read-back unless `--no-verify`, and a flash page that reads back
+(the knock converges there too). `--flash` and `--eeprom` verify by
-wrong is rewritten up to three times before the run stops (see `W` above).
+read-back unless `--no-verify`; images are raw binary, or Intel HEX by
-`--verify-flash` only reports. Images are raw binary, or Intel HEX by
+extension. `--force` overrides the refusable safety checks (today: flashing
-extension. `--force` overrides the refusable safety checks — today, flashing
+application data into a mega's reset walk region).
-application data into a mega's reset walk region.
 
-`--autobaud` opens with the calibration pulse instead of the plain knock, for a
+Readouts come one fact per line: `--info` prints the decoded info block
-loader built `SERIAL autobaud`; the rest of the session is identical, at
+field by field, `--fuses` each fuse byte on its own line — plus, on a
-whatever `--baud` the host chose. Its `--info` adds the **measured clock** —
+boot-sectioned mega, the decoded meaning (where the BOOTSZ section starts,
-the loader's bit-period unit, decoded and multiplied by the session rate —
+what BOOTRST does to reset). Transfers that take wire time — programming,
-which is the number an `OSCCAL` bake or a fixed-baud build for the part is
+reading, erasing, verifying, the update phases — draw a transient progress
-held against; `--clock <hz>` states the drift against a nominal.
+bar on stderr when it is a tty; logs and pipes see only the summary lines.
-
+`-v`/`--verbose` adds the decisions as they happen: knock counts, the
-`--one-wire` marks the link as a shared line (*One-wire* above): the tool
+programming plan (vector-surgery targets, skipped blank pages), update
-reads back and verifies its own echoed bytes, whatever the backend.
+state handling and per-phase page counts.
-It combines with everything, `--scan` included — undiscarded echoes would
-answer every rate a scan probes.
-
-`--scan` is the diagnosis once a fixed-baud loader has gone silent: it walks
-±10 % around `--baud` in 2 % steps, nearest first, one probe per activation
-window — reset the target as each probe announces itself (a board with DTR
-wired to reset is pulsed by the probe's own port-open). A loader
-off-frequency answers at its oscillator's ratio, and the report gives the
-found rate as the session workaround, the offset, the OSCCAL correction's
-direction at ~1 % per step, and the autobaud way out. Standalone — no other
-operation combines with it.
-
-`--peek ADDR[:N]` and `--poke ADDR:HEX` reach the data space (pureboot 5) —
-SRAM, and through the same address space the register file and every I/O
-register. Reading an I/O register can have side effects (reading UDR clears its
-flags), which is the caller's business to know.
-
-Reads are safe anywhere; **two small regions cannot be written without ending the
-session,** because they are what the loader is standing on:
-
-- the **top of SRAM**, where its stack lives — a handful of bytes below RAMEND;
-- on an **autobaud** build, the **measured bit period**: two bytes in
-  GPIOR2:GPIOR1 where the chip has the pair (data `0x32..0x33` on the
-  t25/45/85, `0x4A..0x4B` from the x8 generation on — such a loader has *no*
-  static RAM at all), and the two bytes at RAMSTART on the chips without one
-  (the t13s and classic megas), where they are the whole of the loader's
-  static RAM. Overwrite either home and the next reply is timed against
-  garbage — the symptom is a mangled prompt byte rather than any error; the
-  loader is fine, it simply is no longer speaking the agreed rate.
-
-Both are self-inflicted rather than defects, and a reset clears them. Note also
-that `--poke` can write OSCCAL, which does take effect — but a session can only
-survive a step or two of it before the clock walks the link out of the rate
-autobaud locked to, and OSCCAL reverts on reset regardless.
-
-Readouts come one fact per line: `--info` prints the device's version and
-signature and the geometry that follows from them, `--fuses` each fuse byte
-plus, on a boot-sectioned mega, its decoded meaning. Transfers that take wire time draw a transient progress bar on stderr
-when it is a tty. `-v`/`--verbose` adds the decisions as they happen: knock
-counts, the programming plan, update state handling and per-phase page counts.
 
 ## Tests
 
-libavr rides as the `libavr/` submodule (`git submodule update --init libavr`);
+`tools/check.sh` runs every chip's workflow (`tools/check.sh --full` adds
-`LIBAVR_ROOT` (cache or environment) overrides it for tandem development
+the reflect-mode builds of libavr's spot set; `tools/make_presets.py`
-against a working tree. `tools/check.sh` runs every chip's workflow (`--full`
+regenerates the presets). Per chip preset, `ctest` runs:
-adds the reflect-mode builds of libavr's spot set; `tools/make_presets.py`
-regenerates the presets).
-Per chip preset, `ctest` runs:
 
-- `pureboot.size` — the 510-byte (patched-vector) / 512-byte budget;
+- `pureboot.size` — the 510-byte (tinies) / 512-byte (mega) budget;
-- `pureboot_*.size` — the size matrix: the serial backends × the clock ladder
+- `pureboot_*.size` — the size matrix: the serial backends × the clock
-  (1/8/16 MHz; the t13s' own RC menu), the USART1 instance across that same
+  ladder (1/8/16 MHz; the t13s' own RC menu), plus the USART1 build on the
-  ladder on the x4 chips, and `pureboot_sw_wide`, the slowest ladder rate at
+  x4 chips — every configuration axis that could move the image, each
-  the fastest clock — where a software UART's per-bit spin outgrows its
+  variant against the same slot budget (pins are immediate operands and the
-  one-register delay loop and takes the 16-bit one. That is the largest image
+  timeout is a constant: size-neutral);
-  the configuration space produces, and a shape the ladder default (always the
+- `pureboot.custom` (328P) — the configured-deployment acceptance test: the
-  *fastest* rate a clock reaches) never picks. Pins are an axis for one reason
+  1 MHz software-serial TX=PB1/RX=PB5 build from the configuration example
-  only, and it is enough: a bit-banged link on a USART's own pins has to
+  drives the full protocol suite through the runner's GPIO bridge, fixture
-  release that USART, so `pureboot_{sw,autobaud}_on_usart{0,1}` build there
+  application included;
-  too. The timeout is a constant and is no axis;
+- `pureboot.usart1` (644A) — the same protocol suite over the second
-- `pureboot_autobaud.size` — the clock-free build, which has no clock or baud
+  hardware USART: instance selection is compile-checked everywhere, but
-  axis of its own: one binary per chip has to serve every point the matrix
+  only a live session proves the loader polls the USART it claims;
-  below sweeps. `pureboot*osccal*.size` add the `OSCCAL` trim on the stock
+- `pureboot.pi` — the position-independence lint: no absolute `jmp`/`call`
-  shape and on the tightest image in the space (autobaud on a USART's own
+  in the image, the info block within its first 256 bytes;
-  pins), holding both of the trim write's addressing encodings to the budget;
+- `pureboot.planner` — the host tool's pure logic: programming orders and
-- `pureboot_autobaud.unit` — the measured bit period sits where `--info`
+  their recovery properties, the surgery, the staging composition, the
-  reads it (wire contract, not layout accident): in the GPIOR pair, with no
+  boot-fuse decode, and the update preflight's error/warning matrix over
-  RAM object at all, on the chips that have one; as the loader's only RAM
+  synthetic fuse bytes;
-  object at exactly ram_start elsewhere;
-- `pbm_*.size` — with `PUREBOOT_FULL_MATRIX=1`, the exhaustive cross product
-  replacing that compact matrix, on **every** chip: every plausible oscillator
-  (the internal ones, the CKDIV8 floor, the plain and the UART crystals) ×
-  every rate reachable from it × every backend, unreachable combinations
-  dropping out rather than aborting the configure. Thousands of points per
-  chip, and cheap enough to run rather than reason about;
-- `pureboot.pi` — the position-independence lint: no absolute `jmp`/`call`, no
-  flash-resident section but `.text`, and the image byte-identical when linked
-  at a different base — which is position independence itself rather than a
-  proxy for it;
-- `pureboot.handshake` — the host tool's activation must not hang on a target
-  that never falls quiet: the drain after a prompt is bounded by the handshake
-  deadline, and a well-behaved loader still connects;
-- `pureboot.updatelink` — an update whose image changes the baud or the backend
-  must follow the staging copy onto *its* link, since that copy is the new image;
-  and where nothing was declared, the failure must name the link rather than
-  report a bare activation timeout, because by then the staging slot is written
-  and on a 1 KiB tiny that was the application;
-- `pureboot.planner` — the host tool's pure logic: programming orders and their
-  recovery properties, the surgery, the staging composition, the boot-fuse
-  decode, the update preflight over synthetic fuse bytes, and the repairing
-  verify against a fake device;
-- `pureboot.scan` — `--scan`'s walk and report logic: the probe order, the
-  rate arithmetic, and the trim advice's direction. A pty carries bytes at
-  any termios rate, so the rate physics itself belongs to the hardware
-  harness, and what the wire would arbitrate is pinned as logic;
-- `presets.generated` — CMakePresets.json matches its generator
-  (`tools/make_presets.py --check`), so a hand edit or a generator change
-  cannot drift the pair apart;
 - `pureboot.protocol` — end to end against a simavr device
-  (`test/pureboot_device.cpp`: a hardware USART as a pty, or a cycle-timed
+  (`test/pureboot_device.c` — a hardware USART as a pty, or a cycle-timed
-  GPIO⇄pty bridge for a software-UART build, plus the SPM/NVM module simavr's
+  GPIO⇄pty bridge for a software-UART build, selected with `-l` to match
-  tiny cores lack) driven by the real host tool through knock-from-reset,
+  the loader's link; plus the SPM/NVM module simavr's tiny cores lack)
-  program + verify of both memories, session reconnect, an external reset
+  driven by the real host tool through
-  through the patched vector, and the hand-over to a fixture application whose
+  knock-from-reset, program + verify of both memories, session reconnect, an
-  banner proves the launch — cross-checked against the simulator's
+  external reset through the patched vector, and the hand-over to a fixture
-  ground-truth memory dumps and an independent decode of the surgery;
+  application whose banner proves the launch — cross-checked against the
-- `pureboot.reloc` — the identical image one slot below the resident serves the
+  simulator's ground-truth memory dumps and an independent decode of the
-  complete command set from there;
+  surgery's rjmp words;
-- `pureboot.rehome` (t85) — a loader programmed at address 0 or in the staging
+- `pureboot.reloc` — the identical image installed one slot below the
-  slot re-homes into the top slot through the ordinary update flow;
+  resident serves the complete command set from there (the
-- `pureboot.custom` (328P) — the configuration example's 1 MHz software-serial
+  position-independence acceptance test);
-  build driving the full protocol suite, proving the plumbing produces a
+- `pureboot.update` — the full `--update-loader` flow to a re-timed build,
-  working loader and not just one that fits;
+  then every power-fail phase: the device is killed mid-write, restarted
-- `pureboot.usart1` (644A) — the same suite over the second hardware USART:
+  from its flash dump, and a re-run must complete the update with the
-  instance selection is compile-checked everywhere, but only a live session
+  application intact throughout.
-  proves the loader polls the USART it claims;
-- `pureboot.mute` (328P) — a software link on USART0's own pins, entered from an
-  application that handed over with that USART still enabled: the loader must
-  still answer, which it does only because it releases it. The pin ownership is
-  the runner's, not simavr's — simavr wires a USART through IRQs and never takes
-  the pin from the port, so without that model the state under test could not
-  arise at all;
-- `pureboot.dirty` (328P) — entering the loader from a running application over
-  an SPM buffer it deliberately dirtied, the case the loader declines to guard:
-  a bare verify must see the corruption and the repairing verify must fix it in
-  one rewrite. Hardware forbids the state here, but simavr dispatches SPM from
-  anywhere, which is what makes the path constructible;
-- `pureboot.update` — the full `--update-loader` flow, then every power-fail
-  phase: the device is killed mid-write, restarted from its flash dump, and a
-  re-run must complete the update with the application intact;
-- `pureboot.osccal` (328P, t85) — a loader built with the `OSCCAL` axis holds
-  the trim register at the built byte from its first prompt, observed through
-  the wire on one chip per addressing encoding (`sts` and low-I/O `out`);
-- `pureboot.autobaud` (328P, 1284P) — the clock-free build over the GPIO⇄pty
-  bridge: the calibration handshake, a flash + EEPROM + fuse round trip against
-  the simulator's own memory, a data-space round trip, the hand-over — then the
-  same binary again at double the clock, which is the property the backend
-  exists for. The measured clock `--info` prints is asserted against the
-  simulator's exact clock, inside the unit encoding's own envelope, at both
-  points. A lone calibration pulse with no knock behind it must still let
-  the application boot, so no wait in activation can be unbounded.
 
-`size`, `unit`, `pi`, `planner`, `scan` and `handshake` are host logic and run
+`size`, `pi`, and `planner` are host logic and run anywhere; the
-anywhere; the simulator-driven targets need simavr and a pty, so they are
+simulator-driven targets need simavr and a pty, so they are POSIX-only —
-POSIX-only.
+on Windows the tool is exercised against real hardware.
-
-## Hardware
-
-The suite above proves the protocol on every chip; it cannot prove a *board*.
-Two things live only on silicon: an RC oscillator that is not on its nominal, and
-a reset edge that has to come from somewhere. `tools/pbrig.py` and
-`tools/pbhw.py` cover that, and know nothing per-board — every deployment fact
-is a flag or a `PUREBOOT_*` environment variable.
-
-```sh
-export PUREBOOT_PROGRAMMER=atmelice_isp PUREBOOT_PART=t13 PUREBOOT_PORT=COM6
-tools/pbrig.py backup rig-backup/          # verified, before anything is written
-tools/pbhw.py --autobaud --loader build/ab.bin --app build/pbapp.hex --marker APP
-```
-
-`pbrig.py` is the primitives — `signature`, `reset`, `flash`, `fuses`, `backup`,
-`rate` — and the module `pbhw.py` builds on. Two rig facts are encoded in it
-because neither is guessable: an **ISP access is the reset edge** (the part runs
-the moment the programmer releases it, which is the only edge available when the
-adapter's DTR is not wired to reset, so a session begins with an ISP touch and
-knocks immediately after), and **avrdude splits `-U` on colons**, so a Windows
-path's drive letter breaks the spec and every file is passed as a bare name with
-avrdude run in its own directory.
-
-`pbrig.py rate` is the one that turns "the loader is silent, so the wiring must
-be wrong" into a number. Against a fixture built with `PUREBOOT_HEARTBEAT` — a
-*fixed* cycles-per-bit transmitter — it sweeps the host rate, and the band where
-the marker still decodes brackets the part's true bit rate; with the clock the
-image was built for, that is the clock the part is really running at. No
-instrument beyond the adapter already attached. An ATtiny13A measured this way
-came out at 9.072 MHz against its 9.6 MHz nominal, −5.5 % — inside the
-datasheet's ±10 % and outside what an 8N1 frame survives, which is the whole
-case for the autobaud backend on such a part.
-
-`pbhw.py` takes its bounds from the info block the loader reports, so one run
-covers a 1 KiB tiny and a 128 KiB mega alike: identity, the EEPROM round trip
-and erase, an application flashed and verified and then *seen running*, the
-application region read and erased, the loader slot proven intact across that
-erase by an independent ISP read, and an oversized image refused. It overwrites
-the application flash and EEPROM, which is why `backup` comes first.

pureboot/pureboot.cpp

@@ -1,16 +1,28 @@
-// pureboot — a serial bootloader on libavr: one C++ source, no inline
+// pureboot — a serial bootloader on libavr, pure by constraint: one C++
-// assembly, no global register variables, a 512-byte slot on every chip
+// source with no inline assembly and no global register variables, built for
-// libavr targets. The device speaks primitives; every composite (verify, erase,
+// every chip libavr targets, 512 bytes on each. The device speaks primitives
-// reset-vector surgery, self-update) lives in the host tool. Protocol,
+// — read/program flash, read/write EEPROM, fuse bytes, an info block, a jump
-// deployment and configuration: README.md next to this file.
+// — and everything composite (verify, erase, reset-vector surgery, updating
+// the loader itself) lives in the host tool. Protocol reference: README.md
+// next to this file.
 //
-// The image is position-independent — PC-relative control flow, wire
+// The image is position-independent: control flow is PC-relative, the write
-// addresses in, the write guard and the info block both anchored on the
+// and read paths take wire addresses, the write guard refuses the 512-byte
-// runtime return address — so the identical binary runs from any slot. That
+// slot the code is *running* in (taken from the runtime return address), the
-// is what makes a copy one slot below able to rewrite the resident one, and
+// info block is read relative to that same anchor, and the application jump
-// every change here has to keep it (test/check_pi.py).
+// is an indirect call to an absolute entry. The identical binary therefore
-
+// runs from any 512-byte slot with every command intact: flashed one slot
-#include <chrono>
+// below the resident loader it becomes the staging loader that rewrites the
+// resident — how pureboot updates itself, host-driven, with no other
+// firmware involved.
+//
+// Entry: reset lands in avr::startup::entry below (BOOTRST on the
+// boot-sectioned megas; the patched reset vector — or erased flash walking
+// up into the loader — on the tinies and the boot-section-less m48s). A
+// watchdog reset hands straight to the application. Otherwise the
+// host has one activation window per awaited knock byte ("pb"); an idle line
+// boots the application. A session then stays in the command loop until 'J'
+// jumps away or the chip resets.
 
 #include <libavr/libavr.hpp>
 
@@ -21,145 +33,100 @@ namespace ee = avr::eeprom;
 namespace pureboot {
 namespace {
 
-// Purely polled: every interrupt guard folds to nothing.
+// Purely polled — interrupts stay off, every guard folds to nothing.
 constexpr auto off = avr::irq::guard_policy::unused;
 
 constexpr std::uint8_t ack = '+';
 
-// Deployment parameters come from the build (pureboot_add_loader()). The
+// Per-deployment personality, passed in by the build — pureboot_add_loader()
-// signature is not one of them: the chip database is the only universal
+// (the CMake function next to this file) resolves the defaults: the clock the
-// source — a tiny13A cannot read its own signature row from code. An autobaud
+// board actually runs, the wire baud, the serial backend and its pins. The
-// build carries no clock and no baud at all; it measures both.
+// device signature needs no configuring — it comes from the chip database
-#if !defined(PUREBOOT_AUTOBAUD) && (!defined(PUREBOOT_CLOCK_HZ) || !defined(PUREBOOT_BAUD))
+// (avr::hw::db.signature), the only universal source, since the tiny13A
+// cannot even read its signature row from code.
+#if !defined(PUREBOOT_CLOCK_HZ) || !defined(PUREBOOT_BAUD)
 #error \
-    "PUREBOOT_CLOCK_HZ and PUREBOOT_BAUD select this build's clock and baud — create loader targets with pureboot_add_loader(), or PUREBOOT_AUTOBAUD for a clock-free one (README.md)"
+    "PUREBOOT_CLOCK_HZ and PUREBOOT_BAUD select this build's clock and baud — create loader targets with pureboot_add_loader() (README.md)"
 #endif
 
-#if !defined(PUREBOOT_AUTOBAUD)
 using dev = avr::device<{.clock = avr::hertz_t{PUREBOOT_CLOCK_HZ}}>;
 constexpr avr::baud_t wire_baud{PUREBOOT_BAUD};
-#endif
 
-// The loader owns the top 512 bytes; a staging copy goes in the slot below.
+// The watchdog reset flag's home: MCUSR, or the classic megas' MCUCSR.
-// Chips without a hardware boot section — the tinies and the m48s, whose SPM
+consteval std::int16_t wdrf_field()
-// runs from anywhere (Atmel-8271 §26) — keep the application's relocated
+{
-// reset vector in the word under the slot.
+	auto reg = std::string_view{avr::hw::db.regs[static_cast<std::size_t>(avr::power::detail::reset_reg())].name};
-constexpr std::uint16_t slot_bytes = 512;
+	return avr::hw::db.field_index(reg, "WDRF");
+}
+
+// Geometry: the resident loader owns the top slot of flash — 512 bytes,
+// except on the >64 KiB chips whose own smallest boot sector is 1 KiB (the
+// 1284s): there the slot is 1 KiB, matching the hardware boundary the
+// 512-byte figure comes from everywhere else. The word below the slot is
+// the trampoline (the application's relocated reset vector) on chips
+// without a hardware boot section — the tinies and the m48s, whose SPM
+// runs from anywhere (Atmel-8271 §26). A boot section also means the CPU
+// runs on while the RWW section programs; everywhere else it halts through
+// the operation. The m48s still carry RWWSRE as their temporary-buffer
+// discard (§26.2), so the discard picks by that bit, not by the section.
+constexpr std::uint16_t slot_bytes = spm::flash_bytes > 65536 ? 1024 : 512;
+constexpr std::uint32_t base = spm::flash_bytes - slot_bytes;
 constexpr std::uint16_t page = spm::page_bytes;
 constexpr bool boot_section = avr::hw::curated::has_boot_section();
+constexpr bool rww_discard = spm::detail::has_rww();
 
-// Past 64 KiB one bank of flash does not cover the chip, so a transfer's
+// Past 64 KiB a byte address no longer fits the wire's 16 bits, so on the
-// selector byte carries the bank and the wire address stays a byte address
+// large chips every flash address on the wire — and all slot arithmetic —
-// within it. 'J' is the exception: it is a word address everywhere, because
+// is a word address instead ('J' always was one). A slot spans the same
-// that is what the hardware's own jump takes.
+// wire-high-byte pair in either unit (512 B = 2 x 256 bytes, 1 KiB =
-constexpr bool banked_flash = spm::flash_bytes > 65536;
+// 2 x 256 words), so the slot index is the high byte with its low bit
+// dropped everywhere.
+constexpr bool word_flash = spm::flash_bytes > 65536;
+constexpr std::uint16_t wire_base =
+    word_flash ? static_cast<std::uint16_t>(base / 2) : static_cast<std::uint16_t>(base);
+constexpr std::uint16_t wire_page_mask = word_flash ? (page / 2 - 1) : (page - 1);
 
-// A compile-time window, so the whole EEPROM belongs to the application;
+// The activation window, in seconds, is a compile-time constant (the build
-// re-timing a deployed loader is a self-update with a re-timed build. An
+// may override it): the whole EEPROM belongs to the application, and
-// autobaud build has no clock to convert seconds against and counts polls.
+// re-timing the loader is a bootloader self-update with a re-timed binary.
 #if !defined(PUREBOOT_TIMEOUT)
 #define PUREBOOT_TIMEOUT 8
 #endif
 constexpr std::uint8_t timeout_seconds = PUREBOOT_TIMEOUT;
 
-#if !defined(PUREBOOT_AUTOBAUD_POLLS)
+// The 12-byte info block the host reads with the 'b' command; flash-resident
-#define PUREBOOT_AUTOBAUD_POLLS 4000000
+// (there is no crt to copy a .data image), word-aligned so its wire (word)
-#endif
+// address is exact on the large chips. The page byte is the wire count
-constexpr avr::uint24_t autobaud_budget = PUREBOOT_AUTOBAUD_POLLS;
+// convention: 0 means 256.
-
+[[gnu::progmem]] alignas(2) inline constexpr std::array<std::uint8_t, 12> info_data = {
-// A build may bake a measured oscillator trim (README.md: the RC-oscillator
+    'P',
-// deployment answer); the byte is applied at the top of run(). Orthogonal to
+    'B',
-// the serial backend — an autobaud build may carry it for the application's
+    1, // magic, protocol version
-// benefit alone.
-#if defined(PUREBOOT_OSCCAL)
-static_assert(PUREBOOT_OSCCAL >= 0 && PUREBOOT_OSCCAL <= 0xff, "PUREBOOT_OSCCAL is one OSCCAL byte");
-#endif
-
-// The loader's one identity number. The protocol carries none of its own —
-// a version implies it, and the host tool holds that map (README.md).
-constexpr std::uint8_t version = 8;
-
-// The image's identity stamp, for the host tool rather than for the wire: an
-// update image is a bare 512-byte slot, and without this nothing in it says
-// which chip it was built for. The tool refuses to install an image whose
-// stamp does not match the device — flashing a foreign loader bricks the
-// target, and the loader itself cannot check what has already replaced it.
-//
-// Never read from flash by the loader — 'b' answers out of this array, but at
-// constant indices, so those fold to immediates and no runtime address of it
-// is ever formed. `used` keeps the compiler from dropping the copy the host
-// needs and `retain` keeps --gc-sections from collecting it.
-// clang-format off
-[[gnu::used, gnu::retain, gnu::section(".text.stamp")]]
-inline constexpr std::uint8_t identity_stamp[]{
-    'P', 'B',                      // the magic the host scans an image for
-    version,                       // and from here on, exactly what 'b' answers
     avr::hw::db.signature[0],
     avr::hw::db.signature[1],
     avr::hw::db.signature[2],
+    static_cast<std::uint8_t>(page),
+    wire_base & 0xff,
+    wire_base >> 8, // app flash ends here; resident loader base (a word address on large chips)
+    avr::hw::db.mem.eeprom_size & 0xff,
+    avr::hw::db.mem.eeprom_size >> 8,
+    // bit 0: host must patch the reset vector (no hardware boot section);
+    // bit 1: flash wire addresses are word addresses
+    static_cast<std::uint8_t>((boot_section ? 0 : 1) | (word_flash ? 2 : 0)),
 };
-// clang-format on
-// Where the identity proper starts: past the magic the host scans for.
-constexpr std::uint8_t stamp_identity = 2;
 
-// The address spaces a transfer can name, in a selector byte's low nibble.
+// The serial link. PUREBOOT_USART forces a hardware USART instance,
-// Flash is 0 so it is the cheapest to select.
+// PUREBOOT_SOFT_SERIAL the polled software UART (no vector — the table
-//
+// belongs to the application) on PUREBOOT_RX/PUREBOOT_TX; with neither, the
-// spm_ops is the one that is not memory: a write there hands its byte to
+// chip's first USART where it has one and the software UART elsewhere. Both
-// SPMCSR and fires the instruction at the transfer's address, which is how
+// are class templates on the clock so only the selected backend is ever
-// page erase, page write and RWW re-enable reach the wire without the loader
+// instantiated. pending() is the cheap line test the activation window
-// carrying a command for each. The hardware's four-cycle store-to-SPM window
+// polls; rx() then picks the byte up; drain() holds until the last
-// is why this is one fused primitive and not a poke of SPMCSR — no host can
+// transmitted frame is fully on the wire (the jump hand-over must not let
-// hit that window across a serial link.
+// the target's re-init clip the ack).
-enum : std::uint8_t { sp_flash = 0, sp_eeprom = 1, sp_data = 2, sp_fuse = 3, sp_spm = 4 };
-
-// A selector's high nibble is the flash bank — the address bits above the
-// 16-bit wire address, RAMPZ on the chips that have one. Keeping it here
-// rather than widening the wire address is what lets one 16-bit cursor serve
-// every space: a 24-bit cursor would pay its extra byte on EEPROM and data
-// reads that can never need it.
-[[gnu::always_inline]] inline std::uint8_t space_of(std::uint8_t selector)
-{
-	return selector & 0x0f;
-}
-
-[[gnu::always_inline]] inline std::uint8_t bank_of(std::uint8_t selector)
-{
-	return static_cast<std::uint8_t>(selector >> 4);
-}
-
-// The slot a flash address falls in, as one byte. A slot is half as many words
-// as bytes, so the word address's high byte is exactly this index — which is
-// what lets the write guard compare a single byte, and what the running copy's
-// own return address yields for free.
-constexpr std::uint8_t slot_shift = std::countr_zero(slot_bytes);
-constexpr std::uint8_t bank_shift = 16 - slot_shift;
-
-[[gnu::always_inline]] inline std::uint8_t slot_of([[maybe_unused]] std::uint8_t bank, std::uint16_t at)
-{
-	const auto within = static_cast<std::uint8_t>(at >> slot_shift);
-	if constexpr (banked_flash)
-		return static_cast<std::uint8_t>((bank << bank_shift) | within);
-	else
-		return within;
-}
-
-// The serial link, per the build's PUREBOOT_USART / PUREBOOT_SOFT_SERIAL /
-// PUREBOOT_AUTOBAUD, defaulting to the chip's USART0 where it has one. The
-// software receiver is the polled one: the vector table belongs to the
-// application. Templates on the clock, so only the selected backend
-// instantiates. pending() is the cheap line test the activation window polls;
-// drain() holds until the last frame is off the wire, so a hand-over cannot
-// let the target's re-init clip the ack.
 #if defined(PUREBOOT_SOFT_SERIAL) && defined(PUREBOOT_USART)
 #error "PUREBOOT_SOFT_SERIAL and PUREBOOT_USART select opposing serial backends"
 #endif
-#if defined(PUREBOOT_AUTOBAUD) && defined(PUREBOOT_USART)
-#error "PUREBOOT_AUTOBAUD measures a software link; it cannot drive a hardware USART"
-#endif
-#if defined(PUREBOOT_HALF_DUPLEX) && (defined(PUREBOOT_SOFT_SERIAL) || defined(PUREBOOT_AUTOBAUD))
-#error "PUREBOOT_HALF_DUPLEX is the hardware USART's one-wire mode; a software link goes one-wire by RX == TX"
-#endif
 #if !defined(PUREBOOT_RX)
 #define PUREBOOT_RX pb0
 #endif
@@ -167,50 +134,28 @@ constexpr std::uint8_t bank_shift = 16 - slot_shift;
 #define PUREBOOT_TX pb1
 #endif
 #if defined(PUREBOOT_USART)
-constexpr int usart_unit = PUREBOOT_USART;
+constexpr char usart_digit = '0' + PUREBOOT_USART;
 #else
-constexpr int usart_unit = 0;
+constexpr char usart_digit = '0';
 #endif
 
-// One-wire on the hardware USART (PUREBOOT_HALF_DUPLEX): RXD and TXD tied
+// Whether the chip carries the selected USART: the suffixed instance name,
-// together off-chip, exactly one direction enabled at a time — the library's
+// or — for instance 0 — the classic megas' un-numbered block.
-// .half_duplex turn-around. The activation window is unchanged; only its
+consteval bool usart_exists()
-// poll grows the release-line test rx_ready() carries in this mode.
+{
-constexpr bool hw_half_duplex =
+	const char name[] = {'U', 'S', 'A', 'R', 'T', usart_digit};
-#if defined(PUREBOOT_HALF_DUPLEX)
+	if (avr::hw::db.has_instance(std::string_view{name, sizeof(name)}))
-    true;
+		return true;
-#else
+	return usart_digit == '0' && avr::hw::db.has_instance("USART");
-    false;
+}
-#endif
 
-template <avr::hertz_t C, avr::baud_t B>
+template <avr::hertz_t C>
 struct hardware_link {
-	using uart = avr::uart::usart<usart_unit, C, {.baud = B, .max_baud_error = 2.5_pct, .half_duplex = hw_half_duplex}>;
+	using uart = avr::uart::usart<usart_digit, C, {.baud = wire_baud, .max_baud_error = 2.5_pct}>;
 
-	// The compiled idle poll around the window's narrow (uint24_t) countdown:
+	// The compiled idle poll: lds UCSR0A (2), sbrc skipping the exit (2),
-	// the RXC test, then sbiw + sbci + brne (5). The test's cost follows the
+	// sbiw + sbci + sbci + brne (6).
-	// status register's home — a 2-cycle bit-skip where UCSRnA sits in
+	static constexpr std::uint8_t poll_cycles = 10;
-	// bit-addressable I/O (the classic megas), lds + skip (4) in extended
-	// I/O. Half-duplex polls through rx_ready()'s release-line test, which
-	// -Os outlines: the rcall (3), the UCSR#B read and not-taken skip with
-	// the jump over the write (I/O 3, extended 5), the ret (4) — and the
-	// call in the loop body pushes the countdown into call-saved registers,
-	// where the uint24_t step is ldi+sub+sbc+sbc (4) instead of sbiw+sbci
-	// (3). Measured off the built loops: 18 a poll in bit-addressable I/O,
-	// 22 in extended. A uint32_t countdown pays one more sbci —
-	// window_polls() adds it where the count forces the wide type. Held per
-	// chip by the pureboot.window gates. The lookup rides the baud parameter
-	// so it stays dependent: the trait is an incomplete type on the
-	// USART-less chips, which parse this template without ever instantiating
-	// it.
-	template <avr::baud_t Baud, typename U = avr::hw::usart_of<usart_unit>>
-	static consteval std::uint8_t poll_cost()
-	{
-		if (hw_half_duplex)
-			return U::ucsra::addr < 0x40 ? 18 : 22;
-		return U::ucsra::addr < 0x40 ? 7 : 9;
-	}
-	static constexpr std::uint8_t poll_cycles = poll_cost<B>();
 
 	static void init()
 	{
@@ -234,26 +179,18 @@ struct hardware_link {
 
 	static void drain()
 	{
-		// A drain here always follows this link's own write — the frame is
+		uart::drain();
-		// in flight by construction, so the completion the wait needs is
-		// guaranteed and the bounded default's countdown would be dead bytes.
-		uart::drain_unbounded();
 	}
 };
 
-template <avr::hertz_t C, avr::baud_t B>
+template <avr::hertz_t C>
 struct software_link {
-	// RX == TX is the one-wire deployment: the transmitter becomes a guest
+	using rx_t = avr::uart::software_rx_polled<C, avr::PUREBOOT_RX, wire_baud>;
-	// on the receiver's pull-up line, taking the pin's direction for exactly
+	using tx_t = avr::uart::software_tx<C, avr::PUREBOOT_TX, wire_baud>;
-	// one frame per byte.
-	using rx_t = avr::uart::software_rx_polled<C, avr::PUREBOOT_RX, B>;
-	using tx_t = avr::uart::software_tx<C, avr::PUREBOOT_TX, B, avr::PUREBOOT_RX == avr::PUREBOOT_TX>;
 
-	// The compiled idle poll around the window's narrow (uint24_t) countdown:
+	// The compiled idle poll: sbis skipping the exit (2), sbiw + sbci +
-	// sbis skipping the exit (2), sbiw + sbci + brne (5). A uint32_t
+	// sbci + brne (6).
-	// countdown pays one more sbci — window_polls() adds it where the count
+	static constexpr std::uint8_t poll_cycles = 8;
-	// forces the wide type. Held by the pureboot.window gate.
-	static constexpr std::uint8_t poll_cycles = 7;
 
 	static void init()
 	{
@@ -281,56 +218,21 @@ struct software_link {
 	}
 };
 
-// The clock-free link: the bit period is measured from the host's calibration
+#if defined(PUREBOOT_USART)
-// pulse instead of derived from a clock, so one image serves every F_CPU and
+static_assert(usart_exists(), "PUREBOOT_USART selects a hardware USART this chip does not have");
-// every rate. Activation differs in kind from the other two — there is no
+using link = hardware_link<dev::clock>;
-// clock to time a window against — so this backend brings its own, below.
-struct autobaud_link {
-	// The unit in GPIOR2:GPIOR1 where the chip has them: the loader owns the
-	// whole chip while it runs, and the pair costs one word per access where
-	// the RAM word costs two — six words across the image.
-	using uart = avr::uart::software_autobaud<avr::PUREBOOT_RX, avr::PUREBOOT_TX, avr::uart::unit_home::gpior>;
-
-	static void init()
-	{
-		avr::init<uart>();
-	}
-
-	static std::uint8_t rx()
-	{
-		return uart::template read<off>();
-	}
-
-	static void tx(std::uint8_t byte)
-	{
-		uart::template write<off>(byte);
-	}
-
-	static void drain()
-	{
-		// A drain here always follows this link's own write — the frame is
-		// in flight by construction, so the completion the wait needs is
-		// guaranteed and the bounded default's countdown would be dead bytes.
-		uart::drain_unbounded();
-	}
-};
-
-#if defined(PUREBOOT_AUTOBAUD)
-using link = autobaud_link;
-#elif defined(PUREBOOT_USART)
-static_assert(avr::uart::has_usart<usart_unit>(), "PUREBOOT_USART selects a hardware USART this chip does not have");
-using link = hardware_link<dev::clock, wire_baud>;
 #elif defined(PUREBOOT_SOFT_SERIAL)
-using link = software_link<dev::clock, wire_baud>;
+using link = software_link<dev::clock>;
 #else
-using link = std::conditional_t<avr::uart::has_usart<usart_unit>(), hardware_link<dev::clock, wire_baud>,
+using link = std::conditional_t<usart_exists(), hardware_link<dev::clock>, software_link<dev::clock>>;
-                                software_link<dev::clock, wire_baud>>;
 #endif
 
-// The application's entry, pinned by the linker (--defsym): word 0 on a
+// The application's entry, an absolute address the linker pins (--defsym in
-// boot-sectioned mega, the trampoline at base − 2 elsewhere. Reaching it must
+// CMakeLists.txt): 0x0000 on the mega (word 0 stays the application's own
-// not depend on where this copy runs, so the jump goes through a pointer, and
+// vector — BOOTRST re-vectors a reset into the loader in hardware) and the
-// [[gnu::noipa]] keeps the constant from folding back into a relative call.
+// trampoline word at base - 2 on the tinies. Reaching it must not depend on
+// where this copy runs, so the jump goes through a pointer: [[gnu::noipa]]
+// keeps the constant from folding back into a PC-relative call.
 extern "C" [[noreturn]] void pureboot_app();
 
 [[gnu::noipa, noreturn]] void jump(void (*target)())
@@ -339,65 +241,23 @@ extern "C" [[noreturn]] void pureboot_app();
 	__builtin_unreachable();
 }
 
-[[noreturn]] void run_app()
+[[gnu::noinline, noreturn]] void run_app()
 {
 	jump(pureboot_app);
 }
 
-// Activation: a bounded wait for the host, then the knock. Both forms boot the
+// One activation window is a single 32-bit poll countdown. The divisor is
-// application when the window closes on an idle line, and both bound *every*
+// the backend's counted poll-loop cycles (its own comment reads them off the
-// wait — a knock awaited without a deadline would let one stray edge hold an
+// compiled loop); whole-second precision is all the window promises, so the
-// unattended device in the loader forever.
+// nearest cycle count is plenty.
-#if defined(PUREBOOT_AUTOBAUD)
-// The window is a fixed poll budget: with no clock, whole seconds cannot be
-// timed. A uint24_t holds it — a fourth byte would cost two words at every
-// countdown step for range never used.
-void await_host()
-{
-	for (;;) {
-		if (!link::uart::calibrate(autobaud_budget))
-			run_app();
-		// The calibration pulse has already proven a host is there, so one
-		// byte activates. A knock that never arrives falls back to calibrate(),
-		// whose own budget then boots the application.
-		if (link::uart::template read<off>(autobaud_budget) == 'p')
-			return;
-	}
-}
-#else
-// The window as one countdown, divided by the backend's counted poll-loop
-// cycles. Whole seconds is all it promises. The per-poll cost depends on the
-// countdown's own width (a uint32_t decrement chain is one sbci longer), and
-// the width depends on the poll count — solved narrow-first: a count that
-// fits 24 bits at the narrow cost keeps the narrow loop, anything else takes
-// the wide loop at its own cost. A count fitting 24 bits only at the wide
-// cost stays wide, so the choice cannot oscillate on the boundary.
-consteval std::uint32_t polls_at(std::uint32_t per_poll)
-{
-	// Whole-window cycles first, then the per-poll division: one truncation
-	// instead of one per second. Same instructions either way — only the
-	// countdown's immediate moves.
-	return static_cast<std::uint32_t>(dev::cycles_for<std::chrono::seconds{timeout_seconds}>() / per_poll);
-}
-
-consteval bool narrow_window()
-{
-	return polls_at(link::poll_cycles) <= 0xffffff;
-}
-
 consteval std::uint32_t window_polls()
 {
-	return polls_at(narrow_window() ? link::poll_cycles : link::poll_cycles + 1u);
+	return timeout_seconds * static_cast<std::uint32_t>(dev::clock.hz / link::poll_cycles);
 }
 
-// The countdown in the narrowest type that holds it: a fourth byte would
-// cost a wider decrement chain at every poll for range most windows never
-// use (the autobaud budget makes the same choice).
-using window_t = std::conditional_t<narrow_window(), avr::uint24_t, std::uint32_t>;
-
 bool pending_before_deadline()
 {
-	window_t polls = window_polls();
+	std::uint32_t polls = window_polls();
 	do {
 		if (link::pending())
 			return true;
@@ -405,8 +265,8 @@ bool pending_before_deadline()
 	return false;
 }
 
-// A knock byte under the deadline: an idle window means no host, so the
+// A knock byte under the activation deadline: an idle line means no host is
-// application runs.
+// there, and the application runs.
 std::uint8_t rx_deadline()
 {
 	if (!pending_before_deadline())
@@ -414,194 +274,212 @@ std::uint8_t rx_deadline()
 	return link::rx();
 }
 
-void await_host()
+std::uint16_t rx16()
-{
-	// 'p' then 'b', each under a fresh window; anything else is line noise.
-	while (rx_deadline() != 'p' || rx_deadline() != 'b') {
-	}
-}
-#endif
-
-// Inlined: read across a call, the first byte strands in a call-saved
-// register the caller has to push and pop.
-[[gnu::always_inline]] inline std::uint16_t rx16()
 {
 	std::uint16_t low = link::rx();
 	return static_cast<std::uint16_t>(low | (link::rx() << 8));
 }
 
-// The wire's byte pair as the word it is — AVR is little-endian too, so the
+// The streamers take the count in the wire's 8-bit form: 0 means 256.
-// cast is the identity a shift-and-or spelling makes the compiler rediscover.
+// send_flash stays out of line: its two callers ('b' and 'R') otherwise each
-// Callers read into named variables first: the wire order is a sequence of
+// inline a private copy of the loop. On the large chips the address is a
-// reads, not an argument order.
+// word address and the read goes through ELPM (flash_load_far).
-[[gnu::always_inline]] inline std::uint16_t word_of(std::array<std::uint8_t, 2> pair)
+[[gnu::noinline]] void send_flash(std::uint16_t address, std::uint8_t count)
 {
-	return std::bit_cast<std::uint16_t>(pair);
+	if constexpr (word_flash) {
-}
+		// The 24-bit cursor as the machine holds it: the RAMPZ byte and a
-
+		// 16-bit Z, carried explicitly (the reassembled 32-bit address
-// Out of line: several sites send it, and a call is shorter than a
+		// folds away inside the inlined far load). A single read never
-// load-immediate at each.
+		// crosses a 64 KiB boundary — the protocol forbids it and the host
-[[gnu::noinline]] void tx_ack()
+		// splits its chunks there — so RAMPZ holds for the whole run.
-{
+		std::uint8_t rampz = static_cast<std::uint8_t>(address >> 15);
-	link::tx(ack);
+		std::uint16_t z = static_cast<std::uint16_t>(address << 1);
-}
+		do {
-
+			link::tx(avr::flash_load_far<std::uint8_t>((static_cast<std::uint32_t>(rampz) << 16) | z));
-// A wire address and its selector's bank as the flash address they name.
+			if (++z == 0)
-[[gnu::always_inline]] inline spm::flash_address_t flash_address([[maybe_unused]] std::uint8_t bank, std::uint16_t at)
+				++rampz; // robustness for a host that reads across 64 KiB
-{
+		} while (--count);
-	if constexpr (banked_flash)
+	} else {
-		return (static_cast<spm::flash_address_t>(bank) << 16) | at;
+		do
-	else
+			link::tx(avr::flash_load(reinterpret_cast<const std::uint8_t *>(address++)));
-		return at;
+		while (--count);
-}
-
-// One byte out of any space. Every accessor shares the transfer's cursor, its
-// loop and its call site, so a space costs only its own instruction rather
-// than a body, a loop and a dispatch arm of its own.
-[[gnu::always_inline]] inline std::uint8_t load(std::uint8_t space, [[maybe_unused]] std::uint8_t bank,
-                                                std::uint16_t at)
-{
-	if (space == sp_eeprom)
-		return ee::read(at);
-	if (space == sp_data)
-		return *reinterpret_cast<volatile std::uint8_t *>(at);
-	if (space == sp_fuse)
-		return spm::read_fuse<off>(static_cast<spm::fuse>(at));
-	if constexpr (banked_flash)
-		return avr::flash_load_far<std::uint8_t>(flash_address(bank, at));
-	else
-		return avr::flash_load(reinterpret_cast<const std::uint8_t *>(at));
-}
-
-// One byte into a writable space. Flash is not one of them — it arrives a
-// page at a time through 'W' and is committed through sp_spm — and the fuses
-// are not writable at all: SPM reaches flash and boot lock bits only.
-[[gnu::always_inline]] inline void store(std::uint8_t space, std::uint8_t bank, std::uint16_t at, std::uint8_t value,
-                                         std::uint8_t slot_high)
-{
-	if (space == sp_data) {
-		*reinterpret_cast<volatile std::uint8_t *>(at) = value;
-		return;
 	}
-	if (space == sp_spm) {
+}
-		// The running-slot write guard. An SPM command aimed at the slot this
+
-		// code executes from is dropped, so a broken host cannot brick the
+void send_eeprom(std::uint16_t address, std::uint8_t count)
-		// running loader — while a copy one slot lower may still rewrite the
+{
-		// resident one, which is what a self-update is. Guarding the commit
+	do
-		// rather than the page fill covers erase and write both, and leaves a
+		link::tx(ee::read(address++));
-		// refused page's words in the buffer: harmless, since the next page
+	while (--count);
-		// write auto-erases it (§26.2.1).
+}
-		if (slot_of(bank, at) != slot_high)
+
-			spm::command<off>(value, flash_address(bank, at));
+// EEPROM write, host-paced: each ack goes out once the byte's write has
-		// Only a boot-sectioned mega runs on while its RWW section programs;
+// begun, so the next byte arrives while it completes and the following
-		// everywhere else the CPU halts through erase and write, so the wait
+// write's own ready-wait sees an idle line. Nothing is ever missed, on
-		// is already over by the time it returns.
+// either serial backend, without a buffer.
+void store_eeprom(std::uint16_t address, std::uint8_t count)
+{
+	do {
+		ee::write<off>(address++, link::rx());
+		link::tx(ack);
+	} while (--count);
+}
+
+// One flash page: stream the bytes into the SPM buffer as little-endian
+// words, then erase and program — except the 512-byte slot this code runs
+// in, which is drained but never programmed, so a copy can never erase
+// itself. `slot_high` is the high byte of that running slot's base (run()
+// derives it); a broken host thus cannot brick the running loader, and a
+// copy flashed one slot lower may rewrite the slot above it — how pureboot
+// updates itself. On the mega the RWW section is re-enabled so reads work
+// immediately.
+void program_flash(std::uint16_t wire_address, std::uint8_t slot_high)
+{
+	// A buffer word cannot be loaded twice without an erase (§26.2.1), so a
+	// refused page's drained data must not linger for the next write:
+	// discard the buffer up front — CTPB on the tinies; on the megas
+	// writing RWWSRE aborts a pending load (§26.2.2 — on the m48s that
+	// flush is the bit's whole documented job).
+	if constexpr (rww_discard)
+		spm::rww_enable<off>();
+	else
+		spm::clear_buffer<off>();
+	// One induction either way. On the byte-addressed chips the wire address
+	// itself walks the page (aligned, so the offset bits wrap to zero); on
+	// the word-addressed large chips the wire word address becomes a 32-bit
+	// byte cursor once, and their 256-byte page makes its low byte the whole
+	// in-page offset. The slot index is one high byte of the wire address —
+	// two values on byte-addressed chips (the & ~1), bits 16:9 re-packed on
+	// the large ones.
+	spm::flash_address_t address;
+	std::uint8_t page_high;
+	if constexpr (word_flash) {
+		// Pages are aligned, so one page never crosses a 64 KiB boundary:
+		// RAMPZ is a per-page constant and the fill cursor is a 16-bit Z
+		// whose low byte is the whole in-page offset (256-byte pages). The
+		// slot index is simply the wire word address's high byte.
+		const std::uint8_t rampz = static_cast<std::uint8_t>(wire_address >> 15);
+		const std::uint16_t z0 = static_cast<std::uint16_t>(wire_address << 1);
+		std::uint16_t z = z0;
+		do {
+			std::uint8_t low = link::rx();
+			std::uint8_t high = link::rx();
+			spm::fill<off>((static_cast<spm::flash_address_t>(rampz) << 16) | z,
+			               static_cast<std::uint16_t>(low | (high << 8)));
+			z += 2;
+		} while (static_cast<std::uint8_t>(z));
+		address = (static_cast<spm::flash_address_t>(rampz) << 16) | z0;
+		page_high = static_cast<std::uint8_t>(wire_address >> 8) & 0xfe;
+	} else {
+		address = static_cast<spm::flash_address_t>(wire_address);
+		do {
+			std::uint8_t low = link::rx();
+			std::uint8_t high = link::rx();
+			spm::fill<off>(address, static_cast<std::uint16_t>(low | (high << 8)));
+			address += 2;
+		} while (static_cast<std::uint8_t>(address) & (page - 1));
+		address -= 2; // back inside the page — erase and write ignore the word bits
+		page_high = static_cast<std::uint8_t>(address >> 8) & 0xfe;
+	}
+	if (page_high != slot_high) {
+		// The tinies and the m48s halt the CPU through the erase and the
+		// write, so only the boot-sectioned megas — running on while their
+		// RWW section programs — wait.
+		spm::erase_page<off>(address);
 		if constexpr (boot_section)
 			spm::wait();
-		return;
+		spm::write_page<off>(address);
+		if constexpr (boot_section) {
+			spm::wait();
+			spm::rww_enable<off>();
+		}
 	}
-	// Host-paced: the ack goes out once the write has begun, so the next byte
-	// arrives while it completes and nothing is missed without a buffer.
-	ee::write<off>(at, value);
 }
 
-// One page into the SPM buffer, and only that: the erase and the write that
+// The four fuse/lock bytes in the hardware's own Z order: low, lock,
-// commit it are host-issued sp_spm stores, which reach the same fused
+// extended, high. Writing fuses is not a thing self-programming can do on
-// store-and-SPM pair through the transfer path's own address and data.
+// AVR — SPM reaches flash (and boot lock bits) only.
-//
+void send_fuses()
-// Nothing discards the buffer first: it is write-once per word (§26.2.1), so
-// filling over a refused page or an application's leavings programs stale
-// words — but a page write auto-erases it (§26.2.1; §19.2 on the tinies), so
-// that write clears the condition and the host's read-back rewrites the page.
-void fill_page(std::uint8_t bank, std::uint16_t at)
 {
-	// The address names a page, so its in-page bits are dropped and the walk
+	std::uint8_t which = 0;
-	// starts at the page base; the low byte of the cursor is the whole in-page
+	do
-	// offset, since a page is aligned and never crosses a bank.
+		link::tx(spm::read_fuse<off>(static_cast<spm::fuse>(which)));
-	std::uint16_t z = at & ~static_cast<std::uint16_t>(page - 1);
+	while (++which & 3);
-	do {
-		std::uint8_t low = link::rx();
-		std::uint8_t high = link::rx();
-		spm::fill<off>(flash_address(bank, z), word_of({low, high}));
-		z += 2;
-	} while (static_cast<std::uint8_t>(z) & (page - 1));
 }
 
 [[noreturn]] void run()
 {
-#if defined(PUREBOOT_OSCCAL)
+	// A watchdog reset belongs to the application (whose watchdog stays
-	// The build's oscillator trim, ahead of everything — the WDRF bail
+	// forced on until it clears WDRF) — no activation window in its way.
-	// included — so every path out of reset, the watchdog hand-over to the
+	// The flag register is MCUSR, or the classic megas' MCUCSR.
-	// application first among them, runs on the corrected clock.
+	if (avr::hw::field_impl<wdrf_field()>::test())
-	avr::clock::calibrate(PUREBOOT_OSCCAL);
-#endif
-	// A watchdog reset belongs to the application, whose watchdog stays forced
-	// on until it clears WDRF — no activation window in its way.
-	if (avr::power::peek_reset_cause().watchdog)
 		run_app();
 
 	link::init();
 
-	// The slot this copy runs in, which the write guard follows: the return
+	// The high byte of the 512-byte-aligned base this copy runs at: the
-	// address is a word address and a slot is half as many words as bytes, so
+	// return address is a word address, whose high byte is the 256-word slot
-	// its high byte is the slot index outright. No absolute address is ever
+	// index — on byte-addressed chips doubled back into byte terms.
-	// formed, so the image stays position-independent.
+	// program_flash refuses this one slot and the info block is addressed
-	const auto slot_high = avr::startup::caller_page();
+	// from it, so both follow wherever the code was flashed. The high byte is
+	// spelled as byteswap's low byte: the builtin's value is itself built by
+	// swapping the two stacked bytes, and the double swap folds to the single
+	// byte pick a hand assembler writes — `>> 8` leaves the swap materialized.
+	const std::uint16_t ra_words = reinterpret_cast<std::uint16_t>(__builtin_return_address(0));
+	const std::uint8_t ra_high = static_cast<std::uint8_t>(std::byteswap(ra_words));
+	const std::uint8_t slot_high = word_flash ? ra_high & 0xfe : static_cast<std::uint8_t>(ra_high << 1);
 
-	await_host();
+	// The knock: 'p' then 'b', each under a fresh window; any other byte is
+	// line noise and waits again. Falling out of a window runs the app.
+	while (rx_deadline() != 'p' || rx_deadline() != 'b') {
+	}
 
 	for (;;) {
-		// No prompt while an EEPROM write runs: it blocks SPM and fuse reads
+		// No prompt while an EEPROM write runs: a pending write blocks SPM
-		// (§26.2.1), and the prompt is the previous command's completion ack.
+		// and fuse reads (§26.2.1), and the ack tells the host all is done.
 		ee::wait();
-		tx_ack();
+		link::tx(ack);
 		const std::uint8_t command = link::rx();
 		switch (command) {
-		case 'b': // identity: the version, then the three signature bytes
+		case 'b': { // info block, read relative to the running slot
-			// Straight out of the stamp, so the wire and the image can never
+			// The block sits in the image's first 256 bytes (the build lint
-			// disagree about what this loader is. The indices are constant and
+			// asserts it), and slots are 512-aligned — so the low byte of its
-			// the array is constexpr, so these are immediates, not flash reads:
+			// link address (in wire units: bytes, or words on the large
-			// nothing here needs the stamp's runtime address.
+			// chips) is its offset in any slot, and the high byte of its
-			for (std::uint8_t at = stamp_identity; at != sizeof identity_stamp; ++at)
+			// runtime address is the running slot's. Composed from the two
-				link::tx(identity_stamp[at]);
+			// bytes — the high half is runtime data, so no absolute address
-			break;
+			// is ever materialized.
-		case 'J':   // jump:  sel8 (reserved), addr16 as a wire word address
+			const auto link_low = reinterpret_cast<std::uint16_t>(info_data.data());
-		case 'W':   // fill one flash page buffer: sel8, addr16, then page bytes
+			const std::uint8_t low =
-		case 'G':   // read:  sel8, addr16, n8 (0 = 256)
+			    word_flash ? static_cast<std::uint8_t>(link_low >> 1) : static_cast<std::uint8_t>(link_low);
-		case 'g': { // write: sel8, addr16, n8, then n bytes, each acked
+			send_flash(static_cast<std::uint16_t>(low | (slot_high << 8)), static_cast<std::uint8_t>(info_data.size()));
-			// One decode, one cursor and one loop for every space, both
-			// directions and the jump: a command per memory would carry a copy
-			// of all three each. 'J' — the hand-over and staging transfer —
-			// carries a selector it ignores so its address rides the same two
-			// reads as everything else; 'W' joins the same decode rather than
-			// keeping an address form of its own, so flash addressing is
-			// uniform across every command that names it.
-			const std::uint8_t selector = link::rx();
-			const std::uint8_t space = space_of(selector);
-			const std::uint8_t bank = bank_of(selector);
-			std::uint16_t at = rx16();
-			if (command == 'J') {
-				tx_ack();
-				link::drain();
-				jump(reinterpret_cast<void (*)()>(at));
-			}
-			if (command == 'W') {
-				fill_page(bank, at);
-				break;
-			}
-			std::uint8_t count = link::rx();
-			do {
-				// Read and write are one letter apart in case, so the direction
-				// is a single bit and the loop picks it with a one-word skip.
-				if (command & 0x20) {
-					store(space, bank, at, link::rx(), slot_high);
-					tx_ack();
-				} else
-					link::tx(load(space, bank, at));
-				++at;
-			} while (--count);
 			break;
 		}
+		case 'J': { // jump to a wire word address: hand-over and staging transfer
+			auto target = reinterpret_cast<void (*)()>(rx16());
+			link::tx(ack);
+			link::drain();
+			jump(target);
+		}
+		case 'R':   // read flash: addr16, n8 (0 = 256)
+		case 'r':   // read EEPROM: addr16, n8
+		case 'w': { // write EEPROM: addr16, n8, then n bytes each acked
+			std::uint16_t address = rx16();
+			std::uint8_t count = link::rx();
+			if (command == 'R')
+				send_flash(address, count);
+			else if (command == 'r')
+				send_eeprom(address, count);
+			else
+				store_eeprom(address, count);
+			break;
+		}
+		case 'W': // program one flash page: addr16, page bytes
+			program_flash(rx16(), slot_high);
+			break;
+		case 'F': // fuse and lock bytes
+			send_fuses();
+			break;
 		default: // unknown bytes are ignored; the loop re-acks
 			break;
 		}
@@ -611,7 +489,4 @@ void fill_page(std::uint8_t bank, std::uint16_t at)
 } // namespace
 } // namespace pureboot
 
-// stack::hardware: activation is reset-only, so the reset logic's own
+template struct avr::startup::entry<pureboot::run>;
-// SP = RAMEND stands wherever the datasheet guarantees it (the classic
-// megas still get the write); a 'J' entry runs on the caller's live stack.
-template struct avr::startup::entry<pureboot::run, avr::startup::stack::hardware>;

test/check_pi.py

@@ -1,75 +1,52 @@
 #!/usr/bin/env python3
-"""Position-independence lint: the property that lets the identical image run
+"""Position-independence lint for the pureboot image.
-from any slot, asserted from the built ELF and its object.
 
-  1. No absolute jmp/call — -mrelax normally guarantees it, but a branch that
+The self-staging design lets the identical binary run from any 512-byte
-     grows out of relaxation range would break it silently.
+slot, which holds only if nothing in the image addresses itself absolutely.
-  2. Nothing flash-resident to address: the image is .text alone, so there is
+Two link-time facts guarantee it, both asserted here from the built ELF:
-     no table whose runtime address has to be reconstructed.
-  3. The image is byte-identical when linked at a different base. This is
-     position independence itself rather than a proxy for it — an absolute
-     address anywhere in the image would move with the link and show up as a
-     differing byte.
 
-Usage: check_pi.py <objdump> <objcopy> <cxx> <mcu> <elf> <object> <text_start_hex>
+  1. No absolute jmp/call opcodes — all control flow is PC-relative
+     (rjmp/rcall/ijmp/icall). -mrelax normally guarantees this; a code
+     change that grows a branch out of relaxation range would break it
+     silently.
+  2. The info block sits within the image's first 256 bytes: the 'b'
+     command rebuilds its address as (running slot high byte : low byte of
+     the link address), which needs the offset to fit that low byte.
+
+Usage: check_pi.py <objdump> <nm> <elf> <text_start_hex>
 """
 
-import os
 import re
 import subprocess
 import sys
-import tempfile
-
-
-def fail(message):
-    print(f"FAIL: {message}")
-    sys.exit(1)
 
 
 def main():
-    objdump, objcopy, cxx, mcu, elf, obj, text_start = sys.argv[1:]
+    objdump, nm, elf, text_start = sys.argv[1:]
     text_start = int(text_start, 0)
 
     listing = subprocess.run([objdump, "-d", elf], capture_output=True, text=True, check=True).stdout
-    absolute = [line for line in listing.splitlines() if re.search(r"\t(jmp|call)\t", line)]
+    absolute = [
+        line
+        for line in listing.splitlines()
+        if re.search(r"\t(jmp|call)\t", line)
+    ]
     if absolute:
-        fail("absolute control flow in the image:\n" + "\n".join(absolute))
+        print("FAIL: absolute control flow in the image:")
+        print("\n".join(absolute))
+        sys.exit(1)
 
-    # Allocated flash beyond .text would be data the running copy has to find.
+    symbols = subprocess.run([nm, "-C", elf], capture_output=True, text=True, check=True).stdout
-    # Only ALLOC sections reach the device at all; .comment and the debug
+    info = [line for line in symbols.splitlines() if "info_data" in line]
-    # sections ride along in the ELF container and are never flashed. objdump
+    if len(info) != 1:
-    # prints each section's flags on the line following its header.
+        print(f"FAIL: expected one info-block storage symbol, found {len(info)}")
-    headers = subprocess.run([objdump, "-h", elf], capture_output=True, text=True, check=True).stdout.splitlines()
+        sys.exit(1)
-    for index, line in enumerate(headers):
+    offset = int(info[0].split()[0], 16) - text_start
-        fields = line.split()
+    if not 0 <= offset < 256:
-        if len(fields) < 6 or not fields[0].isdigit():
+        print(f"FAIL: info block at image offset {offset:#x}, must sit in the first 256 bytes")
-            continue
+        sys.exit(1)
-        name, size = fields[1], int(fields[2], 16)
-        flags = headers[index + 1] if index + 1 < len(headers) else ""
-        if "ALLOC" not in flags or not size:
-            continue
-        if name not in (".text", ".noinit", ".bss"):
-            fail(f"flash-resident section {name} ({size} bytes): the image must be .text alone")
 
-    # Relink at a different base and compare the bytes.
+    print(f"PI lint: control flow PC-relative, info block at offset {offset:#x}")
-    with tempfile.TemporaryDirectory() as work:
-        elsewhere = text_start - 0x200 if text_start >= 0x200 else text_start + 0x200
-        images = []
-        for base, tag in ((text_start, "here"), (elsewhere, "there")):
-            relinked = os.path.join(work, f"{tag}.elf")
-            binary = os.path.join(work, f"{tag}.bin")
-            subprocess.run(
-                [cxx, f"-mmcu={mcu}", "-nostartfiles", f"-Wl,--section-start=.text={base:#x}",
-                 "-Wl,--defsym=pureboot_app=0", "-mrelax", obj, "-o", relinked],
-                check=True, capture_output=True)
-            subprocess.run([objcopy, "-O", "binary", relinked, binary], check=True)
-            images.append(open(binary, "rb").read())
-        if images[0] != images[1]:
-            differing = [i for i, (a, b) in enumerate(zip(*images)) if a != b]
-            fail(f"the image changes when linked at {elsewhere:#x} instead of {text_start:#x}: "
-                 f"{len(differing)} byte(s) differ, first at offset {differing[0]:#x}")
-
-    print(f"PI lint: control flow PC-relative, .text only, identical linked at {text_start:#x} and {elsewhere:#x}")
 
 
 if __name__ == "__main__":

test/check_unit.cmake

@@ -1,36 +0,0 @@
-# Asserts the autobaud loader's measured unit sits where the host will read
-# it (--info's measured clock — the address is wire contract). Two homes: on
-# a chip with the GPIOR pair the unit lives there and the image must carry no
-# RAM word for it at all; elsewhere it is the first RAM object at SRAM start.
-# Run as
-#   cmake -DOBJDUMP=... -DELF=... -DRAM_START=<data address> [-DGPIOR=<data address>]
-#         -P check_unit.cmake
-
-execute_process(COMMAND ${OBJDUMP} -t ${ELF} OUTPUT_VARIABLE _syms RESULT_VARIABLE _res)
-if(NOT _res EQUAL 0)
-	message(FATAL_ERROR "${OBJDUMP} -t ${ELF} failed")
-endif()
-
-# The symbol line: "00800100 l     O .noinit  00000002 <mangled>unit_E".
-string(REGEX MATCH "\n0*([0-9a-f]+)[^\n]+[ \t][^ \t\n]*unit_E\n" _line "${_syms}")
-
-if(GPIOR)
-	if(_line)
-		message(FATAL_ERROR "unit_ RAM symbol present although the unit's home is GPIOR ${GPIOR} — "
-		                    "the host peeks the pair, and a RAM copy would be dead weight")
-	endif()
-	message(STATUS "no unit_ RAM object — the unit lives in the GPIOR pair at ${GPIOR}")
-	return()
-endif()
-
-if(NOT _line)
-	message(FATAL_ERROR "no unit_ symbol in ${ELF} — is this the autobaud loader?")
-endif()
-
-# AVR data-space symbols carry the 0x800000 VMA offset.
-math(EXPR _want "0x800000 + ${RAM_START}" OUTPUT_FORMAT HEXADECIMAL)
-math(EXPR _have "0x${CMAKE_MATCH_1}" OUTPUT_FORMAT HEXADECIMAL)
-if(NOT _have STREQUAL _want)
-	message(FATAL_ERROR "unit_ sits at ${_have}, ram_start is ${_want} — the host peeks ram_start")
-endif()
-message(STATUS "unit_ at ${_have} == ram_start")

test/device.c

@@ -7,95 +7,77 @@
 // SPM genuinely writes avr->flash on the mega cores, so on exit (or SIGTERM)
 // we dump the flash image to a file for a ground-truth cross-check against
 // what the client read back through the bootloader.
-#include <csignal>
+#include <signal.h>
-#include <cstdint>
+#include <stdint.h>
-#include <cstdio>
+#include <stdio.h>
-#include <cstdlib>
+#include <stdlib.h>
-#include <cstring>
+#include <string.h>
-#include <print>
-
 #include <unistd.h>
 
-// The parts headers (uart_pty.h) carry no C++ linkage guards of their own,
-// unlike simavr's core headers — the block covers both harmlessly.
-extern "C" {
 #include "avr_uart.h"
 #include "sim_avr.h"
 #include "sim_elf.h"
 #include "uart_pty.h"
-}
 
-namespace {
+static avr_t *avr;
+static uart_pty_t uart_pty;
+static const char *dump_path;
 
-avr_t *avr;
+static void finish(int sig)
-uart_pty_t uart_pty;
-const char *dump_path;
-
-[[noreturn]] void finish(int)
 {
+	(void)sig;
 	if (dump_path) {
-		std::FILE *f = std::fopen(dump_path, "wb");
+		FILE *f = fopen(dump_path, "wb");
 		if (f) {
-			std::fwrite(avr->flash, 1, avr->flashend + 1, f);
+			fwrite(avr->flash, 1, avr->flashend + 1, f);
-			std::fclose(f);
+			fclose(f);
 		}
 	}
 	uart_pty_stop(&uart_pty);
 	_exit(0);
 }
 
-} // namespace
-
 int main(int argc, char *argv[])
 {
 	if (argc < 3) {
-		std::println(stderr, "usage: {} <tsb.elf> <boot_base_hex> [flash_dump.bin]", argv[0]);
+		fprintf(stderr, "usage: %s <tsb.elf> <boot_base_hex> [flash_dump.bin]\n", argv[0]);
 		return 2;
 	}
-	auto boot_base = static_cast<std::uint32_t>(std::strtoul(argv[2], nullptr, 0));
+	uint32_t boot_base = (uint32_t)strtoul(argv[2], NULL, 0);
-	dump_path = argc >= 4 ? argv[3] : nullptr;
+	dump_path = argc >= 4 ? argv[3] : NULL;
 
 	avr = avr_make_mcu_by_name("atmega328p");
 	if (!avr) {
-		std::println(stderr, "device: no ATmega328P core");
+		fprintf(stderr, "device: no ATmega328P core\n");
 		return 1;
 	}
 	avr_init(avr);
 	avr->frequency = 16000000;
 	// Real flash powers up erased (0xff); the app region must look erased
 	// before the bootloader programs it.
-	std::memset(avr->flash, 0xff, avr->flashend + 1);
+	memset(avr->flash, 0xff, avr->flashend + 1);
 
 	// simavr's ELF loader flattens the flash base to 0 (it expects an app at
 	// 0x0), but it hands back the boot code in fw.flash; place it at the boot
 	// section base ourselves and enter there (BOOTRST is not modelled).
-	elf_firmware_t fw{};
+	elf_firmware_t fw = {0};
 	if (elf_read_firmware(argv[1], &fw) != 0) {
-		std::println(stderr, "device: cannot read {}", argv[1]);
+		fprintf(stderr, "device: cannot read %s\n", argv[1]);
 		return 1;
 	}
-	// An image that runs past flash end cannot execute on hardware, and a
+	memcpy(avr->flash + boot_base, fw.flash, fw.flashsize);
-	// naive copy of it would smash the heap beyond avr->flash — after which
-	// the simulation misbehaves in ways that point everywhere but here.
-	// Refuse it loudly instead.
-	if (boot_base + fw.flashsize > avr->flashend + 1) {
-		std::println(stderr, "device: {} B at {:#x} runs past flash end {:#x} — image does not fit its slot",
-		             fw.flashsize, boot_base, avr->flashend);
-		return 1;
-	}
-	std::memcpy(avr->flash + boot_base, fw.flash, fw.flashsize);
 	avr->pc = boot_base;
 	avr->codeend = avr->flashend;
 
 	// Optional: seed the config page (one page below the boot section) with a
 	// hex byte string, so the password gate and emergency erase can be tested.
 	// Layout: [appjump lo][appjump hi][timeout][password...][0xff].
-	const char *cfg = std::getenv("TSB_CONFIG");
+	const char *cfg = getenv("TSB_CONFIG");
 	if (cfg) {
-		std::uint32_t app_end = boot_base - 128; // config page sits directly below the boot code
+		uint32_t app_end = boot_base - 128; // config page sits directly below the boot code
 		for (int i = 0; cfg[i] && cfg[i + 1]; i += 2) {
 			char b[3] = {cfg[i], cfg[i + 1], 0};
-			avr->flash[app_end + i / 2] = static_cast<std::uint8_t>(std::strtoul(b, nullptr, 16));
+			avr->flash[app_end + i / 2] = (uint8_t)strtoul(b, NULL, 16);
 		}
 	}
 
@@ -104,18 +86,18 @@ int main(int argc, char *argv[])
 	// tight-polling loader (one that releases TX between bytes, as one-wire does)
 	// in real time, distorting protocol timing. Clear it so the loader runs at
 	// true cycle speed.
-	std::uint32_t uflags = 0;
+	uint32_t uflags = 0;
 	avr_ioctl(avr, AVR_IOCTL_UART_GET_FLAGS('0'), &uflags);
 	uflags &= ~AVR_UART_FLAG_POLL_SLEEP;
 	avr_ioctl(avr, AVR_IOCTL_UART_SET_FLAGS('0'), &uflags);
 
 	uart_pty_init(avr, &uart_pty);
 	uart_pty_connect(&uart_pty, '0');
-	std::println("TSB_PTY {}", uart_pty.pty.slavename);
+	printf("TSB_PTY %s\n", uart_pty.pty.slavename);
-	std::fflush(stdout);
+	fflush(stdout);
 
-	std::signal(SIGTERM, finish);
+	signal(SIGTERM, finish);
-	std::signal(SIGINT, finish);
+	signal(SIGINT, finish);
 
 	for (;;) {
 		int state = avr_run(avr);
@@ -123,4 +105,5 @@ int main(int argc, char *argv[])
 			break;
 	}
 	finish(0);
+	return 0;
 }

test/pbapp.cpp

@@ -11,10 +11,6 @@
 // reset reaches those loaders through the patched vector (or the runner
 // models BOOTRST), so the application owes them nothing.
 //
-// PUREBOOT_HANDOVER drops the listening and jumps straight in, leaving the
-// USART enabled behind it — the hand-over state a loader bit-banging on that
-// USART's own pins has to survive.
-//
 // The fixture speaks the deployment its loader was built for: the same
 // PUREBOOT_* defines configure it, and without them it assumes the stock
 // deployment (the crystal/RC clock table below, the chip's natural link).
@@ -41,9 +37,6 @@ consteval avr::hertz_t clock()
 #if !defined(PUREBOOT_TX)
 #define PUREBOOT_TX pb1
 #endif
-#if !defined(PUREBOOT_RX)
-#define PUREBOOT_RX pb0
-#endif
 #if !defined(PUREBOOT_USART)
 #define PUREBOOT_USART 0
 #endif
@@ -53,7 +46,7 @@ consteval bool use_hardware()
 #if defined(PUREBOOT_SOFT_SERIAL)
 	return false;
 #else
-	return avr::uart::has_usart<0>();
+	return avr::hw::db.has_instance("USART0") || avr::hw::db.has_instance("USART");
 #endif
 }
 
@@ -66,47 +59,19 @@ struct link {
 #else
 	static constexpr avr::baud_t baud{115200};
 #endif
-	using tx_t = avr::uart::usart<PUREBOOT_USART, C, {.baud = baud, .max_baud_error = 2.5_pct}>;
+	using tx_t = avr::uart::usart<'0' + PUREBOOT_USART, C, {.baud = baud, .max_baud_error = 2.5_pct}>;
-	static void init()
-	{
-		avr::init<tx_t>();
-	}
 	static void tx(char c)
 	{
 		tx_t::write(static_cast<std::uint8_t>(c));
 	}
-	// The loader sits in the top slot — 512 bytes on every chip. The jump
-	// takes a word address, which is what makes the >64 KiB chips' entry
-	// reachable through a 16-bit pointer at all.
-	static void enter_loader()
-	{
-		constexpr std::uint32_t slot = 512;
-		reinterpret_cast<void (*)()>(static_cast<std::uint16_t>((avr::hw::db.mem.flash_size - slot) / 2))();
-	}
-
 	[[noreturn]] static void idle()
 	{
-#if defined(PUREBOOT_HANDOVER)
+		// 'L' hands back to the loader at the top slot — 512 bytes, or the
-		// Hand back at once, with this USART still enabled — the state that
+		// 1 KiB the >64 KiB chips use.
-		// leaves a bit-banged loader on its pins mute unless the loader
+		constexpr std::uint32_t slot = avr::hw::db.mem.flash_size > 65536 ? 1024 : 512;
-		// releases it. Unconditional because there is no command wire to
+		for (;;)
-		// wait on: that loader's link is the pins, not this peripheral.
+			if (tx_t::read_blocking() == 'L')
-		enter_loader();
+				reinterpret_cast<void (*)()>(static_cast<std::uint16_t>((avr::hw::db.mem.flash_size - slot) / 2))();
-		__builtin_unreachable();
-#else
-		for (;;) {
-			auto command = tx_t::read_blocking();
-			if (command == 'L')
-				enter_loader();
-			// 'D' leaves every word of the SPM page buffer dirty, so that a
-			// following 'L' enters the loader with the buffer it never clears.
-			if (command == 'D') {
-				for (std::uint16_t at = 0; at < avr::spm::page_bytes; at += 2)
-					avr::spm::fill(at, 0xdead);
-				tx('D');
-			}
-		}
-#endif
 	}
 };
 
@@ -117,48 +82,15 @@ struct link<C, false> {
 #else
 	static constexpr avr::baud_t baud{57600};
 #endif
-	// A shared-pin deployment (RX == TX) banners as a guest on its own line:
+	using tx_t = avr::uart::software_tx<C, avr::PUREBOOT_TX, baud>;
-	// the pull-up input is the released line, the transmitter takes the pin
-	// for exactly one frame per byte — the shape a real one-wire application
-	// beside this loader uses.
-	static constexpr bool one_wire = avr::PUREBOOT_RX == avr::PUREBOOT_TX;
-	using tx_t = avr::uart::software_tx<C, avr::PUREBOOT_TX, baud, one_wire>;
-	static void init()
-	{
-		// The guest transmitter configures no pin; the released line — the
-		// pull-up input a receiver would own — is established here.
-		if constexpr (one_wire)
-			avr::init<avr::io::input<avr::PUREBOOT_TX, avr::io::pull::up>, tx_t>();
-		else
-			avr::init<tx_t>();
-	}
 	static void tx(char c)
 	{
 		tx_t::write(static_cast<std::uint8_t>(c));
 	}
 	[[noreturn]] static void idle()
 	{
-#if defined(PUREBOOT_HEARTBEAT)
-		// Repeat the banner forever, which turns the fixture into a fixed
-		// cycles-per-bit transmitter: `tools/pbrig.py rate` sweeps the host rate
-		// against it to find the part's true bit rate, and from that the clock
-		// its RC oscillator is really running at. Only the *bit* timing carries
-		// the measurement — the delay merely spaces the lines out, so its own
-		// error does not matter. Software link only: the hardware-link idle owes
-		// the self-update tests a command loop, and a crystal deployment has
-		// nothing to measure.
-		while (true) {
-			tx('A');
-			tx('P');
-			tx('P');
-			tx('\r');
-			tx('\n');
-			dev::delay<50_ms>();
-		}
-#else
 		while (true) {
 		}
-#endif
 	}
 };
 
@@ -166,14 +98,9 @@ struct link<C, false> {
 
 int main()
 {
-	link<dev::clock>::init();
+	avr::init<typename link<dev::clock>::tx_t>();
-#if !defined(PUREBOOT_HANDOVER)
 	link<dev::clock>::tx('A');
 	link<dev::clock>::tx('P');
 	link<dev::clock>::tx('P');
-#endif
-	// The hand-over fixture stays silent: nothing is listening on the USART it
-	// brings up — the loader it hands to speaks those pins directly — so its
-	// banner would be a write into a peer that does not exist.
 	link<dev::clock>::idle();
 }

test/pbautobaud.py

@@ -1,211 +0,0 @@
-#!/usr/bin/env python3
-"""End-to-end autobaud test: drive an autobaud loader in simavr through the
-calibration handshake and a flash + EEPROM + fuse round-trip, cross-checked
-against the simulator's ground-truth memory — then repeat at a second F_CPU with
-the *same* loader binary, which is the property autobaud exists for: one
-clock-agnostic image that locks onto whatever rate the host sends.
-
-Usage: pbautobaud.py <device_bin> <loader_elf> <mcu> <base_hex> <page>
-                     <app_bin> <app_hz> <app_baud> <tool_py> <workdir> [link]
-
-The loader is a software-serial build, driven over the GPIO⇄pty bridge; the
-optional link overrides the default -l sw:B0,B1 — RX == TX in it is the
-one-wire deployment, and every session then runs with the host's echo
-discard on. The app fixture is built for (app_hz, app_baud); the hand-over
-is checked at that point, and a second point at half the clock proves the
-lock is measured, not baked in.
-"""
-
-import os
-import re
-import sys
-import time
-
-
-def fail(message):
-    print(f"FAIL: {message}")
-    sys.exit(1)
-
-
-def main():
-    args = sys.argv[1:]
-    link = args.pop() if len(args) == 11 else "sw:B0,B1"
-    (device_bin, elf, mcu, base_hex, page, app_bin, app_hz, app_baud, tool, workdir) = args
-    base, page, app_hz, app_baud = int(base_hex, 0), int(page), int(app_hz), int(app_baud)
-    one_wire = re.fullmatch(r"sw:([A-H][0-7]),\1(@[01])?", link) is not None
-    extra = ("--one-wire",) if one_wire else ()
-    sys.path.insert(0, os.path.dirname(os.path.abspath(tool)))
-    sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
-    import pbsim
-    import pureboot as pb
-
-    os.makedirs(workdir, exist_ok=True)
-    ee_image = bytes(range(0xA0, 0xB0))
-    ee_path = os.path.join(workdir, "ee.bin")
-    open(ee_path, "wb").write(ee_image)
-
-    # The geometry the surgery planner needs, from the chip class the runner is
-    # told — the same derivation pbtest.py makes: the boot-sectioned megas need
-    # no vector surgery, the tinies and the boot-section-less m48s do, and the
-    # large chips speak word addresses.
-    mega = mcu.startswith("atmega")
-    patch = not mega or mcu.startswith("atmega48")
-    word_flash = base + pb.SLOT > 0x10000
-    wire_base = base // 2 if word_flash else base
-    flags = (1 if patch else 0) | (2 if word_flash else 0)
-    ground_truth = pb.Info(bytes([ord("P"), ord("B"), pb.NEWEST_LOADER, 0, 0, 0, page & 0xFF,
-                                  wire_base & 0xFF, wire_base >> 8, 0, 0, flags]))
-
-    def round_trip(hz, baud, label, hand_over):
-        """One clock point: reset, calibrate + knock, program, verify against the
-        simulator's own flash, and (at the app's point) hand over to the fixture."""
-        dump = os.path.join(workdir, f"flash_{label}.bin")
-        device = pbsim.Device(device_bin, elf, mcu, str(hz), base_hex, page, baud, dump, link=link)
-        try:
-            # The host tool, in autobaud mode, sends the 0xC0 calibration pulse
-            # and a single knock at `baud`; the loader locks to it.
-            out = pbsim.run_tool(tool, device.pty, baud, *extra, "--autobaud", "--info", "--clock", str(hz),
-                                 "--fuses", "--flash", app_bin, "--eeprom", ee_path, "--stay")
-            for needed in ("version", "signature", "fuses", "verify:", "stays"):
-                if needed not in out:
-                    fail(f"{label}: session output lacks {needed!r}\n{out}")
-            # The measured clock, decoded from the unit at whichever home this
-            # version keeps it in. The runner's clock is exact, so the figure
-            # must land inside the
-            # encoding's own envelope: the loader floors the bit period to
-            # 4-cycle spin granules after an 8-cycle discount, and the edge
-            # poll can shave a few cycles more — one granule of slack below
-            # the true clock, none above (in cycles per bit, times the rate).
-            measured = re.search(r"measured\s+(\d+) Hz", out)
-            if not measured:
-                fail(f"{label}: --info lacks the measured clock\n{out}")
-            measured = int(measured.group(1))
-            if not hz - 19 * baud <= measured <= hz + 4 * baud:
-                fail(f"{label}: measured clock {measured} Hz is {measured - hz:+d} off the true {hz}")
-            # Read both memories back over the locked link and check them.
-            read_flash = os.path.join(workdir, f"rf_{label}.bin")
-            read_eeprom = os.path.join(workdir, f"re_{label}.bin")
-            out = pbsim.run_tool(tool, device.pty, baud, *extra, "--autobaud", "--verify-flash", app_bin,
-                                 "--verify-eeprom", ee_path, "--read-flash", read_flash,
-                                 "--read-eeprom", read_eeprom, "--stay")
-            if out.count("verify:") != 2:
-                fail(f"{label}: did not verify both memories\n{out}")
-            if open(read_eeprom, "rb").read()[: len(ee_image)] != ee_image:
-                fail(f"{label}: EEPROM read-back mismatch")
-
-            if hand_over:
-                # Regression: a calibration pulse with no knock behind it must
-                # not wedge the loader. The knock's edge wait used to be
-                # unbudgeted, so one stray low pulse — EMI, or a host that opens
-                # the port and never knocks — held the loader forever and the
-                # application never ran. The whole activation is bounded now, so
-                # the window closes and the app boots; the banner is the proof.
-                # (The pause lets the loader reach its measurement loop, so the
-                # pulse is genuinely seen and the test cannot pass vacuously.)
-                device.reset()
-                port = pb.Port(device.pty, baud)
-                if one_wire:
-                    port = pb.OneWirePort(port)
-                try:
-                    time.sleep(0.2)
-                    port.write(bytes((pb.CALIBRATE,)))
-                    # Accumulate rather than match exactly: the reset leaves the
-                    # idle line a framing artefact ahead of the banner, which is
-                    # noise here — the question is only whether the app ran.
-                    seen = b""
-                    deadline = time.monotonic() + 180.0
-                    while b"APP" not in seen and time.monotonic() < deadline:
-                        seen += port.read_available(1.0)
-                    if b"APP" not in seen:
-                        fail(f"{label}: lone calibration pulse wedged the loader — app never bannered, saw {seen!r}")
-                    print(f"  {label}: lone calibration pulse does not wedge the loader")
-                finally:
-                    port.close()
-
-                device.reset()
-                port = pb.Port(device.pty, baud)
-                if one_wire:
-                    port = pb.OneWirePort(port)
-                try:
-                    loader = pb.Loader(port)
-                    live = loader.connect_autobaud(15)
-                    if not pb.OLDEST_LOADER <= live.version <= pb.NEWEST_LOADER:
-                        fail(f"{label}: loader reports pureboot {live.version}")
-                    if loader.unified:
-                        # pureboot 5's data space. 0x0200 is clear of the
-                        # loader's own .noinit unit at the bottom of SRAM and of
-                        # the stack at the top. Reading it back over the same
-                        # locked link proves both directions of the new space.
-                        probe = bytes(range(0x30, 0x40))
-                        loader.write_ram(0x0200, probe)
-                        if loader.read_ram(0x0200, len(probe)) != probe:
-                            fail(f"{label}: RAM round-trip mismatch")
-                        # The register file and the I/O space share the data
-                        # address space on AVR, so the same command reaches a
-                        # peripheral register. SPMCSR reads back as idle here.
-                        verbose_ram = loader.read_ram(0x0200, 4)
-                        print(f"  {label}: RAM read/write ok ({verbose_ram.hex()})")
-                    loader.run_application()
-                    banner = port.read_exact(3, 5.0)
-                    if banner != b"APP":
-                        fail(f"{label}: application banner was {banner!r}")
-                finally:
-                    port.close()
-        finally:
-            device.stop()
-
-        # Ground truth (read after the runner exits and writes its dump): what
-        # the tool programmed must be what the simulator actually holds.
-        pages = pb.plan_flash(open(app_bin, "rb").read(), ground_truth)
-        flash_true = open(dump, "rb").read()
-        for address, data in pages.items():
-            if flash_true[address : address + page] != data:
-                fail(f"{label}: simulator flash differs from the programmed image at {address:#06x}")
-        print(f"  {label}: locked at {hz} Hz / {baud} Bd, flash+EEPROM verified"
-              + (", hand-over ok" if hand_over else ""))
-
-    def must_lock(hz, baud, label):
-        """The calibration alone, at a tight bit period. Nothing is programmed —
-        the question is only whether the loader can still measure the pulse."""
-        dump = os.path.join(workdir, f"flash_{label}.bin")
-        device = pbsim.Device(device_bin, elf, mcu, str(hz), base_hex, page, baud, dump,
-                              link=link)
-        try:
-            port = pb.Port(device.pty, baud)
-            if one_wire:
-                port = pb.OneWirePort(port)
-            try:
-                live = pb.Loader(port).connect_autobaud(15)
-                if live.version != pb.NEWEST_LOADER:
-                    fail(f"{label}: loader reports pureboot {live.version}")
-            finally:
-                port.close()
-        finally:
-            device.stop()
-        print(f"  {label}: locked at {hz} Hz / {baud} Bd ({hz / baud:.0f} cycles a bit)")
-
-    # The app fixture is built for one clock; the hand-over banners there. A
-    # second point at double that clock, same loader binary, proves the lock is
-    # measured, not baked in — the whole point of autobaud. (Doubling keeps the
-    # bit period healthy; halving would drop it below the software UART's floor.)
-    round_trip(app_hz, app_baud, "clock-a", hand_over=True)
-    round_trip(app_hz * 2, app_baud, "clock-b", hand_over=False)
-
-    # Both points above sit near 100 cycles a bit, which is comfortable. The
-    # calibration's real floor is far tighter, and it is worth a gate: measured
-    # here, the lock is solid down to ~36 cycles a bit and fails outright by ~31
-    # — a sharp edge, not a fraying one. This pins the tightest standard rate the
-    # fixture's clock reaches, so a change that raises the floor is caught.
-    #
-    # It does *not* bound what a real deployment can use. On silicon the
-    # oscillator's own jitter costs roughly a factor of two: an ATtiny13A on its
-    # factory RC trim was reliable at ~118 cycles a bit and already locking only
-    # 1 attempt in 5 by ~59, which no exact-clock simulation can show. The
-    # deployable envelope is a README matter; this is the logic's floor.
-    must_lock(app_hz, app_baud * 2, "tight-bit")
-    print("pbautobaud: calibration lock and flash/EEPROM/fuse round-trip pass at both clocks, "
-          "and the tight bit period still locks")
-
-
-if __name__ == "__main__":
-    main()

test/pbdirty.py

@@ -1,88 +0,0 @@
-#!/usr/bin/env python3
-"""Dirty-page-buffer acceptance test: with no discard in the loader, a page
-filled over words an earlier writer left takes those instead. The whole
-contract is asserted — a bare verify sees the corruption, the repairing
-verify fixes it in one rewrite, and it stays fixed.
-
-The state is reached the one way the loader cannot prevent: an application
-dirties the buffer and jumps in with no reset between. Boot-sectioned megas
-forbid that outright (SPM runs only from the boot section, Atmel-8271 §26.2),
-but simavr dispatches SPM from anywhere, which is what makes it constructible.
-
-Usage: pbdirty.py <device_bin> <pureboot_elf> <mcu> <hz> <base_hex> <page>
-                  <baud> <app_bin> <tool_py> <workdir>
-"""
-
-import os
-import sys
-
-
-def fail(message):
-    print(f"FAIL: {message}")
-    sys.exit(1)
-
-
-def main():
-    device_bin, elf, mcu, hz, base_hex, page, baud, app_bin, tool, workdir = sys.argv[1:]
-    page, baud = int(page), int(baud)
-    sys.path.insert(0, os.path.dirname(os.path.abspath(tool)))
-    sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
-    import pbsim
-    import pureboot as pb
-
-    os.makedirs(workdir, exist_ok=True)
-    dump = os.path.join(workdir, "dump.bin")
-
-    # Reset boots the application on a BOOTRST-unprogrammed mega; its 'L' is
-    # the loader entry this test needs, reached without a reset.
-    device = pbsim.Device(device_bin, elf, mcu, hz, base_hex, page, baud, dump, reset_hex="0")
-    try:
-        port = pb.Port(device.pty, baud)
-        loader = pb.Loader(port)
-        loader.connect(25)
-
-        # Install the application and hand over to it.
-        pb.op_flash(loader, app_bin, erase=False, verify=True)
-        loader.run_application()
-        if port.read_exact(3, 5.0) != b"APP":
-            fail("the application did not start")
-
-        port.write(b"D")
-        if port.read_exact(1, 5.0) != b"D":
-            fail("the application did not acknowledge dirtying the page buffer")
-        port.write(b"L")
-        loader = pb.Loader(port)
-        loader.connect(25)
-
-        # Program by hand, so the corruption is observable before anything
-        # repairs it.
-        pages = pb.plan_flash(open(app_bin, "rb").read(), loader.info)
-        for address in sorted(pages):
-            loader.write_page(address, pages[address])
-        try:
-            pb.verify_pages(loader, pages)
-        except pb.Error as error:
-            if "verify failed" not in str(error):
-                fail(f"the read-back failed, but not at verify: {error}")
-        else:
-            # Either the fixture no longer dirties the buffer, or the loader
-            # clears it again — in which case this test's premise is gone.
-            fail("programming over a dirty page buffer came back clean")
-
-        # What the programming path uses: one rewrite settles it, and it stays
-        # settled.
-        pb.verify_pages(loader, pages, repair=True)
-        pb.verify_pages(loader, pages)
-
-        # Ground truth beyond the loader's own read-back.
-        loader.run_application()
-        if port.read_exact(3, 5.0) != b"APP":
-            fail("the application did not start after the recovered write")
-        port.close()
-    finally:
-        device.stop()
-    print("pbdirty: a dirty page buffer is caught by verify and cleared by the retry")
-
-
-if __name__ == "__main__":
-    main()

test/pbmute.py

@@ -1,79 +0,0 @@
-#!/usr/bin/env python3
-"""Hand-over with a USART left enabled on the loader's own pins.
-
-A software or autobaud link deployed on a USART's TxD is mute if an
-application hands over with that USART still enabled: TXEN keeps the USART
-owning the pin, so the bit-banged transmitter's port writes go nowhere and the
-loader receives and obeys while answering nothing. The link's init releases it.
-
-The state is reached the way silicon reaches it — an application that sets up
-its USART and jumps in with no reset between, so nothing clears UCSRnB for it.
-The pin ownership itself is modelled by the device runner: simavr wires a
-USART through IRQs alone and never takes the pin from the port, so without
-that the mute could not happen here at all (test/pureboot_device.cpp).
-
-Usage: pbmute.py <device_bin> <pureboot_elf> <mcu> <hz> <base_hex> <page>
-                 <baud> <app_bin> <tool_py> <workdir> <link>
-"""
-
-import os
-import re
-import sys
-
-
-def fail(message):
-    print(f"FAIL: {message}")
-    sys.exit(1)
-
-
-def main():
-    device_bin, elf, mcu, hz, base_hex, page, baud, app_bin, tool, workdir, link = sys.argv[1:]
-    page, baud = int(page), int(baud)
-    sys.path.insert(0, os.path.dirname(os.path.abspath(tool)))
-    sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
-    import pbsim
-    import pureboot as pb
-
-    if "@" not in link:
-        fail(f"the link {link} names no owning USART — nothing would be under test")
-    # A shared line (RX == TX) echoes the host's own bytes; discard them the
-    # way the shipped --one-wire mode does.
-    one_wire = re.fullmatch(r"sw:([A-H][0-7]),\1@[01]", link) is not None
-
-    os.makedirs(workdir, exist_ok=True)
-    dump = os.path.join(workdir, "dump.bin")
-
-    device = pbsim.Device(device_bin, elf, mcu, hz, base_hex, page, baud, dump, link=link)
-    try:
-        port = pb.Port(device.pty, baud)
-        if one_wire:
-            port = pb.OneWirePort(port)
-        loader = pb.Loader(port)
-        loader.connect(25)
-        resident = loader.info.version
-
-        # Install the fixture and let it take over. It brings up the USART
-        # that owns these pins and jumps straight back in.
-        pb.op_flash(loader, app_bin, erase=False, verify=True)
-        loader.run_application()
-
-        # The loader is running again with that USART enabled behind it. Only
-        # the release makes it audible; without it the connect times out.
-        loader = pb.Loader(port)
-        try:
-            loader.connect(25)
-        except pb.Error as error:
-            fail(f"the loader never answered after the hand-over — the USART still owns its TX pin ({error})")
-        if loader.info.version != resident:
-            fail(f"identity changed across the hand-over: {resident} then {loader.info.version}")
-
-        # Answering is not enough: it has to still be a working loader.
-        pb.verify_pages(loader, pb.plan_flash(open(app_bin, "rb").read(), loader.info))
-        port.close()
-    finally:
-        device.stop()
-    print("pbmute: a loader on a USART's own pins answers after a hand-over that left it enabled")
-
-
-if __name__ == "__main__":
-    main()

test/pbosccal.py

@@ -1,46 +0,0 @@
-#!/usr/bin/env python3
-"""The build-time OSCCAL trim, observed through the wire: a loader built with
-the OSCCAL axis holds the trim register at the built byte from its first
-prompt on — the write sits at the top of run(), ahead of the WDRF bail, so
-every path out of reset runs on the corrected clock. simavr's clock does not
-follow OSCCAL, which is what makes the value assertable at all: the register
-is plain state there, and the peek must return exactly what the build
-declared rather than whatever the oscillator needed.
-
-Usage: pbosccal.py <device_bin> <pureboot_elf> <mcu> <hz> <base_hex> <page>
-                   <baud> <osccal_addr> <osccal_value> <tool_py> <workdir>
-                   [link]
-"""
-
-import os
-import sys
-
-
-def fail(message):
-    print(f"FAIL: {message}")
-    sys.exit(1)
-
-
-def main():
-    args = sys.argv[1:]
-    link = args.pop() if len(args) == 12 else None
-    (device_bin, elf, mcu, hz, base_hex, page, baud, addr, value, tool, workdir) = args
-    addr, value, baud = int(addr, 0), int(value, 0), int(baud)
-    sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
-    import pbsim
-
-    os.makedirs(workdir, exist_ok=True)
-    dump = os.path.join(workdir, "flash_dump.bin")
-    device = pbsim.Device(device_bin, elf, mcu, hz, base_hex, page, baud, dump, link=link)
-    try:
-        out = pbsim.run_tool(tool, device.pty, baud, "--peek", f"{addr:#x}:1")
-        want = f"{addr:#06x}  {value:02x}"
-        if want not in out:
-            fail(f"OSCCAL at {addr:#x} did not read back {value:#04x}:\n{out}")
-    finally:
-        device.stop()
-    print("OK")
-
-
-if __name__ == "__main__":
-    main()

test/pbrehome.py

@@ -1,13 +1,19 @@
 #!/usr/bin/env python3
-"""Re-homing acceptance test: an image programmed somewhere other than its
+"""Re-homing acceptance test: a pureboot image programmed somewhere other
-canonical slot must still be a working loader, and the ordinary
+than its canonical top slot must still be a working loader —
+position-independent, guarding its accidental slot — and the ordinary
 --update-loader flow must put a build into the top slot from there.
 
-Two positions. Address 0, a raw .bin handed to a programmer: the staging
+Two positions are exercised. Address 0 (a raw .bin handed to a programmer,
-install and the word-0 redirect run from copies outside page 0's slot, so the
+which defaults to offset 0): the staging install and the word-0 redirect
-running-slot guard never blocks them. And the staging slot itself, where a
+both run from copies whose slots are not page 0's, so the running-slot
-loader already sitting there IS the staging copy — recognized by its embedded
+guard never blocks the flow. The staging slot itself: a loader already
-block and left in place, then streaming the new resident like any staged copy.
+sitting there IS the installed staging copy — the tool recognizes it by
+its embedded info block and leaves it in place instead of tripping the
+copy's own guard on the composed through-word — and that (older) copy
+streams the new resident like any staged copy. In both cases flashing an
+application through the healed resident overwrites the stale copy, vector
+surgery included, and the banner proves the launch.
 
 Usage: pbrehome.py <device_bin> <pureboot_elf> <update_bin> <mcu> <hz>
                    <base_hex> <page> <baud> <app_bin> <tool_py> <workdir>
@@ -84,7 +90,7 @@ def main():
     # The staging slot: erased flash with the loader sitting exactly where
     # a staging copy would — the tool must leave it in place and let it
     # stream the (different) update build into the resident slot.
-    stage = base - pb.SLOT
+    stage = base - 512
     rehome_from(pbsim, pb, device_bin, elf, hex(stage), hex(stage), update_bin, base, page, baud, app_bin, workdir,
                 mcu, hz)
     print("re-home from the staging slot: converged")

test/pbreloc.py

@@ -1,9 +1,11 @@
 #!/usr/bin/env python3
-"""Position-independence acceptance test: the identical binary, flashed one
+"""Position-independence acceptance test: the identical pureboot binary,
-slot below the resident, must serve the complete command set from there. The
+flashed one slot below the resident loader, must serve the complete command
-info block must come back byte-identical, the write guard must refuse the
+set from there. The resident installs it (through-word composed by the host
-staged copy's own slot and permit the resident's, and the staged copy must be
+layer), 'J' transfers control, and every command is exercised against the
-able to rewrite the resident verbatim.
+staged copy — the info block must come back byte-identical, the write guard
+must protect the staged copy's own slot and permit the resident's, and the
+staged copy must be able to rewrite the resident slot verbatim.
 
 Usage: pbreloc.py <device_bin> <pureboot_elf> <mcu> <hz> <base_hex> <page>
                   <baud> <tool_py> <workdir>
@@ -64,10 +66,8 @@ def main():
         if loader.read_eeprom(0, len(pattern)) != pattern:
             fail("EEPROM round-trip through the staged copy")
 
-        # The guard, both ways: its own slot refused (drained, unchanged), the
+        # The guard, both ways: its own slot refused (drained, unchanged),
-        # resident slot writable. The refusal leaves its drained words in the
+        # the resident slot writable.
-        # SPM buffer, so the write that follows may take them — and clears
-        # them by writing, so the retry must not.
         before = loader.read_flash(stage, page)
         loader.write_page(stage, bytes(page))
         if loader.read_flash(stage, page) != before:
@@ -75,13 +75,11 @@ def main():
         marker = bytes((i * 3) & 0xFF for i in range(page))
         loader.write_page(base, marker)
         if loader.read_flash(base, page) != marker:
-            loader.write_page(base, marker)
+            fail("the staged copy could not write the resident slot")
-            if loader.read_flash(base, page) != marker:
-                fail("the staged copy could not write the resident slot, even on retry")
 
         # Restore the resident image through the staged copy, then 'J' back
         # into it and prove it lives.
-        resident = image + b"\xff" * (pb.SLOT - len(image))
+        resident = image + b"\xff" * (info.slot - len(image))
         pb.write_differing(loader, base, resident)
         back_info = loader.enter_copy(base, 25)
         if back_info.raw != resident_info:

test/pbsim.py

@@ -8,13 +8,10 @@ import subprocess
 
 
 class Device:
-    def __init__(self, binary, elf, mcu, hz, base_hex, page, baud, dump, reset_hex=None, resume=None, link=None,
+    def __init__(self, binary, elf, mcu, hz, base_hex, page, baud, dump, reset_hex=None, resume=None, link=None):
-                 window=False):
         cmd = [binary]
         if link:
             cmd += ["-l", link]
-        if window:
-            cmd.append("-w")  # report the first-transmit cycle, free-run idle
         cmd += [elf, mcu, hz, base_hex, str(page), str(baud), dump]
         if reset_hex is not None or resume is not None:
             # Chips without a hardware boot section — the tinies and the

test/pbtest.py

@@ -1,17 +1,18 @@
 #!/usr/bin/env python3
-"""End-to-end protocol test: drive the simavr device with the real host tool
+"""End-to-end pureboot protocol test: spawn the simavr device, then drive it
-over its pty through flash, EEPROM, fuse and hand-over scenarios, and
+with the real host tool (pureboot.py, as a subprocess over the device's pty)
-cross-check the tool's view against the simulator's ground-truth dumps.
+through flash + EEPROM + fuse + hand-over scenarios, and cross-check
+the tool's view against the simulator's ground-truth memory dumps.
 
 Usage: pbtest.py <device_bin> <pureboot_elf> <mcu> <hz> <base_hex> <page>
                  <baud> <eeprom_size> <app_bin> <tool_py> <workdir> [link]
 
-The optional link is the runner's -l spec (usart1, sw:B5,B1, ...), for a
+The optional link is the runner's -l spec (usart1, sw:B5,B1, ...) for a
 loader built off the chip's natural serial default.
+Exits 0 if every scenario passes.
 """
 
 import os
-import re
 import sys
 
 
@@ -56,49 +57,30 @@ def main():
     # the page byte is the wire's 0-means-256.
     mega = mcu.startswith("atmega")
     patch = not mega or mcu.startswith("atmega48")
-    # Where SRAM begins: the x8 and x4 megas push it past their extended I/O
+    word_flash = base + 512 > 0x10000
-    # space, everything else starts right after the plain I/O registers. The
-    # loader keeps no statics and its stack sits at RAMEND, so the first SRAM
-    # byte is free for the data-space probe below.
-    classic = mcu in ("atmega8", "atmega8a", "atmega16", "atmega16a", "atmega32", "atmega32a")
-    ram_base = 0x0100 if mega and not classic else 0x0060
-    word_flash = base + pb.SLOT > 0x10000
     wire_base = base // 2 if word_flash else base
     flags = (1 if patch else 0) | (2 if word_flash else 0)
     info = pb.Info(
-        bytes([ord("P"), ord("B"), pb.NEWEST_LOADER, 0, 0, 0, page & 0xFF])
+        bytes([ord("P"), ord("B"), 1, 0, 0, 0, page & 0xFF])
         + bytes([wire_base & 0xFF, wire_base >> 8, eeprom_size & 0xFF, eeprom_size >> 8])
         + bytes([flags])
     )
 
-    # A shared-line link (RX == TX in the -l spec) makes the host read every
-    # byte it sends back off the line; all sessions then discard the echo.
-    one_wire = bool(link) and re.fullmatch(r"sw:([A-H][0-7]),\1(@[01])?", link) is not None
-    extra = ("--one-wire",) if one_wire else ()
-
     device = pbsim.Device(device_bin, elf, mcu, hz, base_hex, page, baud, dump, link=link)
     try:
         # Session 1: knock from reset, identify, program everything, stay.
-        out = pbsim.run_tool(tool, device.pty, baud, *extra, "--info", "--fuses", "--flash", app_bin,
+        out = pbsim.run_tool(tool, device.pty, baud, "--info", "--fuses", "--flash", app_bin,
                              "--eeprom", ee_path, "--stay")
-        for needed in ("version", "signature", "fuses", "verify:", "stays"):
+        for needed in ("signature", "fuses", "verify:", "stays"):
             if needed not in out:
                 fail(f"session 1 output lacks {needed!r}")
 
-        # Session 2: reconnect into the live session, verify, dump, exercise
+        # Session 2: reconnect into the live session, verify, dump, hand over
-        # the data space; hand over is deferred — the pty must be reopened for
+        # is deferred — the pty must be reopened for the APP banner first.
-        # the APP banner first.
+        out = pbsim.run_tool(tool, device.pty, baud, "--verify-flash", app_bin, "--verify-eeprom", ee_path,
-        probe = "c0ffee"
+                             "--read-flash", read_flash, "--read-eeprom", read_eeprom, "--stay")
-        out = pbsim.run_tool(tool, device.pty, baud, *extra, "--verify-flash", app_bin, "--verify-eeprom", ee_path,
-                             "--read-flash", read_flash, "--read-eeprom", read_eeprom,
-                             "--poke", f"{ram_base:#x}:{probe}", "--peek", f"{ram_base:#x}:3", "--stay")
         if out.count("verify:") != 2:
             fail("session 2 did not verify both memories")
-        # What went into SRAM must come back out of it: the data space is one
-        # more selector on the same transfer as flash and EEPROM, so a wrong
-        # selector decode would show up here and nowhere else.
-        if probe not in out.replace(" ", ""):
-            fail(f"data-space round trip at {ram_base:#x} did not read back {probe}\n{out}")
 
         eeprom_back = open(read_eeprom, "rb").read()
         if eeprom_back[: len(ee_image)] != ee_image:
@@ -117,34 +99,9 @@ def main():
         # land in the application, which banners on the same link.
         device.reset()
         port = pb.Port(device.pty, baud)
-        if one_wire:
-            port = pb.OneWirePort(port)
         try:
             loader = pb.Loader(port)
-            live = loader.connect(15)
+            loader.connect(15)
-            # The loader built from this tree must report a version the tool
-            # beside it speaks — a bump the tool was never told about is a
-            # loader it would refuse to talk to. Not equality with the newest:
-            # the tool now spans two loader generations, the fixed-baud one
-            # here and the unified autobaud loader that follows it.
-            if not pb.OLDEST_LOADER <= live.version <= pb.NEWEST_LOADER:
-                fail(f"loader reports pureboot {live.version}, the tool speaks "
-                     f"{pb.OLDEST_LOADER}..{pb.NEWEST_LOADER}")
-
-            # A W addressed inside a page rather than at its base must still
-            # consume exactly one page and prompt. The loader's own slot is the
-            # target — the guard refuses to commit it — and the payload is
-            # erased-state bytes, so the probe can disturb neither the image nor
-            # the page buffer it leaves behind. Hand-built rather than through
-            # write_page(), which would follow the fill with its erase and
-            # write; the point here is that the fill alone consumes exactly one
-            # page whatever the address's low bits say.
-            wire = base + 1
-            port.write(bytes((ord("W"), pb.selector(pb.SP_FLASH, wire), wire & 0xFF, (wire >> 8) & 0xFF))
-                       + b"\xff" * page)
-            if port.read_exact(1, 5.0) != pb.PROMPT:
-                fail("unaligned W did not return to the prompt")
-
             loader.run_application()
             banner = port.read_exact(3, 5.0)
             if banner != b"APP":
@@ -165,7 +122,7 @@ def main():
     # loader, the trampoline on the application's own entry (patched-vector
     # chips only — a boot-sectioned mega's word 0 stays the application's).
     if patch:
-        flash_words = (base + pb.SLOT) // 2
+        flash_words = (base + 512) // 2
         app = open(app_bin, "rb").read()
         word0 = flash_true[0] | (flash_true[1] << 8)
         if rjmp_decode(word0, 0, flash_words) != base // 2:

test/pbupdate.py

@@ -1,12 +1,17 @@
 #!/usr/bin/env python3
-"""Self-update end-to-end: an application is flashed, the loader replaces
+"""Self-update end-to-end: an application is flashed, then the loader
-itself with a re-timed build, and every power-fail phase is rehearsed by
+replaces itself with a re-timed build through the host tool's
-killing the device mid-write, restarting it from its flash dump, and letting
+--update-loader — and the power-fail phases of that update are rehearsed by
-a re-run complete the update.
+killing the simulated device mid-write, restarting it from its flash dump,
+and letting a re-run complete the update.
 
-The boot-sectioned megas run the BOOTRST-unprogrammed profile — reset boots
+The boot-sectioned megas run the BOOTRST-unprogrammed profile (reset boots
-the application, whose 'L' is the application-owned loader entry — with
+the application; the fixture application's 'L' jump is the application-owned
---assume-fuses standing in for the fuse read simavr cannot model.
+loader entry), with --assume-fuses standing in for the fuse read simavr
+cannot model. The patched-vector chips — the tinies and the m48s — reset
+into a loader at every phase by construction: the t13a because its staging
+slot carries the reset vector itself, the others through the word-0 redirect
+the tool plants around the resident rewrite.
 
 Usage: pbupdate.py <device_bin> <pureboot_elf> <update_elf> <mcu> <hz>
                    <base_hex> <page> <baud> <app_bin> <tool_py> <workdir>
@@ -45,7 +50,7 @@ def assumed_fuses(pb, image):
     image's embedded signature."""
     info = pb.image_info(image)
     which, ladder = pb.BOOT_FUSE[bytes(info.signature[1:3])]
-    bits = min((b for b in ladder if ladder[b] * 2 >= 2 * pb.SLOT), key=lambda b: ladder[b])
+    bits = min((b for b in ladder if ladder[b] * 2 >= 2 * info.slot), key=lambda b: ladder[b])
     fuses = bytearray((0xFF, 0xFF, 0xFF, 0xFF))
     fuses[which] = 0xF8 | (bits << 1) | 1
     return bytes(fuses)
@@ -88,13 +93,16 @@ def main():
     # The m48s are megas without a boot section: patched vector, no fuse
     # preflight, and the same reset-to-0 the tinies get.
     patch = not mega or mcu.startswith("atmega48")
+    # Word-addressed (>64 KiB) chips use the 1 KiB slot; their loader base
+    # itself sits beyond the 16-bit byte space — the 644's base + slot only
+    # touches the 64 KiB boundary and stays byte-addressed.
+    slot = 1024 if base >= 0x10000 and mega else 512
     reset_hex = "0" if mega else None  # the boot-sectioned mega runs BOOTRST-unprogrammed here
     sys.path.insert(0, os.path.dirname(os.path.abspath(tool)))
     sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
     import pbsim
     import pureboot as pb
 
-    slot = pb.SLOT
     os.makedirs(workdir, exist_ok=True)
     objcopy = os.environ.get("PB_OBJCOPY", "avr-objcopy")
     images = {}

test/pbwindow.py

@@ -1,137 +0,0 @@
-#!/usr/bin/env python3
-"""The activation window as a behavioral duration gate.
-
-The loader's window is a counted poll loop whose per-poll cost is hand-counted
-in the source (`link::poll_cycles`) — but the loop compiles in consumer
-context, so only the running image can prove the count. This test installs a
-real application beside the loader (the host tool's own `plan_flash` supplies
-the reset-vector surgery), starts the simulator with the line idle, and reads
-the cycle of the first transmit activity: nothing talks until the window
-closes and the application banners, so that cycle *is* the window, give or
-take a banner lead measured in microseconds. Asserted at ±2 % — one
-mis-counted cycle per poll shifts a window by 10 % and more.
-
-Fixed-baud loaders declare their window in seconds (--seconds, the build's
-TIMEOUT). The autobaud loader's window is its calibration poll budget
-(--autobaud-polls); the seconds it amounts to are budget × 9 / f_cpu, the
-measured cost of the calibrate() wait loop this gate pins.
-"""
-import argparse
-import importlib.util
-import pathlib
-import select
-import sys
-import time
-
-sys.path.insert(0, str(pathlib.Path(__file__).resolve().parent))
-from pbsim import Device
-
-# The calibrate() budget loop's cycles per poll in the built image — what the
-# README's window arithmetic rests on, verified here. A measured fact, not a
-# design constant: the wait's exit branches land where the compiler's block
-# layout puts them, and the bounded-calibration rework moved the loop from
-# ten cycles to nine.
-AUTOBAUD_POLL_CYCLES = 9
-
-
-def load_tool(path):
-    spec = importlib.util.spec_from_file_location("pureboot", path)
-    module = importlib.util.module_from_spec(spec)
-    spec.loader.exec_module(module)
-    return module
-
-
-def compose_flash(pb, loader_bytes, app_bytes, mcu, base, page):
-    """The flash image a completed programming session leaves: application
-    (with the tinies' vector surgery), loader at base — built through the
-    host tool's own planner so the surgery is the shipped one, not a copy."""
-    flash_size = base + pb.SLOT
-    patch = not mcu.startswith("atmega") or mcu.startswith("atmega48")
-    word_flash = flash_size > 0x10000
-    wire_base = base // 2 if word_flash else base
-    flags = (1 if patch else 0) | (2 if word_flash else 0)
-    raw = bytes((ord("P"), ord("B"), 5, 0, 0, 0, page & 0xFF,
-                 wire_base & 0xFF, wire_base >> 8, 0, 0, flags))
-    info = pb.Info(raw)
-
-    flash = bytearray(b"\xff" * flash_size)
-    for address, content in pb.plan_flash(app_bytes, info).items():
-        flash[address:address + len(content)] = content
-    flash[base:base + len(loader_bytes)] = loader_bytes
-    return bytes(flash)
-
-
-def first_tx_cycle(device, deadline):
-    """The PB_WINDOW_TX report, or None. The runner prints it once."""
-    stream = device.proc.stdout
-    while True:
-        remaining = deadline - time.monotonic()
-        if remaining <= 0:
-            return None
-        ready, _, _ = select.select([stream], [], [], remaining)
-        if not ready:
-            return None
-        line = stream.readline()
-        if not line:
-            return None
-        if line.startswith("PB_WINDOW_TX"):
-            return int(line.split()[1])
-
-
-def main():
-    parser = argparse.ArgumentParser()
-    parser.add_argument("--device", required=True)
-    parser.add_argument("--loader", required=True)
-    parser.add_argument("--mcu", required=True)
-    parser.add_argument("--hz", type=int, required=True)
-    parser.add_argument("--base", required=True)
-    parser.add_argument("--page", type=int, required=True)
-    parser.add_argument("--baud", type=int, required=True)
-    parser.add_argument("--app", required=True)
-    parser.add_argument("--tool", required=True)
-    parser.add_argument("--workdir", required=True)
-    parser.add_argument("--link", default=None)
-    parser.add_argument("--seconds", type=float, default=None)
-    parser.add_argument("--autobaud-polls", type=int, default=None)
-    args = parser.parse_args()
-    if (args.seconds is None) == (args.autobaud_polls is None):
-        parser.error("exactly one of --seconds / --autobaud-polls")
-
-    pb = load_tool(args.tool)
-    base = int(args.base, 0)
-    expected = (args.seconds if args.seconds is not None
-                else args.autobaud_polls * AUTOBAUD_POLL_CYCLES / args.hz)
-
-    work = pathlib.Path(args.workdir)
-    work.mkdir(parents=True, exist_ok=True)
-    # Every loader target objcopies its slot content beside the ELF (.bin).
-    loader_bytes = pathlib.Path(args.loader + ".bin").read_bytes()
-    app_bytes = pathlib.Path(args.app).read_bytes()
-    flash_file = work / "window-flash.bin"
-    flash_file.write_bytes(compose_flash(pb, loader_bytes, app_bytes, args.mcu, base, args.page))
-
-    device = Device(args.device, args.loader, args.mcu, str(args.hz), args.base, args.page,
-                    args.baud, str(work / "window-dump.bin"), resume=str(flash_file),
-                    link=args.link, window=True)
-    try:
-        # Simulation speed is machine-dependent; a few hundred thousand
-        # cycles per wall second is the pessimistic floor.
-        budget = max(60.0, expected * args.hz / 300000)
-        cycle = first_tx_cycle(device, time.monotonic() + budget)
-    finally:
-        device.stop()
-
-    if cycle is None:
-        print(f"  [FAIL] no transmit activity within {budget:.0f} s wall "
-              f"(expected a {expected:.2f} s window)")
-        return 1
-    measured = cycle / args.hz
-    error = (measured - expected) / expected
-    ok = abs(error) <= 0.02
-    print(f"  [{'PASS' if ok else 'FAIL'}] window {measured:.3f} s vs declared "
-          f"{expected:.3f} s ({error:+.1%}, gate ±2%)")
-    return 0 if ok else 1
-
-
-if __name__ == "__main__":
-    raise SystemExit(main())

test/pureboot_device.c

@@ -0,0 +1,461 @@
+// simavr "device" for the pureboot protocol tests, every chip. Loads the
+// boot-linked ELF at the loader base, starts execution there (BOOTRST / the
+// patched vector are not what is under test), and exposes the loader's
+// serial link as a pty for the real host tool:
+//
+//   - Hardware USART builds: simavr's uart_pty on the selected instance.
+//   - Software UART builds: an 8N1 bridge between a pty and the GPIO pins,
+//     timed against the simulated cycle counter (drives the loader's RX,
+//     decodes its TX).
+//
+// The link follows the chip's natural default (USART0 on the megas, the
+// software UART on PB0/PB1 elsewhere) unless -l overrides it: `-l usart1`
+// for the second instance, `-l sw:B5,B1` for a software build's RX,TX pins.
+//
+// simavr's tiny cores decode the SPM opcode but attach no NVM module — SPM
+// is a silent no-op (the mega's boot section has one, avr_flash). The
+// missing module is supplied here: the SPM ioctl reads SPMCSR/Z/r1:r0 and
+// implements buffer fill, page erase, page write, and CTPB, completing
+// instantly. RFLB's LPM diversion (fuse readout) stays unmodeled, so the
+// 'F' command answers with flash bytes — the tests assert transport only.
+//
+// On exit (or SIGTERM) the flash and EEPROM are dumped to files for a
+// ground-truth cross-check against what the host read back.
+#include <fcntl.h>
+#include <pty.h>
+#include <signal.h>
+#include <stdint.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <termios.h>
+#include <unistd.h>
+
+#include "avr_eeprom.h"
+#include "avr_flash.h"
+#include "avr_ioport.h"
+#include "avr_uart.h"
+#include "sim_avr.h"
+#include "sim_elf.h"
+#include "sim_io.h"
+#include "uart_pty.h"
+
+static avr_t *avr;
+static uart_pty_t uart_pty;
+static int link_software;
+static char uart_digit = '0';
+static char sw_rx_port = 'B', sw_tx_port = 'B';
+static int sw_rx_bit = 0, sw_tx_bit = 1;
+static const char *dump_path;
+static uint32_t reset_pc;
+static volatile sig_atomic_t reset_requested;
+
+static int parse_link(const char *spec)
+{
+	if (strcmp(spec, "usart0") == 0 || strcmp(spec, "usart1") == 0) {
+		link_software = 0;
+		uart_digit = spec[5];
+		return 0;
+	}
+	if (strncmp(spec, "sw", 2) == 0) {
+		link_software = 1;
+		if (spec[2] == '\0')
+			return 0;
+		if (sscanf(spec + 2, ":%c%d,%c%d", &sw_rx_port, &sw_rx_bit, &sw_tx_port, &sw_tx_bit) == 4)
+			return 0;
+	}
+	return -1;
+}
+
+// simavr 1.6's avr_flash PGERS handler erases spm_pagesize bytes starting at
+// Z & ~1 instead of the page containing Z (its PGWRT path masks correctly) —
+// hardware ignores the in-page bits (§26.8.1), so an erase issued with Z
+// anywhere inside the page wipes half the neighbouring page in simulation
+// only. Wrap the mega's registered flash ioctl and re-dispatch page erases
+// with Z forced to the page boundary; everything else passes through.
+//
+// A second gap on the boot-section-less m48s: their RWWSRE bit is the
+// temporary-buffer discard (Atmel-8271 §26.2/§26.3.1), but the stock model
+// gates its RWWSRE branch on AVR_SELFPROG_HAVE_RWW — absent on the m48
+// core — so the discard store falls through into the buffer-fill branch and
+// plants whatever Z/R1:R0 happen to hold. Perform the silicon's discard
+// here instead.
+static avr_flash_t *mega_flash;
+static int (*mega_flash_ioctl)(avr_io_t *io, uint32_t ctl, void *param);
+
+static int fixed_flash_ioctl(avr_io_t *io, uint32_t ctl, void *param)
+{
+	if (ctl == AVR_IOCTL_FLASH_SPM && avr_regbit_get(io->avr, mega_flash->pgers)) {
+		uint16_t z = (uint16_t)(io->avr->data[30] | (io->avr->data[31] << 8));
+		uint16_t masked = (uint16_t)(z & ~(mega_flash->spm_pagesize - 1));
+		io->avr->data[30] = (uint8_t)masked;
+		io->avr->data[31] = (uint8_t)(masked >> 8);
+		int result = mega_flash_ioctl(io, ctl, param);
+		io->avr->data[30] = (uint8_t)z;
+		io->avr->data[31] = (uint8_t)(z >> 8);
+		return result;
+	}
+	if (ctl == AVR_IOCTL_FLASH_SPM && !(mega_flash->flags & AVR_SELFPROG_HAVE_RWW) &&
+	    (io->avr->data[mega_flash->r_spm] & 0x11) == 0x11) { // RWWSRE|SELFPRGEN: the m48 buffer discard
+		for (int i = 0; i < mega_flash->spm_pagesize / 2; i++) {
+			mega_flash->tmppage[i] = 0xffff;
+			mega_flash->tmppage_used[i] = 0;
+		}
+		avr_regbit_clear(io->avr, mega_flash->selfprgen);
+		return 0;
+	}
+	return mega_flash_ioctl(io, ctl, param);
+}
+
+static void fix_mega_flash_erase(void)
+{
+	for (avr_io_t *io = avr->io_port; io; io = io->next) {
+		if (io->kind && strcmp(io->kind, "flash") == 0) {
+			mega_flash = (avr_flash_t *)io;
+			mega_flash_ioctl = io->ioctl;
+			io->ioctl = fixed_flash_ioctl;
+			return;
+		}
+	}
+	fprintf(stderr, "device: no flash module to fix — SPM page erases may misalign\n");
+}
+
+static void request_reset(int sig)
+{
+	(void)sig;
+	reset_requested = 1;
+}
+
+// ------------------------------------------------------------- tiny NVM ---
+
+typedef struct {
+	avr_io_t io;
+	uint8_t buffer[128];
+	uint8_t used[128]; // a buffer word loads once until erased — like silicon
+	unsigned page;
+} tiny_nvm_t;
+
+static tiny_nvm_t nvm;
+
+static int nvm_ioctl(avr_io_t *io, uint32_t ctl, void *param)
+{
+	(void)param;
+	if (ctl != AVR_IOCTL_FLASH_SPM)
+		return -1;
+	tiny_nvm_t *n = (tiny_nvm_t *)io;
+	avr_t *mcu = io->avr;
+	uint8_t command = mcu->data[0x57] & 0x1f; // SPMCSR, both tinies
+	uint16_t z = (uint16_t)(mcu->data[30] | (mcu->data[31] << 8));
+	uint32_t page_base = (uint32_t)(z & ~(n->page - 1)) % (mcu->flashend + 1);
+	if (command == 0x01) { // SPMEN alone: buffer fill from r1:r0
+		unsigned offset = z & (n->page - 1) & ~1u;
+		if (!n->used[offset]) { // first write wins until the buffer clears
+			n->buffer[offset] = mcu->data[0];
+			n->buffer[offset + 1] = mcu->data[1];
+			n->used[offset] = 1;
+		}
+	} else if (command == 0x03) { // PGERS
+		memset(mcu->flash + page_base, 0xff, n->page);
+	} else if (command == 0x05) { // PGWRT: programming only clears bits
+		for (unsigned i = 0; i < n->page; i++)
+			mcu->flash[page_base + i] &= n->buffer[i];
+		memset(n->buffer, 0xff, n->page);
+		memset(n->used, 0, n->page);
+	} else if (command == 0x11) { // CTPB
+		memset(n->buffer, 0xff, n->page);
+		memset(n->used, 0, n->page);
+	}
+	mcu->data[0x57] &= (uint8_t)~0x1f; // the operation completes instantly
+	return 0;
+}
+
+// ----------------------------------------------------------- GPIO bridge ---
+
+static int pty_master = -1;
+static avr_irq_t *rx_pin; // the loader's RX (PB0), driven from the pty
+static avr_cycle_count_t bit_cycles;
+
+static int tx_level = 1, tx_active, tx_bit;
+static uint8_t tx_shift;
+
+static avr_cycle_count_t tx_sample(avr_t *mcu, avr_cycle_count_t when, void *param)
+{
+	(void)mcu;
+	(void)param;
+	tx_shift = (uint8_t)((tx_shift >> 1) | (tx_level ? 0x80 : 0));
+	if (++tx_bit < 8)
+		return when + bit_cycles;
+	if (write(pty_master, &tx_shift, 1) != 1)
+		fprintf(stderr, "device: pty write lost a byte\n");
+	tx_active = 0;
+	return 0;
+}
+
+static void tx_hook(avr_irq_t *irq, uint32_t value, void *param)
+{
+	(void)irq;
+	(void)param;
+	int level = value & 1;
+	if (!tx_active && tx_level == 1 && level == 0) { // start edge
+		tx_active = 1;
+		tx_bit = 0;
+		avr_cycle_timer_register(avr, bit_cycles + bit_cycles / 2, tx_sample, NULL);
+	}
+	tx_level = level;
+}
+
+static uint8_t rx_queue[8192];
+static unsigned rx_head, rx_tail; // ring: head = next to send
+static int rx_active, rx_bit;
+static uint8_t rx_byte;
+
+static void rx_start_next(void);
+
+static avr_cycle_count_t rx_step(avr_t *mcu, avr_cycle_count_t when, void *param)
+{
+	(void)mcu;
+	(void)param;
+	if (rx_bit < 8) {
+		avr_raise_irq(rx_pin, (rx_byte >> rx_bit) & 1);
+		rx_bit++;
+		return when + bit_cycles;
+	}
+	if (rx_bit == 8) { // stop bit, plus one idle bit of margin
+		avr_raise_irq(rx_pin, 1);
+		rx_bit++;
+		return when + 2 * bit_cycles;
+	}
+	rx_active = 0;
+	rx_start_next();
+	return 0;
+}
+
+static void rx_start_next(void)
+{
+	if (rx_active || rx_head == rx_tail)
+		return;
+	rx_byte = rx_queue[rx_head];
+	rx_head = (rx_head + 1) % sizeof(rx_queue);
+	rx_active = 1;
+	rx_bit = 0;
+	avr_raise_irq(rx_pin, 0); // start bit
+	avr_cycle_timer_register(avr, bit_cycles, rx_step, NULL);
+}
+
+// A reset abandons whatever the bridge was mid-transfer: bytes still queued
+// for a chip that no longer has the context to receive them meaningfully,
+// and a decode in progress on a TX line the reset may have already changed.
+// The pending cycle timers must go with the state: avr_reset drops the TX
+// output latch, whose falling edge starts a spurious decode before this
+// runs, and a stale tx_sample would then interleave with the loader's first
+// real answer through the shared shift state, corrupting it.
+static void bridge_reset(void)
+{
+	avr_cycle_timer_cancel(avr, tx_sample, NULL);
+	avr_cycle_timer_cancel(avr, rx_step, NULL);
+	rx_head = rx_tail = 0;
+	rx_active = 0;
+	tx_active = 0;
+	tx_level = 1;
+	avr_raise_irq(rx_pin, 1); // idle line
+}
+
+static void poll_pty(void)
+{
+	uint8_t chunk[256];
+	ssize_t got = read(pty_master, chunk, sizeof(chunk));
+	for (ssize_t i = 0; i < got; i++) {
+		unsigned next = (rx_tail + 1) % sizeof(rx_queue);
+		if (next == rx_head)
+			break; // full: the host will retry on timeout
+		rx_queue[rx_tail] = chunk[i];
+		rx_tail = next;
+	}
+	if (got > 0)
+		rx_start_next();
+}
+
+// ------------------------------------------------------------------ main ---
+
+static void finish(int sig)
+{
+	(void)sig;
+	if (dump_path) {
+		FILE *f = fopen(dump_path, "wb");
+		if (f) {
+			fwrite(avr->flash, 1, avr->flashend + 1, f);
+			fclose(f);
+		}
+		avr_eeprom_desc_t ee = {.ee = NULL, .offset = 0, .size = 0};
+		if (avr_ioctl(avr, AVR_IOCTL_EEPROM_GET, &ee) == 0 && ee.ee && ee.size) {
+			char path[512];
+			snprintf(path, sizeof(path), "%s.eeprom", dump_path);
+			f = fopen(path, "wb");
+			if (f) {
+				fwrite(ee.ee, 1, ee.size, f);
+				fclose(f);
+			}
+		}
+	}
+	if (!link_software)
+		uart_pty_stop(&uart_pty);
+	_exit(0);
+}
+
+int main(int argc, char *argv[])
+{
+	int link_given = 0;
+	for (int opt; (opt = getopt(argc, argv, "l:")) != -1;) {
+		if (opt != 'l' || parse_link(optarg) != 0) {
+			fprintf(stderr, "device: bad link spec (usart0, usart1, sw, or sw:B0,B1 as RX,TX)\n");
+			return 2;
+		}
+		link_given = 1;
+	}
+	int args = argc - optind;
+	if (args < 7 || args > 9) {
+		fprintf(stderr,
+		        "usage: %s [-l link] <pureboot.elf> <mcu> <hz> <base_hex> <page> <baud> <flash_dump>"
+		        " [reset_hex] [resume_flash]\n"
+		        "  -l link:      usart0 | usart1 | sw[:B0,B1] (RX,TX); default: the chip's own\n"
+		        "  reset_hex:    reset vector (default: base with a boot section, else 0)\n"
+		        "  resume_flash: raw full-flash image loaded instead of the ELF — a prior\n"
+		        "                run's dump, for power-fail resume tests\n",
+		        argv[0]);
+		return 2;
+	}
+	argv += optind - 1; // argv[1] is the ELF again, whatever was parsed
+	const char *mcu_name = argv[2];
+	uint32_t base = (uint32_t)strtoul(argv[4], NULL, 0);
+	unsigned page = (unsigned)atoi(argv[5]);
+	unsigned baud = (unsigned)atoi(argv[6]);
+	dump_path = argv[7];
+	int is_mega = strncmp(mcu_name, "atmega", 6) == 0;
+	if (!link_given)
+		link_software = !is_mega; // the chips' natural links: USART0, or PB0/PB1
+
+	avr = avr_make_mcu_by_name(mcu_name);
+	if (!avr) {
+		fprintf(stderr, "device: no %s core\n", mcu_name);
+		return 1;
+	}
+	avr_init(avr);
+	avr->frequency = (uint32_t)strtoul(argv[3], NULL, 0);
+	memset(avr->flash, 0xff, avr->flashend + 1); // real flash powers up erased
+
+	if (args > 8) {
+		// Resume: the full flash image of an interrupted prior run.
+		FILE *f = fopen(argv[9], "rb");
+		if (!f || fread(avr->flash, 1, avr->flashend + 1, f) == 0) {
+			fprintf(stderr, "device: cannot read %s\n", argv[9]);
+			return 1;
+		}
+		fclose(f);
+	} else {
+		elf_firmware_t fw = {0};
+		if (elf_read_firmware(argv[1], &fw) != 0) {
+			fprintf(stderr, "device: cannot read %s\n", argv[1]);
+			return 1;
+		}
+		memcpy(avr->flash + base, fw.flash, fw.flashsize);
+	}
+	// The boot-sectioned megas enter the loader in hardware (BOOTRST, not
+	// modeled — the argument picks the modeled fuse's target); the tinies
+	// and the boot-section-less m48s reset to word 0 like silicon — erased
+	// flash walks up into the loader, and after the host's surgery the
+	// patched vector routes there.
+	int boot_section = is_mega && strncmp(mcu_name, "atmega48", 8) != 0;
+	reset_pc = args > 7 ? (uint32_t)strtoul(argv[8], NULL, 0) : (boot_section ? base : 0);
+	avr->pc = reset_pc;
+	avr->codeend = avr->flashend;
+
+	// Erased EEPROM, as hardware powers up (simavr zeroes it).
+	uint8_t blank[1024];
+	memset(blank, 0xff, sizeof(blank));
+	avr_eeprom_desc_t seed = {.ee = blank, .offset = 0, .size = 0};
+	if (avr_ioctl(avr, AVR_IOCTL_EEPROM_GET, &seed) == 0 && seed.size <= sizeof(blank)) {
+		seed.ee = blank;
+		avr_ioctl(avr, AVR_IOCTL_EEPROM_SET, &seed);
+	}
+
+	// The megas carry simavr's avr_flash module (and its two gaps the wrap
+	// above fixes); the tinies get the NVM module simavr lacks. Which serial
+	// bridge runs is the link's business, not the chip class's.
+	if (is_mega) {
+		fix_mega_flash_erase();
+	} else {
+		nvm.page = page;
+		memset(nvm.buffer, 0xff, sizeof(nvm.buffer));
+		nvm.io.kind = "tiny_nvm";
+		nvm.io.ioctl = nvm_ioctl;
+		avr_register_io(avr, &nvm.io);
+	}
+
+	if (!link_software) {
+		// POLL_SLEEP paces an idle-polling loader in host real time (a
+		// no-hardware CPU-saving hack); clear it so cycles run free.
+		uint32_t flags = 0;
+		avr_ioctl(avr, AVR_IOCTL_UART_GET_FLAGS(uart_digit), &flags);
+		flags &= ~AVR_UART_FLAG_POLL_SLEEP;
+		avr_ioctl(avr, AVR_IOCTL_UART_SET_FLAGS(uart_digit), &flags);
+		uart_pty_init(avr, &uart_pty);
+		uart_pty_connect(&uart_pty, uart_digit);
+		printf("PB_PTY %s\n", uart_pty.pty.slavename);
+	} else {
+		bit_cycles = (avr->frequency + baud / 2) / baud; // matches uart.hpp's own rounding exactly
+		rx_pin = avr_io_getirq(avr, AVR_IOCTL_IOPORT_GETIRQ(sw_rx_port), (unsigned)sw_rx_bit);
+		avr_irq_register_notify(avr_io_getirq(avr, AVR_IOCTL_IOPORT_GETIRQ(sw_tx_port), (unsigned)sw_tx_bit), tx_hook,
+		                        NULL);
+		avr_raise_irq(rx_pin, 1); // idle line
+
+		int slave;
+		struct termios raw;
+		cfmakeraw(&raw);
+		if (openpty(&pty_master, &slave, NULL, &raw, NULL) != 0) {
+			fprintf(stderr, "device: openpty failed\n");
+			return 1;
+		}
+		fcntl(pty_master, F_SETFL, O_NONBLOCK);
+		printf("PB_PTY %s\n", ttyname(slave));
+	}
+	fflush(stdout);
+
+	signal(SIGTERM, finish);
+	signal(SIGINT, finish);
+	signal(SIGUSR1, request_reset); // an external reset line, for the tests
+
+	long since_poll = 0;
+	for (;;) {
+		int state = avr_run(avr);
+		if (state == cpu_Done || state == cpu_Crashed)
+			break;
+		if (reset_requested) {
+			reset_requested = 0;
+			avr_reset(avr);
+			avr->pc = reset_pc;
+			if (!link_software) { // reset restores the pacing hack; re-clear it
+				uint32_t flags = 0;
+				avr_ioctl(avr, AVR_IOCTL_UART_GET_FLAGS(uart_digit), &flags);
+				flags &= ~AVR_UART_FLAG_POLL_SLEEP;
+				avr_ioctl(avr, AVR_IOCTL_UART_SET_FLAGS(uart_digit), &flags);
+			} else {
+				bridge_reset();
+			}
+		}
+		if (link_software && ++since_poll >= 2000) {
+			since_poll = 0;
+			poll_pty();
+			// An unthrottled idle simulation runs the activation window out
+			// from under the host's real-time knock cadence: a 1 MHz build's
+			// 8 s window is 8 M cycles — tens of wall milliseconds — so a
+			// first knock lost to an in-flight reset misses the window
+			// entirely. Pace the simulation only while the bridge is fully
+			// quiet (nothing decoding, nothing queued); transfers keep full
+			// speed, and a quiet window stretches toward real time.
+			if (!rx_active && !tx_active && rx_head == rx_tail)
+				usleep(200);
+		}
+	}
+	finish(0);
+	return 0;
+}

test/pureboot_device.cpp

@@ -1,668 +0,0 @@
-// simavr "device" for the pureboot protocol tests, every chip. Loads the
-// boot-linked ELF at the loader base, starts execution there (BOOTRST / the
-// patched vector are not what is under test), and exposes the loader's
-// serial link as a pty for the real host tool:
-//
-//   - Hardware USART builds: simavr's uart_pty on the selected instance.
-//   - Software UART builds: an 8N1 bridge between a pty and the GPIO pins,
-//     timed against the simulated cycle counter (drives the loader's RX,
-//     decodes its TX).
-//
-// The link follows the chip's natural default (USART0 on the megas, the
-// software UART on PB0/PB1 elsewhere) unless -l overrides it: `-l usart1`
-// for the second instance, `-l sw:B5,B1` for a software build's RX,TX pins,
-// and `-l sw:D0,D1@0` where those pins are a USART's own — see the pin
-// ownership the bridge models below.
-//
-// simavr's tiny cores decode the SPM opcode but attach no NVM module — SPM
-// is a silent no-op (the mega's boot section has one, avr_flash). The
-// missing module is supplied here: the SPM ioctl reads SPMCSR/Z/r1:r0 and
-// implements buffer fill, page erase, page write, and CTPB, completing
-// instantly. RFLB's LPM diversion (fuse readout) stays unmodeled, so the
-// 'F' command answers with flash bytes — the tests assert transport only.
-//
-// On exit (or SIGTERM) the flash and EEPROM are dumped to files for a
-// ground-truth cross-check against what the host read back.
-#include <csignal>
-#include <cstdint>
-#include <cstdio>
-#include <cstdlib>
-#include <cstring>
-#include <print>
-#include <string_view>
-
-#include <fcntl.h>
-#include <pty.h>
-#include <termios.h>
-#include <unistd.h>
-
-// The parts headers (uart_pty.h) carry no C++ linkage guards of their own,
-// unlike simavr's core headers — the block covers both harmlessly.
-extern "C" {
-#include "avr_eeprom.h"
-#include "avr_flash.h"
-#include "avr_ioport.h"
-#include "avr_uart.h"
-#include "sim_avr.h"
-#include "sim_elf.h"
-#include "sim_io.h"
-#include "uart_pty.h"
-}
-
-namespace {
-
-avr_t *avr;
-uart_pty_t uart_pty;
-bool link_software;
-avr_uart_t *hw_uart; // the pty-driven USART, for the datasheet-reset fix below
-char uart_digit = '0';
-char sw_rx_port = 'B', sw_tx_port = 'B';
-int sw_rx_bit = 0, sw_tx_bit = 1;
-char sw_tx_owner = 0; // the USART whose TXD the software link sits on
-const char *dump_path;
-std::uint32_t reset_pc;
-volatile std::sig_atomic_t reset_requested;
-
-// -w: report the cycle of the first transmit activity, once. What the
-// activation-window gate reads — with an idle line and an application
-// installed, the first thing that ever talks is the application's banner,
-// so this cycle *is* the loader's window plus a banner lead measured in
-// microseconds. Idle pacing is skipped in this mode: there is no real-time
-// host in the loop, and a paced multi-second window would take hours.
-bool window_report;
-bool window_tx_seen;
-
-void window_first_tx()
-{
-	if (!window_report || window_tx_seen)
-		return;
-	window_tx_seen = true;
-	std::println("PB_WINDOW_TX {}", avr->cycle);
-	std::fflush(stdout);
-}
-
-void window_uart_hook(avr_irq_t *, std::uint32_t, void *)
-{
-	window_first_tx();
-}
-
-// One-wire (RX == TX in the link spec): both directions on one GPIO line
-// idling on the firmware's pull-up. The bridge then follows the pin's
-// direction the way the real wiring does: it drives only while the
-// firmware's DDR bit reads input, decodes transitions as the firmware's
-// transmit only while the firmware owns the line, ignores its own raises
-// coming back through the shared irq — and echoes every byte it drives back
-// to the pty, which is what the host-side FTDI tie does and what the host
-// tool's --one-wire mode reads back and discards.
-bool link_one_wire;
-bool mcu_owns_line;
-bool self_drive;
-
-int parse_link(std::string_view spec)
-{
-	if (spec == "usart0" || spec == "usart1") {
-		link_software = false;
-		uart_digit = spec[5];
-		return 0;
-	}
-	if (spec.starts_with("sw")) {
-		link_software = true;
-		if (spec.size() == 2)
-			return 0;
-		char owner = 0;
-		int fields =
-		    std::sscanf(spec.data() + 2, ":%c%d,%c%d@%c", &sw_rx_port, &sw_rx_bit, &sw_tx_port, &sw_tx_bit, &owner);
-		if (fields == 4 || fields == 5) {
-			sw_tx_owner = owner;
-			link_one_wire = sw_rx_port == sw_tx_port && sw_rx_bit == sw_tx_bit;
-			return 0;
-		}
-	}
-	return -1;
-}
-
-// simavr 1.6's avr_flash PGERS handler erases spm_pagesize bytes starting at
-// Z & ~1 instead of the page containing Z (its PGWRT path masks correctly) —
-// hardware ignores the in-page bits (§26.8.1), so an erase issued with Z
-// anywhere inside the page wipes half the neighbouring page in simulation
-// only. Wrap the mega's registered flash ioctl and re-dispatch page erases
-// with Z forced to the page boundary; everything else passes through.
-//
-// A second gap on the boot-section-less m48s: their RWWSRE bit is the
-// temporary-buffer discard (Atmel-8271 §26.2/§26.3.1), but the stock model
-// gates its RWWSRE branch on AVR_SELFPROG_HAVE_RWW — absent on the m48
-// core — so the discard store falls through into the buffer-fill branch and
-// plants whatever Z/R1:R0 happen to hold. Perform the silicon's discard
-// here instead.
-avr_flash_t *mega_flash;
-int (*mega_flash_ioctl)(avr_io_t *io, std::uint32_t ctl, void *param);
-
-int fixed_flash_ioctl(avr_io_t *io, std::uint32_t ctl, void *param)
-{
-	if (ctl == AVR_IOCTL_FLASH_SPM && avr_regbit_get(io->avr, mega_flash->pgers)) {
-		auto z = static_cast<std::uint16_t>(io->avr->data[30] | (io->avr->data[31] << 8));
-		auto masked = static_cast<std::uint16_t>(z & ~(mega_flash->spm_pagesize - 1));
-		io->avr->data[30] = static_cast<std::uint8_t>(masked);
-		io->avr->data[31] = static_cast<std::uint8_t>(masked >> 8);
-		int result = mega_flash_ioctl(io, ctl, param);
-		io->avr->data[30] = static_cast<std::uint8_t>(z);
-		io->avr->data[31] = static_cast<std::uint8_t>(z >> 8);
-		return result;
-	}
-	if (ctl == AVR_IOCTL_FLASH_SPM && !(mega_flash->flags & AVR_SELFPROG_HAVE_RWW) &&
-	    (io->avr->data[mega_flash->r_spm] & 0x11) == 0x11) { // RWWSRE|SELFPRGEN: the m48 buffer discard
-		for (int i = 0; i < mega_flash->spm_pagesize / 2; i++) {
-			mega_flash->tmppage[i] = 0xffff;
-			mega_flash->tmppage_used[i] = 0;
-		}
-		avr_regbit_clear(io->avr, mega_flash->selfprgen);
-		return 0;
-	}
-	return mega_flash_ioctl(io, ctl, param);
-}
-
-void fix_mega_flash_erase()
-{
-	for (avr_io_t *io = avr->io_port; io; io = io->next) {
-		if (io->kind && std::string_view{io->kind} == "flash") {
-			mega_flash = reinterpret_cast<avr_flash_t *>(io);
-			mega_flash_ioctl = io->ioctl;
-			io->ioctl = fixed_flash_ioctl;
-			return;
-		}
-	}
-	std::println(stderr, "device: no flash module to fix — SPM page erases may misalign");
-}
-
-void request_reset(int)
-{
-	reset_requested = 1;
-}
-
-// ------------------------------------------------------------- tiny NVM ---
-
-struct tiny_nvm_t {
-	avr_io_t io;
-	std::uint8_t buffer[128];
-	std::uint8_t used[128]; // a buffer word loads once until erased — like silicon
-	unsigned page;
-};
-
-tiny_nvm_t nvm;
-
-int nvm_ioctl(avr_io_t *io, std::uint32_t ctl, void *)
-{
-	if (ctl != AVR_IOCTL_FLASH_SPM)
-		return -1;
-	auto *n = reinterpret_cast<tiny_nvm_t *>(io);
-	avr_t *mcu = io->avr;
-	std::uint8_t command = mcu->data[0x57] & 0x1f; // SPMCSR, both tinies
-	auto z = static_cast<std::uint16_t>(mcu->data[30] | (mcu->data[31] << 8));
-	std::uint32_t page_base = static_cast<std::uint32_t>(z & ~(n->page - 1)) % (mcu->flashend + 1);
-	if (command == 0x01) { // SPMEN alone: buffer fill from r1:r0
-		unsigned offset = z & (n->page - 1) & ~1u;
-		if (!n->used[offset]) { // first write wins until the buffer clears
-			n->buffer[offset] = mcu->data[0];
-			n->buffer[offset + 1] = mcu->data[1];
-			n->used[offset] = 1;
-		}
-	} else if (command == 0x03) { // PGERS
-		std::memset(mcu->flash + page_base, 0xff, n->page);
-	} else if (command == 0x05) { // PGWRT: programming only clears bits
-		for (unsigned i = 0; i < n->page; i++)
-			mcu->flash[page_base + i] &= n->buffer[i];
-		std::memset(n->buffer, 0xff, n->page);
-		std::memset(n->used, 0, n->page);
-	} else if (command == 0x11) { // CTPB
-		std::memset(n->buffer, 0xff, n->page);
-		std::memset(n->used, 0, n->page);
-	}
-	mcu->data[0x57] &= static_cast<std::uint8_t>(~0x1f); // the operation completes instantly
-	return 0;
-}
-
-// ----------------------------------------------------------- GPIO bridge ---
-
-int pty_master = -1;
-avr_irq_t *rx_pin; // the loader's RX (PB0), driven from the pty
-avr_cycle_count_t bit_cycles;
-
-int tx_level = 1, tx_active, tx_bit;
-std::uint8_t tx_shift;
-
-avr_cycle_count_t tx_sample(avr_t *, avr_cycle_count_t when, void *)
-{
-	if (tx_bit < 0) {
-		// Half a bit into the start bit: a real receiver re-samples here and
-		// abandons a false start. The device's own init produces one — DDR
-		// drives the pin low for the instructions until the idle level is
-		// written — and without this check that glitch decodes as a stray
-		// byte (and would read as first transmit activity under -w).
-		if (tx_level) {
-			tx_active = 0;
-			return 0;
-		}
-		window_first_tx();
-		tx_bit = 0;
-		return when + bit_cycles;
-	}
-	if (tx_bit < 8) {
-		tx_shift = static_cast<std::uint8_t>((tx_shift >> 1) | (tx_level ? 0x80 : 0));
-		if (++tx_bit < 8)
-			return when + bit_cycles;
-		// The byte is delivered at the stop bit's sampling point (9.5 bit
-		// times), where a hardware receiver raises its RXC — not sooner: a
-		// host answering before the stop bit would put its start bit on the
-		// wire while the device is still driving, which the device,
-		// transmitting, is not watching for.
-		return when + bit_cycles;
-	}
-	if (write(pty_master, &tx_shift, 1) != 1)
-		std::println(stderr, "device: pty write lost a byte");
-	tx_active = 0;
-	return 0;
-}
-
-// A USART owns its TxD pin whenever its transmitter is enabled, and the port
-// register cannot drive it (§20.2 / Atmel-8271 §19.2) — which is why a
-// bit-banged link deployed on those pins is mute until it clears UCSRnB.
-// simavr wires a USART entirely through IRQs and never touches the port pin
-// model, so the ownership does not exist there and the mute cannot happen:
-// supply it, or the very state this models is untestable. The link spec's
-// trailing @n names the USART; without one the pins are nobody's.
-avr_uart_t *tx_owner;
-
-bool tx_pin_taken()
-{
-	if (!tx_owner)
-		return false;
-	if (avr_regbit_get(avr, tx_owner->txen))
-		return true;
-	// One-wire on the USART's RXD: RXEN forces the shared pin's direction to
-	// input (§20.7.3), so the firmware's drive goes nowhere until the
-	// release — the receive-side twin of the TXD hold.
-	return link_one_wire && avr_regbit_get(avr, tx_owner->rxen);
-}
-
-// simavr leaves TXEN set in UCSRnB out of reset, where silicon clears the
-// whole register (§20.11.3) — which would hand the pin to a USART no code has
-// enabled, making a freshly reset chip mute for reasons hardware does not
-// have. Reset it the way the datasheet does, so the ownership starts from
-// nobody's and only an application that really enables the USART takes it.
-void reset_tx_owner()
-{
-	if (tx_owner)
-		avr_regbit_clear(avr, tx_owner->txen);
-}
-
-void find_tx_owner()
-{
-	for (avr_io_t *io = avr->io_port; io; io = io->next)
-		if (io->kind && std::string_view{io->kind} == "uart" &&
-		    reinterpret_cast<avr_uart_t *>(io)->name == sw_tx_owner) {
-			tx_owner = reinterpret_cast<avr_uart_t *>(io);
-			reset_tx_owner();
-			return;
-		}
-	std::println(stderr, "device: no USART{} to own the software link's TX pin", sw_tx_owner);
-}
-
-void tx_hook(avr_irq_t *, std::uint32_t value, void *)
-{
-	if (link_one_wire && (self_drive || !mcu_owns_line)) {
-		// The bridge's own drive coming back through the shared irq, or a
-		// transition while the line is the bridge's — either way not the
-		// firmware talking: the decoder sees an idle line.
-		tx_level = 1;
-		return;
-	}
-	if (tx_pin_taken()) { // the USART holds the line; the port write goes nowhere
-		tx_level = 1;
-		return;
-	}
-	int level = value & 1;
-	if (!tx_active && tx_level == 1 && level == 0) { // start edge, confirmed mid-bit
-		tx_active = 1;
-		tx_bit = -1;
-		avr_cycle_timer_register(avr, bit_cycles / 2, tx_sample, nullptr);
-	}
-	tx_level = level;
-}
-
-std::uint8_t rx_queue[8192];
-unsigned rx_head, rx_tail; // ring: head = next to send
-int rx_active, rx_bit;
-std::uint8_t rx_byte;
-
-void rx_start_next();
-
-// Every level the bridge itself puts on the line goes through here, so the
-// shared-pin decoder can tell its own drive from the firmware's.
-void bridge_drive(int level)
-{
-	self_drive = true;
-	avr_raise_irq(rx_pin, static_cast<std::uint32_t>(level));
-	self_drive = false;
-}
-
-avr_cycle_count_t rx_step(avr_t *, avr_cycle_count_t when, void *)
-{
-	if (rx_bit < 8) {
-		bridge_drive((rx_byte >> rx_bit) & 1);
-		rx_bit++;
-		return when + bit_cycles;
-	}
-	if (rx_bit == 8) { // stop bit, plus one idle bit of margin
-		bridge_drive(1);
-		// The host-side tie: an FTDI adapter on a one-wire line reads every
-		// byte it transmits — supply that echo, which the host tool's
-		// --one-wire mode consumes as its wiring check.
-		if (link_one_wire && write(pty_master, &rx_byte, 1) != 1)
-			std::println(stderr, "device: pty echo lost a byte");
-		rx_bit++;
-		return when + 2 * bit_cycles;
-	}
-	rx_active = 0;
-	rx_start_next();
-	return 0;
-}
-
-void rx_start_next()
-{
-	if (rx_active || rx_head == rx_tail)
-		return;
-	// The firmware is answering on the shared line: hold the byte — a real
-	// host's transmission waits out the reply on the wire too. The next
-	// poll_pty tick retries once the line is handed back.
-	if (link_one_wire && mcu_owns_line)
-		return;
-	rx_byte = rx_queue[rx_head];
-	rx_head = (rx_head + 1) % sizeof(rx_queue);
-	rx_active = 1;
-	rx_bit = 0;
-	bridge_drive(0); // start bit
-	avr_cycle_timer_register(avr, bit_cycles, rx_step, nullptr);
-}
-
-// The shared pin's direction is the line's ownership: DDR-out is the
-// firmware driving a frame, DDR-in hands the line back to the bridge.
-void on_ddr(avr_irq_t *, std::uint32_t value, void *)
-{
-	const bool owns = (value >> sw_rx_bit) & 1;
-	if (mcu_owns_line && !owns)
-		bridge_drive(1); // hand-back: a turn-based host idles here, and the cache stays truthful
-	mcu_owns_line = owns;
-	// A byte held back while the firmware answered starts from the next
-	// poll_pty tick, never from inside the DDR write itself — the port
-	// model's own pull-up re-derivation runs right after this notify and
-	// would erase a start edge raised here.
-}
-
-// A reset abandons whatever the bridge was mid-transfer: bytes still queued
-// for a chip that no longer has the context to receive them meaningfully,
-// and a decode in progress on a TX line the reset may have already changed.
-// The pending cycle timers must go with the state: avr_reset drops the TX
-// output latch, whose falling edge starts a spurious decode before this
-// runs, and a stale tx_sample would then interleave with the loader's first
-// real answer through the shared shift state, corrupting it.
-void bridge_reset()
-{
-	avr_cycle_timer_cancel(avr, tx_sample, nullptr);
-	avr_cycle_timer_cancel(avr, rx_step, nullptr);
-	rx_head = rx_tail = 0;
-	rx_active = 0;
-	tx_active = 0;
-	tx_level = 1;
-	mcu_owns_line = false; // avr_reset zeroed DDR: every pin reads input again
-	// Re-drive the idle line through a forced transition: ioport pin irqs are
-	// IRQ_FLAG_FILTERED, and avr_reset zeroes the port latch while the irq
-	// keeps its pre-reset cached value — so a plain raise(1) against a cached
-	// 1 is dropped and the device reads the line stuck low. A loader entering
-	// calibration on that line measures reset-to-first-edge as one giant
-	// pulse and mis-locks or boots the application on the first real knock.
-	// No cycles run between the two raises, so the device only ever sees the
-	// final idle-high.
-	bridge_drive(0);
-	bridge_drive(1);
-}
-
-void poll_pty()
-{
-	std::uint8_t chunk[256];
-	ssize_t got = read(pty_master, chunk, sizeof(chunk));
-	for (ssize_t i = 0; i < got; i++) {
-		unsigned next = (rx_tail + 1) % sizeof(rx_queue);
-		if (next == rx_head)
-			break; // full: the host will retry on timeout
-		rx_queue[rx_tail] = chunk[i];
-		rx_tail = next;
-	}
-	// Unconditional: a byte held back while the firmware owned a shared
-	// line restarts from here once the hand-back has happened.
-	rx_start_next();
-}
-
-// ------------------------------------------------------------------ main ---
-
-[[noreturn]] void finish(int)
-{
-	if (dump_path) {
-		std::FILE *f = std::fopen(dump_path, "wb");
-		if (f) {
-			std::fwrite(avr->flash, 1, avr->flashend + 1, f);
-			std::fclose(f);
-		}
-		avr_eeprom_desc_t ee = {.ee = nullptr, .offset = 0, .size = 0};
-		if (avr_ioctl(avr, AVR_IOCTL_EEPROM_GET, &ee) == 0 && ee.ee && ee.size) {
-			char path[512];
-			std::snprintf(path, sizeof(path), "%s.eeprom", dump_path);
-			f = std::fopen(path, "wb");
-			if (f) {
-				std::fwrite(ee.ee, 1, ee.size, f);
-				std::fclose(f);
-			}
-		}
-	}
-	if (!link_software)
-		uart_pty_stop(&uart_pty);
-	_exit(0);
-}
-
-} // namespace
-
-int main(int argc, char *argv[])
-{
-	bool link_given = false;
-	for (int opt; (opt = getopt(argc, argv, "l:w")) != -1;) {
-		if (opt == 'w') {
-			window_report = true;
-			continue;
-		}
-		if (opt != 'l' || parse_link(optarg) != 0) {
-			std::println(stderr, "device: bad link spec (usart0, usart1, sw, or sw:B0,B1 as RX,TX)");
-			return 2;
-		}
-		link_given = true;
-	}
-	int args = argc - optind;
-	if (args < 7 || args > 9) {
-		std::print(stderr,
-		           "usage: {} [-l link] [-w] <pureboot.elf> <mcu> <hz> <base_hex> <page> <baud> <flash_dump>"
-		           " [reset_hex] [resume_flash]\n"
-		           "  -l link:      usart0 | usart1 | sw[:B0,B1[@0]] (RX,TX, then the USART owning\n"
-		           "                them); default: the chip's own\n"
-		           "  -w:           print PB_WINDOW_TX <cycle> at the first transmit activity and\n"
-		           "                free-run idle time (window measurement mode)\n"
-		           "  reset_hex:    reset vector (default: base with a boot section, else 0)\n"
-		           "  resume_flash: raw full-flash image loaded instead of the ELF — a prior\n"
-		           "                run's dump, for power-fail resume tests\n",
-		           argv[0]);
-		return 2;
-	}
-	argv += optind - 1; // argv[1] is the ELF again, whatever was parsed
-	const std::string_view mcu_name = argv[2];
-	auto base = static_cast<std::uint32_t>(std::strtoul(argv[4], nullptr, 0));
-	auto page = static_cast<unsigned>(std::atoi(argv[5]));
-	auto baud = static_cast<unsigned>(std::atoi(argv[6]));
-	dump_path = argv[7];
-	const bool is_mega = mcu_name.starts_with("atmega");
-	if (!link_given)
-		link_software = !is_mega; // the chips' natural links: USART0, or PB0/PB1
-
-	avr = avr_make_mcu_by_name(mcu_name.data());
-	if (!avr) {
-		std::println(stderr, "device: no {} core", mcu_name);
-		return 1;
-	}
-	avr_init(avr);
-	avr->frequency = static_cast<std::uint32_t>(std::strtoul(argv[3], nullptr, 0));
-	std::memset(avr->flash, 0xff, avr->flashend + 1); // real flash powers up erased
-
-	if (args > 8) {
-		// Resume: the full flash image of an interrupted prior run.
-		std::FILE *f = std::fopen(argv[9], "rb");
-		if (!f || std::fread(avr->flash, 1, avr->flashend + 1, f) == 0) {
-			std::println(stderr, "device: cannot read {}", argv[9]);
-			return 1;
-		}
-		std::fclose(f);
-	} else {
-		elf_firmware_t fw{};
-		if (elf_read_firmware(argv[1], &fw) != 0) {
-			std::println(stderr, "device: cannot read {}", argv[1]);
-			return 1;
-		}
-		// An image past flash end would smash the simulator's heap and turn
-		// into phantom peripheral behavior (lessons: believe the size gate
-		// first) — refuse it loudly instead.
-		if (base + fw.flashsize > avr->flashend + 1) {
-			std::println(stderr, "device: {} B at {:#x} runs past flash end {:#x} — image does not fit its slot",
-			             fw.flashsize, base, avr->flashend);
-			return 1;
-		}
-		std::memcpy(avr->flash + base, fw.flash, fw.flashsize);
-	}
-	// The boot-sectioned megas enter the loader in hardware (BOOTRST, not
-	// modeled — the argument picks the modeled fuse's target); the tinies
-	// and the boot-section-less m48s reset to word 0 like silicon — erased
-	// flash walks up into the loader, and after the host's surgery the
-	// patched vector routes there.
-	const bool boot_section = is_mega && !mcu_name.starts_with("atmega48");
-	reset_pc = args > 7 ? static_cast<std::uint32_t>(std::strtoul(argv[8], nullptr, 0)) : (boot_section ? base : 0);
-	avr->pc = reset_pc;
-	avr->codeend = avr->flashend;
-
-	// Erased EEPROM, as hardware powers up (simavr zeroes it).
-	std::uint8_t blank[1024];
-	std::memset(blank, 0xff, sizeof(blank));
-	avr_eeprom_desc_t seed = {.ee = blank, .offset = 0, .size = 0};
-	if (avr_ioctl(avr, AVR_IOCTL_EEPROM_GET, &seed) == 0 && seed.size <= sizeof(blank)) {
-		seed.ee = blank;
-		avr_ioctl(avr, AVR_IOCTL_EEPROM_SET, &seed);
-	}
-
-	// The megas carry simavr's avr_flash module (and its two gaps the wrap
-	// above fixes); the tinies get the NVM module simavr lacks. Which serial
-	// bridge runs is the link's business, not the chip class's.
-	if (is_mega) {
-		fix_mega_flash_erase();
-	} else {
-		nvm.page = page;
-		std::memset(nvm.buffer, 0xff, sizeof(nvm.buffer));
-		nvm.io.kind = "tiny_nvm";
-		nvm.io.ioctl = nvm_ioctl;
-		avr_register_io(avr, &nvm.io);
-	}
-
-	if (!link_software) {
-		// POLL_SLEEP paces an idle-polling loader in host real time (a
-		// no-hardware CPU-saving hack); clear it so cycles run free.
-		std::uint32_t flags = 0;
-		avr_ioctl(avr, AVR_IOCTL_UART_GET_FLAGS(uart_digit), &flags);
-		flags &= ~AVR_UART_FLAG_POLL_SLEEP;
-		avr_ioctl(avr, AVR_IOCTL_UART_SET_FLAGS(uart_digit), &flags);
-		// simavr leaves TXEN set out of reset where silicon clears the whole
-		// UCSR#B (§20.11.3). Harmless to a loader that enables TXEN itself —
-		// but a half-duplex build's receiver-only init then *drops* TXEN,
-		// and this uart model clears UDRE on that edge and never re-raises
-		// it on a later enable: the first transmitter after the hand-over
-		// waits UDRE forever, a wedge silicon does not have. Start from the
-		// datasheet's zero, as the software bridge's tx-owner model does.
-		for (avr_io_t *io = avr->io_port; io; io = io->next)
-			if (io->kind && std::string_view{io->kind} == "uart" &&
-			    reinterpret_cast<avr_uart_t *>(io)->name == uart_digit)
-				hw_uart = reinterpret_cast<avr_uart_t *>(io);
-		if (hw_uart)
-			avr_regbit_clear(avr, hw_uart->txen);
-		uart_pty_init(avr, &uart_pty);
-		uart_pty_connect(&uart_pty, uart_digit);
-		if (window_report)
-			avr_irq_register_notify(avr_io_getirq(avr, AVR_IOCTL_UART_GETIRQ(uart_digit), UART_IRQ_OUTPUT),
-			                        window_uart_hook, nullptr);
-		std::println("PB_PTY {}", uart_pty.pty.slavename);
-	} else {
-		bit_cycles = (avr->frequency + baud / 2) / baud; // matches uart.hpp's own rounding exactly
-		if (sw_tx_owner)
-			find_tx_owner();
-		rx_pin = avr_io_getirq(avr, AVR_IOCTL_IOPORT_GETIRQ(sw_rx_port), static_cast<unsigned>(sw_rx_bit));
-		avr_irq_register_notify(
-		    avr_io_getirq(avr, AVR_IOCTL_IOPORT_GETIRQ(sw_tx_port), static_cast<unsigned>(sw_tx_bit)), tx_hook,
-		    nullptr);
-		if (link_one_wire)
-			avr_irq_register_notify(avr_io_getirq(avr, AVR_IOCTL_IOPORT_GETIRQ(sw_rx_port), IOPORT_IRQ_DIRECTION_ALL),
-			                        on_ddr, nullptr);
-		bridge_drive(1); // idle line
-
-		int slave;
-		struct termios raw;
-		cfmakeraw(&raw);
-		if (openpty(&pty_master, &slave, nullptr, &raw, nullptr) != 0) {
-			std::println(stderr, "device: openpty failed");
-			return 1;
-		}
-		fcntl(pty_master, F_SETFL, O_NONBLOCK);
-		std::println("PB_PTY {}", ttyname(slave));
-	}
-	std::fflush(stdout);
-
-	std::signal(SIGTERM, finish);
-	std::signal(SIGINT, finish);
-	std::signal(SIGUSR1, request_reset); // an external reset line, for the tests
-
-	long since_poll = 0;
-	for (;;) {
-		int state = avr_run(avr);
-		if (state == cpu_Done || state == cpu_Crashed)
-			break;
-		if (reset_requested) {
-			reset_requested = 0;
-			avr_reset(avr);
-			avr->pc = reset_pc;
-			if (!link_software) { // reset restores the pacing hack; re-clear it
-				std::uint32_t flags = 0;
-				avr_ioctl(avr, AVR_IOCTL_UART_GET_FLAGS(uart_digit), &flags);
-				flags &= ~AVR_UART_FLAG_POLL_SLEEP;
-				avr_ioctl(avr, AVR_IOCTL_UART_SET_FLAGS(uart_digit), &flags);
-				if (hw_uart) // and simavr's bogus reset TXEN (§20.11.3: zero)
-					avr_regbit_clear(avr, hw_uart->txen);
-			} else {
-				bridge_reset();
-				reset_tx_owner();
-			}
-		}
-		if (link_software && ++since_poll >= 2000) {
-			since_poll = 0;
-			poll_pty();
-			// An unthrottled idle simulation runs the activation window out
-			// from under the host's real-time knock cadence: a 1 MHz build's
-			// 8 s window is 8 M cycles — tens of wall milliseconds — so a
-			// first knock lost to an in-flight reset misses the window
-			// entirely. Pace the simulation only while the bridge is fully
-			// quiet (nothing decoding, nothing queued); transfers keep full
-			// speed, and a quiet window stretches toward real time.
-			if (!window_report && !rx_active && !tx_active && rx_head == rx_tail)
-				usleep(200);
-		}
-	}
-	finish(0);
-}

test/test_handshake.py

@@ -1,212 +0,0 @@
-#!/usr/bin/env python3
-"""Host-tool activation handshake: bounded against a line that misbehaves.
-
-`_handshake` drains the line after it sees a prompt, to absorb a real loader's
-trailing bytes before it asks for the identity. That drain must be bounded: a
-target that never falls quiet — a board stuck in a reset loop presents exactly
-this, ~60 reboots/s of UART-reset garbage in which a stray 0x2b reads as a
-prompt — otherwise spins the tool forever. Regression for that hang, plus a
-control that a well-behaved loader still connects.
-
-The handshake must also survive its own leftovers: after `--stay` the loader's
-final prompt can still be in the USB pipeline when the next invocation opens
-the port, and on a board wired to reset on open, that opening starts a fresh
-activation window the stale prompt then betrays — the tool commits to an
-identity read against a device that never heard its knock, and what it finally
-collects is the application's banner. StaleDTRPort is that moment as a port.
-
-Stdlib only, no device: host-tool logic, so it runs on every chip's preset
-beside pureboot.planner.
-"""
-import importlib.util
-import pathlib
-import threading
-import time
-
-PB = pathlib.Path(__file__).resolve().parents[1] / "pureboot" / "pureboot.py"
-_spec = importlib.util.spec_from_file_location("pureboot", PB)
-pb = importlib.util.module_from_spec(_spec)
-_spec.loader.exec_module(pb)
-
-P = F = 0
-
-
-def check(name, ok):
-    global P, F
-    P, F = P + (1 if ok else 0), F + (0 if ok else 1)
-    print(f"  [{'PASS' if ok else 'FAIL'}] {name}")
-
-
-class FloodPort:
-    """A line that never falls quiet: read_available always returns bytes, and
-    they contain a prompt. No identity ever completes."""
-
-    def flush_input(self):
-        pass
-
-    def write(self, data):
-        pass
-
-    def read_available(self, wait):
-        time.sleep(0.01)  # a real read waits; keep the busy loop off a core
-        return b"+\x00\xff"
-
-    def read_exact(self, count, timeout):
-        raise pb.Error("no identity")
-
-
-class LoaderPort:
-    """A well-behaved pureboot 5: a prompt to the knock, then quiet, then the
-    slim identity (version 5 + m328p signature) and a closing prompt."""
-
-    def __init__(self):
-        self.pending = b""
-        self.exacts = 0
-
-    def flush_input(self):
-        self.pending = b""
-
-    def write(self, data):
-        if b"p" in data:
-            self.pending = b"+"  # the prompt answers the knock, nothing else
-
-    def read_available(self, wait):
-        data, self.pending = self.pending, b""
-        return data
-
-    def read_exact(self, count, timeout):
-        self.exacts += 1
-        return b"\x05\x1e\x95\x0f" if self.exacts == 1 else b"+"  # identity, then prompt
-
-
-class StaleDTRPort:
-    """`--stay`, then a fresh invocation on a board that resets when its port
-    opens. Three facts of that moment, all timed from the open: the previous
-    session's final prompt is still in transit and lands only after the
-    opening flush has already run; the reset holds the device off the line
-    at first, eating anything written before it completes; and the fresh
-    window is finite — once it expires the application boots and prints a
-    banner whose bytes are what a pending identity read collects. A
-    handshake that trusts the stale prompt spends the whole window waiting
-    on a device that never heard its knock; one that drains the line first
-    knocks into the real window and connects."""
-
-    STALE_AT = 0.02   # the leftover prompt becomes visible (post-flush)
-    READY_AT = 0.05   # reset complete, activation window opens
-    WINDOW = 1.0      # window length; expiry boots the application
-
-    def __init__(self):
-        self.t0 = time.monotonic()
-        # (visible-from, bytes): the line as a timed queue.
-        self.queue = [(self.t0 + self.STALE_AT, b"+")]
-        self.armed = False  # a 'p' heard inside the window arms 'b'
-        self.booted = False
-
-    def _boot_check(self):
-        if not self.booted and time.monotonic() > self.t0 + self.READY_AT + self.WINDOW:
-            self.booted = True
-            self.queue.append((self.t0 + self.READY_AT + self.WINDOW,
-                               b"W r libavr tempmon\r\n"))
-
-    def _visible(self):
-        self._boot_check()
-        now = time.monotonic()
-        return b"".join(d for t, d in self.queue if t <= now)
-
-    def _consume(self, n):
-        now = time.monotonic()
-        left = []
-        for t, d in self.queue:
-            if t <= now and n:
-                take = min(n, len(d))
-                d = d[take:]
-                n -= take
-            if d:
-                left.append((t, d))
-        self.queue = left
-
-    def flush_input(self):
-        self._consume(len(self._visible()))
-
-    def write(self, data):
-        self._boot_check()
-        now = time.monotonic()
-        if now < self.t0 + self.READY_AT or self.booted:
-            return  # still in reset, or the application owns the line
-        if b"p" in data:
-            self.armed = True
-            self.queue.append((now + 0.01, b"+"))
-        if b"b" in data and self.armed:
-            # The slim identity (version 5 + m328p signature) and a prompt.
-            self.queue.append((now + 0.01, b"\x05\x1e\x95\x0f+"))
-
-    def read_available(self, wait):
-        deadline = time.monotonic() + wait
-        while True:
-            data = self._visible()
-            if data:
-                self._consume(len(data))
-                return data
-            if time.monotonic() >= deadline:
-                return b""
-            time.sleep(0.005)
-
-    def read_exact(self, count, timeout):
-        deadline = time.monotonic() + timeout
-        data = b""
-        while len(data) < count:
-            visible = self._visible()
-            if visible:
-                take = visible[:count - len(data)]
-                self._consume(len(take))
-                data += take
-            elif time.monotonic() >= deadline:
-                raise pb.Error(f"timeout: got {len(data)} of {count} bytes")
-            else:
-                time.sleep(0.005)
-        return data
-
-
-def terminates(port, wait, budget):
-    """Run connect_autobaud in a thread; True if it returns/raises within
-    `budget` seconds rather than hanging."""
-    done = threading.Event()
-
-    def run():
-        try:
-            pb.Loader(port).connect_autobaud(wait)
-        except Exception:
-            pass
-        finally:
-            done.set()
-
-    threading.Thread(target=run, daemon=True).start()
-    return done.wait(budget)
-
-
-def main():
-    # the hang: a flooding target must not spin the drain forever. With wait=0.5
-    # the whole handshake has to give up well inside a few seconds.
-    check("flooding target: handshake terminates, drain is bounded",
-          terminates(FloodPort(), wait=0.5, budget=4.0))
-
-    # the control: a real loader still connects and reads identity.
-    info = pb.Loader(LoaderPort()).connect_autobaud(2.0)
-    check("well-behaved loader still connects (version 5)", info.version == 5)
-
-    # the stale prompt: a --stay leftover plus reset-on-open must not burn the
-    # fresh window — the pre-knock drain absorbs it and the first real knock
-    # lands inside the window.
-    try:
-        stale_ok = pb.Loader(StaleDTRPort()).connect(2.5).version == 5
-    except pb.Error as failed:
-        print(f"        ({failed})")
-        stale_ok = False
-    check("stale --stay prompt + reset-on-open: connects in the fresh window", stale_ok)
-
-    print(f"\n  {P} passed, {F} failed")
-    return 1 if F else 0
-
-
-if __name__ == "__main__":
-    raise SystemExit(main())

test/test_planner.py

@@ -1,8 +1,9 @@
 #!/usr/bin/env python3
-"""Host-tool unit tests — the planning and policy logic, no simulator:
+"""Host-tool unit tests — the pure planning and policy logic, no simulator:
-programming orders and their recovery properties, the reset-vector surgery,
+the flash-programming orders and their recovery properties, the reset-vector
-the staging composition, the boot-fuse decode, and the update preflight over
+surgery, the staging-slot composition, the mega boot-fuse decode, and the
-fuse combinations simavr cannot model.
+update preflight's error/warning matrix (fuse combinations simavr cannot
+model reach it here as synthetic bytes).
 
 Usage: test_planner.py <tool_py>
 """
@@ -26,21 +27,14 @@ def expect_error(what, fn, *needles):
     fail(f"{what}: no error raised")
 
 
-def info_of(pb, base, page, patch, flash, signature=(0x1E, 0x93, 0x0B), word_flash=False, version=None):
+def info_of(pb, base, page, patch, flash, signature=(0x1E, 0x93, 0x0B), word_flash=False):
     scale = 2 if word_flash else 1
     wire_base = base // scale
     flags = (1 if patch else 0) | (2 if word_flash else 0)
-    # The EEPROM size comes from the signature, as it must: pureboot 5 derives
+    raw = bytes((0x50, 0x42, 1, *signature, page & 0xFF, wire_base & 0xFF, wire_base >> 8,
-    # the whole geometry from the signature rather than sending it, so a
+                 0, 2, flags))
-    # synthetic block that disagreed with its own signature would describe a
-    # chip that cannot exist.
-    eeprom = pb.CHIP_GEOMETRY[signature][2]
-    raw = bytes((0x50, 0x42, pb.NEWEST_LOADER if version is None else version,
-                 *signature, page & 0xFF, wire_base & 0xFF, wire_base >> 8,
-                 eeprom & 0xFF, eeprom >> 8, flags))
     info = pb.Info(raw)
-    if info.flash_size != flash:
+    assert info.flash_size == flash
-        fail(f"info_of({base:#x}) decodes to {info.flash_size:#x} of flash, not {flash:#x}")
     return info
 
 
@@ -61,21 +55,6 @@ def main():
     tiny = info_of(pb, 0x1E00, 64, True, 0x2000)
     mega = info_of(pb, 0x7E00, 128, False, 0x8000, signature=(0x1E, 0x95, 0x0F))
 
-    # Versioning: the block's third byte is the loader's version, and the tool
-    # speaks a window of them. Every version in the window decodes, so an older
-    # deployed loader stays usable; one above the window is refused by name,
-    # since which version changed the protocol is knowledge only the tool
-    # holds, and it holds none about a version it has never heard of.
-    for version in range(pb.OLDEST_LOADER, pb.NEWEST_LOADER + 1):
-        if info_of(pb, 0x1E00, 64, True, 0x2000, version=version).version != version:
-            fail(f"pureboot {version} does not decode")
-    expect_error(
-        "unknown loader version",
-        lambda: info_of(pb, 0x1E00, 64, True, 0x2000, version=pb.NEWEST_LOADER + 1),
-        f"pureboot {pb.NEWEST_LOADER + 1}",
-        "newer tool",
-    )
-
     # mega_boot: BOOTSZ words and the BOOTRST sense per chip — the fuse byte
     # index (EXTENDED on the x8 line except the m328s' HIGH, HIGH elsewhere)
     # and the per-family ladders (Atmel-2486/2466/2503/2545/8271/DS40002065/
@@ -100,7 +79,9 @@ def main():
         ((0x1E, 0x97, 0x05), 0x20000, 3, {0b11: 0x1FC00, 0b10: 0x1F800, 0b01: 0x1F000, 0b00: 0x1E000}),  # 1284P
     )
     for signature, flash, which, ladder in cases:
-        chip = info_of(pb, flash - pb.SLOT, 128 if flash < 0x20000 else 0, False, flash,
+        # Word-addressed chips carry the 1 KiB slot (their smallest boot sector).
+        slot = 1024 if flash > 0x10000 else 512
+        chip = info_of(pb, flash - slot, 128 if flash < 0x20000 else 0, False, flash,
                        signature=signature, word_flash=flash > 0x10000)
         for bits, start in ladder.items():
             fuses = bytearray((0xFF, 0xFF, 0xFF, 0xFF))
@@ -113,12 +94,10 @@ def main():
             if prog or at != start:
                 fail(f"mega_boot {signature[1]:02x}{signature[2]:02b} unprogrammed: {prog} {at:#07x}")
 
-    # Word-addressed info decode: the 1284P's base and page ride the wire
+    # Word-addressed info decode: the 1284P's base/page ride the wire scaled,
-    # scaled — a 17-bit base halved into the block's two bytes, a 256-byte page
+    # and its slot is 1 KiB.
-    # spelled 0 — and its slot is the same 512 bytes as everywhere else, so its
+    big = info_of(pb, 0x1FC00, 0, False, 0x20000, signature=(0x1E, 0x97, 0x05), word_flash=True)
-    # staging slot lands inside the 1 KiB minimum boot section.
+    if big.page != 256 or big.base != 0x1FC00 or big.stage != 0x1F800 or big.slot != 1024:
-    big = info_of(pb, 0x1FE00, 0, False, 0x20000, signature=(0x1E, 0x97, 0x05), word_flash=True)
-    if big.page != 256 or big.base != 0x1FE00 or big.stage != 0x1FC00:
         fail(f"word-addressed info decode: page {big.page}, base {big.base:#x}, stage {big.stage:#x}")
 
     # Surgery: word 0 lands on the loader, the trampoline on the original
@@ -168,24 +147,13 @@ def main():
         fail("mega staging content should be the bare image")
     expect_error("mega staging size", lambda: pb.staging_content(image + b"!", mega), "512")
 
-    # The image stamp: found in a synthetic binary, absent in noise. pureboot
+    # The embedded info block: found in a synthetic binary, absent in noise.
-    # 5 stamps the magic, its version and the signature, and the geometry is
+    binary = bytes((0xAA,)) * 10 + tiny.raw + bytes((0xBB,)) * 10
-    # looked up from there — so what comes back must equal what a live device
-    # of the same chip reports.
-    stamp = bytes((0x50, 0x42, pb.NEWEST_LOADER)) + bytes(tiny.signature)
-    binary = bytes((0xAA,)) * 10 + stamp + bytes((0xBB,)) * 10
     found = pb.image_info(binary)
     if found is None or found.raw != tiny.raw:
-        fail(f"image_info misreads the v{pb.NEWEST_LOADER} stamp: "
+        fail("image_info misses the embedded block")
-             f"{found.raw.hex() if found else None} != {tiny.raw.hex()}")
     if pb.image_info(bytes((0xAA,)) * 40) is not None:
         fail("image_info invents a block")
-    # An older loader's image stays readable, so a deployed build can be
-    # identified and installed like any other.
-    old = info_of(pb, 0x1E00, 64, True, 0x2000, version=pb.OLDEST_LOADER)
-    found_old = pb.image_info(bytes((0xAA,)) * 10 + old.raw)
-    if found_old is None or found_old.version != pb.OLDEST_LOADER:
-        fail("image_info misses an older loader's block")
 
     # loader_image must peel a padded image down to the slot content: a raw
     # .bin padded from address 0 (or a whole-flash read-back with the loader
@@ -226,15 +194,6 @@ def main():
     if pb.update_preflight(bytes((0xAA,)) * 8 + tiny.raw, tiny, None) != []:
         fail("tiny preflight should pass without fuses")
 
-    # The 1284s' smallest boot section (512 words) is exactly the resident
-    # slot plus its staging slot, so self-update is possible at the minimum
-    # BOOTSZ — no fuse step up, the 644's geometry. That holds only while a
-    # slot is 512 B: at 1 KiB the staging slot would fall outside the section
-    # and the preflight would refuse.
-    notes = pb.update_preflight(bytes((0xAA,)) * 8 + big.raw, big, fuses(0xFE))
-    if not any("staging slot" in n for n in notes):
-        fail(f"1284 minimum-BOOTSZ notes: {notes}")
-
     # The walk-region refusal: BOOTRST aimed below the loader plus app data
     # in the walk span errors without --force; erased spans and unprogrammed
     # BOOTRST pass.
@@ -245,114 +204,6 @@ def main():
     pb.check_walk_region({0x7800: bytes((0xFF,)) * 128}, mega, fuses(0xFA), False)
     pb.check_walk_region(deep, mega, None, False)  # fuses unknown: no check
 
-    # The repairing verify: a mismatched page is rewritten rather than raised,
-    # bounded so a fault that is not self-clearing cannot spin.
-    class FakeLoader:
-        """A device whose first `bad` writes of any page land wrong."""
-
-        def __init__(self, info, bad):
-            self.info = info
-            self.bad = bad
-            self.flash = {}
-            self.writes = 0
-
-        def write_page(self, address, data):
-            self.writes += 1
-            self.flash[address] = bytes(len(data)) if self.bad > 0 else bytes(data)
-            self.bad -= 1
-
-        def read_flash(self, address, count):
-            return self.flash.get(address, bytes(count))
-
-    want = {0: bytes((i * 5) & 0xFF for i in range(128))}
-
-    # One bad write, then good: repaired in place, and the caller never sees
-    # an error. The rewrite is counted, so a silent no-op cannot pass.
-    device = FakeLoader(info_of(pb, 0x7E00, 128, False, 0x8000), bad=1)
-    device.write_page(0, want[0])
-    pb.verify_pages(device, want, repair=True)
-    if device.writes != 2:
-        fail(f"repairing verify made {device.writes} writes, expected 2")
-
-    # Without repair the same state raises, so the repair is what fixed it.
-    device = FakeLoader(info_of(pb, 0x7E00, 128, False, 0x8000), bad=1)
-    device.write_page(0, want[0])
-    expect_error("verify without repair", lambda: pb.verify_pages(device, want), "verify failed")
-
-    # A page that never comes good stops after RETRIES rewrites, and says so.
-    device = FakeLoader(info_of(pb, 0x7E00, 128, False, 0x8000), bad=99)
-    device.write_page(0, want[0])
-    expect_error(
-        "unrepairable page",
-        lambda: pb.verify_pages(device, want, repair=True),
-        "verify failed",
-        f"after {pb.RETRIES} retries",
-    )
-    if device.writes != pb.RETRIES + 1:
-        fail(f"unrepairable page took {device.writes} writes, expected {pb.RETRIES + 1}")
-
-    # The knock handshake against a device that is not listening yet — the
-    # state a port open leaves behind: it resets the chip into a fresh
-    # activation window while the previous session's prompt is still in
-    # flight, so the first knock is lost and a prompt arrives anyway.
-    class FakePort:
-        """A loader in its activation window, plus `lost` leading writes the
-        reset swallows and one stale prompt still on the wire."""
-
-        def __init__(self, info_raw, lost=0, stale=b"", active=False):
-            self.info_raw = info_raw
-            self.lost = lost
-            self.inflight = bytearray(stale)
-            self.rx = bytearray()
-            self.active = active
-            self.last = None
-
-        def flush_input(self):
-            self.rx.clear()
-
-        def write(self, data):
-            if self.lost:
-                self.lost -= 1
-                return
-            for byte in bytes(data):
-                if not self.active:
-                    self.active = self.last == ord("p") and byte == ord("b")
-                    self.last = byte
-                    if self.active:
-                        self.rx += pb.PROMPT
-                elif byte == ord("b"):
-                    self.rx += self.info_raw + pb.PROMPT
-                else:
-                    self.rx += pb.PROMPT
-
-        def read_available(self, wait):
-            self.rx = self.inflight + self.rx  # the stale prompt lands late
-            self.inflight.clear()
-            out, self.rx = bytes(self.rx), bytearray()
-            return out
-
-        def read_exact(self, count, timeout):
-            if len(self.rx) < count:
-                raise pb.Error(f"timeout: got {len(self.rx)} of {count} bytes")
-            out, self.rx = bytes(self.rx[:count]), self.rx[count:]
-            return out
-
-    raw = info_of(pb, 0x7E00, 128, False, 0x8000).raw
-    for what, port in (
-        ("clean window", FakePort(raw)),
-        ("stale prompt over a lost knock", FakePort(raw, lost=1, stale=pb.PROMPT)),
-        ("live session", FakePort(raw, active=True)),
-    ):
-        info = pb.Loader(port).connect(5)
-        if info.raw != raw:
-            fail(f"connect ({what}) returned {info.raw.hex()}")
-
-    # A device that never answers still says so, and a version the tool cannot
-    # speak is reported as such rather than retried into a timeout.
-    expect_error("dead device", lambda: pb.Loader(FakePort(raw, lost=99)).connect(0), "no answer")
-    old = bytes(raw[:2]) + bytes((pb.NEWEST_LOADER + 1,)) + bytes(raw[3:])
-    expect_error("unspeakable version", lambda: pb.Loader(FakePort(old)).connect(5), "needs a newer tool")
-
     print("test_planner: all planner and policy checks pass")
 
 

test/test_scan.py

@@ -1,71 +0,0 @@
-#!/usr/bin/env python3
-"""--scan's walk and report logic, no simulator: the probe order, the rate
-arithmetic, and the advice's direction. The rate physics itself is not
-sim-testable — a pty carries bytes at any termios rate — so what the wire
-would arbitrate is pinned here as logic instead.
-
-Usage: test_scan.py <tool_py>
-"""
-
-import os
-import sys
-
-
-def fail(message):
-    print(f"FAIL: {message}")
-    sys.exit(1)
-
-
-def main():
-    sys.path.insert(0, os.path.dirname(os.path.abspath(sys.argv[1])))
-    import pureboot as pb
-
-    walk = pb.scan_ratios()
-    if walk != [0, -2, 2, -4, 4, -6, 6, -8, 8, -10, 10]:
-        fail(f"probe walk is not built-rate-first, nearest-out: {walk}")
-
-    if pb.scan_rate(9600, 4) != 9984 or pb.scan_rate(9600, -4) != 9216:
-        fail("probe rate arithmetic")
-    if pb.scan_rate(115200, 0) != 115200:
-        fail("the built rate must probe unchanged")
-
-    # A loader answering fast means a fast oscillator: the trim goes down.
-    report = "\n".join(pb.scan_report(9600, 4, 6))
-    for needle in ("9984", "+4 %", "--baud 9984", "4 steps lower", "pureboot 6"):
-        if needle not in report:
-            fail(f"+4 % report lacks {needle!r}:\n{report}")
-    report = "\n".join(pb.scan_report(9600, -6, 6))
-    if "6 steps higher" not in report:
-        fail(f"-6 % report advises the wrong direction:\n{report}")
-
-    report = "\n".join(pb.scan_report(9600, 0, 6))
-    if "none" not in report or "steps" in report:
-        fail(f"an on-rate answer must advise no trim:\n{report}")
-
-    report = "\n".join(pb.scan_report(9600, 4, 6, clock=9600000))
-    if "9984000" not in report:
-        fail(f"the absolute clock must scale with the found ratio:\n{report}")
-
-    # The walk's rates mostly have no termios B-constant, so the POSIX port
-    # must set them through termios2 — probed on a pty, which accepts the
-    # ioctl without caring about the speed. Without this every off-nominal
-    # probe would abort the walk on the platform --scan matters most on.
-    if os.name == "posix":
-        import pty
-
-        master, slave = pty.openpty()
-        try:
-            port = pb.Port(os.ttyname(slave), pb.scan_rate(9600, 4))
-            port.set_baud(pb.scan_rate(9600, -4))
-            port.close()
-        except pb.Error as error:
-            fail(f"PosixPort refused an off-nominal probe rate: {error}")
-        finally:
-            os.close(master)
-            os.close(slave)
-
-    print("OK")
-
-
-if __name__ == "__main__":
-    main()

test/test_update_link.py

@@ -1,167 +0,0 @@
-#!/usr/bin/env python3
-"""Self-update across a link change: the host must follow the staging copy.
-
-`--update-loader` installs the new image in the staging slot and then *enters
-it* to have it rewrite the resident. That copy is the new image, so it speaks the
-new image's baud and backend — but the host was talking to the *resident*. Where
-the two differ, the host kept knocking at the old rate in the old mode, the
-staging copy never answered, and the update stranded: staging installed, resident
-untouched, and on a 1 KiB tiny the application region (which *is* the staging
-slot there) already gone.
-
-The wire cannot be probed for this — 512 bytes of position-independent code carry
-no header saying what rate they were built for — so the operator declares it, and
-a mismatch with nothing declared has to say so instead of reporting a bare
-timeout.
-
-Stdlib only, no device: host-tool logic, so it runs on every chip's preset beside
-pureboot.planner.
-"""
-import importlib.util
-import pathlib
-
-PB = pathlib.Path(__file__).resolve().parents[1] / "pureboot" / "pureboot.py"
-_spec = importlib.util.spec_from_file_location("pureboot", PB)
-pb = importlib.util.module_from_spec(_spec)
-_spec.loader.exec_module(pb)
-
-IDENTITY = b"\x05\x1e\x95\x0f"  # pureboot 5 + m328p signature
-P = F = 0
-
-
-def check(name, ok, detail=""):
-    global P, F
-    P, F = P + (1 if ok else 0), F + (0 if ok else 1)
-    print(f"  [{'PASS' if ok else 'FAIL'}] {name}" + (f" — {detail}" if detail else ""))
-
-
-class TwoLinkPort:
-    """A board whose resident and staging copy answer on different links.
-
-    Only the rate currently set decides who can be heard, which is the physical
-    truth: a loader's bit timing is a cycle count, so a copy built for another
-    rate is unreadable until the host retunes. The knock bytes carry the mode, so
-    a backend mismatch is caught the same way.
-    """
-
-    def __init__(self, resident=(57600, False), staged=(38400, False)):
-        self.resident, self.staged = resident, staged
-        self.baud = resident[0]
-        self.entered = False        # a 'J' has handed control to the staging copy
-        self.switches = []          # every retune the host asked for
-        self.pending = bytearray()  # what the device has queued to send
-
-    # --- the part under test needs this to exist at all
-    def set_baud(self, baud):
-        self.baud = baud
-        self.switches.append(baud)
-
-    def flush_input(self):
-        self.pending.clear()
-
-    def _audible(self, knock=None):
-        baud, autobaud = self.staged if self.entered else self.resident
-        if self.baud != baud:
-            return False
-        if knock is None:
-            return True
-        return knock == (bytes((pb.CALIBRATE, ord("p"))) if autobaud else b"pb")
-
-    def write(self, data):
-        data = bytes(data)
-        if data[:1] == b"J" and len(data) == 3:
-            # The resident acks the jump, then control moves to the copy.
-            if self._audible():
-                self.pending += pb.PROMPT
-            self.entered = True
-        elif data in (b"pb", bytes((pb.CALIBRATE, ord("p")))):
-            if self._audible(data):
-                self.pending += pb.PROMPT
-        elif data == b"b":
-            if self._audible():
-                self.pending += IDENTITY + pb.PROMPT
-
-    def read_available(self, wait):
-        out, self.pending = bytes(self.pending), bytearray()
-        return out
-
-    def read_exact(self, count, timeout):
-        if len(self.pending) < count:
-            raise pb.Error(f"timeout: got {len(self.pending)} of {count} bytes")
-        out, self.pending = bytes(self.pending[:count]), self.pending[count:]
-        return out
-
-
-def connected(port):
-    """A Loader already in session with the resident."""
-    loader = pb.Loader(port)
-    loader.connect(2.0)
-    return loader
-
-
-def main():
-    # The control first: where the staged image keeps the resident's link, the
-    # flow works and needs no retune. This is the case that always passed, and
-    # it is what made the bug look like "self-update is broken" rather than
-    # "self-update cannot change the link".
-    port = TwoLinkPort(resident=(57600, False), staged=(57600, False))
-    loader = connected(port)
-    try:
-        loader.enter_copy(0x7C00, 2.0)
-        check("same link: staging copy entered", True)
-    except pb.Error as error:
-        check("same link: staging copy entered", False, str(error))
-
-    # A baud change, declared. The host must retune before knocking.
-    port = TwoLinkPort(resident=(57600, False), staged=(38400, False))
-    loader = connected(port)
-    try:
-        loader.enter_copy(0x7C00, 2.0, link=(38400, False))
-        check("baud change declared: entered after retuning", 38400 in port.switches,
-              f"switches={port.switches}")
-    except (pb.Error, TypeError) as error:
-        check("baud change declared: entered after retuning", False, repr(error))
-
-    # A backend change, declared: the knock itself has to become the calibration
-    # pulse, or an autobaud staging copy never hears a thing.
-    port = TwoLinkPort(resident=(57600, False), staged=(57600, True))
-    loader = connected(port)
-    try:
-        loader.enter_copy(0x7C00, 2.0, link=(57600, True))
-        check("backend change declared: entered as autobaud", True)
-    except (pb.Error, TypeError) as error:
-        check("backend change declared: entered as autobaud", False, repr(error))
-
-    # Nothing declared against a changed link: it still cannot work, but the
-    # error has to name the cause. A bare "no answer" sent the operator looking
-    # at the wiring while the application region sat erased.
-    port = TwoLinkPort(resident=(57600, False), staged=(38400, False))
-    loader = connected(port)
-    try:
-        loader.enter_copy(0x7C00, 0.3)
-        check("undeclared mismatch: reported", False, "unexpectedly succeeded")
-    except pb.Error as error:
-        text = str(error).lower()
-        check("undeclared mismatch: error names the link, not just a timeout",
-              "link" in text or "baud" in text or "backend" in text, str(error))
-    except TypeError as error:
-        check("undeclared mismatch: error names the link, not just a timeout",
-              False, repr(error))
-
-    # The resident's own link must be restored for the caller: a declared
-    # staging link is for the copy, and the tool talks to the new resident after.
-    port = TwoLinkPort(resident=(57600, False), staged=(38400, False))
-    loader = connected(port)
-    try:
-        loader.enter_copy(0x7C00, 2.0, link=(38400, False))
-        check("session records the link it is now speaking", loader.baud == 38400,
-              f"loader.baud={getattr(loader, 'baud', None)}")
-    except (pb.Error, TypeError, AttributeError) as error:
-        check("session records the link it is now speaking", False, repr(error))
-
-    print(f"\n  {P} passed, {F} failed")
-    return 1 if F else 0
-
-
-if __name__ == "__main__":
-    raise SystemExit(main())

tools/check.sh

@@ -2,14 +2,12 @@
 # The port's gate: every chip's generated workflow — build, size matrix, and
 # the simulator-driven protocol suites. --full adds the reflect-spot builds
 # (libavr's rule: reflect compiles are bounded to its spot set, never the
-# full matrix) and swaps the compact size matrix for the exhaustive
+# full matrix). LIBAVR_ROOT must point at the libavr checkout.
-# clock × baud × backend cross product. libavr resolves from the `libavr/`
-# submodule; LIBAVR_ROOT overrides it for a working tree.
 set -e
 cd "$(dirname "$0")/.."
 
 full=0
-[[ "$1" == "--full" ]] && { full=1; shift; export PUREBOOT_FULL_MATRIX=1; }
+[[ "$1" == "--full" ]] && { full=1; shift; }
 
 CHIPS=(attiny13 attiny13a attiny25 attiny45 attiny85
        atmega8 atmega8a atmega16 atmega16a atmega32 atmega32a
@@ -36,10 +34,4 @@ if ((full)); then
 	done
 fi
 
-# Every tree is freshly built now — the one moment the README's size table
-# can be held to what the images measure (a per-preset ctest sees only its
-# own chip; the table needs all of them, and ungated it drifts: a
-# common-code shave moves every row at once with nothing over budget).
-python3 tools/sizes.py check-readme
-
 echo "check: every chip green"

tools/make_presets.py

@@ -7,14 +7,11 @@ port's TUs compile identically; the sims prove nothing new there) exist for
 libavr's reflect spot set only, mirroring its rule: the full reflect matrix
 is never built, one chip per hardware class and pack vintage is.
 
-Run from the repo root: tools/make_presets.py — or with --check, which
+Run from the repo root: tools/make_presets.py
-verifies the committed file matches this generator and edits nothing (the
-ctest entry `presets.generated` runs that, so drift reds the gate).
 """
 
 import json
 import os
-import sys
 
 CHIPS = [
     "attiny13", "attiny13a", "attiny25", "attiny45", "attiny85",
@@ -44,7 +41,7 @@ def main():
         "hidden": True,
         "generator": "Ninja",
         "binaryDir": "${sourceDir}/build/${presetName}",
-        "toolchainFile": "${sourceDir}/libavr/cmake/avr-toolchain.cmake",
+        "toolchainFile": "$env{LIBAVR_ROOT}/cmake/avr-toolchain.cmake",
         "cacheVariables": {
             "CMAKE_BUILD_TYPE": "Release",
             "CMAKE_EXPORT_COMPILE_COMMANDS": "ON",
@@ -75,9 +72,6 @@ def main():
     for chip in REFLECT_SPOT:
         add(chip, "reflect")
 
-    # CMake rejects unknown fields in the presets root, $comment included, so
-    # the file cannot carry a generated-file marker; the --check ctest is the
-    # whole of rule 10's guard here.
     presets = {
         "version": 8,
         "configurePresets": configure,
@@ -85,19 +79,12 @@ def main():
         "testPresets": test,
         "workflowPresets": workflows,
     }
-    rendered = json.dumps(presets, indent=1) + "\n"
     path = os.path.join(os.path.dirname(os.path.abspath(__file__)), "..", "CMakePresets.json")
-    if "--check" in sys.argv[1:]:
-        current = open(path).read() if os.path.exists(path) else ""
-        if current != rendered:
-            print("CMakePresets.json does not match its generator — run tools/make_presets.py")
-            return 1
-        return 0
     with open(path, "w") as f:
-        f.write(rendered)
+        json.dump(presets, f, indent=1)
+        f.write("\n")
     print(f"{len(CHIPS)} chips, {len(REFLECT_SPOT)} reflect: {os.path.normpath(path)}")
-    return 0
 
 
 if __name__ == "__main__":
-    sys.exit(main())
+    main()

tools/pbhw.py

@@ -1,265 +0,0 @@
-#!/usr/bin/env python3
-"""Hardware acceptance suite for a pureboot deployment.
-
-`tools/check.sh` proves the protocol under simavr on every chip. This proves one
-*board*: that the loader actually installed on it answers, that the memories
-round-trip over the real link, that the application it flashes runs afterwards,
-and that the refusals which keep a 512-byte slot alive still fire. Run it once
-when a board is brought up, and again whenever the deployment moves — a new
-clock, a new backend, new pins.
-
-Every check derives its bounds from the info block the loader itself reports, so
-nothing here is per-chip: the same run covers a 1 KiB tiny whose application
-region is 510 usable bytes and a 128 KiB mega whose flash needs a bank in the
-selector.
-
-**This overwrites the board's application flash and EEPROM.** Capture them first
-with `pbrig.py backup`, which verifies what it captured.
-
-    tools/pbhw.py --programmer atmelice_isp --part t13 --port COM6 \
-                  --autobaud --loader build/ab.bin --app build/pbapp.hex \
-                  --marker APP
-"""
-
-from __future__ import annotations
-
-import argparse
-import pathlib
-import sys
-import tempfile
-
-sys.path.insert(0, str(pathlib.Path(__file__).resolve().parent))
-import pbrig  # noqa: E402
-
-
-class Suite:
-    def __init__(self, rig: pbrig.Rig, work: pathlib.Path):
-        self.rig = rig
-        self.work = work
-        self.results: list[tuple[str, bool, str]] = []
-
-    def check(self, name: str, ok: bool, detail: str = "") -> bool:
-        self.results.append((name, ok, detail))
-        print(f"  {'PASS' if ok else 'FAIL'}  {name}" + (f"   {detail}" if detail else ""))
-        return ok
-
-    @staticmethod
-    def _brief(text: str, limit: int = 78) -> str:
-        return " | ".join(l.strip() for l in text.splitlines() if l.strip())[:limit]
-
-    # ----------------------------------------------------------------- checks
-
-    def identity(self) -> object | None:
-        """The info block, which every later check takes its bounds from."""
-        module = pbrig.load_pureboot(self.rig.d.pureboot)
-        self.rig.reset()
-        port = self.rig.open_port()  # wrapped for the echo where the line is shared
-        try:
-            loader = module.Loader(port)
-            if self.rig.d.autobaud:
-                loader.connect_autobaud(self.rig.d.wait)
-            else:
-                loader.connect(self.rig.d.wait)
-            info = loader.info
-            self.check("identity read", True, info.describe())
-            return info
-        except Exception as error:  # noqa: BLE001 — a dead link is a result
-            self.check("identity read", False, str(error)[:70])
-            return None
-        finally:
-            try:
-                port.close()
-            except Exception:
-                pass
-
-    def scan(self) -> None:
-        """The --scan walk against real termios and a real oscillator: every
-        probe rate must open a port (the off-nominal rates exist only through
-        termios2), and one probe must answer — the nominal on a healthy board,
-        a neighbor on a drifted one. The rig injects the one reset per probe
-        the operator supplies in the field; this is the rate physics the
-        simulator cannot arbitrate (a pty carries bytes at any rate), pinned
-        on silicon."""
-        module = pbrig.load_pureboot(self.rig.d.pureboot)
-        found = None
-        try:
-            for pct in module.scan_ratios():
-                rate = module.scan_rate(self.rig.d.baud, pct)
-                self.rig.reset()
-                try:
-                    # Same wrap as identity(): on a shared line an undiscarded
-                    # echo answers every rate a scan probes, so the walk would
-                    # report the first one it tried.
-                    port = self.rig.open_port(rate)
-                except module.Error as error:
-                    self.check("scan opens every probe rate", False, f"{rate} Bd: {error}")
-                    return
-                try:
-                    module.Loader(port).connect(min(self.rig.d.wait, 6.0))
-                    found = pct
-                    break
-                except module.Error:
-                    continue
-                finally:
-                    port.close()
-        except Exception as error:  # noqa: BLE001 — a rig hiccup is a result
-            self.check("scan walks the probe ladder", False, str(error)[:70])
-            return
-        self.check("scan finds the board's rate", found is not None,
-                   "no probe answered" if found is None else f"{found:+d} % of {self.rig.d.baud} Bd")
-
-    def eeprom(self, info) -> None:
-        size = info.eeprom_size
-        if not size:
-            print("  skip  EEPROM (this part has none)")
-            return
-        # A pattern no erase or partial write could produce by accident.
-        pattern = bytes((i * 7 + 3) & 0xFF for i in range(size))
-        image = self.work / "ee.bin"
-        image.write_bytes(pattern)
-
-        rc, out = self.rig.pureboot("--eeprom", str(image), "--verify-eeprom", str(image))
-        self.check(f"EEPROM write + verify ({size} B)", rc == 0, self._brief(out))
-
-        back = self.work / "ee-back.bin"
-        rc, out = self.rig.pureboot("--read-eeprom", str(back))
-        got = back.read_bytes() if back.exists() else b""
-        self.check("EEPROM reads back what was written", got == pattern, f"{len(got)} B")
-
-        self.rig.pureboot("--erase-eeprom")
-        erased = self.work / "ee-erased.bin"
-        self.rig.pureboot("--read-eeprom", str(erased))
-        got = erased.read_bytes() if erased.exists() else b""
-        self.check("EEPROM erase leaves 0xff", got == b"\xff" * size, f"{len(got)} B")
-
-    def application(self, info, app: pathlib.Path, marker: str,
-                    marker_wait: float = 2.5) -> None:
-        rc, out = self.rig.pureboot("--flash", str(app), "--verify-flash", str(app))
-        self.check(f"application flash + verify ({app.name})", rc == 0, self._brief(out))
-
-        if marker:
-            # The tool hands over as it ends its session, so the application is
-            # already running — but only on a board whose DTR is unwired, where
-            # opening a port simply listens. Where DTR *is* wired to reset (an
-            # Arduino, most USB-serial dev boards), this open resets the part
-            # and the activation window comes first, so a marker emitted once at
-            # startup happens on the far side of a wait this cannot know the
-            # length of: the window is a compile-time constant and nothing on
-            # the wire reports it. Hence --marker-wait, and a fixture that
-            # repeats its banner (PUREBOOT_HEARTBEAT) rather than saying it once.
-            data = self.rig.capture(seconds=marker_wait)
-            seen = marker.encode() in data
-            sample = "".join(chr(b) if 32 <= b < 127 else "." for b in data[:40])
-            self.check(f"application runs (emits {marker!r})", seen,
-                       f"|{sample}|" if seen or data else
-                       f"nothing in {marker_wait:g} s — if this board resets when its port "
-                       f"opens, that wait has to outlast the activation window")
-
-        back = self.work / "app-back.bin"
-        rc, out = self.rig.pureboot("--read-flash", str(back))
-        got = back.read_bytes() if back.exists() else b""
-        self.check("application flash reads back", rc == 0 and len(got) == info.base,
-                   f"{len(got)} B of {info.base}")
-
-    def erase_and_guard(self, info, loader_image: pathlib.Path | None) -> None:
-        rc, out = self.rig.pureboot("--erase-flash")
-        self.check("application region erases", rc == 0, self._brief(out))
-
-        # The slot must be untouched by an application erase, which only an
-        # independent read can show — so this one goes over ISP, not the link.
-        whole = self.work / "whole.bin"
-        if not self.rig.read_memory("flash", whole, "r"):
-            self.check("loader slot survives the erase", False, "ISP read failed")
-            return
-        image = whole.read_bytes()
-        image += b"\xff" * (info.flash_size - len(image))
-
-        # Erased application flash, up to the trampoline word the host composes
-        # on a patched-vector part.
-        limit = info.base - 2 if info.patch_vector else info.base
-        self.check("erased application region is 0xff",
-                   set(image[0:limit]) <= {0xFF}, f"0x0000..{limit:#06x}")
-
-        if loader_image and loader_image.exists():
-            want = loader_image.read_bytes()
-            got = image[info.base:info.base + len(want)]
-            self.check("loader slot survives the erase", got == want,
-                       f"{len(want)} B at {info.base:#06x}")
-        else:
-            print("  skip  loader slot comparison (pass --loader <image.bin>)")
-
-    def refusals(self, info) -> None:
-        # One word too many: a patched-vector part spends the slot's last word
-        # on the trampoline, so its application stops two bytes short.
-        limit = info.base - 2 if info.patch_vector else info.base
-        oversized = self.work / "oversized.bin"
-        oversized.write_bytes(bytes(limit + 2))
-        rc, out = self.rig.pureboot("--flash", str(oversized))
-        self.check(f"image over {limit} B refused", rc != 0, self._brief(out))
-
-    # ------------------------------------------------------------------- run
-
-    def run(self, app: pathlib.Path | None, loader_image: pathlib.Path | None,
-            marker: str, marker_wait: float = 2.5) -> int:
-        print("identity")
-        info = self.identity()
-        if info is None:
-            print("\nthe loader never answered; nothing below can be trusted")
-            return 1
-
-        if not self.rig.d.autobaud:
-            print("\nscan")
-            self.scan()
-
-        print("\nEEPROM")
-        self.eeprom(info)
-
-        if app:
-            print("\napplication")
-            self.application(info, app, marker, marker_wait)
-        else:
-            print("\nskip  application checks (pass --app <image.hex>)")
-
-        print("\nerase and the write guard")
-        self.erase_and_guard(info, loader_image)
-
-        print("\nrefusals")
-        self.refusals(info)
-
-        passed = sum(1 for _, ok, _ in self.results if ok)
-        print(f"\n{passed}/{len(self.results)} passed")
-        return 0 if passed == len(self.results) else 1
-
-
-def main(argv: list[str] | None = None) -> int:
-    parser = argparse.ArgumentParser(
-        description="hardware acceptance suite for one pureboot deployment",
-        epilog="overwrites the board's application flash and EEPROM — back them up first")
-    pbrig.Deployment.add_arguments(parser)
-    parser.add_argument("--app", type=pathlib.Path,
-                        help="application image to flash (test/pbapp.cpp built for this deployment)")
-    parser.add_argument("--loader", type=pathlib.Path,
-                        help="the resident loader's .bin, to prove the slot survives an erase")
-    parser.add_argument("--marker", default="",
-                        help="text the application emits when it runs, e.g. APP")
-    parser.add_argument("--marker-wait", type=float, default=2.5,
-                        help="seconds to listen for it. On a board whose DTR is wired to "
-                             "reset, opening the port resets the part, so this must outlast "
-                             "the activation window (default 2.5)")
-    args = parser.parse_args(argv)
-
-    rig = pbrig.Rig(pbrig.Deployment.from_args(args))
-    print(f"rig: {args.part} on {args.programmer}, link {args.port} at {args.baud} Bd"
-          f"{' (autobaud)' if args.autobaud else ''}")
-    print("this overwrites the application flash and EEPROM\n")
-    with tempfile.TemporaryDirectory(prefix="pbhw-") as temporary:
-        return Suite(rig, pathlib.Path(temporary)).run(args.app, args.loader, args.marker,
-                                                      args.marker_wait)
-
-
-if __name__ == "__main__":
-    try:
-        sys.exit(main())
-    except pbrig.Error as error:
-        print(f"error: {error}", file=sys.stderr)
-        sys.exit(2)

tools/pbrig.py

@@ -1,450 +0,0 @@
-#!/usr/bin/env python3
-"""Hardware rig driver for pureboot: an ISP programmer beside a serial link.
-
-The simulated suites (`test/pb*.py`) prove the protocol; this drives the same
-loader on real silicon, where the things a cycle-exact simulator cannot model
-live — an RC oscillator off its nominal, a reset edge that has to come from
-somewhere, a serial bridge with its own idea of what a baud is.
-
-Nothing here knows a port name, a part or a programmer. Every deployment fact
-arrives from the command line or the environment, so the same script serves any
-board: see `Deployment`. As a module it is the reset/flash/talk primitives that
-`pbhw.py` builds its acceptance suite from; as a command it is the handful of
-one-shot operations worth having on a rig — most importantly `backup`, which is
-the only thing standing between a fuse experiment and an unrecoverable part.
-
-Two rig facts are encoded here because they are not guessable and cost a
-session each to learn:
-
-* **An ISP access resets the part**, and it runs again the moment the programmer
-  releases it. That is the only reset edge available when the serial adapter's
-  DTR is not wired to reset — so a loader session begins with an ISP touch and
-  knocks immediately after, which is what `Rig.pureboot()` does.
-* **avrdude splits `-U memory:op:file:format` on colons**, so a Windows path's
-  drive letter breaks the spec. Every file argument is therefore passed as a
-  bare filename with avrdude run in that file's own directory.
-"""
-
-from __future__ import annotations
-
-import argparse
-import dataclasses
-import importlib.util
-import os
-import pathlib
-import subprocess
-import sys
-import time
-
-HERE = pathlib.Path(__file__).resolve().parent
-DEFAULT_PUREBOOT = HERE.parent / "pureboot" / "pureboot.py"
-
-# Memories worth capturing before an experiment, and the format each is read in.
-# Fuses and lock are per-part: a part without an extended fuse simply fails that
-# one read, which `backup` reports and steps over rather than aborting on.
-BACKUP_MEMORIES = (
-    ("flash", "i", "hex"),
-    ("flash", "r", "bin"),
-    ("eeprom", "i", "hex"),
-    ("eeprom", "r", "bin"),
-    ("lfuse", "h", "hex"),
-    ("hfuse", "h", "hex"),
-    ("efuse", "h", "hex"),
-    ("lock", "h", "hex"),
-    ("calibration", "h", "hex"),
-)
-
-
-class Error(Exception):
-    pass
-
-
-def bitclock_for(hz: int) -> str:
-    """A safe ISP bitclock for a part *currently running* at `hz`.
-
-    SCK must stay under a quarter of the target clock, so the bitclock follows
-    the clock in force — not the one about to be fused in. Halving that ceiling
-    again costs nothing on a link that moves a few hundred bytes and buys margin
-    against an oscillator that is already known to be off its nominal.
-    """
-    ceiling = hz // 8
-    for candidate in (1000, 4000, 8000, 32000, 125000, 400000):
-        if candidate <= ceiling:
-            best = candidate
-        else:
-            break
-    else:
-        best = 400000
-    if ceiling < 1000:
-        raise Error(f"a part at {hz} Hz is too slow to reach over ISP safely")
-    return f"{best // 1000}kHz"
-
-
-@dataclasses.dataclass
-class Deployment:
-    """Everything about one board. No default names a real device."""
-
-    port: str = ""              # serial device the loader speaks on
-    baud: int = 57600           # host rate; for autobaud, the rate to drive
-    autobaud: bool = False      # send the calibration pulse instead of p+b
-    one_wire: bool = False      # shared line: the host discards its own echo
-    programmer: str = ""        # avrdude -c
-    part: str = ""              # avrdude -p
-    avrdude: str = "avrdude"
-    bitclock: str = "125kHz"    # see bitclock_for()
-    pureboot: pathlib.Path = DEFAULT_PUREBOOT
-    wait: int = 12              # seconds the host keeps knocking
-
-    @classmethod
-    def from_env(cls) -> "Deployment":
-        """Environment defaults, so a rig's facts live in one place per machine."""
-        return cls(
-            port=os.environ.get("PUREBOOT_PORT", ""),
-            baud=int(os.environ.get("PUREBOOT_BAUD", "57600")),
-            autobaud=os.environ.get("PUREBOOT_AUTOBAUD", "") not in ("", "0"),
-            one_wire=os.environ.get("PUREBOOT_ONE_WIRE", "") not in ("", "0"),
-            programmer=os.environ.get("PUREBOOT_PROGRAMMER", ""),
-            part=os.environ.get("PUREBOOT_PART", ""),
-            avrdude=os.environ.get("AVRDUDE", "avrdude"),
-            bitclock=os.environ.get("PUREBOOT_BITCLOCK", "125kHz"),
-            pureboot=pathlib.Path(os.environ.get("PUREBOOT_TOOL", str(DEFAULT_PUREBOOT))),
-        )
-
-    @staticmethod
-    def add_arguments(parser: argparse.ArgumentParser) -> None:
-        """Deployment flags, shared by this tool and pbhw.py."""
-        env = Deployment.from_env()
-        parser.add_argument("--port", default=env.port, help="serial device the loader speaks on")
-        parser.add_argument("--baud", type=int, default=env.baud,
-                            help="host rate (for autobaud, the rate to drive)")
-        parser.add_argument("--autobaud", action="store_true", default=env.autobaud,
-                            help="send the calibration pulse instead of the p+b knock")
-        parser.add_argument("--one-wire", action="store_true", default=env.one_wire,
-                            help="shared line: pass the tool its echo discard")
-        parser.add_argument("--programmer", default=env.programmer, help="avrdude -c, e.g. atmelice_isp")
-        parser.add_argument("--part", default=env.part, help="avrdude -p, e.g. t13 or m328p")
-        parser.add_argument("--avrdude", default=env.avrdude, help="path to avrdude")
-        parser.add_argument("--bitclock", default=env.bitclock, help="ISP bitclock, e.g. 125kHz or 8kHz")
-        parser.add_argument("--pureboot", type=pathlib.Path, default=env.pureboot,
-                            help="path to pureboot.py")
-        parser.add_argument("--wait", type=int, default=env.wait, help="seconds to keep knocking")
-
-    @classmethod
-    def from_args(cls, args: argparse.Namespace) -> "Deployment":
-        return cls(port=args.port, baud=args.baud, autobaud=args.autobaud,
-                   one_wire=args.one_wire,
-                   programmer=args.programmer, part=args.part, avrdude=args.avrdude,
-                   bitclock=args.bitclock, pureboot=args.pureboot, wait=args.wait)
-
-
-def load_pureboot(path: pathlib.Path = DEFAULT_PUREBOOT):
-    """The host tool as a module — its Port and Loader, not a subprocess.
-
-    Used where a subprocess cannot express what is needed: a poke followed by a
-    peek in the *same* session, or a raw read at an arbitrary baud.
-    """
-    spec = importlib.util.spec_from_file_location("pureboot", path)
-    if spec is None or spec.loader is None:
-        raise Error(f"cannot load the host tool from {path}")
-    module = importlib.util.module_from_spec(spec)
-    spec.loader.exec_module(module)
-    return module
-
-
-class Rig:
-    """One board: its programmer on one side, its serial link on the other."""
-
-    def __init__(self, deployment: Deployment):
-        self.d = deployment
-        if not deployment.programmer or not deployment.part:
-            raise Error("a rig needs --programmer and --part")
-
-    # ------------------------------------------------------------- programmer
-
-    def avrdude(self, *args: str, cwd: pathlib.Path | None = None,
-                bitclock: str | None = None, timeout: int = 300) -> subprocess.CompletedProcess:
-        command = [self.d.avrdude, "-c", self.d.programmer, "-p", self.d.part,
-                   "-B", bitclock or self.d.bitclock, *args]
-        return subprocess.run(command, capture_output=True, text=True,
-                              cwd=None if cwd is None else str(cwd), timeout=timeout)
-
-    @staticmethod
-    def _ok(result: subprocess.CompletedProcess) -> bool:
-        return result.returncode == 0
-
-    def reset(self, bitclock: str | None = None) -> None:
-        """An ISP access, which resets the part; it runs when avrdude exits."""
-        self.avrdude("-U", "signature:r:-:h", bitclock=bitclock)
-
-    def signature(self, bitclock: str | None = None) -> str:
-        result = self.avrdude("-U", "signature:r:-:h", bitclock=bitclock)
-        for line in reversed(result.stdout.splitlines()):
-            if line.strip().startswith("0x"):
-                return line.strip()
-        raise Error(f"no signature read: {(result.stderr or result.stdout).strip()[:200]}")
-
-    def read_memory(self, memory: str, destination: pathlib.Path, fmt: str = "r",
-                    bitclock: str | None = None) -> bool:
-        """Read `memory` into `destination`, whose directory avrdude runs in."""
-        destination = pathlib.Path(destination).resolve()
-        destination.parent.mkdir(parents=True, exist_ok=True)
-        result = self.avrdude("-U", f"{memory}:r:{destination.name}:{fmt}",
-                              cwd=destination.parent, bitclock=bitclock)
-        # A memory the part does not have (a tiny's extended fuse) leaves avrdude
-        # happy and the file empty. An empty capture is a miss, not a backup.
-        return self._ok(result) and destination.exists() and destination.stat().st_size > 0
-
-    def write_memory(self, memory: str, source: pathlib.Path, fmt: str = "i",
-                     erase: bool = False, bitclock: str | None = None) -> bool:
-        source = pathlib.Path(source).resolve()
-        args = ["-U", f"{memory}:w:{source.name}:{fmt}"]
-        if erase:
-            args.insert(0, "-e")
-        result = self.avrdude(*args, cwd=source.parent, bitclock=bitclock)
-        return "verified" in (result.stdout + result.stderr)
-
-    def flash_hex(self, image: pathlib.Path, erase: bool = True,
-                  bitclock: str | None = None) -> bool:
-        return self.write_memory("flash", image, "i", erase=erase, bitclock=bitclock)
-
-    def read_fuses(self, bitclock: str | None = None) -> dict[str, str]:
-        out: dict[str, str] = {}
-        for fuse in ("lfuse", "hfuse", "efuse", "lock"):
-            result = self.avrdude("-U", f"{fuse}:r:-:h", bitclock=bitclock)
-            values = [l.strip() for l in result.stdout.splitlines() if l.strip().startswith("0x")]
-            if values:
-                out[fuse] = values[-1]
-        return out
-
-    def write_fuses(self, bitclock: str | None = None, **fuses: str) -> bool:
-        """Write named fuses. A fuse change moves the clock the *next* access is
-        timed against, so pass a bitclock safe for both sides of the change."""
-        args: list[str] = []
-        for name, value in fuses.items():
-            args += ["-U", f"{name}:w:{value}:m"]
-        if not args:
-            return True
-        result = self.avrdude(*args, bitclock=bitclock)
-        text = result.stdout + result.stderr
-        return "verified" in text or "written" in text
-
-    # ------------------------------------------------------------ backup
-
-    def backup(self, directory: pathlib.Path, prefix: str = "") -> dict[str, bool]:
-        """Capture every memory worth keeping, then prove it by a second read.
-
-        A backup nobody verified is a guess. Each memory is read twice and the
-        two reads compared; a mismatch is reported rather than quietly stored.
-        """
-        directory = pathlib.Path(directory).resolve()
-        directory.mkdir(parents=True, exist_ok=True)
-        stem = prefix or self.d.part
-        status: dict[str, bool] = {}
-        for memory, fmt, extension in BACKUP_MEMORIES:
-            name = f"{stem}-{memory}.{extension}"
-            if not self.read_memory(memory, directory / name, fmt):
-                status[f"{memory}.{extension}"] = False
-                continue
-            if extension == "bin":  # only the raw form is worth comparing byte-wise
-                again = directory / f".{name}.again"
-                self.read_memory(memory, again, fmt)
-                same = again.exists() and again.read_bytes() == (directory / name).read_bytes()
-                again.unlink(missing_ok=True)
-                status[f"{memory}.{extension}"] = same
-            else:
-                status[f"{memory}.{extension}"] = True
-        return status
-
-    # ------------------------------------------------------------ serial link
-
-    def pureboot(self, *args: str, reset_first: bool = True, baud: int | None = None,
-                 autobaud: bool | None = None, timeout: int = 300,
-                 bitclock: str | None = None) -> tuple[int, str]:
-        """Reset, then knock immediately — see the module docstring.
-
-        Returns the host tool's exit status and its combined output, so a caller
-        can assert on what it printed as well as on whether it succeeded.
-        """
-        if reset_first:
-            self.reset(bitclock=bitclock)
-        command = [sys.executable, str(self.d.pureboot), "--port", self.d.port,
-                   "--baud", str(self.d.baud if baud is None else baud),
-                   "--wait", str(self.d.wait)]
-        if self.d.autobaud if autobaud is None else autobaud:
-            command.append("--autobaud")
-        if self.d.one_wire:
-            command.append("--one-wire")
-        command += [str(a) for a in args]
-        try:
-            result = subprocess.run(command, capture_output=True, text=True, timeout=timeout)
-        except subprocess.TimeoutExpired as expired:
-            return 99, f"TIMEOUT after {timeout}s\n{expired.stdout or ''}{expired.stderr or ''}"
-        return result.returncode, (result.stdout or "") + (result.stderr or "")
-
-    def open_port(self, baud: int | None = None):
-        """A port opened the way this deployment says to speak to the board.
-
-        Everything the rig runs as a *subprocess* gets its flags from
-        `pureboot()` above; anything that drives the protocol in-process has
-        to reach the same facts, and until this existed only the subprocess
-        path could. A shared line is the one where that gap is fatal rather
-        than untidy: the host reads back every byte it writes, so an
-        undiscarded echo answers the knock before the device does. Open
-        through here and a one-wire deployment cannot be silently driven as
-        a two-wire one.
-        """
-        module = load_pureboot(self.d.pureboot)
-        port = module.Port(self.d.port, self.d.baud if baud is None else baud)
-        return module.OneWirePort(port) if self.d.one_wire else port
-
-    def capture(self, seconds: float = 2.0, baud: int | None = None) -> bytes:
-        """Listen to whatever the board is saying, at an arbitrary rate.
-
-        Opening the port does not reset a board whose DTR is unwired, so this can
-        sample a running application repeatedly without disturbing it — which is
-        what makes the rate sweep below possible.
-        """
-        module = load_pureboot(self.d.pureboot)
-        port = module.Port(self.d.port, self.d.baud if baud is None else baud)
-        try:
-            data = b""
-            deadline = time.monotonic() + seconds
-            while time.monotonic() < deadline:
-                chunk = port.read_available(0.2)
-                if chunk:
-                    data += chunk
-            return data
-        finally:
-            try:
-                port.close()
-            except Exception:
-                pass
-
-
-def measure_rate(rig: Rig, marker: bytes, built_baud: int, nominal_hz: int | None = None,
-                 span_percent: float = 12.0, step_percent: float = 0.5,
-                 seconds: float = 0.75) -> dict:
-    """Find a transmitting board's true bit rate, using only the serial port.
-
-    The board must be emitting something recognisable at a *fixed* cycles-per-bit
-    — `test/pbapp.cpp` built with PUREBOOT_HEARTBEAT does. Since its bit timing is
-    a cycle count, its wire rate scales with its actual clock, so the host rates
-    at which `marker` still decodes bracket that rate; the centre of the band is
-    the answer, and with the clock the image was built for it gives the real one.
-
-    This is the measurement that turns "the loader is silent, so the wiring must
-    be wrong" into a number, and it needs no instrument beyond the adapter
-    already attached.
-    """
-    steps = int(span_percent / step_percent)
-    clean: list[int] = []
-    samples: list[tuple[int, int, bool]] = []
-    for index in range(-steps, steps + 1):
-        baud = int(round(built_baud * (1 + index * step_percent / 100.0)))
-        if baud <= 0:
-            continue
-        data = rig.capture(seconds=seconds, baud=baud)
-        hit = marker in data
-        samples.append((baud, len(data), hit))
-        if hit:
-            clean.append(baud)
-    result: dict = {"samples": samples, "clean": clean, "built_baud": built_baud}
-    if clean:
-        low, high = min(clean), max(clean)
-        centre = (low + high) / 2.0
-        result |= {"low": low, "high": high, "centre": centre,
-                   "half_width_percent": (high - low) / 2.0 / centre * 100.0,
-                   "error_percent": (centre / built_baud - 1.0) * 100.0}
-        if nominal_hz:
-            result["measured_hz"] = nominal_hz * centre / built_baud
-    return result
-
-
-# ------------------------------------------------------------------- command
-
-
-def main(argv: list[str] | None = None) -> int:
-    parser = argparse.ArgumentParser(
-        description="pureboot hardware rig: ISP reset/flash beside the serial link")
-    Deployment.add_arguments(parser)
-    sub = parser.add_subparsers(dest="command", required=True)
-
-    sub.add_parser("signature", help="read the part signature over ISP")
-    sub.add_parser("reset", help="reset the part (an ISP access) and let it run")
-    sub.add_parser("fuses", help="read the fuse and lock bytes")
-
-    p = sub.add_parser("flash", help="program a hex image over ISP")
-    p.add_argument("image", type=pathlib.Path)
-    p.add_argument("--no-erase", action="store_true", help="do not chip-erase first")
-
-    p = sub.add_parser("backup", help="capture and verify every memory")
-    p.add_argument("directory", type=pathlib.Path)
-    p.add_argument("--prefix", default="", help="filename stem (default: the part name)")
-
-    p = sub.add_parser("rate", help="measure the board's true bit rate and clock")
-    p.add_argument("--marker", default="APP", help="text the board emits (default: APP)")
-    p.add_argument("--built-baud", type=int, required=True,
-                   help="the baud the running image was built for")
-    p.add_argument("--nominal-hz", type=int, default=0,
-                   help="the clock the image was built for, to report the real one")
-    p.add_argument("--span", type=float, default=12.0, help="sweep +-this many percent")
-    p.add_argument("--step", type=float, default=0.5, help="sweep step in percent")
-    p.add_argument("--verbose", action="store_true", help="print every step")
-
-    p = sub.add_parser("bitclock", help="a safe ISP bitclock for a clock in force")
-    p.add_argument("hz", type=int)
-
-    args = parser.parse_args(argv)
-
-    if args.command == "bitclock":
-        print(bitclock_for(args.hz))
-        return 0
-
-    rig = Rig(Deployment.from_args(args))
-
-    if args.command == "signature":
-        print(rig.signature())
-    elif args.command == "reset":
-        rig.reset()
-        print("reset")
-    elif args.command == "fuses":
-        for name, value in rig.read_fuses().items():
-            print(f"{name:<6} {value}")
-    elif args.command == "flash":
-        ok = rig.flash_hex(args.image, erase=not args.no_erase)
-        print(f"{args.image.name}: {'verified' if ok else 'FAILED'}")
-        return 0 if ok else 1
-    elif args.command == "backup":
-        status = rig.backup(args.directory, args.prefix)
-        for name, ok in status.items():
-            print(f"  {'ok  ' if ok else 'FAIL'} {name}")
-        missing = [n for n, ok in status.items() if not ok]
-        # Fuses a part does not have are expected misses, not failures.
-        fatal = [n for n in missing if not n.startswith(("efuse", "calibration"))]
-        print(f"\n{len(status) - len(missing)}/{len(status)} captured into {args.directory}")
-        return 1 if fatal else 0
-    elif args.command == "rate":
-        result = measure_rate(rig, args.marker.encode(), args.built_baud,
-                              args.nominal_hz or None, args.span, args.step)
-        if args.verbose:
-            for baud, size, hit in result["samples"]:
-                print(f"  {baud:7d} Bd  {size:5d} B  {'MARKER' if hit else ''}")
-        if not result["clean"]:
-            print(f"no capture contained {args.marker!r} at any rate — is the board "
-                  f"transmitting, and on the pin this port is wired to?")
-            return 1
-        print(f"clean band  {result['low']}..{result['high']} Bd")
-        print(f"centre      {result['centre']:.0f} Bd "
-              f"(+-{result['half_width_percent']:.1f} %)")
-        print(f"vs built    {result['built_baud']} Bd ({result['error_percent']:+.1f} %)")
-        if "measured_hz" in result:
-            print(f"true clock  {result['measured_hz'] / 1e6:.3f} MHz")
-    return 0
-
-
-if __name__ == "__main__":
-    try:
-        sys.exit(main())
-    except Error as error:
-        print(f"error: {error}", file=sys.stderr)
-        sys.exit(2)

tools/sizes.py

@@ -1,187 +0,0 @@
-#!/usr/bin/env python3
-"""What the loader images actually measure, and whether the README still agrees.
-
-The size matrix asserts every image fits its slot; it says nothing about the
-numbers the README prints, and those drift. Every row of that table was eight
-bytes stale once `startup::caller_page()` landed — common code, so every build
-moved at once and no test noticed, because none of them was over budget.
-
-Two questions, both answered from built trees:
-
-  sizes.py max            the largest image per chip, and anything over budget
-  sizes.py check-readme   the README's per-chip table against what is built
-
-Nothing here knows a chip's geometry. The (image, budget) pairs come from each
-build's own `CTestTestfile.cmake` — the same values the gate checks — so the
-slot rules stay where they belong, in `pureboot/CMakeLists.txt`, and a chip
-added or a budget changed needs no edit here. Only trees a configure preset
-still owns are read: a stale directory keeps its last build, and a loader built
-before a slot changed will happily report a size that was true once
-(`tools/prune-build-trees.sh` in libavr removes them).
-
-Sizes come from `avr-size`, and a target is only as current as its last build —
-run the gate first if you want the table checked against today's source.
-"""
-
-from __future__ import annotations
-
-import argparse
-import pathlib
-import re
-import shutil
-import subprocess
-import sys
-
-ROOT = pathlib.Path(__file__).resolve().parents[1]
-# add_test(<name>.size ... -DELF=<path> ... -DLIMIT=<n> ...) — the gate's own
-# pairing of an image with the budget it must fit.
-# ctest writes the name as a bracket argument ([=[name.size]=]) and quotes the
-# rest, so the name starts after the bracket and the path ends at the quote.
-SIZE_TEST = re.compile(r'add_test\(\s*\[=\[(?P<name>[^\]]+?)\.size\]=\][^\n]*?'
-                       r'-DELF=(?P<elf>[^"\s]+)[^\n]*?-DLIMIT=(?P<limit>\d+)')
-
-
-def avr_size() -> str:
-    for env in (ROOT / "../../toolchain").resolve().glob("avr-gcc-*/bin/avr-size"):
-        if env.is_file():
-            return str(env)
-    found = shutil.which("avr-size")
-    if not found:
-        sys.exit("no avr-size found (build the toolchain, or put it on PATH)")
-    return found
-
-
-def preset_dirs() -> list[pathlib.Path]:
-    """Build trees a configure preset still owns, newest-listed first."""
-    listing = subprocess.run(["cmake", "--list-presets"], cwd=ROOT, capture_output=True, text=True)
-    names = re.findall(r'^\s*"(.+)"$', listing.stdout, re.MULTILINE)
-    if not names:
-        sys.exit("cmake --list-presets returned nothing — run from a configured checkout")
-    return [d for d in (ROOT / "build" / n for n in names) if (d / "CTestTestfile.cmake").is_file()]
-
-
-def measure(paths: list[str], tool: str) -> dict[str, int]:
-    """.text per ELF, in one avr-size call per batch."""
-    sizes: dict[str, int] = {}
-    for start in range(0, len(paths), 400):
-        batch = [p for p in paths[start:start + 400] if pathlib.Path(p).is_file()]
-        if not batch:
-            continue
-        out = subprocess.run([tool, *batch], capture_output=True, text=True).stdout
-        for line in out.splitlines()[1:]:
-            fields = line.split()
-            if len(fields) >= 6 and fields[0].isdigit():
-                sizes[fields[5]] = int(fields[0])
-    return sizes
-
-
-def collect() -> dict[str, list[tuple[str, int, int]]]:
-    """chip -> [(target, text, limit)], from every owned build tree."""
-    tool = avr_size()
-    found: dict[str, list[tuple[str, str, int]]] = {}
-    for tree in preset_dirs():
-        chip = tree.name.split("-")[0]
-        for match in SIZE_TEST.finditer((tree / "CTestTestfile.cmake").read_text()):
-            found.setdefault(chip, []).append((match["name"], match["elf"], int(match["limit"])))
-    sizes = measure([elf for rows in found.values() for _, elf, _ in rows], tool)
-    # A chip's generated and reflect trees must answer with the same bytes
-    # (the identity invariant), so the same target measuring two sizes means
-    # a stale tree — or an identity breach. Either is a finding; picking one
-    # silently is how a gate reports another build's numbers as today's.
-    for chip, rows in found.items():
-        seen: dict[str, tuple[int, str]] = {}
-        for name, elf, _ in rows:
-            if elf not in sizes:
-                continue
-            if name in seen and seen[name][0] != sizes[elf]:
-                sys.exit(f"{chip} {name}: {seen[name][0]} B in {seen[name][1]} but "
-                         f"{sizes[elf]} B in {elf} — a stale tree (rebuild or remove it) "
-                         f"or a cross-mode identity breach")
-            seen.setdefault(name, (sizes[elf], elf))
-    measured = {
-        chip: sorted(((name, sizes[elf], limit) for name, elf, limit in rows if elf in sizes),
-                     key=lambda row: -row[1])
-        for chip, rows in sorted(found.items())
-    }
-    # A configured-but-unbuilt preset registers its tests with no images behind
-    # them; it is not a chip with nothing to say, it is a chip not built yet.
-    return {chip: rows for chip, rows in measured.items() if rows}
-
-
-def cmd_max(args) -> int:
-    measured = collect()
-    if not measured:
-        sys.exit("nothing built — configure and build a preset first")
-    over = []
-    print(f"{'chip':<13} {'largest image':<34} {'.text':>6} {'budget':>7}  headroom")
-    for chip, rows in measured.items():
-        name, text, limit = rows[0]
-        flag = "OVER" if text > limit else f"{limit - text:>5} B"
-        print(f"{chip:<13} {name:<34} {text:>6} {limit:>7}  {flag}")
-        over += [(chip, n, t, l) for n, t, l in rows if t > l]
-    total = sum(len(rows) for rows in measured.values())
-    print(f"\n{total} images across {len(measured)} chips")
-    if over:
-        print("\nOVER BUDGET:")
-        for chip, name, text, limit in over:
-            print(f"  {chip} {name}: {text} > {limit}")
-        return 1
-    tightest = min(((chip, n, t, l) for chip, rows in measured.items() for n, t, l in rows),
-                   key=lambda row: row[3] - row[2])
-    chip, name, text, limit = tightest
-    print(f"tightest fit: {chip} {name} — {text} of {limit}, {limit - text} B spare")
-    return 0
-
-
-def cmd_check_readme(args) -> int:
-    """The README's per-chip table, against the stock build and the worst
-    autobaud configuration (OSCCAL baked, plus the USART-pin release where
-    the chip has a USART; the one-wire fold of the same build is its twin
-    and competes for the same cell) — the config the Autobaud column
-    documents."""
-    readme = (ROOT / "pureboot" / "README.md").read_text()
-    measured = collect()
-    rows = re.findall(r"^\|\s*(AT\w+[^|]*?)\s*\|[^|]*\|[^|]*\|[^|]*\|\s*(\d+) B\s*\|\s*(\d+) B\s*\|$",
-                      readme, re.MULTILINE)
-    if not rows:
-        sys.exit("no size table found in pureboot/README.md")
-    bad = skipped = 0
-    for chips, stock_doc, auto_doc in rows:
-        # "ATmega48, 48A, 48P, 48PA †" — the first name is the family's base.
-        chip = re.sub(r"[^a-z0-9]", "", chips.split(",")[0].strip().lower())
-        built = {name: text for name, text, _ in measured.get(chip, [])}
-        # The on-USART pair defines the column where the chip has a USART;
-        # the default-pin pair is the whole space elsewhere. Whichever twin
-        # measures larger is the number the cell must state.
-        candidates = [name for name in ("pureboot_autobaud_osccal_on_usart0",
-                                        "pureboot_1w_autobaud_osccal_on_usart0") if name in built]
-        if not candidates:
-            candidates = [name for name in ("pureboot_autobaud_osccal",
-                                            "pureboot_1w_autobaud_osccal") if name in built]
-        worst = max(candidates, key=lambda name: built[name], default="pureboot_autobaud_osccal")
-        for target, documented in (("pureboot", stock_doc), (worst, auto_doc)):
-            if target not in built:
-                skipped += 1
-                continue
-            if built[target] != int(documented):
-                print(f"  {chip:<12} {target:<18} README says {documented} B, built is {built[target]} B")
-                bad += 1
-    if bad:
-        print(f"\n{bad} row(s) stale — update pureboot/README.md")
-        return 1
-    print(f"README size table matches every built image ({len(rows)} rows"
-          + (f", {skipped} not built" if skipped else "") + ")")
-    return 0
-
-
-def main() -> int:
-    parser = argparse.ArgumentParser(description=__doc__.splitlines()[0])
-    subs = parser.add_subparsers(dest="cmd", required=True)
-    subs.add_parser("max", help="largest image per chip, and anything over budget")
-    subs.add_parser("check-readme", help="the README's size table against what is built")
-    args = parser.parse_args()
-    return {"max": cmd_max, "check-readme": cmd_check_readme}[args.cmd](args)
-
-
-if __name__ == "__main__":
-    raise SystemExit(main())

tsb/tsb_asm.cpp

@@ -65,7 +65,7 @@ constexpr std::uint8_t comm_window = 200;
 constexpr std::uint16_t build_date = 26 * 512 + 7 * 32 + 20;
 
 // Fixed 115200 8N1; the library solves UBRR + U2X from clock and baud.
-constexpr auto baud = avr::uart::solve_baud(16_MHz, 115200_Bd);
+constexpr auto baud = avr::uart::detail::solve_baud(16_MHz, 115200_Bd);
 
 // The 16-byte device-info block, streamed out on activation.
 // clang-format off
@@ -263,7 +263,7 @@ extern "C" [[noreturn]] void tsb_app(); // the application's reset vector: --def
 	// 0, and rx()/tx() raise RXEN0/TXEN0 on first use — only the divisor low
 	// byte and U2X0 need a store. The library still does the datasheet work.
 	static_assert(baud.u2x && baud.ubrr < 256, "lean bring-up writes UBRR0L only, with U2X0");
-	hw::ubrr0::write(static_cast<std::uint8_t>(baud.ubrr));
+	hw::reg<"UBRR0">::write(static_cast<std::uint8_t>(baud.ubrr));
 	hw::ucsr0a::write(hw::ucsr0a::u2x0(1));
 	// General-purpose registers are undefined at power-on (no crt zeroes them);
 	// the direction latch must start "not receiving" so the first rx() enables

tsb/tsb_policy.cpp

@@ -1,302 +0,0 @@
-// TinySafeBoot on libavr — the policy floor: pureboot's rules, measured.
-//
-// The full TinySafeBoot feature set — watchdog bail, one-wire half-duplex,
-// config-page activation timeout, password gate, emergency erase, and
-// config/flash/EEPROM read-write — under philosophy #5 exactly as pureboot
-// obeys it: no assembly, no register variables; code, attributes, and flags
-// only. Every lesson pureboot's development produced is applied — the
-// library's half-duplex serial and startup entry, lean bring-up from reset
-// state, one merged send loop over both memories, oracle-shaped loop bounds,
-// locals threaded through noinline primitives, pureboot's codegen flags —
-// and the result is 638 bytes: 198 below the idiomatic tier, and 126 above
-// the 512 B boot section the tricks/asm tiers reach with the banned
-// mechanisms (526/510). This tier exists to keep that number an artifact
-// rather than a claim: the gap to 512 is the rent of policy-clean C++ —
-// helpers that hold a cursor across rx()/tx() pay push/pop and argument
-// threading where a global-register protocol pays nothing, and both
-// control-flow merges tried (a parametrized paged session, a merged store
-// loop) measured larger than the split cases they replaced. TSB's wire fixes
-// the per-command loop shapes on the device, so pureboot 5's one-transfer-
-// loop collapse has no purchase here.
-//
-// The wire protocol is strict request/response, which is what makes the
-// shared line safe: the device drives it only between a received command and
-// its reply, and releases it (the library's half-duplex choreography)
-// whenever it waits.
-
-#include <libavr/libavr.hpp>
-
-using namespace avr::literals;
-namespace spm = avr::spm;
-namespace ee = avr::eeprom;
-
-using dev = avr::device<{.clock = 16_MHz}>;
-// One-wire: RX and TX share the line, exactly as the native-UART TSB expects.
-using serial_t = dev::uart0<{.baud = 115200_Bd, .max_baud_error = 3_pct, .half_duplex = true}>;
-inline constexpr serial_t serial{};
-
-namespace tsb {
-namespace {
-
-// The loader is purely polled — it never enables interrupts — so every SPM and
-// EEPROM lock folds to nothing under this posture.
-constexpr auto off = avr::irq::guard_policy::unused;
-
-// The handshake bytes, identical across every TSB host.
-constexpr std::uint8_t confirm = '!';
-constexpr std::uint8_t request = '?';
-constexpr std::uint8_t knock = '@';
-
-// Boot geometry for the 1 KB boot section (BOOTSZ=10); the page size and the
-// flash/EEPROM extents are the chip database's to know. app_end is the config
-// page (TSB's LASTPAGE), one page below the boot section.
-constexpr std::uint16_t page = spm::page_bytes;
-constexpr std::uint16_t boot_bytes = 1024;
-constexpr std::uint16_t app_end = spm::flash_bytes - boot_bytes - page;
-constexpr std::uint16_t eeprom_end = avr::hw::db.mem.eeprom_size - 1;
-
-// Lockout-proof floor for the activation window (the oracle's F_CPU/1MHz).
-constexpr std::uint8_t act_min = 16;
-// Post-activation window: the host gets seconds, not milliseconds, mid-session.
-constexpr std::uint8_t comm_window = 200;
-
-// Firmware version stamp: YY*512 + MM*32 + DD, the encoding the host decodes.
-constexpr std::uint16_t build_date = 26 * 512 + 7 * 32 + 27;
-
-// The 16-byte device-info block, streamed out on activation.
-// clang-format off
-[[gnu::progmem]] constexpr std::uint8_t info[16] = {
-    'T', 'S', 'B',
-    build_date & 0xFF, build_date >> 8,
-    0xF3,             // status: native-UART fixed-baud lineage
-    avr::hw::db.signature[0], avr::hw::db.signature[1], avr::hw::db.signature[2],
-    page / 2,         // page size in words
-    (app_end / 2) & 0xFF, (app_end / 2) >> 8, // app-flash boundary, words
-    eeprom_end & 0xFF, eeprom_end >> 8,
-    0xAA, 0xAA,       // ATmega processor-type marker (bytes 14 == 15)
-};
-// clang-format on
-
-// The receive window, pre-floored where it is set. In .noinit: there is no
-// crt to clear a .bss image, and run() stores it before the first receive.
-[[gnu::section(".noinit")]] std::uint8_t window;
-
-const std::uint8_t *flash_ptr(std::uint16_t addr)
-{
-	return reinterpret_cast<const std::uint8_t *>(addr);
-}
-
-// Bounded byte receive: poll under nested countdowns, 0 on silence. The 0
-// then falls through every compare — not a knock, not a confirm, not a
-// command — so a silent host unwinds the loader to the application from
-// anywhere, and a mid-session cable pull cannot wedge it. The line release on
-// a direction change is the serial backend's.
-[[gnu::noinline]] std::uint8_t rx()
-{
-	std::uint16_t outer = static_cast<std::uint16_t>(window) << 8;
-	do {
-		std::uint8_t fine = 0;
-		do {
-			if (auto byte = serial.read())
-				return *byte;
-		} while (--fine);
-	} while (--outer);
-	return 0;
-}
-
-// One-wire transmit: the backend takes the line with a turn-around guard and
-// holds it until the whole frame is out.
-[[gnu::noinline]] void tx(std::uint8_t byte)
-{
-	serial.write(byte);
-}
-
-// '?', then hand back the host's reply for the callers' one-byte compare.
-[[gnu::noinline]] std::uint8_t rcnf()
-{
-	tx(request);
-	return rx();
-}
-
-// The one send loop: the info block, the config page, application flash and
-// EEPROM pages all stream through here.
-[[gnu::noinline]] void send_block(bool eep, std::uint16_t at, std::uint8_t count)
-{
-	do {
-		tx(eep ? ee::read(at) : avr::flash_load(flash_ptr(at)));
-		++at;
-	} while (--count);
-}
-
-// One EEPROM byte in — shared by the emergency wipe and the 'E' stream.
-[[gnu::noinline]] void eeput(std::uint16_t at, std::uint8_t value)
-{
-	ee::write<off>(at, value);
-}
-
-// Wait out a running SPM op, then re-open the RWW section — after every page
-// op and before handing over, as the oracle does.
-[[gnu::noinline]] void settle()
-{
-	spm::wait();
-	spm::rww_enable<off>();
-}
-
-// One host page straight into the erased flash page at `at` — through the SPM
-// word buffer (low byte then high), no SRAM staging — then committed. `at`
-// names a page base, so the cursor's low byte reaching the boundary ends the
-// walk.
-[[gnu::noinline]] void store_flash_page(std::uint16_t at)
-{
-	do {
-		std::uint8_t low = rx();
-		std::uint8_t high = rx();
-		spm::fill<off>(at, std::bit_cast<std::uint16_t>(std::array{low, high}));
-		at += 2;
-	} while (static_cast<std::uint8_t>(at) & (page - 1));
-	spm::write_page<off>(at - page);
-	settle();
-}
-
-extern "C" [[noreturn]] void tsb_app(); // the application's reset vector: --defsym=tsb_app=0
-
-[[noreturn]] void appjump()
-{
-	settle();
-	tsb_app();
-}
-
-// Step one page down and erase it — the erase shared by the whole-app walk,
-// the config rewrite and the emergency wipe; hands the stepped address back.
-[[gnu::noinline]] std::uint16_t erase_below(std::uint16_t at)
-{
-	at -= page;
-	spm::erase_page<off>(at);
-	settle();
-	return at;
-}
-
-// Erase the whole application, top-down like the oracle: the loop bound is a
-// compare with zero, and the returned 0 is the address every caller wants
-// next.
-[[gnu::noinline]] std::uint16_t erase_application()
-{
-	std::uint16_t at = app_end;
-	do {
-		at = erase_below(at);
-	} while (at != 0);
-	return at;
-}
-
-[[noreturn]] void run()
-{
-	// A watchdog reset hands straight back to the application, as the
-	// reference loader does, rather than re-entering the bootloader.
-	if (avr::hw::mcusr::wdrf.test())
-		appjump();
-
-	// Lean bring-up from reset state: UCSR0C already reads 8N1, UBRR0H reads
-	// 0, and the half-duplex write()/read() raise TXEN0/RXEN0 on first use —
-	// only the divisor low byte and U2X0 need a store. The solver still does
-	// the datasheet work; the asserts pin the reset-state assumptions.
-	{
-		constexpr auto sol = avr::uart::solve_baud(dev::clock, 115200_Bd);
-		static_assert(sol.u2x && sol.ubrr < 256, "lean bring-up writes UBRR0L only, with U2X0");
-		avr::hw::ubrr0::write(static_cast<std::uint8_t>(sol.ubrr));
-		avr::hw::ucsr0a::write(avr::hw::ucsr0a::u2x0(1));
-	}
-
-	// Activation: 3×'@', each inside the config page's timeout window
-	// (floored so a corrupt page cannot lock the loader out); anything else —
-	// including silence — hands over.
-	window = avr::flash_load(flash_ptr(app_end + 2)) | act_min;
-	for (std::uint8_t k = 3; k; --k)
-		if (rx() != knock)
-			appjump();
-	window = comm_window;
-
-	// Password gate (config page from app_end+3, 0xff-terminated; a blank
-	// page is no password). A wrong byte blanks the comparison and drains the
-	// line forever, so a wrong password can never fall through; a 0 requests
-	// emergency erase behind two confirms. On pass the info block goes out;
-	// the emergency path skips it and drops into the command loop.
-	std::uint16_t at = app_end + 3;
-	std::uint8_t mask = 0xff;
-	for (;;) {
-		std::uint8_t expected = avr::flash_load(flash_ptr(at)) & mask;
-		++at;
-		if (expected == 0xff) {
-			send_block(false, reinterpret_cast<std::uint16_t>(&info[0]), sizeof info);
-			break;
-		}
-		std::uint8_t got = rx();
-		if (got == 0) {
-			if (mask == 0)
-				continue;
-			if (rcnf() != confirm || rcnf() != confirm)
-				appjump();
-			std::uint16_t a = erase_application();
-			do {
-				eeput(a, 0xff);
-			} while (++a <= eeprom_end);
-			erase_below(app_end + page);
-			break;
-		}
-		if (got != expected)
-			mask = 0;
-	}
-
-	for (;;) {
-		tx(confirm); // Mainloop ready
-		const std::uint8_t command = rx();
-		switch (command) {
-		case 'f': // read application flash, one page per host '!'
-			for (std::uint16_t a = 0; a < app_end; a += page) {
-				if (rx() != confirm)
-					break;
-				send_block(false, a, page);
-			}
-			break;
-		case 'e': // read EEPROM, one page per host '!', until the host stops
-			for (std::uint16_t a = 0;; a += page) {
-				if (rx() != confirm)
-					break;
-				send_block(true, a, page);
-			}
-			break;
-		case 'F': { // erase the application, then take pages behind '?'
-			std::uint16_t a = erase_application();
-			for (; rcnf() == confirm; a += page)
-				store_flash_page(a);
-			break;
-		}
-		case 'E': // take EEPROM pages behind '?', each write host-paced
-			for (std::uint16_t a = 0; rcnf() == confirm;) {
-				std::uint8_t count = page;
-				do {
-					eeput(a, rx());
-					++a;
-				} while (--count);
-			}
-			break;
-		case 'c': // read the config page
-		read_config:
-			send_block(false, app_end, page);
-			break;
-		case 'C': // replace the config page, then echo it back to verify
-			if (rcnf() != confirm)
-				break;
-			store_flash_page(erase_below(app_end + page));
-			goto read_config;
-		default: // 'q' or any other byte runs the application
-			appjump();
-		}
-	}
-}
-
-} // namespace
-} // namespace tsb
-
-// Reset lands at the boot section base (BOOTRST): the entry stub in .vectors
-// is laid first and does the one line of crt a crt-less image needs.
-template struct avr::startup::entry<tsb::run>;

tsb/tsb_tricks.cpp

@@ -65,7 +65,7 @@ constexpr std::uint8_t comm_window = 200;
 constexpr std::uint16_t build_date = 26 * 512 + 7 * 32 + 20;
 
 // Fixed 115200 8N1; the library solves UBRR + U2X from clock and baud.
-constexpr auto baud = avr::uart::solve_baud(16_MHz, 115200_Bd);
+constexpr auto baud = avr::uart::detail::solve_baud(16_MHz, 115200_Bd);
 
 // The 16-byte device-info block, streamed out on activation.
 // clang-format off
@@ -240,7 +240,7 @@ extern "C" [[noreturn]] void tsb_app(); // the application's reset vector: --def
 	// 0, and rx()/tx() raise RXEN0/TXEN0 on first use — only the divisor low
 	// byte and U2X0 need a store. The library still does the datasheet work.
 	static_assert(baud.u2x && baud.ubrr < 256, "lean bring-up writes UBRR0L only, with U2X0");
-	hw::ubrr0::write(static_cast<std::uint8_t>(baud.ubrr));
+	hw::reg<"UBRR0">::write(static_cast<std::uint8_t>(baud.ubrr));
 	hw::ucsr0a::write(hw::ucsr0a::u2x0(1));
 	// General-purpose registers are undefined at power-on (no crt zeroes them);
 	// the direction latch must start "not receiving" so the first rx() enables