Recent development (unreleased)¶
This document describes the major changes that are expected to be included in the next release of flashrom and which are currently only available by source code checkout (see Building from source). These changes may be further revised before the next release.
CLI updates¶
Colored errors and warnings¶
The classic CLI prints error messages in red and warnings in yellow when the
output is a terminal. Set the NO_COLOR environment variable to a non-empty
value to disable it. The log file written by -o is never colored. Color
output is not implemented on Windows.
Platform updates¶
FreeBSD¶
The internal programmer now finds the SMBIOS tables on UEFI systems by asking
the loader for the entry point, so board matching and laptop detection work
there. MSR access uses cpuctl(4) from the base system instead of the
retired devcpu port. The internal programmer is offered on non-x86
FreeBSD hosts, matching Linux.
Programmers updates¶
ft2232_spi: specify device by USB port path¶
It was previously not possible to specify a device in case of collision in serial/description, for example between multiples of the same device with blank EEPROM. now it is possible to select by USB bus & port path.
Example:
flashrom -p ft2232_spi:type=4232H,usbpath=1-1.4.2 -r -c W25Q80BV/W25Q80DV -r dump.bin
ni845x_spi programmer is deleted¶
The driver was gated to 32-bit Windows and linked against a proprietary National Instruments library that ships for no other platform, so the hardware has never been usable anywhere else.
It has been disabled by default all the time and was not built or tested by CI scripts.
Removed programmers¶
The rayer_spi programmer has been removed. It bit-banged SPI over a PC
parallel port (RayeR cable, Altera ByteBlasterMV, Atmel STK200/300, Macraigor
Wiggler, Xilinx Parallel Cable III, and SPI Tiny Tools hardware). Parallel
ports are absent from modern hardware, and the driver required raw x86 I/O port
access, restricting it to legacy x86 hosts.
The atapromise programmer has been removed. It reflashed the option ROM
on Promise PDC2026x ATA/RAID PCI cards via raw x86 port I/O and was capped at
32 kB by the tested card wiring. The hardware is obsolete and legacy hosts
already ship a flashrom that supports it.
The gfxnvidia programmer has been removed. It reflashed the parallel
flash on old NVIDIA graphics cards through a memory-mapped PCI BAR. The
hardware is obsolete and legacy hosts already ship a flashrom that supports
it.
The satasii programmer has been removed. It reflashed the parallel flash
on Silicon Image SATA/IDE PCI controllers through a memory-mapped PCI BAR.
The hardware is obsolete and legacy hosts already ship a flashrom that
supports it.
The nicnatsemi programmer has been removed. It reflashed the socketed
parallel boot ROM on National Semiconductor DP838* PCI network cards through
raw x86 port I/O. The hardware is obsolete and legacy hosts already ship a
flashrom that supports it.
The drkaiser programmer has been removed. It reflashed the parallel
flash on Dr. Kaiser PC-Waechter PCI cards through a memory-mapped PCI BAR.
The hardware is obsolete and legacy hosts already ship a flashrom that
supports it.
The satamv programmer has been removed. It reflashed the parallel flash
on Marvell 88SX70xx/88SX60xx SATA PCI controllers through raw x86 port I/O.
The hardware is obsolete and legacy hosts already ship a flashrom that
supports it.
The atahpt programmer has been removed. It reflashed the option ROM on
HighPoint ATA/RAID PCI controllers through raw x86 port I/O. The hardware is
obsolete and legacy hosts already ship a flashrom that supports it.
The it8212 programmer has been removed. It reflashed the flash on ITE
IT8212F ATA/RAID PCI controllers through a memory-mapped PCI BAR. The
hardware is obsolete and legacy hosts already ship a flashrom that supports
it.
The developerbox_spi programmer has been removed. It drove the on-board
SPI flash of the Socionext SynQuacer DeveloperBox for recovery. The board is
end-of-life with a negligible install base, so nothing in current use depends
on it.
Platform support¶
DOS support is removed¶
flashrom no longer builds for DOS. The DJGPP target needed its own DPMI physical memory mapping, port I/O wrappers and delay loop, none of which was shared with any other platform or covered by CI.