Enables the USB SPI transfer retry mechanism when the error code
USB_SPI_WRITE_COUNT_INVALID is returned. This error code can
indicate a recoverable USB transfer failure.
BUG=b:153887087
BRANCH=none
TEST=Manual testing of ServoMicro and Flashrom when performing
reads, writes, and verification of the EC firmware on Nami.
TEST=Modified ServoMicro to randomly corrupt USB packets when
reading the packet length to replicate bad packets and the
verify recovery is successful.
Change-Id: I9e6b2ccec0b06aab0d6920f1bddf108058e5d6b1
Signed-off-by: Brian J. Nemec <bnemec@chromium.com>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/41152
Reviewed-by: Edward O'Callaghan <quasisec@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
The Chromium flashrom fork has very poor dispatch logic whereas
upstream has proper inversion of control with a generic 'data'
void * member to stuff long-lived state in. Leverage the member
to store the USB descriptor state in during the life-time of the
spi master.
V.2: Remove unnecessary indirection as is the case in
commit a25c13cdb6.
BUG=b:140394053
BRANCH=none
TEST=builds && detects flashchip name.
Change-Id: Ida9dce97fef2c6dfd68a278c879917fdd3ff7fef
Signed-off-by: Edward O'Callaghan <quasisec@google.com>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/40105
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Reviewed-by: Sam McNally <sammc@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
The ChromiumOS flashrom fork has since const'ify flashctx
in a few places. This aligns the function signatures to
match with downstream to ease forward porting patches
out of downstream back into mainline flashrom.
This patch is minimum viable alignment and so feedback is
welcome.
Change-Id: Iff6dbda13cb0d941481c0d204b9c30895630fbd1
Signed-off-by: Edward O'Callaghan <quasisec@google.com>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/40324
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
The raiden_debug programmer is of type USB. However, it does not set the
field `devs.dev`, which will result in a segfault when trying to print
the devices of the non-existing table.
Fix that by replacing `devs.note` with `devs.dev` and adding an empty
device table. Since Device IDs are not used to match programmers,
nothing could be added to the table.
TEST=Running `flashrom -L` no longer segfaults and returns normally.
Change-Id: Ie4171a11384c34abb102d1aadf86aa1b8829fc04
Signed-off-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/39826
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Edward O'Callaghan <quasisec@chromium.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
This overcomes a problem with the ServoMicro where USB packets can be
ack'd by the device without triggering interrupts or loading data into
the USB endpoints. The retry mechanism attempts the USB read 3 times
before reattempting the write call to avoid performing multiple SPI
transfers due to a USB problem. This process repeats 3 times before we
return the last error code. Intermediary problems are reported in the
status code.
Based off the downstream commit:
https://chromium-review.googlesource.com/c/chromiumos/third_party/flashrom/+/2038271
Change-Id: I76cde68852fa4963582d57c7dcb9f24de32c6da8
Signed-off-by: Edward O'Callaghan <quasisec@google.com>
Reviewed-on: https://review.coreboot.org/c/flashrom/+/39310
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
