On 01.12.2010 11:12, Michael Karcher wrote:
Am Dienstag, den 30.11.2010, 01:42 +0100 schrieb Carl-Daniel Hailfinger:
Handle erase failure in partial write. Clean up erase function checking. Update a few comments and messages to improve readability.
Reviewing would be easier of course if the first one would be a separate patch from the second and third one, but as it seems they don't overlap, it's OK, too.
Noted.
Acked-by: Michael Karcher flashrom@mkarcher.dialup.fu-berlin.de
Thanks for your review! I split the patch as suggested.
Handle erase failure in partial write.
Signed-off-by: Carl-Daniel Hailfinger c-d.hailfinger.devel.2006@gmx.net Acked-by: Michael Karcher flashrom@mkarcher.dialup.fu-berlin.de
Index: flashrom-handle_failed_erase_or_write/flashrom.c =================================================================== --- flashrom-handle_failed_erase_or_write/flashrom.c (Revision 1236) +++ flashrom-handle_failed_erase_or_write/flashrom.c (Arbeitskopie) @@ -1468,11 +1468,20 @@ /* If everything is OK, don't try another erase function. */ if (!ret) break; - /* FIXME: Reread the whole chip here so we know the current - * chip contents? curcontents might be up to date, but this - * code is only reached if something failed, and then we don't - * know exactly what failed, and how. + /* Write/erase failed, so try to find out what the current chip + * contents are. If no usable erase functions remain, we could + * abort the loop instead of continuing, the effect is the same. + * The only difference is whether the reason for other unusable + * functions is printed or not. If in doubt, verbosity wins. */ + if (flash->read(flash, curcontents, 0, size)) { + /* Now we are truly screwed. Read failed as well. */ + msg_cerr("Can't read anymore!\n"); + /* We have no idea about the flash chip contents, so + * retrying with another erase function is pointless. + */ + break; + } } /* Free the scratchpad. */ free(curcontents);