sysupgrade failed on Atheros PB44 too. 2010/6/2 <openwrt-devel-requ...@lists.openwrt.org>
> Send openwrt-devel mailing list submissions to > openwrt-devel@lists.openwrt.org > > To subscribe or unsubscribe via the World Wide Web, visit > https://lists.openwrt.org/mailman/listinfo/openwrt-devel > or, via email, send a message with subject or body 'help' to > openwrt-devel-requ...@lists.openwrt.org > > You can reach the person managing the list at > openwrt-devel-ow...@lists.openwrt.org > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of openwrt-devel digest..." > > > Today's Topics: > > 1. ar71xx sysupgrade fails sometimes on routerstation-pro > (Bruno Randolf) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Wed, 2 Jun 2010 16:27:51 +0900 > From: Bruno Randolf <b...@einfach.org> > To: n...@openwrt.org, juh...@openwrt.org > Cc: openwrt-devel@lists.openwrt.org, ?? ??? <matsuy...@thinktube.com> > Subject: [OpenWrt-Devel] ar71xx sysupgrade fails sometimes on > routerstation-pro > Message-ID: <201006021627.51382....@einfach.org> > Content-Type: Text/Plain; charset="us-ascii" > > hi! > > on the routerstation-pro the mtd command of the sysupgrade script > (ar71xx/base-files/lib/upgrade/platform.sh) sometimes fails because of > problems with the FIS table. this sometimes results in a segfault, > sometimes > in an error message. i see this about every third time i try to flash an > image. > > i tracked it down to these lines in mtd fis.c, both in fis_validate() and > fis_remap(): > > if (!desc->hdr.name[0] || (desc->hdr.name[0] == 0xff)) > break; > > when the problem happens, desc->hdr.name[0] is zero, and we break out > early > from the loop, causing other things to be uninitialized. > > in the case of fis_validate() this returns the error "Failed to validate > the > new FIS partition table", later. but in fis_remap() the uninitialized > (NULL) > pointers 'first' and 'last' are accessed later, causing the segfault. > obviously the segfault could and should be avoided by adding something like > the following lines to fis_remap(): > > if (!first || !last) > return -1; > > but the real problem must be somewhere else. and this is where i am stuck > now... i wonder why we only sometimes get a 0 for desc->hdr.name[0]? can > it be > a problem with the flash driver for the ar71xx? > > bruno > > ps: i'm not the only one seeing this problem: > https://forum.openwrt.org/viewtopic.php?id=24188 > > > ------------------------------ > > _______________________________________________ > openwrt-devel mailing list > openwrt-devel@lists.openwrt.org > https://lists.openwrt.org/mailman/listinfo/openwrt-devel > > > End of openwrt-devel Digest, Vol 54, Issue 4 > ******************************************** >
_______________________________________________ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/mailman/listinfo/openwrt-devel
