On 28.08.2007 18:19, ron minnich wrote:
On 8/28/07, Carl-Daniel Hailfinger c-d.hailfinger.devel.2006@gmx.net wrote:
lib/lar.c and util/lar/example.c differ in subtle ways in find_file. Inverted logic in one file, bogus calculations in the other one. We might want to make sure they behave the same way.
This still applies. However, it can be fixed in another commit.
walk += (ntohl(header->len) + ntohl(header->offset) - 1) & 0xfffffff0;ARGH! Shouldn't that be
walk += (ntohl(header->len) + ntohl(header->offset) + 15) & 0xfffffff0;
OK, I added this to the code: if (strcmp(&walk[0], "LARCHIVE") != 0) continue; before the other fix I created. It dies: LinuxBIOS-3.0.0 Tue Aug 28 08:21:43 PDT 2007 starting... Choosing fallback boot. LAR: Attempting to open 'fallback/initram'. LAR: Start 0xfff00000 len 0x100000 LAR: search for normal/payload LAR: search for normal/option_table LAR: search for normal/stage2 LAR: search for normal/initram LAR: search for %s @ %p
So I put in your fix (replace -1 with +15)
Yes, I just reread that code and have to agree my +15 solution was not the right one. Maybe we could add some debug printing to find out where the code is looking for MAGIC, but I'd leave that to a later patch.
And it is worse: LinuxBIOS-3.0.0 Tue Aug 28 08:21:43 PDT 2007 starting... Choosing fallback boot. LAR: Attempting to open 'fallback/initram'. LAR: Start 0xfff00000 len 0x100000 LAR: search for normal/payload LAR: search for normal/stage2 LAR: search for %s @ %p
So, how about we leave my patch in for now while I try to track this nasty bug down?
Agreed.
Regards, Carl-Daniel