Skip to contentSkip to conversation

iosapic.c NULL pointer deref

Linux-ia64 [email protected] 2 participants 2 messages Oct 18, 2000 – Oct 19, 2000

lore
Share & tools
lore
#0 lore raw

In the 2.4.0-test8 and -test9 IA-64 patch, the file arch/ia64/dig/iosapic.c, function pci_for_each_dev(), the bridge pointer can be NULL, but gets dereferenced anyhow, causing a nice tombstone. I see this with an AMI MegaRAID 471 card that happens to have two bridges on it. Putting in tests against NULL, that routine doesn't blow up any longer, but interrupts are not properly assigned to that card then, and eventually all SCSI reads/writes time out.

I'm happy to work with whomever is responsible for this code to have it properly assign interrupts to these cards.

Thanks,
Matt

#1 lore raw

Matt,

This code was put in specifically for AMI MegaRaid controller and bridge cards. Mike from Intel did the patch. Anyway, I am working on the IOSAPIC code to fix a problem and would like to fix this problem also. Please let me know if you have any suggestion.

Thanks,
Asit

Quoted 32 lines

-----Original Message-----
From: [email protected] [mailto:[email protected]]
Sent: Wednesday, October 18, 2000 3:12 PM
To: [email protected]
Cc: [email protected]
Subject: [Linux-ia64] iosapic.c NULL pointer deref

In the 2.4.0-test8 and -test9 IA-64 patch, the file arch/ia64/dig/iosapic.c,
function pci_for_each_dev(), the bridge pointer can be NULL, but gets dereferenced anyhow, causing a nice tombstone. I see this with an AMI MegaRAID 471 card that happens to have two bridges on it. Putting in tests
against NULL, that routine doesn't blow up any longer, but interrupts are
not properly assigned to that card then, and eventually all SCSI reads/writes time out.

I'm happy to work with whomever is responsible for this code to have it
properly assign interrupts to these cards.

Thanks,
Matt

_______________________________________________
Linux-IA64 mailing list
[email protected]
http://lists.linuxia64.org/lists/listinfo/linux-ia64