Bug 23244 – Kernel panic w/ SiI3112A on 2.4.21-pfeifer_pre1-r1

• Status: RESOLVED LATER
• Product: Gentoo Linux

[Description Zack Gilburd (RETIRED) 2003-06-21 13:36:28 UTC]

I will let the panic explain itself.  This is 100% reproducable for me.  It
panics every time I boot this kernel right when it's about to check the partitions.

http://tehunlose.com/tmp/satakernelpanic/

[Comment 1 Zack Gilburd (RETIRED) 2003-06-21 13:37:08 UTC]

Created attachment 13666 [details]
the panic

panic

[Comment 2 Zack Gilburd (RETIRED) 2003-06-21 13:38:24 UTC]

Created attachment 13667 [details]
system.map

[Comment 3 Jay Pfeifer (RETIRED) 2003-06-22 23:11:08 UTC]

please attach the .config used to compile this kernel. 
 
Thanks, 
 
Jay 

[Comment 4 Zack Gilburd (RETIRED) 2003-06-23 01:03:03 UTC]

Created attachment 13714 [details]
kernel config

[Comment 5 Jay Pfeifer (RETIRED) 2003-07-18 06:29:19 UTC]

fyi - this isn't specific to my sources, this is looking like the preempt atatraid highmem issue. turn off highmem and let me know how it goes.

[Comment 6 Zack Gilburd (RETIRED) 2003-07-18 09:30:50 UTC]

I don't understand how this could be "unspecific" to your sources if your 2.4.21_pre1-r1 is the 
only kernel that produces this effect for me.  I use highmem on all of my kernels because...well... I 
have a high ammount of memory :). 
 
On a side note (for others that may be following this bug), the 2.4.22_pre2-gss sources work 
perfectly with both the SiI3112A and my Adaptec 29160N. 

[Comment 7 Jay Pfeifer (RETIRED) 2003-07-18 11:48:28 UTC]

take a look here: 
http://members.optusnet.com.au/ckolivas/kernel/ 
 
Notice the part towards the bottom: 
"I have Highmem and IDE Raid and I get a bug on booting your kernel! There seems to be a 
real problem with IDE and highmem in 2.4 and the preempt patch brings it out. The only 
solution is to not enable highmem (limiting you to 900Mb) or not using my patches at all. I'm 
looking into this but it's not just my patchset, it's any 2.4 with preempt." 
 
I am looking @ fix as well, but again this is not specific to my patchset. 

[Comment 8 Jay Pfeifer (RETIRED) 2003-07-18 11:50:25 UTC]

oh, and i use highmem on all mine as well too (all my systems have 1 - 4 GB RAM), highmem 
works great it is just the ataraid stuff and preempt with highmem that causes the issue. 

[Comment 9 Adrien Tay Pamart 2003-08-12 07:34:54 UTC]

Could that help a bit ?

I am encountering a very similar bug using a different version of the kernel sources (2.4.22_pre2-gss) w/ highmem, preempt with IDE hardware.

http://www.ussg.iu.edu/hypermail/linux/kernel/0303.1/1188.html

I looked up in the source tree and the patch has not made it yet to highmem.h