Under kernel-2.4.2-0.1.29 running the buildinstall from anaconda-runtime from the qa0319 build hung while creating one of the boot disk images. At first I thought it was the loop hangs returning, but the oops in dmesg shows that it was waiting for a umount of the image which caused an oops. ksymoops gives: Oops: 0002 CPU: 0 EIP: 0010:[<cd565dc8>] Using defaults from ksymoops -t elf32-i386 -a i386 EFLAGS: 00210246 eax: 00000000 ebx: c0746204 ecx: c0746204 edx: 00000000 esi: c4141f14 edi: c0a13048 ebp: 00000000 esp: c4141edc ds: 0018 es: 0018 ss: 0018 Process umount (pid: 22591, stackpage=c4141000) Stack: c0746204 c014d8b2 c0746204 c014d090 cbfea4a0 c1f20204 c1f20204 c0746204 c014d920 c0746204 c4141f14 c4141f14 c014da38 c4141f14 c352b20c ca93440c c0a13000 c8859900 cd56ad40 cd4556c0 c013dac9 c0a13000 c012de5e c195b290 Call Trace: [<c014d8b2>] [<c014d090>] [<c014d920>] [<c014da38>] [<cd56ad40>] [<cd4556c0>] [<c013dac9>] [<c012de5e>] [<c013df81>] [<c013e0a3>] [<c013e11c>] [<c010910b>] Code: 89 50 04 89 02 5b c3 90 53 8b 5c 24 08 8b 83 5c 01 00 00 8b >>EIP; cd565dc8 <[fat]fat_clear_inode+18/20> <===== Trace; c014d8b2 <clear_inode+102/130> Trace; c014d090 <destroy_inode+30/40> Trace; c014d920 <dispose_list+40/60> Trace; c014da38 <invalidate_inodes+48/60> Trace; cd56ad40 <[fat]fat_sops+0/38> Trace; cd4556c0 <[msdos]msdos_fs_type+0/17> Trace; c013dac9 <kill_super+79/170> Trace; c012de5e <kfree+20e/2b0> Trace; c013df81 <do_umount+1d1/1e0> Trace; c013e0a3 <sys_umount+113/180> Trace; c013e11c <sys_oldumount+c/10> Trace; c010910b <system_call+33/38> Code; cd565dc8 <[fat]fat_clear_inode+18/20> 00000000 <_EIP>: Code; cd565dc8 <[fat]fat_clear_inode+18/20> <===== 0: 89 50 04 mov %edx,0x4(%eax) <===== Code; cd565dcb <[fat]fat_clear_inode+1b/20> 3: 89 02 mov %eax,(%edx) Code; cd565dcd <[fat]fat_clear_inode+1d/20> 5: 5b pop %ebx Code; cd565dce <[fat]fat_clear_inode+1e/20> 6: c3 ret Code; cd565dcf <[fat]fat_clear_inode+1f/20> 7: 90 nop Code; cd565dd0 <[fat]fat_put_super+0/b0> 8: 53 push %ebx Code; cd565dd1 <[fat]fat_put_super+1/b0> 9: 8b 5c 24 08 mov 0x8(%esp,1),%ebx Code; cd565dd5 <[fat]fat_put_super+5/b0> d: 8b 83 5c 01 00 00 mov 0x15c(%ebx),%eax Code; cd565ddb <[fat]fat_put_super+b/b0> 13: 8b 00 mov (%eax),%eax
Have you tested with a more recent kernel?
I'm not seeing it now, but since I'm not overly certain of the test parameters, I'm still a little leery... I'll keep trying to see if I can trigger it with 2.4.2-2 since I'm sure I'll be running buildinstall several times over the next little bit.
If you ever reproduce with 2.4.3-12 or later please re-open, thanks.