im using agpgart as a module with my i810 if i start an x session then shut it down and try to start it again i get a error GARTInit: AGPIOC_INFO failed (Invalid Arguement) I810(0): AGP GART support is not available. Make sure yer kernel has agpgart support or that the agpgart kernel module is available. wont work again till i reboot and cant unload the module btw the drm works now thnks to whoever fixed it using the gentoo kernel with xfree r11 and i810 chipset
Hmm, ok try the following: emerge --clean rsync Then delete the following file: /usr/portage/x11-base/xfree/files/4.2.0/XFree86-4.2.0-agpgart-load.patch.bz2 then: # emerge xfree Which should merge in -r12 without the agpgart-load patch that was also in -r11. Let me know if this fixes things. As for the drm patch, you should actually thank "Sottek, Matthew J" <matthew.j.sottek@intel.com>, as he is the author, and Mandrake for having (allowing me to snatch it from them) ;)
hi! thanks for writing back unfortunately, that didnt work :( one thing i noticed, is when i shut down an x session, it doesnt return to either kdm or tty1 like it ussually does, just hangs on tty7 with a blackscreen and blinking cursor and if im reading lsmod right, it sez being used by 6 processes.. so its just not shutting down right ive talked to other peeple with an i810 and im the only one thats getting this error. so i don't know...
azarah / todd: the same problem happens with me. I havent't deleted /usr/portage/x11- base/xfree/files/4.2.0/XFree86-4.2.0-agpgart-load.patch.bz2 and emerged xfree again but if i comment out the Load "dri" line in my /etc/XF86Config, the problem goes away. Unfortunately, so does DRI. Kream - Aniruddha Karim Shankar
kream azarah yep, commenting out dri does stop the GARTInit error this is what i get in dmesg when i try to do it with dri: Linux agpgart interface v0.99 (c) Jeff Hartmann agpgart: Maximum main memory to use for agp memory: 148M agpgart: Detected an Intel i810 Chipset. agpgart: AGP aperture is 64M @ 0xd8000000 memory : cb9df6e0 memory : 00000000 memory : cb9df720 [drm] AGP 0.99 on Intel i810 @ 0xd8000000 64MB [drm] Initialized i810 1.2.0 20010920 on minor 0 mtrr: base(0xd8000000) is not aligned on a size(0x180000) boundary /* the above is before i shutdown and restart x x, then this pops up: Unable to handle kernel paging request at virtual address 0100001b printing eip: c0127a60 *pde = 00000000 Oops: 0000 CPU: 0 EIP: 0010:[<c0127a60>] Not tainted EFLAGS: 00013246 eax: 01000000 ebx: 01000000 ecx: c1205ff4 edx: 00000000 esi: c9f62000 edi: cb345800 ebp: 00000000 esp: cb0f1ee0 ds: 0018 es: 0018 ss: 0018 Process X (pid: 3259, stackpage=cb0f1000) Stack: c1205ff4 c9f62000 cc9a528f c1205ff4 cb3f7640 cb612800 cc9a52df cb345800 c9f62000 cb0f1f30 bffffb24 cb0f1f70 cc9a58ca cb345800 00000040 cb345800 cb9df6e0 40406440 00000000 cb345800 00000002 00000000 00000000 00000000 Call Trace: [<cc9a528f>] [<cc9a52df>] [<cc9a58ca>] [<cc9a0ff2>] [<c0145429>] [<c010871b>] Code: 0f b6 43 1b 8b 14 85 e4 7d 33 c0 69 c3 b1 79 37 9e 8b 8a c0 <3>X[3259] exited with preempt_count 1 so it seems to be a dri thing... too bad the dri.sourceforge drivers wont compile :( the mtrr errors i got under slack too, but never had any problems with dri or agpgart, so i dont think they are related
Can any of you guys please try the fix I suggested ? It is the only thing that I can think of which may cause this, thanks.
I followed the steps you laid out in http://bugs.gentoo.org/show_bug.cgi?id=3551#c1 . The problem still remains. ls -l /usr/portage/x11-base/xfree/files/4.2.0/XFree86-4.2.0-agpgart-load.patch.bz2 ls: /usr/portage/x11-base/xfree/files/4.2.0/XFree86-4.2.0-agpgart-load.patch.bz2: No such file or directory appended are relevant sections from dmesg and /var/log/XFree86.0.log XFree86.0.log ===> (WW) xf86AcquireGART: AGPIOC_ACQUIRE failed (Device or resource busy) (EE) GARTInit: AGPIOC_INFO failed (Invalid argument) (EE) I810(0): AGP GART support is not available. Make sure your kernel has agpgart support or that the agpgart kernel module is loaded. (II) UnloadModule: "i810" (II) UnloadModule: "ddc" (II) Unloading /usr/X11R6/lib/modules/libddc.a (II) UnloadModule: "int10" (II) Unloading /usr/X11R6/lib/modules/linux/libint10.a (II) UnloadModule: "vbe" (II) Unloading /usr/X11R6/lib/modules/libvbe.a (II) UnloadModule: "vgahw" (II) Unloading /usr/X11R6/lib/modules/libvgahw.a (EE) Screen(s) found, but none have a usable configuration. dmesg ===> devfs_register(1): could not append to parent, err: -17 devfs_register(a1): could not append to parent, err: -17 devfs_register(1): could not append to parent, err: -17 devfs_register(a1): could not append to parent, err: -17 nfsd[4900] exited with preempt_count 1 nfsd[4899] exited with preempt_count 1 nfsd[4898] exited with preempt_count 1 nfsd[4897] exited with preempt_count 1 nfsd[4896] exited with preempt_count 1 nfsd[4895] exited with preempt_count 1 nfsd[4894] exited with preempt_count 1 Unable to handle kernel paging request at virtual address 0100001b printing eip: c0129c18 *pde = 00000000 Oops: 0000 CPU: 0 EIP: 0010:[<c0129c18>] Not tainted EFLAGS: 00013246 eax: 01000000 ebx: 01000000 ecx: c10ea894 edx: 00000000 esi: c482a000 edi: c11f5800 ebp: 00000000 esp: c6acdee0 ds: 0018 es: 0018 ss: 0018 Process X (pid: 5061, stackpage=c6acd000) Stack: c10ea894 c482a000 c01fc3bf c10ea894 c630e740 c6be5000 c01fc40f c11f5800 c482a000 c6acdf30 bffffcbc c6acdf70 c01fc9fa c11f5800 00000040 c11f5800 c6d76860 40406440 00000000 c11f5800 00000002 00000000 00000000 00000000 Call Trace: [<c01fc3bf>] [<c01fc40f>] [<c01fc9fa>] [<c01f8122>] [<c0147309>] [<c010871b>] Code: 0f b6 43 1b 8b 14 85 c4 a5 31 c0 69 c3 b1 79 37 9e 8b 8a c0 <3>X[5061] exited with preempt_count 1
I wanted to ask this last night already, but forgot :( To be honest, I also get kernel oops sometimes on X quit, and this started even before the new xfree .... Think it was somewhere with 2.4.19_pre and rmap/preempt kernels. Thus I want to ask you guys if it will be possible to try a vanilla 2.4.18 kernel and see if that fixes things (yeah, I know .. lot slower, etc). Thanks.
ya know i did that, and didnt have any agpgart problems with the vanilla kernel unfortunately, the tulip and 8139too network card modules it made wouldnt load on the bright side, my all in wonder card i stuck in works great so far
Guess we will have to wait for the kernel guys to fix this, if not already.
everything works fine when I compile the 2.4.18-6mdk kernel, available from ftp://ftp.cse.buffalo.edu/pub/Linux/Mandrake/mandrake/8.2/i586/Mandrake/RPMS.cooker/kernel-source-2.4.18-6mdk.i586.rpm *sigh*
suggest we close this as obsolete. I'm very happy with the xorg-x11 DRI :)