LinuxLists.cc - CONFIG_FB_3DFX_ACCEL is unusable (in 2.6.7)

2004-06-16 19:24:04

Subject: CONFIG_FB_3DFX_ACCEL is unusable (in 2.6.7)

I suggest marking this option as BROKEN or remove at all for now.
It looks like some artifact of old hwcursor code which never worked in
2.6.x.
It uses some fields (namely par->hwcursor.*) never initialized,
uninitialized timer, contains obvious bugs (using "&&" for testing
bitfields) and finally causes oopses crashing system - like this:

kernel: Console: switching consoles 1-1 to colour frame buffer device 100x37
kernel: Console: switching consoles 2-2 to colour frame buffer device 100x37
kernel: Console: switching consoles 3-3 to colour frame buffer device 100x37
kernel: Console: switching consoles 4-4 to colour frame buffer device 100x37
kernel: Console: switching consoles 5-5 to colour frame buffer device 100x37
kernel: Console: switching consoles 6-6 to colour frame buffer device 100x37
kernel: Console: switching consoles 7-7 to colour frame buffer device 100x37
kernel: Uninitialised timer!
kernel: This is just a warning. Your computer is OK
kernel: function=0x00000000, data=0x0
kernel: [<c011b09f>] check_timer_failed+0x5f/0x70
kernel: [<c011b3c2>] del_timer+0x12/0x80
kernel: [<d88c4555>] tdfxfb_cursor+0x315/0x380 [tdfxfb]
kernel: [<c01e5c63>] fbcon_cursor+0x273/0x380
kernel: [<c01e59d3>] fbcon_putcs+0x93/0xb0
kernel: [<c01bee01>] do_update_region+0xc1/0x180
kernel: [<c01bf893>] set_cursor+0x63/0x80
kernel: [<c01bfa6c>] redraw_screen+0xfc/0x1e0
kernel: [<c01bf132>] update_attr+0xb2/0xd0
kernel: [<c01c3662>] take_over_console+0x2f2/0x420
kernel: [<c01e4931>] set_con2fb_map+0x41/0x50
kernel: [<c01e93bc>] fb_ioctl+0x2cc/0x330
kernel: [<c012c28d>] filemap_nopage+0x1ed/0x330
kernel: [<c0139009>] do_no_page+0x189/0x2e0
kernel: [<c013932e>] handle_mm_fault+0xbe/0x150
kernel: [<c010fbc9>] do_page_fault+0x2e9/0x4c6
kernel: [<c014e32f>] cdev_get+0x4f/0xb0
kernel: [<c014e06c>] chrdev_open+0xec/0x210
kernel: [<c01451e9>] dentry_open+0xf9/0x210
kernel: [<c01450ec>] filp_open+0x4c/0x50
kernel: [<c01568c9>] sys_ioctl+0x109/0x280
kernel: [<c0103c1b>] syscall_call+0x7/0xb
kernel:
kernel: Console: switching consoles 8-8 to colour frame buffer device 100x37
kernel: Console: switching consoles 9-9 to colour frame buffer device 100x37
kernel: Console: switching consoles 13-13 to colour frame buffer device 100x37
kernel: Console: switching consoles 14-14 to colour frame buffer device 100x37
kernel: Console: switching consoles 15-15 to colour frame buffer device 100x37
kernel: Console: switching consoles 16-16 to colour frame buffer device 100x37
kernel: Console: switching consoles 17-17 to colour frame buffer device 100x37
kernel: Console: switching consoles 18-18 to colour frame buffer device 100x37
kernel: Console: switching consoles 19-19 to colour frame buffer device 100x37
kernel: Console: switching consoles 20-20 to colour frame buffer device 100x37
kernel: Console: switching consoles 21-21 to colour frame buffer device 100x37
kernel: Unable to handle kernel NULL pointer dereference at virtual address 0000001f
kernel: printing eip:
kernel: d88c42dd
kernel: *pde = 00000000
kernel: Oops: 0000 [#1]
kernel: PREEMPT
kernel: Unable to handle kernel NULL pointer dereference at virtual address 0000002f
kernel: printing eip:
kernel: d88c42dd
kernel: *pde = 00000000
kernel: Oops: 0000 [#2]
kernel: PREEMPT
kernel: Modules linked in: binfmt_misc af_packet via686a i2c_sensor i2c_isa i2c_core tdfxfb cfbimgblt snd_ens1371 snd_rawmidi snd_seq_device snd_pcm snd_page_alloc snd_timer snd_ac97_codec snd ne2k_pci 8390 crc32 parport_pc parport via_agp agpgart evdev 8250 serial_core joydev analog es1371 soundcore gameport ac97_codec pcspkr uhci_hcd usbcore rtc
kernel: CPU: 0
kernel: EIP: 0060:[<d88c42dd>] Not tainted
kernel: EFLAGS: 00010086 (2.6.7)
kernel: EIP is at tdfxfb_cursor+0x9d/0x380 [tdfxfb]
kernel: eax: c01e59d3 ebx: d7fc8e01 ecx: 0000000f edx: 00000010
kernel: esi: c028ab10 edi: d265bb90 ebp: 0000000c esp: d265bb24
kernel: ds: 007b es: 007b ss: 0068
kernel: Process tput (pid: 5644, threadinfo=d265a000 task=d22bac90)
kernel: Stack: c01e4c93 00000008 d795866c d265bb90 d7958400 00000010 00000002

kernel: d7fc8e01 c028ab10 c01e59d3 ffffffff ffffffff d7fc8e00 00000002 d7958400
kernel: 0000000c c01e5d45 0000ffff 00000019 00000001 c0342708 c01e00ff 0000000f
kernel: Call Trace:
kernel: [<c01e4c93>] accel_putcs+0x253/0x2d0
kernel: [<c01e59d3>] fbcon_putcs+0x93/0xb0
kernel: [<c01e5d45>] fbcon_cursor+0x355/0x380
kernel: [<c01e00ff>] proc_ide_write_settings+0x14f/0x260
kernel: [<c01e59d3>] fbcon_putcs+0x93/0xb0
kernel: [<c01bf81d>] hide_cursor+0x1d/0x30
kernel: [<c01c2d94>] vt_console_print+0x2c4/0x2d0
kernel: [<c01bf893>] set_cursor+0x63/0x80
kernel: [<c01c2ad0>] vt_console_print+0x0/0x2d0
kernel: [<c0114762>] __call_console_drivers+0x42/0x50
kernel: [<c0114855>] call_console_drivers+0x75/0x100
kernel: [<c0114baa>] release_console_sem+0x4a/0xe0
kernel: [<c0114aac>] printk+0xfc/0x160
kernel: [<c01043f0>] die+0x80/0x100
kernel: [<c010fa97>] do_page_fault+0x1b7/0x4c6
kernel: [<c01ac422>] copy_to_user+0x32/0x50
kernel: [<c016a7d3>] load_elf_binary+0x513/0xc00
kernel: [<c010f8e0>] do_page_fault+0x0/0x4c6
kernel: [<c0103dc5>] error_code+0x2d/0x38
kernel: [<d88c42dd>] tdfxfb_cursor+0x9d/0x380 [tdfxfb]
kernel: [<c01e5c63>] fbcon_cursor+0x273/0x380
kernel: [<c01bf893>] set_cursor+0x63/0x80
kernel: [<c01c30e1>] con_flush_chars+0x31/0x40
kernel: [<c01c2f54>] con_write+0x24/0x40
kernel: [<c01b6186>] opost_block+0xc6/0x170
kernel: [<c012c247>] filemap_nopage+0x1a7/0x330
kernel: [<c0139009>] do_no_page+0x189/0x2e0
kernel: [<c01b834d>] write_chan+0x1ad/0x230
kernel: [<c0111360>] default_wake_function+0x0/0x10
kernel: [<c010fbc9>] do_page_fault+0x2e9/0x4c6
kernel: [<c0111360>] default_wake_function+0x0/0x10
kernel: [<c013a621>] vma_merge+0x121/0x190
kernel: [<c01b3291>] tty_write+0x1c1/0x260
kernel: [<c013ac72>] do_mmap_pgoff+0x512/0x650
kernel: [<c01b81a0>] write_chan+0x0/0x230
kernel: [<c0145f8d>] vfs_write+0xcd/0x120
kernel: [<c0146078>] sys_write+0x38/0x60
kernel: [<c0103c1b>] syscall_call+0x7/0xb
kernel:
kernel: Code: 0f b7 04 51 c1 e0 10 89 44 24 04 0f b7 04 53 c1 e0 08 09 44
kernel: <6>note: tput[5644] exited with preempt_count 2

--
Jakub Bogusz http://cyber.cs.net.pl/~qboosh/

2004-06-17 15:11:18

by Luiz Fernando N. Capitulino

[permalink] [raw]

Subject: Re: CONFIG_FB_3DFX_ACCEL is unusable (in 2.6.7)

Hello Jakub,

Em Wed, Jun 16, 2004 at 09:24:33PM +0200, Jakub Bogusz escreveu:

| I suggest marking this option as BROKEN or remove at all for now.

I enabled this option. This code was generating warnings, and I
prefer to enable it, otherwise people can forget it (I think
mark it as EXPERIMENTAL is enough).

This code is under development, anyway. Actually, I think people is
not working in it because there is no people to test. So, I
think you is welcome. :-)

I'll forward to fb-devel list.

| It looks like some artifact of old hwcursor code which never worked in
| 2.6.x.
| It uses some fields (namely par->hwcursor.*) never initialized,
| uninitialized timer, contains obvious bugs (using "&&" for testing
| bitfields) and finally causes oopses crashing system - like this:
|
| kernel: Console: switching consoles 1-1 to colour frame buffer device 100x37
| kernel: Console: switching consoles 2-2 to colour frame buffer device 100x37
| kernel: Console: switching consoles 3-3 to colour frame buffer device 100x37
| kernel: Console: switching consoles 4-4 to colour frame buffer device 100x37
| kernel: Console: switching consoles 5-5 to colour frame buffer device 100x37
| kernel: Console: switching consoles 6-6 to colour frame buffer device 100x37
| kernel: Console: switching consoles 7-7 to colour frame buffer device 100x37
| kernel: Uninitialised timer!
| kernel: This is just a warning. Your computer is OK
| kernel: function=0x00000000, data=0x0
| kernel: [<c011b09f>] check_timer_failed+0x5f/0x70
| kernel: [<c011b3c2>] del_timer+0x12/0x80
| kernel: [<d88c4555>] tdfxfb_cursor+0x315/0x380 [tdfxfb]
| kernel: [<c01e5c63>] fbcon_cursor+0x273/0x380
| kernel: [<c01e59d3>] fbcon_putcs+0x93/0xb0
| kernel: [<c01bee01>] do_update_region+0xc1/0x180
| kernel: [<c01bf893>] set_cursor+0x63/0x80
| kernel: [<c01bfa6c>] redraw_screen+0xfc/0x1e0
| kernel: [<c01bf132>] update_attr+0xb2/0xd0
| kernel: [<c01c3662>] take_over_console+0x2f2/0x420
| kernel: [<c01e4931>] set_con2fb_map+0x41/0x50
| kernel: [<c01e93bc>] fb_ioctl+0x2cc/0x330
| kernel: [<c012c28d>] filemap_nopage+0x1ed/0x330
| kernel: [<c0139009>] do_no_page+0x189/0x2e0
| kernel: [<c013932e>] handle_mm_fault+0xbe/0x150
| kernel: [<c010fbc9>] do_page_fault+0x2e9/0x4c6
| kernel: [<c014e32f>] cdev_get+0x4f/0xb0
| kernel: [<c014e06c>] chrdev_open+0xec/0x210
| kernel: [<c01451e9>] dentry_open+0xf9/0x210
| kernel: [<c01450ec>] filp_open+0x4c/0x50
| kernel: [<c01568c9>] sys_ioctl+0x109/0x280
| kernel: [<c0103c1b>] syscall_call+0x7/0xb
| kernel:
| kernel: Console: switching consoles 8-8 to colour frame buffer device 100x37
| kernel: Console: switching consoles 9-9 to colour frame buffer device 100x37
| kernel: Console: switching consoles 13-13 to colour frame buffer device 100x37
| kernel: Console: switching consoles 14-14 to colour frame buffer device 100x37
| kernel: Console: switching consoles 15-15 to colour frame buffer device 100x37
| kernel: Console: switching consoles 16-16 to colour frame buffer device 100x37
| kernel: Console: switching consoles 17-17 to colour frame buffer device 100x37
| kernel: Console: switching consoles 18-18 to colour frame buffer device 100x37
| kernel: Console: switching consoles 19-19 to colour frame buffer device 100x37
| kernel: Console: switching consoles 20-20 to colour frame buffer device 100x37
| kernel: Console: switching consoles 21-21 to colour frame buffer device 100x37
| kernel: Unable to handle kernel NULL pointer dereference at virtual address 0000001f
| kernel: printing eip:
| kernel: d88c42dd
| kernel: *pde = 00000000
| kernel: Oops: 0000 [#1]
| kernel: PREEMPT
| kernel: Unable to handle kernel NULL pointer dereference at virtual address 0000002f
| kernel: printing eip:
| kernel: d88c42dd
| kernel: *pde = 00000000
| kernel: Oops: 0000 [#2]
| kernel: PREEMPT
| kernel: Modules linked in: binfmt_misc af_packet via686a i2c_sensor i2c_isa i2c_core tdfxfb cfbimgblt snd_ens1371 snd_rawmidi snd_seq_device snd_pcm snd_page_alloc snd_timer snd_ac97_codec snd ne2k_pci 8390 crc32 parport_pc parport via_agp agpgart evdev 8250 serial_core joydev analog es1371 soundcore gameport ac97_codec pcspkr uhci_hcd usbcore rtc
| kernel: CPU: 0
| kernel: EIP: 0060:[<d88c42dd>] Not tainted
| kernel: EFLAGS: 00010086 (2.6.7)
| kernel: EIP is at tdfxfb_cursor+0x9d/0x380 [tdfxfb]
| kernel: eax: c01e59d3 ebx: d7fc8e01 ecx: 0000000f edx: 00000010
| kernel: esi: c028ab10 edi: d265bb90 ebp: 0000000c esp: d265bb24
| kernel: ds: 007b es: 007b ss: 0068
| kernel: Process tput (pid: 5644, threadinfo=d265a000 task=d22bac90)
| kernel: Stack: c01e4c93 00000008 d795866c d265bb90 d7958400 00000010 00000002
|
| kernel: d7fc8e01 c028ab10 c01e59d3 ffffffff ffffffff d7fc8e00 00000002 d7958400
| kernel: 0000000c c01e5d45 0000ffff 00000019 00000001 c0342708 c01e00ff 0000000f
| kernel: Call Trace:
| kernel: [<c01e4c93>] accel_putcs+0x253/0x2d0
| kernel: [<c01e59d3>] fbcon_putcs+0x93/0xb0
| kernel: [<c01e5d45>] fbcon_cursor+0x355/0x380
| kernel: [<c01e00ff>] proc_ide_write_settings+0x14f/0x260
| kernel: [<c01e59d3>] fbcon_putcs+0x93/0xb0
| kernel: [<c01bf81d>] hide_cursor+0x1d/0x30
| kernel: [<c01c2d94>] vt_console_print+0x2c4/0x2d0
| kernel: [<c01bf893>] set_cursor+0x63/0x80
| kernel: [<c01c2ad0>] vt_console_print+0x0/0x2d0
| kernel: [<c0114762>] __call_console_drivers+0x42/0x50
| kernel: [<c0114855>] call_console_drivers+0x75/0x100
| kernel: [<c0114baa>] release_console_sem+0x4a/0xe0
| kernel: [<c0114aac>] printk+0xfc/0x160
| kernel: [<c01043f0>] die+0x80/0x100
| kernel: [<c010fa97>] do_page_fault+0x1b7/0x4c6
| kernel: [<c01ac422>] copy_to_user+0x32/0x50
| kernel: [<c016a7d3>] load_elf_binary+0x513/0xc00
| kernel: [<c010f8e0>] do_page_fault+0x0/0x4c6
| kernel: [<c0103dc5>] error_code+0x2d/0x38
| kernel: [<d88c42dd>] tdfxfb_cursor+0x9d/0x380 [tdfxfb]
| kernel: [<c01e5c63>] fbcon_cursor+0x273/0x380
| kernel: [<c01bf893>] set_cursor+0x63/0x80
| kernel: [<c01c30e1>] con_flush_chars+0x31/0x40
| kernel: [<c01c2f54>] con_write+0x24/0x40
| kernel: [<c01b6186>] opost_block+0xc6/0x170
| kernel: [<c012c247>] filemap_nopage+0x1a7/0x330
| kernel: [<c0139009>] do_no_page+0x189/0x2e0
| kernel: [<c01b834d>] write_chan+0x1ad/0x230
| kernel: [<c0111360>] default_wake_function+0x0/0x10
| kernel: [<c010fbc9>] do_page_fault+0x2e9/0x4c6
| kernel: [<c0111360>] default_wake_function+0x0/0x10
| kernel: [<c013a621>] vma_merge+0x121/0x190
| kernel: [<c01b3291>] tty_write+0x1c1/0x260
| kernel: [<c013ac72>] do_mmap_pgoff+0x512/0x650
| kernel: [<c01b81a0>] write_chan+0x0/0x230
| kernel: [<c0145f8d>] vfs_write+0xcd/0x120
| kernel: [<c0146078>] sys_write+0x38/0x60
| kernel: [<c0103c1b>] syscall_call+0x7/0xb
| kernel:
| kernel: Code: 0f b7 04 51 c1 e0 10 89 44 24 04 0f b7 04 53 c1 e0 08 09 44
| kernel: <6>note: tput[5644] exited with preempt_count 2
|
|
| --
| Jakub Bogusz http://cyber.cs.net.pl/~qboosh/
| -
| To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
| the body of a message to [email protected]
| More majordomo info at http://vger.kernel.org/majordomo-info.html
| Please read the FAQ at http://www.tux.org/lkml/

--
Luiz Fernando N. Capitulino
<http://www.telecentros.sp.gov.br>

2004-06-17 15:56:04

by Luiz Fernando N. Capitulino

[permalink] [raw]