http://www.linuxdiskcert.org/ide-2.4.19-p6.all.convert.3.patch.bz2
Spinlock issues address and 99% fixed.
6 new chipsets.
Taskfile IO should be stable.
Borrowed cleanups from 2.5.5 credit given to <martin d>
Bunches of patchs that have been submitted but not added until now.
Thanks for all the support gang ... this should make life a little better.
AEC6280 U100
AEC6880 U133
SI680 U133
CSB6 U100
HPT374 U100(U133??)
HPT372
Intel ICH3
VIA latest (via vojtech)
Super Taskfile special vender modes, and transparent to API.
Things which can not be fixed yet.
Spinlock sleep during probe to get.
Spinlock on pci-proc operations.
Process to convert codebase to as single filter/mask ruleset for timings.
Cheers,
Andre Hedrick
LAD Storage Consulting Group
On approximately Mon, Apr 15, 2002 at 11:39:14AM -0700, Andre Hedrick wrote:
>
> http://www.linuxdiskcert.org/ide-2.4.19-p6.all.convert.3.patch.bz2
>
Just got this when compiling...
drivers/ide/idedriver.o: In function `task_no_data_intr':
drivers/ide/idedriver.o(.text+0x9f1): undefined reference to `ide__sti'
drivers/ide/idedriver.o: In function `flagged_task_no_data_intr':
drivers/ide/idedriver.o(.text+0x14c7): undefined reference to `ide__sti'
drivers/ide/idedriver.o: In function `ide_cmd_ioctl':
drivers/ide/idedriver.o(.text+0x23e0): undefined reference to `set_transfer'
drivers/ide/idedriver.o(.text+0x23fc): undefined reference to `ide_ata66_check'
drivers/ide/idedriver.o(.text+0x24a7): undefined reference to `ide_driveid_update'
drivers/ide/idedriver.o: In function `check_dma_crc':
drivers/ide/idedriver.o(.text+0x2fc3): undefined reference to `ide_auto_reduce_xfer'
drivers/ide/idedriver.o: In function `ide_dump_status':
drivers/ide/idedriver.o(.text+0x34c8): undefined reference to `ide__sti'
drivers/ide/idedriver.o: In function `drive_cmd_intr':
drivers/ide/idedriver.o(.text+0x3b21): undefined reference to `ide__sti'
drivers/ide/idedriver.o: In function `ide_wait_stat':
drivers/ide/idedriver.o(.text+0x3cd5): undefined reference to `ide__sti'
drivers/ide/idedriver.o: In function `ide_do_request':
drivers/ide/idedriver.o(.text+0x42c9): undefined reference to `ide__sti'
drivers/ide/idedriver.o: In function `ide_intr':
drivers/ide/idedriver.o(.text+0x48f3): undefined reference to `ide__sti'
drivers/ide/idedriver.o: In function `report_drive_dmaing':
drivers/ide/idedriver.o(.text+0x784a): undefined reference to `eighty_ninty_three'
drivers/ide/idedriver.o(.text+0x789e): undefined reference to `eighty_ninty_three'
drivers/ide/idedriver.o: In function `config_drive_for_dma':
drivers/ide/idedriver.o(.text+0x79ce): undefined reference to `eighty_ninty_three'
drivers/ide/idedriver.o(.text+0x79fe): undefined reference to `eighty_ninty_three'
drivers/ide/idedriver.o: In function `via_set_drive':
drivers/ide/idedriver.o(.text+0x91a8): undefined reference to `ide_config_drive_speed'
drivers/ide/idedriver.o: In function `actual_try_to_identify':
drivers/ide/idedriver.o(.text+0xaaec): undefined reference to `ide__sti'
make: *** [vmlinux] Error 1
This comes at the final linking stage.
Thanks for the great stuff Andre.
Josh McKinney
Please send me your .config cause this built clean ...
On Mon, 15 Apr 2002, Josh McKinney wrote:
> On approximately Mon, Apr 15, 2002 at 11:39:14AM -0700, Andre Hedrick wrote:
> >
> > http://www.linuxdiskcert.org/ide-2.4.19-p6.all.convert.3.patch.bz2
> >
> Just got this when compiling...
>
> drivers/ide/idedriver.o: In function `task_no_data_intr':
> drivers/ide/idedriver.o(.text+0x9f1): undefined reference to `ide__sti'
> drivers/ide/idedriver.o: In function `flagged_task_no_data_intr':
> drivers/ide/idedriver.o(.text+0x14c7): undefined reference to `ide__sti'
> drivers/ide/idedriver.o: In function `ide_cmd_ioctl':
> drivers/ide/idedriver.o(.text+0x23e0): undefined reference to `set_transfer'
> drivers/ide/idedriver.o(.text+0x23fc): undefined reference to `ide_ata66_check'
> drivers/ide/idedriver.o(.text+0x24a7): undefined reference to `ide_driveid_update'
> drivers/ide/idedriver.o: In function `check_dma_crc':
> drivers/ide/idedriver.o(.text+0x2fc3): undefined reference to `ide_auto_reduce_xfer'
> drivers/ide/idedriver.o: In function `ide_dump_status':
> drivers/ide/idedriver.o(.text+0x34c8): undefined reference to `ide__sti'
> drivers/ide/idedriver.o: In function `drive_cmd_intr':
> drivers/ide/idedriver.o(.text+0x3b21): undefined reference to `ide__sti'
> drivers/ide/idedriver.o: In function `ide_wait_stat':
> drivers/ide/idedriver.o(.text+0x3cd5): undefined reference to `ide__sti'
> drivers/ide/idedriver.o: In function `ide_do_request':
> drivers/ide/idedriver.o(.text+0x42c9): undefined reference to `ide__sti'
> drivers/ide/idedriver.o: In function `ide_intr':
> drivers/ide/idedriver.o(.text+0x48f3): undefined reference to `ide__sti'
> drivers/ide/idedriver.o: In function `report_drive_dmaing':
> drivers/ide/idedriver.o(.text+0x784a): undefined reference to `eighty_ninty_three'
> drivers/ide/idedriver.o(.text+0x789e): undefined reference to `eighty_ninty_three'
> drivers/ide/idedriver.o: In function `config_drive_for_dma':
> drivers/ide/idedriver.o(.text+0x79ce): undefined reference to `eighty_ninty_three'
> drivers/ide/idedriver.o(.text+0x79fe): undefined reference to `eighty_ninty_three'
> drivers/ide/idedriver.o: In function `via_set_drive':
> drivers/ide/idedriver.o(.text+0x91a8): undefined reference to `ide_config_drive_speed'
> drivers/ide/idedriver.o: In function `actual_try_to_identify':
> drivers/ide/idedriver.o(.text+0xaaec): undefined reference to `ide__sti'
> make: *** [vmlinux] Error 1
>
> This comes at the final linking stage.
>
> Thanks for the great stuff Andre.
>
> Josh McKinney
> -
> 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/
>
Andre Hedrick
LAD Storage Consulting Group
http://www.linuxdiskcert.org/ide-2.4.19-p6.all.convert.3a.patch.bz2
Lets try it again, I diffed the wrong tree :-/
On Mon, 15 Apr 2002, Josh McKinney wrote:
> On approximately Mon, Apr 15, 2002 at 11:39:14AM -0700, Andre Hedrick wrote:
> >
> > http://www.linuxdiskcert.org/ide-2.4.19-p6.all.convert.3a.patch.bz2
> >
> Just got this when compiling...
>
> drivers/ide/idedriver.o: In function `task_no_data_intr':
> drivers/ide/idedriver.o(.text+0x9f1): undefined reference to `ide__sti'
> drivers/ide/idedriver.o: In function `flagged_task_no_data_intr':
> drivers/ide/idedriver.o(.text+0x14c7): undefined reference to `ide__sti'
> drivers/ide/idedriver.o: In function `ide_cmd_ioctl':
> drivers/ide/idedriver.o(.text+0x23e0): undefined reference to `set_transfer'
> drivers/ide/idedriver.o(.text+0x23fc): undefined reference to `ide_ata66_check'
> drivers/ide/idedriver.o(.text+0x24a7): undefined reference to `ide_driveid_update'
> drivers/ide/idedriver.o: In function `check_dma_crc':
> drivers/ide/idedriver.o(.text+0x2fc3): undefined reference to `ide_auto_reduce_xfer'
> drivers/ide/idedriver.o: In function `ide_dump_status':
> drivers/ide/idedriver.o(.text+0x34c8): undefined reference to `ide__sti'
> drivers/ide/idedriver.o: In function `drive_cmd_intr':
> drivers/ide/idedriver.o(.text+0x3b21): undefined reference to `ide__sti'
> drivers/ide/idedriver.o: In function `ide_wait_stat':
> drivers/ide/idedriver.o(.text+0x3cd5): undefined reference to `ide__sti'
> drivers/ide/idedriver.o: In function `ide_do_request':
> drivers/ide/idedriver.o(.text+0x42c9): undefined reference to `ide__sti'
> drivers/ide/idedriver.o: In function `ide_intr':
> drivers/ide/idedriver.o(.text+0x48f3): undefined reference to `ide__sti'
> drivers/ide/idedriver.o: In function `report_drive_dmaing':
> drivers/ide/idedriver.o(.text+0x784a): undefined reference to `eighty_ninty_three'
> drivers/ide/idedriver.o(.text+0x789e): undefined reference to `eighty_ninty_three'
> drivers/ide/idedriver.o: In function `config_drive_for_dma':
> drivers/ide/idedriver.o(.text+0x79ce): undefined reference to `eighty_ninty_three'
> drivers/ide/idedriver.o(.text+0x79fe): undefined reference to `eighty_ninty_three'
> drivers/ide/idedriver.o: In function `via_set_drive':
> drivers/ide/idedriver.o(.text+0x91a8): undefined reference to `ide_config_drive_speed'
> drivers/ide/idedriver.o: In function `actual_try_to_identify':
> drivers/ide/idedriver.o(.text+0xaaec): undefined reference to `ide__sti'
> make: *** [vmlinux] Error 1
>
> This comes at the final linking stage.
>
> Thanks for the great stuff Andre.
>
> Josh McKinney
> -
> 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/
>
Andre Hedrick
LAD Storage Consulting Group
On approximately Mon, Apr 15, 2002 at 12:59:20PM -0700, Andre Hedrick wrote:
>
> http://www.linuxdiskcert.org/ide-2.4.19-p6.all.convert.3a.patch.bz2
>
> Lets try it again, I diffed the wrong tree :-/
>
Well it compiled fine this time, but it seems to have problems with gcc-3.0.4 I am using to
compile kernels. This gcc has been fine for all other uses, but it must be something with this
particular build because gcc-2.95.4 works without a problem. Anyways, here is the oops report
if it may interest you.
<1>Unable to handle kernel paging request at virtual address 3fff432a
c01c2c90
*pde = 00000000
Oops: 0000
CPU: 0
EIP: 0010:[<c01c2c90>] Not tainted
Using defaults from ksymoops -t elf32-i386 -a i386
EFLAGS: 00010202
eax: 3fff427a ebx: c02fc928 ecx: 00000001 edx: dda11e04
esi: c02fc928 edi: dda11e98 ebp: bffff7f0 esp: dda11e04
ds: 0018 es: 0018 ss: 0018
Process hdparm (pid: 492, stackpage=dda11000)
Stack: 3fff427a 00100000 0258e100 0010003f 000e0000 2d575744 39504d41 30353136
31003238 00000000 10000003 30320028 35422e30 57443032 57444320 30423630
3332422d 41304358 20202020 20202020 20202020 20202020 20202020 80102020
Code: 0f b7 80 b0 00 00 00 8b 93 f0 00 00 00 66 89 82 b0 00 00 00
>>EIP; c01c2c90 <ide_driveid_update+30/90> <=====
>>eax; 3fff427a Before first symbol
>>ebx; c02fc928 <ide_hwifs+388/20a8>
>>edx; dda11e04 <_end+1d70bb50/2253ad4c>
>>esi; c02fc928 <ide_hwifs+388/20a8>
>>edi; dda11e98 <_end+1d70bbe4/2253ad4c>
>>ebp; bffff7f0 Before first symbol
>>esp; dda11e04 <_end+1d70bb50/2253ad4c>
Code; c01c2c90 <ide_driveid_update+30/90>
00000000 <_EIP>:
Code; c01c2c90 <ide_driveid_update+30/90> <=====
0: 0f b7 80 b0 00 00 00 movzwl 0xb0(%eax),%eax <=====
Code; c01c2c97 <ide_driveid_update+37/90>
7: 8b 93 f0 00 00 00 mov 0xf0(%ebx),%edx
Code; c01c2c9d <ide_driveid_update+3d/90>
d: 66 89 82 b0 00 00 00 mov %ax,0xb0(%edx)
Yep,
Edit ide-taskfile.c find ide_driveid_update
Changes the #if 1 to #if 0
Attempting to make the code compress but it appears that entry for
updating is more than a simple direct call. Still banging it out.
Cheers,
On Mon, 15 Apr 2002, Josh McKinney wrote:
> On approximately Mon, Apr 15, 2002 at 12:59:20PM -0700, Andre Hedrick wrote:
> >
> > http://www.linuxdiskcert.org/ide-2.4.19-p6.all.convert.3a.patch.bz2
> >
> > Lets try it again, I diffed the wrong tree :-/
> >
>
> Well it compiled fine this time, but it seems to have problems with gcc-3.0.4 I am using to
> compile kernels. This gcc has been fine for all other uses, but it must be something with this
> particular build because gcc-2.95.4 works without a problem. Anyways, here is the oops report
> if it may interest you.
>
> <1>Unable to handle kernel paging request at virtual address 3fff432a
> c01c2c90
> *pde = 00000000
> Oops: 0000
> CPU: 0
> EIP: 0010:[<c01c2c90>] Not tainted
> Using defaults from ksymoops -t elf32-i386 -a i386
> EFLAGS: 00010202
> eax: 3fff427a ebx: c02fc928 ecx: 00000001 edx: dda11e04
> esi: c02fc928 edi: dda11e98 ebp: bffff7f0 esp: dda11e04
> ds: 0018 es: 0018 ss: 0018
> Process hdparm (pid: 492, stackpage=dda11000)
> Stack: 3fff427a 00100000 0258e100 0010003f 000e0000 2d575744 39504d41 30353136
> 31003238 00000000 10000003 30320028 35422e30 57443032 57444320 30423630
> 3332422d 41304358 20202020 20202020 20202020 20202020 20202020 80102020
> Code: 0f b7 80 b0 00 00 00 8b 93 f0 00 00 00 66 89 82 b0 00 00 00
>
>
> >>EIP; c01c2c90 <ide_driveid_update+30/90> <=====
>
> >>eax; 3fff427a Before first symbol
> >>ebx; c02fc928 <ide_hwifs+388/20a8>
> >>edx; dda11e04 <_end+1d70bb50/2253ad4c>
> >>esi; c02fc928 <ide_hwifs+388/20a8>
> >>edi; dda11e98 <_end+1d70bbe4/2253ad4c>
> >>ebp; bffff7f0 Before first symbol
> >>esp; dda11e04 <_end+1d70bb50/2253ad4c>
>
> Code; c01c2c90 <ide_driveid_update+30/90>
> 00000000 <_EIP>:
> Code; c01c2c90 <ide_driveid_update+30/90> <=====
> 0: 0f b7 80 b0 00 00 00 movzwl 0xb0(%eax),%eax <=====
> Code; c01c2c97 <ide_driveid_update+37/90>
> 7: 8b 93 f0 00 00 00 mov 0xf0(%ebx),%edx
> Code; c01c2c9d <ide_driveid_update+3d/90>
> d: 66 89 82 b0 00 00 00 mov %ax,0xb0(%edx)
>
> -
> 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/
>
Andre Hedrick
LAD Storage Consulting Group
On Mon, Apr 15 2002, Andre Hedrick wrote:
>
> http://www.linuxdiskcert.org/ide-2.4.19-p6.all.convert.3.patch.bz2
seems that you haven't fixed the case where you cannot enable dma on an
atapi drive if CONFIG_IDEDMA_ONLYDISK has been selected? to me this is a
bug -- I don't want auto-enable of dma on my atapi drives, but I surely
want to be able to hdparm -d1 them at will later on.
--
Jens Axboe
It is 99% usage for distro to ship stable booting kernels given atapi is
flakey. However if you have a suggestion I am accepting options
--andre
On Tue, 16 Apr 2002, Jens Axboe wrote:
> On Mon, Apr 15 2002, Andre Hedrick wrote:
> >
> > http://www.linuxdiskcert.org/ide-2.4.19-p6.all.convert.3.patch.bz2
>
> seems that you haven't fixed the case where you cannot enable dma on an
> atapi drive if CONFIG_IDEDMA_ONLYDISK has been selected? to me this is a
> bug -- I don't want auto-enable of dma on my atapi drives, but I surely
> want to be able to hdparm -d1 them at will later on.
>
> --
> Jens Axboe
>
On Mon, Apr 15 2002, Andre Hedrick wrote:
>
> It is 99% usage for distro to ship stable booting kernels given atapi is
> flakey. However if you have a suggestion I am accepting options
I agree with the option, just not the way it is implemented. If I want
to enable dma after boot, I must be able to do so.
I have a solution, the current SuSE kernel gets it right. I'll backport
that to 2.4.19-pre once this patch is in.
--
Jens Axboe
This patch is not going anywhere for a while, until all the archs have
consistant local_*_*() spinlock but I noted what needs to be fixed and who
broke what.
I have another method but it was not reasonable or predictable until I
finish the driver conversion to fully modular and delete 95% of the
duplicated code mess I made over the years.
Your are free to copy it into 2.5 as always.
regards,
--andre
On Tue, 16 Apr 2002, Jens Axboe wrote:
> On Mon, Apr 15 2002, Andre Hedrick wrote:
> >
> > It is 99% usage for distro to ship stable booting kernels given atapi is
> > flakey. However if you have a suggestion I am accepting options
>
> I agree with the option, just not the way it is implemented. If I want
> to enable dma after boot, I must be able to do so.
>
> I have a solution, the current SuSE kernel gets it right. I'll backport
> that to 2.4.19-pre once this patch is in.
>
> --
> Jens Axboe
>