Well now that 2.5 is open I thought I'd add this to the wish list.
1) I wish that ECC/ECP/IEEE 1294 would be fixed in 2.4 and 2.5 and work right as this is causing problems in some webcams that worked in 2.2.x but now no longer work in 2.4. And since just about every major distribution is shipping 2.4 it makes it a little harder to use a 2.2 kernel if you get a dsitro as many are tuned to 2.4 and often don't work right with 2.2. This actually killed my web cam somewhere around 2.4.10~13.14?
2) The nasty VIA ide-floppy / iomega zip 100 drive bugs would get fixed as well as this prevents people from using their zip drives under Linux and forces them to use them under another OS.
just my .02 cents
On Wed, 5 Dec 2001 [email protected] wrote:
> 2) The nasty VIA ide-floppy / iomega zip 100 drive bugs would get fixed
> as well as this prevents people from using their zip drives under Linux
> and forces them to use them under another OS.
Last I checked, mine worked fine. Details please?
Dave.
--
| Dave Jones. http://www.codemonkey.org.uk
| SuSE Labs
[email protected] wrote:
> Well now that 2.5 is open I thought I'd add this to the wish list.
>
> 1) I wish that ECC/ECP/IEEE 1294 would be fixed in 2.4 and 2.5 and work right as this is causing problems in some webcams that worked in 2.2.x but now no longer work in 2.4. And since just about every major distribution is shipping 2.4 it makes it a little harder to use a 2.2 kernel if you get a dsitro as many are tuned to 2.4 and often don't work right with 2.2. This actually killed my web cam somewhere around 2.4.10~13.14?
>
> 2) The nasty VIA ide-floppy / iomega zip 100 drive bugs would get fixed as well as this prevents people from using their zip drives under Linux and forces them to use them under another OS.
I have a patch out, and am trying to convince Marcelo to include it. I
*DO* know about this as I get lots of the mail complaining about it. If
you are suffering, please try with the patch
Marcelo, can we put the ide-floppy patch in the next 2.4.17-pre? I'll
happily send you the patch again.
> just my .02 cents
and my 0.02 euros :-)
--
Paul
ide-floppy maintainer
Email:
[email protected]
Web:
http://paulbristow.net
ICQ:
11965223
On Thu, 6 Dec 2001, Paul Bristow wrote:
> [email protected] wrote:
>
> > Well now that 2.5 is open I thought I'd add this to the wish list.
> >
> > 1) I wish that ECC/ECP/IEEE 1294 would be fixed in 2.4 and 2.5 and work right as this is causing problems in some webcams that worked in 2.2.x but now no longer work in 2.4. And since just about every major distribution is shipping 2.4 it makes it a little harder to use a 2.2 kernel if you get a dsitro as many are tuned to 2.4 and often don't work right with 2.2. This actually killed my web cam somewhere around 2.4.10~13.14?
> >
> > 2) The nasty VIA ide-floppy / iomega zip 100 drive bugs would get fixed as well as this prevents people from using their zip drives under Linux and forces them to use them under another OS.
>
>
> I have a patch out, and am trying to convince Marcelo to include it. I
> *DO* know about this as I get lots of the mail complaining about it. If
> you are suffering, please try with the patch
>
> Marcelo, can we put the ide-floppy patch in the next 2.4.17-pre? I'll
> happily send you the patch again.
Send it again together with a full description of what the patch
changes, please.
ABIT KT7A MB
zip 100 ide drive on /dev/hdd
52x Blaster cdrom on /dev/hdc
WD 6Gig HD on /dev/hdb
WD 30Gig HD on /dev/hda
Athlon 1.2Ghz 266Mhz fsb
512Meg RAM
I believe this is the via686 chipset but would have top grep dmesg for I am not sure if it is A or B. the website where I got it says B, but I think in dmesg I have seen A, along with a message via workaround bug
isa modem /dev/tty03
pci SB live
pci dec tulip network card
pci sym53c810 scsi card && HP cd-rw
pci ieee1394 firewire card
agp GForce2 video card
ups on /dev/usb/ttyUSB0 using usb to serial converter
palm on com2 /dev/tty02
/dev/tty01 disabled MB only handles 2 com ports either built in or like my setup
I get messages like the following:
hdd lost interrupt (zip)
I also occasionally get hdc lost interrupt as well (cdrom) but have not seen them since I removed the zip drive
my hdb also had problems until I reformated ext2 it (loosing ALL data on it) (TGFBU)
hda is ext3 (built into kernel)
I work with some large files, oftern megs in size sometimes gigs.
If I mount the drive and do large data transfers to it forget it. 30Megs cp hangs. sometimes just doing touch will hang it
kill -9 on the cp process does not do anything.
umount wont umount.
shutdown becomes a defunct process and it is rather interesting to be able to do ps -ef and see 10 shutdown processes in the process table.
The system is usually still usable but the logs and console get message about lot interrupt.
The maintainer knows and I tried the patch he had against 2.4.14 and it helped but did not fix the problem.
fdisk against the drive works just fine, but using it actually seems to not work.
moved the drive into another system and the drive works just fine under other OS
put another hard drive on hdd and the other drive seemed to work just fine with data transfer of 1Gig file to the drive / umount / mount etc
I don't think the power is overloaded as if it was I think I'd have kernel compile problems especially doing make -j 5 on UP with X running as well as other progs.
Am now running 2.4.16 without the zip ini the system and all is okay except for the IEEE parport bug which breaks ecp/epp to my web cam. However since it worked in 2.4.13 I found the line of code and just patched back that one line.
parport is also VIA.. thus it must be a buggy chipset
Joe
Dave Jones <[email protected]> wrote:
> On Wed, 5 Dec 2001 [email protected] wrote:
> 2) The nasty VIA ide-floppy / iomega zip 100 drive bugs would get fixed
> as well as this prevents people from using their zip drives under Linux
> and forces them to use them under another OS.
Last I checked, mine worked fine. Details please?
Dave.
--
| Dave Jones. http://www.codemonkey.org.uk
| SuSE Labs
I tried the patch on your site against 2.4.14, it helped, system still hung...
hdd lost interrupt
is there buffereing done in data transfers to disks?
I.E. if I cp to a drive is the data transfered right away or is there a delay?
J
>I have a patch out, and am trying to convince Marcelo to include it. I
> *DO* know about this as I get lots of the mail complaining about it. > If
> you are suffering, please try with the patch
----- Original Message -----
From: <[email protected]>
To: "Dave Jones" <[email protected]>
Cc: <[email protected]>
Sent: Thursday, December 06, 2001 4:00 PM
Subject: Re: Re: 2.4.x 2.5.x wishlist
> ABIT KT7A MB
> zip 100 ide drive on /dev/hdd
> 52x Blaster cdrom on /dev/hdc
> WD 6Gig HD on /dev/hdb
> WD 30Gig HD on /dev/hda
> Athlon 1.2Ghz 266Mhz fsb
> 512Meg RAM
>
> I believe this is the via686 chipset but would have top grep dmesg for I
am not sure if it is A or B. the website where I got it says B, but I think
in dmesg I have seen A, along with a message via workaround bug
>
The ABIT KT7A uses the 686b southbridge.
--Wilson.
[email protected] wrote:
> I tried the patch on your site against 2.4.14, it helped, system still hung...
Damn. Lost interrupt is outside my code. This lives in the ide driver
proper and is probably relative toi the via82cxxx specific controller
code that is deep voodoo from Andre. I guess Andre/Jens have not found
all the *funnies* in the buggy chipset.
> hdd lost interrupt
>
> is there buffereing done in data transfers to disks?
Yes, but at the FS level, not in clever driver level stuff.
> I.E. if I cp to a drive is the data transfered right away or is there a delay?
>
> J
>
>
>>I have a patch out, and am trying to convince Marcelo to include it. I
>>*DO* know about this as I get lots of the mail complaining about it. > If
>>you are suffering, please try with the patch
>>
>
--
Paul
Email:
[email protected]
Web:
http://paulbristow.net
ICQ:
11965223
Thus my wish for 2.5.x. It seems that the problem I am having maybe in the ide driver itself. I searched the web on this a lot before I sent this to the list, and found a post from Alan Cox saying that this was a low priority.
Joe
> I tried the patch on your site against 2.4.14, it helped, system still hung...
Damn. Lost interrupt is outside my code. This lives in the ide driver
proper and is probably relative toi the via82cxxx specific controller
code that is deep voodoo from Andre. I guess Andre/Jens have not found
all the *funnies* in the buggy chipset.
> hdd lost interrupt
>
> is there buffereing done in data transfers to disks?
Yes, but at the FS level, not in clever driver level stuff.
> I.E. if I cp to a drive is the data transfered right away or is there a delay?
>
> J
>
>
>>I have a patch out, and am trying to convince Marcelo to include it. I
>>*DO* know about this as I get lots of the mail complaining about it. > If
>>you are suffering, please try with the patch
>>
>
--
Paul
Email:
[email protected]
Web:
http://paulbristow.net
ICQ:
11965223
On Thu, 6 Dec 2001, Paul Bristow wrote:
> [email protected] wrote:
>
> > I tried the patch on your site against 2.4.14, it helped, system still hung...
>
>
> Damn. Lost interrupt is outside my code. This lives in the ide driver
> proper and is probably relative toi the via82cxxx specific controller
> code that is deep voodoo from Andre. I guess Andre/Jens have not found
> all the *funnies* in the buggy chipset.
Paul I have to finish a packet_taskfile_ioctl for seeing if we have to
correct data-transport rules in place. For the longest time it has been a
pig in a poke. The ATAPI diagnostics toolkit will be significantly
tougher but still doable.
Regards,
Andre Hedrick
Linux ATA Development
Andre Hedrick wrote:
> > Damn. Lost interrupt is outside my code. This lives in the ide driver
> > proper and is probably relative toi the via82cxxx specific controller
> > code that is deep voodoo from Andre. I guess Andre/Jens have not found
> > all the *funnies* in the buggy chipset.
>
> Paul I have to finish a packet_taskfile_ioctl for seeing if we have to
> correct data-transport rules in place. For the longest time it has been a
> pig in a poke. The ATAPI diagnostics toolkit will be significantly
> tougher but still doable.
that's why I am hoping to get it into the wish list for 2.5...