Linux 2.4.18-pre9-ac1
[+ indicates stuff that went to Marcelo, o stuff that has not,
* indicates stuff that is merged in mainstream now, X stuff that proved bad and was dropped out]
Linux 2.4.18pre9-ac1
o Initial merge of DVD card driver (Christian Wolff,Marcus Metzler)
| This is just an initial testing piece. DVB needs merging | properly and this is only a first bit of testing
o Random number generator support for AMD768 (me)
o Add AMD768 to i810 driver pci ident list (me)
o Initial AMD768 power management work (me)
| Unfinished pending some docs clarifications
o Fix bugbuf mishandling for modular es1370 (me)
o Fix up i2o readl abuse, post_wait race, and (me, Arjan van de Ven)
some deadlock cases
o Added cpu_relax to yam driver (me)
o Fixup AMD762 if the BIOS apparently got it wrong(me)
(eg ASUS boards)
o MP1.4 alignment fixup
o pcwd cleanup, backport of fixes from 2.5 (Rob Radez)
o Add support for more Moxa cards to mxser (Damian Wrobel)
o Add remaining missing MODULE_LICENSE tags (Hubert Mantel)
o Fix floppy reservation ranges (Anton Altaparmakov)
o Fix max file size setup (Andi Kleen)
Linux 2.4.18pre7-ac3
o Fix a wrong error return in the megaraid driver (Arjan van de Ven)
* FreeVXFS update (Christoph Hellwig)
+ Qnxfs update (Anders Larsen)
o Fix non compile with PCI=n (Adrian Bunk)
o Fix DRM 4.0 non compile in i810 (me)
o Drop out now dead CLONE thread/parent fixup (Dave McCracken)
* Make NetROM incoming frame check stricter (Tomi Manninen)
* Use sock_orphan in AX.25/NetROM (Jeroen PE1RXQ)
o Pegasus update (Petko Manolov)
o Make reparent_to_init and exec_usermodehelper (Andrew Morton)
use set_user, fix a tiny set_user SMP race
o Mark framebuffer mappings VM_IO (Andrew Morton)
o Neomagic frame buffer driver (Denis Kropp)
- Needs FPU code fixing before it can be merged
o Hyperthreading awareness for MTRR driver
o Correct NR_IRQ with no apic support (Brian Gerst)
* Fix missing includes in sound drivers (Michal Jaegermann)
Linux 2.4.18pre7-ac2
* i810 audio driver update (Doug Ledford)
o Early ioremap for x86 specific code (Mikael Pettersson)
| This is needed to do things like apic/dmi detect early enough
o Pentium IV APIC/NMI watchdog (Mikael Pettersson)
* Add C1MRX support to sonypi driver (Junichi Morita)
* Fix "make rpm" with two '-' in extraversion (Gerald Britton)
o Fix aacraid hang/irq storm on i960 boards (Chris Pascoe)
* Fix isdn audio compiler behaviour dependancy (Urs Thuermann)
* YAM driver fixes (Jean-Paul Roubelat)
* ROSE protocol stack update/fixes (Jean-Paul Roubelat)
o Fix UFS/CDROM oops (Zwane Mwaikambo)
o Fix nm256 hang on Dell Latitude (origin unknown)
| Please test this tree with other NM256 based boxes and check | those still work...
o Merge PnPBIOS patch (Thomas Hood, David Hinds, Tom Lees,
Christian Schmidt, ..)
o Merge new sis frame buffer drivers (Thomas Winischhofer)
* cs46xx oops fix (Mike Gorse)
* Fix a second cs46xx bug related to this (me)
o Fix acpitable oopses on boot and other problems (James Cleverdon)
o Fix io port type on the hpt366 driver (Pete Popov)
o Updated matrox drivers (Petr Vandrovec)
* IPchains fixes needed for 2.4.18pre7
o IDE config text updates for the IDE patches (Anton Altaparmakov)
o Merge the first bits of ZV support (Marcus Metzler)
o Add initial ZV support to yenta socket driver (me)
for TI cards
o Fix pirq routing on the CS5530 (me)
| Finally the palmax pcmcia/cardbus works properly
Linux 2.4.18pre7-ac1
o Merge with 2.4.18pre7 (Arjan van de Ven)
| + some quota fixups redone by me
| several 18pre7 netfilter bugs left unfixed for now
o Rmap-12a (Rik van Riel and co)
Linux 2.4.18pre3-ac2
o Re-merge the IDE patches (Andre Hedrick and co)
* Fix check/request region in ali_ircc and lowcomx(Steven Walter)
com90xx, sealevel, sb1000
* Remove unused message from 6pack driver (Adrian Bunk)
* Fix unused variable warning in i60scsi (Adrian Bunk)
* Fix off by one floppy oops (Keith Owens)
o Fix i2o_config use of undefined C (Andreas Dilger)
* Fix fdomain scsi oopses (Per Larsson)
* Fix sf16fmi hang on boot (me)
o Add bridge resources to the resource tree (Ivan Kokshaysky)
* Fix iphase ATM oops on close in on case (Till Immanuel Patzschke)
* Enable OOSTORE on winchip processors (Dave Jones, me)
| Worth about 10-20% performance
* Code Page 1250 support (Petr Titera)
* Fix sdla and hpfs doc typos (Sven Vermeulen)
o Document /proc/stat (Sven Heinicke)
* Update cs4281 drivers (Tom Woller)
| Fixes xmms stutter, remove wrapper code | handle tosh boxes, allow record device change | trigger wakeups on ioctl triggered changes
+/o Fix locking of file struct stuff found by ibm (Dipankar Sarma)
audit
o Use spin_lock_init in serial.c (Dave Miller)
* Fix AF_UNIX shutdown bug (Dave Miller)
Linux 2.4.18pre3-ac1
o 32bit uid quota
o rmap-11b VM (Rik van Riel,
William Irwin etc)
* Make scsi printer visible (Stefan Wieseckel)
* Report Hercules Fortissimo card (Minya Sorakinu)
* Fix O_NDELAY close mishandling on the following (me)
sound cards: cmpci, cs46xx, es1370, es1371, esssolo1, sonicvibes
* tdfx pixclock handling fix (Jurriaan)
o Fix mishandling of file system size limiting (Andrea Arcangeli)
* generic_serial cleanups (Rasmus Andersen)
o serial.c locking fixes for SMP - move from cli (Kees)
too
* Truncate fixes from old -ac tree (Andrew Morton)
* Hopefully fix the i2o oops (me)
| Not the right fix but it'll do till I rewrite this
* Fix non blocking tty blocking bug (Peter Benie)
o IRQ routing workaround for problem HP laptops (Cory Bell)
* Fix the rcpci driver (Pete Popov)
* Fix documentation of aedsp location (Adrian Bunk)
* Fix the worst of the APM ate my cpu problems (Andreas Steinmetz)
* Correct icmp documentation (Pierre Lombard)
* Multiple mxser crash on boot fix (Stephan von Krawczynski)
o ldm header fix (Anton Altaparmakov)
* Fix unchecked kmalloc in i2c_proc (Ragnar Hojland Espinosa)
* Fix unchecked kmalloc in airo_cs (Ragnar Hojland Espinosa)
* Fix unchecked kmalloc in btaudio (Ragnar Hojland Espinosa)
* Fix unchecked kmalloc in qnx4/inode.c (Ragnar Hojland Espinosa)
* Disable DRM4.1 GMX2000 driver (4.0 required) (me)
* Fix sb16 lower speed limit bug (Jori Liesenborgs)
o Fix compilation of orinoco driver (Ben Herrenschmidt)
* ISAPnP init fix (Chris Rankin)
o Export release_console_sem (Andrew Morton)
* Output nat crash fix (Rusty Russell)
* Fix PLIP (Niels Jensen)
o Natsemi driver hang fix (Manfred Spraul)
* Add mono/stereo reporting to gemtek pci radio (Jonathan Hudson)
Quoted 5 lines · Alan Cox
[+ indicates stuff that went to Marcelo, o stuff that has not,
* indicates stuff that is merged in mainstream now, X stuff that proved bad and was dropped out]Linux 2.4.18pre9-ac1
I notice that in linux/drivers/scsi/scsi_merge.c you seem to be reverting the MO drive clustering fix from Jens:
http://www.uwsg.indiana.edu/hypermail/linux/kernel/0202.0/1321.html
Was this intentional? If so, why?
-Erik
--
Erik B. Andersen http://codepoet-consulting.com/
--This message was written using 73% post-consumer electrons--
Quoted 5 lines
I notice that in
linux/drivers/scsi/scsi_merge.cyou seem to be reverting the MO drive clustering fix from Jens:http://www.uwsg.indiana.edu/hypermail/linux/kernel/0202.0/1321.htmlWas this intentional? If so, why?
I want to find out why it was done first and then test it. Leaving it out will ensure it bugs me until I test it
Quoted 8 lines · Alan Cox
Quoted 5 lines
I notice that in
linux/drivers/scsi/scsi_merge.cyou seem to be reverting the MO drive clustering fix from Jens:http://www.uwsg.indiana.edu/hypermail/linux/kernel/0202.0/1321.htmlWas this intentional? If so, why?
I want to find out why it was done first and then test it. Leaving it out will ensure it bugs me until I test it
If you leave it out, you surely want to make sure that the other request init and re-init paths agree on the clustering for MO devices. Because they don't.
As far as I'm concerned, removing the MO conditional wrt clustering is the right fix.
Signature · 1 line
Jens Axboe
Quoted 10 lines · Jens Axboe
Quoted 8 lines · Alan Cox
Quoted 5 lines
I notice that in
linux/drivers/scsi/scsi_merge.cyou seem to be reverting the MO drive clustering fix from Jens:http://www.uwsg.indiana.edu/hypermail/linux/kernel/0202.0/1321.htmlWas this intentional? If so, why?
I want to find out why it was done first and then test it. Leaving it out will ensure it bugs me until I test it
Quoted 6 lines
If you leave it out, you surely want to make sure that the other request init and re-init paths agree on the clustering for MO devices. Because they don't.
As far as I'm concerned, removing the MO conditional wrt clustering is the right fix.
BTW, if you are concerned with the write/read vs seek latencies of MO drives, then the disable clustering hack was definitely the wrong way to try and limit request sizes. In fact it achieved absolutely _nothing_. Clustering at this level is completely device independent, too.
In short, the old code made no sense whatsoever.
Now, disabling request merging for MO devices might make a whole lot more sense. That might be worth while trying, and I'd be happy to give you a patch to try that out instead.
Signature · 1 line
Jens Axboe
Quoted 6 lines
Quoted 6 lines
Quoted 2 lines
I want to find out why it was done first and then test it. Leaving it out will ensure it bugs me until I test it
If you leave it out, you surely want to make sure that the other request init and re-init paths agree on the clustering for MO devices. Because they don't.
No - I want to run a test set with an M/O drive before and after the change and see what it shows in real life. I suspect nothing much.
Quoted 3 lines
Now, disabling request merging for MO devices might make a whole lot more sense. That might be worth while trying, and I'd be happy to give you a patch to try that out instead.
I don't think that should be required actually. The killer on M/O disks is seek time, and to an extent rotational latency (its 3 trips round a cheaper M/O disk to rewrite a sector). If anything clustering writes to the same track should be a big win.
Alan
Quoted 4 lines · Alan Cox
I don't think that should be required actually. The killer on M/O disks is seek time, and to an extent rotational latency (its 3 trips round a cheaper M/O disk to rewrite a sector). If anything clustering writes to the same track should be a big win.
I believe the impetus to the cluster patch is not to address parformance, but because without it a media error on the MO causes a system failure. That seems a good reason to put in the patch, and you can certainly test it with and without, just be sure to sync() before trying the standard code ;-)
Signature · 3 lines
bill davidsen <[email protected]> CTO, TMR Associates, Inc Doing interesting things with little computers since 1979.
Quoted 9 lines · Alan Cox
Quoted 6 lines
Quoted 6 lines
Quoted 2 lines
I want to find out why it was done first and then test it. Leaving it out will ensure it bugs me until I test it
If you leave it out, you surely want to make sure that the other request init and re-init paths agree on the clustering for MO devices. Because they don't.
No - I want to run a test set with an M/O drive before and after the change and see what it shows in real life. I suspect nothing much.
That completely ignores that there is a _bug_ there currently. There was a reason I removed disabled clustering, you know... 2.4 before the change oopses, 2.4.18-preX (forget when Marcelo took the patch) has it fixed.
Quoted 8 lines
Quoted 3 lines
Now, disabling request merging for MO devices might make a whole lot more sense. That might be worth while trying, and I'd be happy to give you a patch to try that out instead.
I don't think that should be required actually. The killer on M/O disks is seek time, and to an extent rotational latency (its 3 trips round a cheaper M/O disk to rewrite a sector). If anything clustering writes to the same track should be a big win.
You are probably right. Can't make pigs fly :-)
Signature · 1 line
Jens Axboe
Quoted 9 lines · Alan Cox
Quoted 6 lines
Quoted 6 lines
Quoted 2 lines
I want to find out why it was done first and then test it. Leaving it out will ensure it bugs me until I test it
If you leave it out, you surely want to make sure that the other request init and re-init paths agree on the clustering for MO devices. Because they don't.
No - I want to run a test set with an M/O drive before and after the change and see what it shows in real life. I suspect nothing much.
I was able to hang the kernel several times while talking to my MO drive prior to the fix...
-Erik
--
Erik B. Andersen http://codepoet-consulting.com/
--This message was written using 73% post-consumer electrons--
Quoted 2 lines
Linux 2.4.18pre9-ac1
...
Does it contain rmap-12e or still only 12a (as per the changelog)?
Signature · 1 line
Balazs Pozsar
Quoted 4 lines
Quoted 1 line
Linux 2.4.18pre9-ac1
...
Does it contain rmap-12e or still only 12a (as per the changelog)?
Still 12a - I've not done the 12a->e merge and I've seen problems with b-d
Quoted 5 lines · Pozsar Balazs
Quoted 2 lines
Linux 2.4.18pre9-ac1
...
Does it contain rmap-12e or still only 12a (as per the changelog)?
my patches 2.4.18-pre9 -> pre9-rmap12e & 2.4.18-pre9-ac1 -> ac1-rmap12e available at
http://den.st/nick/2.4.18-pre9-rmap12e.bz2
http://den.st/nick/2.4.18-pre9-ac1-rmap12e.bz2
should be applied against pre9 & pre9-ac1 respectively.
Signature · 2 lines
With best wishes, Nick Orlov.
Just another fast question:
Is IDE Taskfile Access and Taskfile IO safe to use?
Thanks,
Signature · 1 line
Balazs Pozsar.
Quoted 5 lines · Alan Cox
[+ indicates stuff that went to Marcelo, o stuff that has not,
* indicates stuff that is merged in mainstream now, X stuff that proved bad and was dropped out]Linux 2.4.18pre9-ac1
I'm sure lot of people would like to know. Are you considering using bk ?
Are you waiting for linus to make a final decision, will you never use it, do you prefer some other tools ? do you prefer the way it's now ?
Oh, and does RH use any source management software internally ?
best wishes,
Thomas
Quoted 3 lines
I'm sure lot of people would like to know. Are you considering using bk ?
Are you waiting for linus to make a final decision, will you never use it, do you prefer some other tools ? do you prefer the way it's now ?
For what I'm doing it doesnt really make much sense unless Marcelo does. If Marcelo did I might play with it
Quoted 1 line
Oh, and does RH use any source management software internally ?
yes