Linux 2.4.17-rc1

Linux 2.4.17-rc1

[Marcelo Tosatti]

Hi, 

I've just copied 2.4.17-rc1 to ftp.kernel.org... Its mirroring yet,
probably.

Well, I want people with the "unfreeable" buffer/cache problem to confirm
with me that 2.4.17-rc1 is working ok.

The same change which should fix that problem also should make 2.4 a bit
less "swap happy".


rc1: 

- Finish MODULE_LICENSE fixups for fs/nls 	(Mark Hymers)
- Console race fix				(Andrew Morton/Robert Love)
- Configure.help update				(Eric S. Raymond)
- Correctly fix Direct IO bug			(Linus Benedict Torvalds)
- Turn off aacraid debugging			(Alan Cox)
- Added missing spinlocking in do_loopback()	(Alexander Viro)
- Added missing __devexit_p() in i82092 
  pcmcia driver					(Keith Owens)
- ns83820 zerocopy bugfix			(Benjamin LaHaise)
- Fix VM problems where cache/buffers didn't get
  freed						(me)

pre8:

- ext3 quota fix 				(Neil Brown)
- Add __devexit_p() to ISDN driver		(Kai Germaschewski)
- Declare missing function on fdomain.h		(Eyal Lebedinsky)
- Add Sony Vaio PCG-Z600NE to broken APM 
  reporting blacklist				(Kai Germaschewski)
- ns83820 driver update				(Benjamin LaHaise)
- pas16 driver cleanup				(Alan Cox)
- disable console flush on secondary CPUs on
  IA64						(Andrew Morton)
- fix typo on parport's ChangeLog		(Tim Waugh)
- fix use count for multiple queued requests on 
  closed fd					(Douglas Gilbert)
- Check return value of get_user() on 
  set_vesa_blanking				(Jeff Garzik)
- Remove asm/segment.h include from nbd 	(Jeff Garzik)
- Guard sysrq.h against multiple inclusion 	(Jeff Garzik)
- Minor PCI skeleton changes			(Jeff Garzik)
- Add via rhine MMIO to Configure.help		(Jeff Garzik)
- Jeff Garzik is not the via82cxxx driver 
  maintainer anymore: "No time, no hardware".	(Jeff Garzik)
- Remove old tulip documentation		(Jeff Garzik)
- Avoid direct IO's "misunderstanding" of which 
  block device it should use			(Masanori Goto)
- Remove mcheck_init() call from processor
  dependant code and put it in unified codepath	(Dave Jones)
- Netfilter bugfixes				(Harald Welte)


pre7:

- More USB updates				(Greg KH)
- Add missing checks on shmat()			(Christoph Rohland)
- ymfpci update					(Pete Zaitcev)
- Add aacraid driver 				(Alan Cox)
- Actually apply some of the Alan's changes
  which were on pre6 changelog.			(silly me)
- Clean up t128 SCSI driver			(Alan Cox)
- Clean up dtc SCSI driver			(Alan Cox) 
- Undo lcall patch from -pre6			(me)
- More ISDN updates				(Kai Germaschewski)

pre6:

- ISDN fixes					(Kai Germaschewski)
- Eicon driver updates				(Kai Germaschewski)
- ymfpci update					(Pete Zaitcev)
- Fix multithread coredump deadlock		(Manfred Spraul)
- Support /dev/kmem access to vmalloc space	(Marc Boucher)
- ext3 fixes/enhancements			(Andrew Morton)	
- Add IT8172G driver to Config.in/Makefile	(Giacomo Catenazzi)
- Configure.help update				(Eric S. Raymond)
- Create __devexit_p() function and use that on 
  drivers which need it to make it possible to 
  use newer binutils				(Keith Owens) 
- Make PCMCIA compile without PCI support	(Paul Mackerras)
- Use copy_user_highpage instead copy_highpage
  on COW path.					(David S. Miller)
- Cacheline align some more performance
  critical spinlocks				(Anton Blanchard)
- sonypi driver update				(Michael C.B. Ashley/Bob Donnelly)
- direct render for some SiS cards		(Torsten Duwe/Alan Cox)
- full handling of the NFSv3 'jukebox' feature  (Trond Myklebust)
- NFS performance improvements			(Trond Myklebust)
- More parport fixes				(Tim Waugh)
- Fix lots of core NCR5380 bugs			(Alan Cox)
- NCR5380/PAS driver update			(Alan Cox)
- Add aacraid to the SCSI list			(Alan Cox)
- fdomain driver fixes				(Alan Cox)

pre5:

- 8139too fixes					(Andreas Dilger)
- sym53c8xx_2 update				(Gerard Roudier)
- loopback deadlock bugfix			(Jan Kara)
- Yet another devfs update			(Richard Gooch)	
- Enable K7 SSE					(John Clemens)
- Make grab_cache_page return NULL instead 
  ERR_PTR: callers expect NULL on failure	(Christoph Hellwig)
- Make ide-{disk-floppy} compile without 
  PROCFS support				(Robert Love)
- Another ymfpci update				(Pete Zaitcev)
- indent NCR5380.{c,h}, g_NCR5380.{c,h}, plus 
  NCR5380 fix					(Alan Cox)
- SPARC32/64 update				(David S. Miller)
- Fix atyfb warnings				(David S. Miller)
- Make bootmem init code correctly align 
  bootmem data					(David S. Miller)
- Networking updates				(David S. Miller)
- Fix scanning luns > 7 on SCSI-3 devices 	(Michael Clark)
- Add sparse lun hint for Chaparral G8324 
	Fibre-SCSI controller			(Michael Clark)
- Really apply sg changes			(me)
- Parport updates				(Tim Waugh)
- ReiserFS updates				(Vladimir V. Saveliev)
- Make AGP code scan all kinds of devices:
  they are not always video ones		(Alan Cox)
- EXPORT_NO_SYMBOLS in floppy.c			(Alan Cox)
- Pentium IV Hyperthreading support		(Alan Cox)

pre4:

- Added missing tcp_diag.c and tcp_diag.h	(me)

pre3:

- Enable ppro errata workaround                 (Dave Jones)
- Update tmpfs documentation                    (Christoph Rohland)
- Fritz!PCIv2 ISDN card support                 (Kai Germaschewski)
- Really apply ymfpci changes                   (Pete Zaitcev)
- USB update                                    (Greg KH)
- Adds detection of more eepro100 cards         (Troy A. Griffitts)
- Make ftruncate64() compliant with SuS         (Andrew Morton)
- ATI64 fb driver update                        (Geert Uytterhoeven)
- Coda fixes                                    (Jan Harkes)
- devfs update                                  (Richard Gooch)
- Fix ad1848 breakage in -pre2                  (Alan Cox)
- Network updates                               (David S. Miller)
- Add cramfs locking                            (Christoph Hellwig)
- Move locking of page_table_lock on expand_stack
  before accessing any vma field                (Manfred Spraul)
- Make time monotonous with gettimeofday        (Andi Kleen)
- Add MODULE_LICENSE(GPL) to ide-tape.c         (Mikael Pettersson)
- Minor cs46xx ioctl fix                        (Thomas Woller)

pre2:

- Remove userland header from bonding driver	(David S. Miller)
- Create a SLAB for page tables on i386		(Christoph Hellwig)
- Unregister devices at shaper unload time	(David S. Miller)
- Remove several unused variables from various
  places in the kernel				(David S. Miller)
- Fix slab code to not blindly trust cc_data():
  it may be not valid on some platforms		(David S. Miller)
- Fix RTC driver bug				(David S. Miller)
- SPARC 32/64 update				(David S. Miller)
- W9966 V4L driver update			(Jakob Jemi)
- ad1848 driver fixes				(Alan Cox/Daniel T. Cobra)
- PCMCIA update					(David Hinds)
- Fix PCMCIA problem with multiple PCI busses 	(Paul Mackerras)
- Correctly free per-process signal struct	(Dave McCracken)
- IA64 PAL/signal headers cleanup		(Nathan Myers)
- ymfpci driver cleanup 			(Pete Zaitcev)
- Change NLS "licenses" to be "GPL/BSD" instead 
  only BSD.					(Robert Love)
- Fix serial module use count			(Russell King)
- Update sg to 3.1.22				(Douglas Gilbert)
- ieee1394 update				(Ben Collins)
- ReiserFS fixes				(Nikita Danilov)
- Update ACPI documentantion			(Patrick Mochel)
- Smarter atime update				(Andrew Morton)
- Correctly mark ext2 sb as dirty and sync it	(Andrew Morton) 
- IrDA update					(Jean Tourrilhes)
- Count locked buffers at
  balance_dirty_state(): Helps interactivity under
  heavy IO workloads				(Andrew Morton)
- USB update					(Greg KH)
- ide-scsi locking fix                          (Christoph Hellwig)

pre1:

- Change USB maintainer 			(Greg Kroah-Hartman)
- Speeling fix for rd.c				(From Ralf Baechle's tree)
- Updated URL for bigphysmem patch in v4l docs  (Adrian Bunk)
- Add buggy 440GX to broken pirq blacklist 	(Arjan Van de Ven)
- Add new entry to Sound blaster ISAPNP list	(Arjan Van de Ven)
- Remove crap character from Configure.help	(Niels Kristian Bech Jensen)
- Backout erroneous change to lookup_exec_domain (Christoph Hellwig)
- Update osst sound driver to 1.65		(Willem Riede)
- Fix i810 sound driver problems		(Andris Pavenis)
- Add AF_LLC define in network headers		(Arnaldo Carvalho de Melo)
- block_size cleanup on some SCSI drivers	(Erik Andersen)
- Added missing MODULE_LICENSE("GPL") in some   (Andreas Krennmair)
  modules
- Add ->show_options() to super_ops and 
  implement NFS method				(Alexander Viro)
- Updated i8k driver				(Massimo Dal Zoto)
- devfs update  				(Richard Gooch)

Re: Linux 2.4.17-rc1

[Ken Brownfield]

MINOR issue.  I'm getting this under -pre8 and -rc1.  Not sure about
earlier, I only started enabling this at -pre8.  Maybe since -pre6?

# depmod -ae
depmod: *** Unresolved symbols in /lib/modules/2.4/kernel/drivers/char/drm/sis.o
depmod:         sis_free
depmod:         sis_malloc

My .config is at http://web.irridia.com/info/linux/config".

I'm taking a look at the VM tweak ASAP.

Thanks,
-- 
Ken.
brownfld@irridia.com

On Thu, Dec 13, 2001 at 06:44:16PM -0200, Marcelo Tosatti wrote:
| 
| Hi, 
| 
| I've just copied 2.4.17-rc1 to ftp.kernel.org... Its mirroring yet,
| probably.
| 
| Well, I want people with the "unfreeable" buffer/cache problem to confirm
| with me that 2.4.17-rc1 is working ok.
| 
| The same change which should fix that problem also should make 2.4 a bit
| less "swap happy".
| 
| 
| rc1: 
| 
| - Finish MODULE_LICENSE fixups for fs/nls 	(Mark Hymers)
| - Console race fix				(Andrew Morton/Robert Love)
| - Configure.help update				(Eric S. Raymond)
| - Correctly fix Direct IO bug			(Linus Benedict Torvalds)
| - Turn off aacraid debugging			(Alan Cox)
| - Added missing spinlocking in do_loopback()	(Alexander Viro)
| - Added missing __devexit_p() in i82092 
|   pcmcia driver					(Keith Owens)
| - ns83820 zerocopy bugfix			(Benjamin LaHaise)
| - Fix VM problems where cache/buffers didn't get
|   freed						(me)
| 
| pre8:
| 
| - ext3 quota fix 				(Neil Brown)
| - Add __devexit_p() to ISDN driver		(Kai Germaschewski)
| - Declare missing function on fdomain.h		(Eyal Lebedinsky)
| - Add Sony Vaio PCG-Z600NE to broken APM 
|   reporting blacklist				(Kai Germaschewski)
| - ns83820 driver update				(Benjamin LaHaise)
| - pas16 driver cleanup				(Alan Cox)
| - disable console flush on secondary CPUs on
|   IA64						(Andrew Morton)
| - fix typo on parport's ChangeLog		(Tim Waugh)
| - fix use count for multiple queued requests on 
|   closed fd					(Douglas Gilbert)
| - Check return value of get_user() on 
|   set_vesa_blanking				(Jeff Garzik)
| - Remove asm/segment.h include from nbd 	(Jeff Garzik)
| - Guard sysrq.h against multiple inclusion 	(Jeff Garzik)
| - Minor PCI skeleton changes			(Jeff Garzik)
| - Add via rhine MMIO to Configure.help		(Jeff Garzik)
| - Jeff Garzik is not the via82cxxx driver 
|   maintainer anymore: "No time, no hardware".	(Jeff Garzik)
| - Remove old tulip documentation		(Jeff Garzik)
| - Avoid direct IO's "misunderstanding" of which 
|   block device it should use			(Masanori Goto)
| - Remove mcheck_init() call from processor
|   dependant code and put it in unified codepath	(Dave Jones)
| - Netfilter bugfixes				(Harald Welte)
| 
| 
| pre7:
| 
| - More USB updates				(Greg KH)
| - Add missing checks on shmat()			(Christoph Rohland)
| - ymfpci update					(Pete Zaitcev)
| - Add aacraid driver 				(Alan Cox)
| - Actually apply some of the Alan's changes
|   which were on pre6 changelog.			(silly me)
| - Clean up t128 SCSI driver			(Alan Cox)
| - Clean up dtc SCSI driver			(Alan Cox) 
| - Undo lcall patch from -pre6			(me)
| - More ISDN updates				(Kai Germaschewski)
| 
| pre6:
| 
| - ISDN fixes					(Kai Germaschewski)
| - Eicon driver updates				(Kai Germaschewski)
| - ymfpci update					(Pete Zaitcev)
| - Fix multithread coredump deadlock		(Manfred Spraul)
| - Support /dev/kmem access to vmalloc space	(Marc Boucher)
| - ext3 fixes/enhancements			(Andrew Morton)	
| - Add IT8172G driver to Config.in/Makefile	(Giacomo Catenazzi)
| - Configure.help update				(Eric S. Raymond)
| - Create __devexit_p() function and use that on 
|   drivers which need it to make it possible to 
|   use newer binutils				(Keith Owens) 
| - Make PCMCIA compile without PCI support	(Paul Mackerras)
| - Use copy_user_highpage instead copy_highpage
|   on COW path.					(David S. Miller)
| - Cacheline align some more performance
|   critical spinlocks				(Anton Blanchard)
| - sonypi driver update				(Michael C.B. Ashley/Bob Donnelly)
| - direct render for some SiS cards		(Torsten Duwe/Alan Cox)
| - full handling of the NFSv3 'jukebox' feature  (Trond Myklebust)
| - NFS performance improvements			(Trond Myklebust)
| - More parport fixes				(Tim Waugh)
| - Fix lots of core NCR5380 bugs			(Alan Cox)
| - NCR5380/PAS driver update			(Alan Cox)
| - Add aacraid to the SCSI list			(Alan Cox)
| - fdomain driver fixes				(Alan Cox)
| 
| pre5:
| 
| - 8139too fixes					(Andreas Dilger)
| - sym53c8xx_2 update				(Gerard Roudier)
| - loopback deadlock bugfix			(Jan Kara)
| - Yet another devfs update			(Richard Gooch)	
| - Enable K7 SSE					(John Clemens)
| - Make grab_cache_page return NULL instead 
|   ERR_PTR: callers expect NULL on failure	(Christoph Hellwig)
| - Make ide-{disk-floppy} compile without 
|   PROCFS support				(Robert Love)
| - Another ymfpci update				(Pete Zaitcev)
| - indent NCR5380.{c,h}, g_NCR5380.{c,h}, plus 
|   NCR5380 fix					(Alan Cox)
| - SPARC32/64 update				(David S. Miller)
| - Fix atyfb warnings				(David S. Miller)
| - Make bootmem init code correctly align 
|   bootmem data					(David S. Miller)
| - Networking updates				(David S. Miller)
| - Fix scanning luns > 7 on SCSI-3 devices 	(Michael Clark)
| - Add sparse lun hint for Chaparral G8324 
| 	Fibre-SCSI controller			(Michael Clark)
| - Really apply sg changes			(me)
| - Parport updates				(Tim Waugh)
| - ReiserFS updates				(Vladimir V. Saveliev)
| - Make AGP code scan all kinds of devices:
|   they are not always video ones		(Alan Cox)
| - EXPORT_NO_SYMBOLS in floppy.c			(Alan Cox)
| - Pentium IV Hyperthreading support		(Alan Cox)
| 
| pre4:
| 
| - Added missing tcp_diag.c and tcp_diag.h	(me)
| 
| pre3:
| 
| - Enable ppro errata workaround                 (Dave Jones)
| - Update tmpfs documentation                    (Christoph Rohland)
| - Fritz!PCIv2 ISDN card support                 (Kai Germaschewski)
| - Really apply ymfpci changes                   (Pete Zaitcev)
| - USB update                                    (Greg KH)
| - Adds detection of more eepro100 cards         (Troy A. Griffitts)
| - Make ftruncate64() compliant with SuS         (Andrew Morton)
| - ATI64 fb driver update                        (Geert Uytterhoeven)
| - Coda fixes                                    (Jan Harkes)
| - devfs update                                  (Richard Gooch)
| - Fix ad1848 breakage in -pre2                  (Alan Cox)
| - Network updates                               (David S. Miller)
| - Add cramfs locking                            (Christoph Hellwig)
| - Move locking of page_table_lock on expand_stack
|   before accessing any vma field                (Manfred Spraul)
| - Make time monotonous with gettimeofday        (Andi Kleen)
| - Add MODULE_LICENSE(GPL) to ide-tape.c         (Mikael Pettersson)
| - Minor cs46xx ioctl fix                        (Thomas Woller)
| 
| pre2:
| 
| - Remove userland header from bonding driver	(David S. Miller)
| - Create a SLAB for page tables on i386		(Christoph Hellwig)
| - Unregister devices at shaper unload time	(David S. Miller)
| - Remove several unused variables from various
|   places in the kernel				(David S. Miller)
| - Fix slab code to not blindly trust cc_data():
|   it may be not valid on some platforms		(David S. Miller)
| - Fix RTC driver bug				(David S. Miller)
| - SPARC 32/64 update				(David S. Miller)
| - W9966 V4L driver update			(Jakob Jemi)
| - ad1848 driver fixes				(Alan Cox/Daniel T. Cobra)
| - PCMCIA update					(David Hinds)
| - Fix PCMCIA problem with multiple PCI busses 	(Paul Mackerras)
| - Correctly free per-process signal struct	(Dave McCracken)
| - IA64 PAL/signal headers cleanup		(Nathan Myers)
| - ymfpci driver cleanup 			(Pete Zaitcev)
| - Change NLS "licenses" to be "GPL/BSD" instead 
|   only BSD.					(Robert Love)
| - Fix serial module use count			(Russell King)
| - Update sg to 3.1.22				(Douglas Gilbert)
| - ieee1394 update				(Ben Collins)
| - ReiserFS fixes				(Nikita Danilov)
| - Update ACPI documentantion			(Patrick Mochel)
| - Smarter atime update				(Andrew Morton)
| - Correctly mark ext2 sb as dirty and sync it	(Andrew Morton) 
| - IrDA update					(Jean Tourrilhes)
| - Count locked buffers at
|   balance_dirty_state(): Helps interactivity under
|   heavy IO workloads				(Andrew Morton)
| - USB update					(Greg KH)
| - ide-scsi locking fix                          (Christoph Hellwig)
| 
| pre1:
| 
| - Change USB maintainer 			(Greg Kroah-Hartman)
| - Speeling fix for rd.c				(From Ralf Baechle's tree)
| - Updated URL for bigphysmem patch in v4l docs  (Adrian Bunk)
| - Add buggy 440GX to broken pirq blacklist 	(Arjan Van de Ven)
| - Add new entry to Sound blaster ISAPNP list	(Arjan Van de Ven)
| - Remove crap character from Configure.help	(Niels Kristian Bech Jensen)
| - Backout erroneous change to lookup_exec_domain (Christoph Hellwig)
| - Update osst sound driver to 1.65		(Willem Riede)
| - Fix i810 sound driver problems		(Andris Pavenis)
| - Add AF_LLC define in network headers		(Arnaldo Carvalho de Melo)
| - block_size cleanup on some SCSI drivers	(Erik Andersen)
| - Added missing MODULE_LICENSE("GPL") in some   (Andreas Krennmair)
|   modules
| - Add ->show_options() to super_ops and 
|   implement NFS method				(Alexander Viro)
| - Updated i8k driver				(Massimo Dal Zoto)
| - devfs update  				(Richard Gooch)
| 

Re: Linux 2.4.17-rc1

[Roger Larsson]

On Thursday den 13 December 2001 21.44, Marcelo Tosatti wrote:
> Hi,
>
> I've just copied 2.4.17-rc1 to ftp.kernel.org... Its mirroring yet,
> probably.
>
> Well, I want people with the "unfreeable" buffer/cache problem to confirm
> with me that 2.4.17-rc1 is working ok.
>
> The same change which should fix that problem also should make 2.4 a bit
> less "swap happy".
>
>

Hi,

I have run some "files bigger than memory" (streaming) tests.
Some significant differences with earlier kernel:

* write - lowered throughput (26 MB/s => 22 MB/s)

* copy - throughput better by 2/3 (16 MB/s => 25 MB/s) !

* dbench 32 - back down to lower than 2.4.12 (due to increased fairness?)
	2.4.11				18.9 MB/s
	2.4.12				23.3 MB/s
	2.4.16				34.9 MB/s
	2.4.17-rc1			20.3 MB/s
	2.4.17-rc1 (file-readahead:1000)	24.5 MB/s
	[lets forget about them now...]

* diff - usage of "file-readahead" more than doubles efficency
	(most-kernels: 11 MB/s => 2.4.17-rc1 with readahead: 25 MB/s)

The nicest thing is that it is the first kernel where readahead tuning is not 
necessary for the copy operation.
Now it is only multiple-big-concurrent-reads that _needs_ "file-readahead"

/RogerL

-- 
Roger Larsson
Skellefteå
Sweden

Re: [BUG] 2.4.17-rc1 still has I/O freezing bug

[Ken Brownfield]

This reproduces for me as well under -rc1 (SMP, ext3, aic7xxx,
1xCeleron).  In my case, dd gets to about 912 MB before stopping, and
during the write up to that point the machines' interactive response
stalls occasionally for a few seconds.  Sync etc stuck in device-wait.

Hopefully whoever wrote the patch can comment.

-- 
Ken.
brownfld@irridia.com

On Thu, Dec 13, 2001 at 10:18:42PM -0800, Barry K. Nathan wrote:
| 2.4.17-rc1 still has the bug that I reported for -pre8, where dd from
| /dev/zero to a file on a loopback-mounted filesystem still freezes after
| a few seconds. (I confirmed this by testing it on my systems here.)
| 
| Instructions to reproduce on a regular system (as opposed to the weird
| setup on my notebook, which I detailed in a previous post to the
| linux-kernel mailing list, that made me notice this problem in real-world
| use):
| 
| 1. dd if=/dev/zero bs=1024k count=1024 of=/tmp/loopf (put the file on
| any non-loopback filesystem -- this works with both IDE and SCSI)
| 
| 2. mke2fs /tmp/loopf
| 
| 3. mount -o loop /tmp/loopf /mnt/disk (or just /mnt, or wherever)
| 
| 4. log into another virtual console or whatever, so you can do
| 'ls -l /mnt/disk' or the equivalent (this is especially important if
| your disk doesn't have a visible activity light)
| 
| 5. dd if=/dev/zero bs=1024k of=/mnt/disk/zero [normally this should fill
| the disk up then exit with an error]
| 
| 6. Watch the activity light, or wait 10 seconds and issue ls commands in
| the other terminal. The activity light will turn off in a few seconds,
| and the /mnt/disk/zero file will stop growing after a few hundred
| megabytes or so, well ahead of the disk actually running out of space.
| Killing the dd command with ^C at the console is impossible, and a kill -9
| also fails. ps shows the dd command and the [loop0] kernel thread both
| stuck in D state, despite the lack of physical disk activity.
| 
| This is 100% reproducible for me, on both of the systems that I have
| tried it on. With -rc1, as with -pre8, *reverting* the following patch
| fixes the problem:
| 
| diff -urN linux-2.4.17-pre1/fs/buffer.c linux/fs/buffer.c
| --- linux-2.4.17-pre1/fs/buffer.c	Wed Nov 21 14:40:17 2001
| +++ linux/fs/buffer.c	Sat Dec  1 00:35:16 2001
| @@ -1036,6 +1036,7 @@
|  	unsigned long dirty, tot, hard_dirty_limit, soft_dirty_limit;
|  
|  	dirty = size_buffers_type[BUF_DIRTY] >> PAGE_SHIFT;
| +	dirty += size_buffers_type[BUF_LOCKED] >> PAGE_SHIFT;
|  	tot = nr_free_buffer_pages();
|  
|  	dirty *= 100;
| 
| -Barry K. Nathan <barryn@pobox.com>

Re: Linux 2.4.17-rc1

[Daniel Phillips]

On December 13, 2001 09:44 pm, Marcelo Tosatti wrote:
> rc1: 
> 
> - Finish MODULE_LICENSE fixups for fs/nls 	(Mark Hymers)
> - Console race fix				(Andrew Morton/Robert Love)
> - Configure.help update				(Eric S. Raymond)
> - Correctly fix Direct IO bug			(Linus Benedict Torvalds)
> - Turn off aacraid debugging			(Alan Cox)
> - Added missing spinlocking in do_loopback()	(Alexander Viro)
> - Added missing __devexit_p() in i82092 
>   pcmcia driver					(Keith Owens)
> - ns83820 zerocopy bugfix			(Benjamin LaHaise)
> - Fix VM problems where cache/buffers didn't get
>   freed						(me)

Will there be a rc2?

--
Daniel

Re: Linux 2.4.17-rc1

[Marcelo Tosatti]

On Sun, 16 Dec 2001, Daniel Phillips wrote:

> On December 13, 2001 09:44 pm, Marcelo Tosatti wrote:
> > rc1: 
> > 
> > - Finish MODULE_LICENSE fixups for fs/nls 	(Mark Hymers)
> > - Console race fix				(Andrew Morton/Robert Love)
> > - Configure.help update				(Eric S. Raymond)
> > - Correctly fix Direct IO bug			(Linus Benedict Torvalds)
> > - Turn off aacraid debugging			(Alan Cox)
> > - Added missing spinlocking in do_loopback()	(Alexander Viro)
> > - Added missing __devexit_p() in i82092 
> >   pcmcia driver					(Keith Owens)
> > - ns83820 zerocopy bugfix			(Benjamin LaHaise)
> > - Fix VM problems where cache/buffers didn't get
> >   freed						(me)
> 
> Will there be a rc2?

Yes there will.

There have been reiserfs bug reports (I'm waiting for the fix for -rc2),
and I'm waiting for Richard's patch to fix a devfs update issue.

I want to test those before 2.4.17. 

Re: Linux 2.4.17-rc1

[Richard Gooch]

Marcelo Tosatti writes:
> On Sun, 16 Dec 2001, Daniel Phillips wrote:
> > Will there be a rc2?
> 
> Yes there will.
> 
> There have been reiserfs bug reports (I'm waiting for the fix for
> -rc2), and I'm waiting for Richard's patch to fix a devfs update
> issue.

I've got the devfs patch ready, I'm just waiting for followups on bug
reports from people. I was also hoping to get feedback from Al about
the blkdev+devfs races, but he hasn't followed up. I was going to send
it to you once I've gotten those followups. But if you want to apply
the patch anyway (and thus defer some testing to -rc2), let me know
and I'll send it off to you. I'm assuming that Al's silence means he
hasn't managed to poke holes in my solution :-)

Either way, I'm pretty confident that my current patch is an
improvement, even it it doesn't fix everything.

				Regards,

					Richard....
Permanent: rgooch@atnf.csiro.au
Current:   rgooch@ras.ucalgary.ca

Re: Linux 2.4.17-rc1

[Marcelo Tosatti]

On Mon, 17 Dec 2001, Richard Gooch wrote:

> Marcelo Tosatti writes:
> > On Sun, 16 Dec 2001, Daniel Phillips wrote:
> > > Will there be a rc2?
> > 
> > Yes there will.
> > 
> > There have been reiserfs bug reports (I'm waiting for the fix for
> > -rc2), and I'm waiting for Richard's patch to fix a devfs update
> > issue.
> 
> I've got the devfs patch ready, I'm just waiting for followups on bug
> reports from people. 

Ok, as soon as you get the reports from people, please send me the patch
or tell me its broken :)

Re: Linux 2.4.17-rc1

[David Relson]

IMHO, 2.4.17-rc1 seems to be ready to be promoted to 2.4.17. It's passed a 
suitable "release candidate" test - available for a couple of days and 
nobody has found any major problems.

David

At 11:59 PM 12/15/01, Daniel Phillips wrote:
>On December 13, 2001 09:44 pm, Marcelo Tosatti wrote:
> > rc1:
> >
> > - Finish MODULE_LICENSE fixups for fs/nls     (Mark Hymers)
> > - Console race fix                            (Andrew Morton/Robert Love)
> > - Configure.help update                               (Eric S. Raymond)
> > - Correctly fix Direct IO bug                 (Linus Benedict Torvalds)
> > - Turn off aacraid debugging                  (Alan Cox)
> > - Added missing spinlocking in do_loopback()  (Alexander Viro)
> > - Added missing __devexit_p() in i82092
> >   pcmcia driver                                       (Keith Owens)
> > - ns83820 zerocopy bugfix                     (Benjamin LaHaise)
> > - Fix VM problems where cache/buffers didn't get
> >   freed                                               (me)
>
>Will there be a rc2?
>
>--
>Daniel
>-
>To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
>the body of a message to majordomo@vger.kernel.org
>More majordomo info at  http://vger.kernel.org/majordomo-info.html
>Please read the FAQ at  http://www.tux.org/lkml/

Re: Linux 2.4.17-rc1

[Rik van Riel]

On Sun, 16 Dec 2001, David Relson wrote:

> IMHO, 2.4.17-rc1 seems to be ready to be promoted to 2.4.17. It's
> passed a suitable "release candidate" test - available for a couple of
> days and nobody has found any major problems.

I guess you must have missed the deadlock with the loop
device driver, then ;)

Rik
-- 
Shortwave goes a long way:  irc.starchat.net  #swl

http://www.surriel.com/		http://distro.conectiva.com/

Re: Linux 2.4.17-rc1

[Dave Jones]

On Sun, 16 Dec 2001, David Relson wrote:

> IMHO, 2.4.17-rc1 seems to be ready to be promoted to 2.4.17. It's passed a
> suitable "release candidate" test - available for a couple of days and
> nobody has found any major problems.

Except for loop deadlock, sysvfs oops, and a glut of __devexit
non-compiles. Whilst the sysvfs oops shouldn't affect many, loop
is used by a lot of people, and the __devexit patches would save
us another month of debian sid users who don't bother to read archives.

regards,
Dave.

-- 
| Dave Jones.        http://www.codemonkey.org.uk
| SuSE Labs

Re: Linux 2.4.17-rc1

[Ed Tomlinson]

Dave Jones wrote:

> On Sun, 16 Dec 2001, David Relson wrote:
> 
>> IMHO, 2.4.17-rc1 seems to be ready to be promoted to 2.4.17. It's passed
>> a suitable "release candidate" test - available for a couple of days and
>> nobody has found any major problems.
> 
> Except for loop deadlock, sysvfs oops, and a glut of __devexit
> non-compiles. Whilst the sysvfs oops shouldn't affect many, loop
> is used by a lot of people, and the __devexit patches would save
> us another month of debian sid users who don't bother to read archives.

Also there are some O_DIRECT problems.  Also Chris Mason has confirmed a
opps occuring when finish_unfinished is called when remounting from ro to 
rw...

In my books this makes 2 nasties (loop, opps), 1 pita (__devexit stuff) and
2 that would nice to have working.

There are probably others.

Ed Tomlinson

Ed Tomlinson

Re: Linux 2.4.17-rc1

[David Relson]

Rik & Dave,

(as I sheepishly wipes egg off my face), guess it's time for "rc2", not 
"final".

I guess I skipped too many lkml messages before posting mine.  'Sorry about 
that.

David

At 09:26 AM 12/16/01, Dave Jones wrote:
>On Sun, 16 Dec 2001, David Relson wrote:
>
> > IMHO, 2.4.17-rc1 seems to be ready to be promoted to 2.4.17. It's passed a
> > suitable "release candidate" test - available for a couple of days and
> > nobody has found any major problems.
>
>Except for loop deadlock, sysvfs oops, and a glut of __devexit
>non-compiles. Whilst the sysvfs oops shouldn't affect many, loop
>is used by a lot of people, and the __devexit patches would save
>us another month of debian sid users who don't bother to read archives.
>
>regards,
>Dave.
>
>--
>| Dave Jones.        http://www.codemonkey.org.uk
>| SuSE Labs
>
>-
>To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
>the body of a message to majordomo@vger.kernel.org
>More majordomo info at  http://vger.kernel.org/majordomo-info.html
>Please read the FAQ at  http://www.tux.org/lkml/

Re: Linux 2.4.17-rc1

[John Alvord]

You shouldn't fix every single defect, just the ones with wide
impact. Otherwise the rcX series could go on a long time. john

On Sun, 16 Dec 2001, David Relson wrote:

> Rik & Dave,
> 
> (as I sheepishly wipes egg off my face), guess it's time for "rc2", not 
> "final".
> 
> I guess I skipped too many lkml messages before posting mine.  'Sorry about 
> that.
> 
> David
> 
> At 09:26 AM 12/16/01, Dave Jones wrote:
> >On Sun, 16 Dec 2001, David Relson wrote:
> >
> > > IMHO, 2.4.17-rc1 seems to be ready to be promoted to 2.4.17. It's passed a
> > > suitable "release candidate" test - available for a couple of days and
> > > nobody has found any major problems.
> >
> >Except for loop deadlock, sysvfs oops, and a glut of __devexit
> >non-compiles. Whilst the sysvfs oops shouldn't affect many, loop
> >is used by a lot of people, and the __devexit patches would save
> >us another month of debian sid users who don't bother to read archives.
> >
> >regards,
> >Dave.
> >
> >--
> >| Dave Jones.        http://www.codemonkey.org.uk
> >| SuSE Labs
> >
> >-
> >To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> >the body of a message to majordomo@vger.kernel.org
> >More majordomo info at  http://vger.kernel.org/majordomo-info.html
> >Please read the FAQ at  http://www.tux.org/lkml/
> 
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
> 

Re: Linux 2.4.17-rc1

[David Relson]

At 11:10 AM 12/16/01, John Alvord wrote:
>You shouldn't fix every single defect, just the ones with wide
>impact. Otherwise the rcX series could go on a long time. john

John,

Well said.  Fix the show stoppers in the rcX series for 2.4.17.  Leave the 
small ones for 2.4.18-preX.

... repeat above steps for 2.4.18-rcX and 2.4.19
... repeat for ...

... iterate until time for 2.6.0 :-)

David

Re: Linux 2.4.17-rc1

[David Weinehall]

On Sun, Dec 16, 2001 at 11:22:53AM -0500, David Relson wrote:
> At 11:10 AM 12/16/01, John Alvord wrote:
> >You shouldn't fix every single defect, just the ones with wide
> >impact. Otherwise the rcX series could go on a long time. john
> 
> John,
> 
> Well said.  Fix the show stoppers in the rcX series for 2.4.17.  Leave the 
> small ones for 2.4.18-preX.
> 
> ... repeat above steps for 2.4.18-rcX and 2.4.19
> ... repeat for ...
> 
> ... iterate until time for 2.6.0 :-)

Uhm, hopefully the release of 2.6.0 won't mean that maintainance of
2.4.xx will cease. While it *might* be possible that 2.4.xx is bugfree
enough, some smaller fixes and driver-backports will probably still
be needed.


Regards: David Weinehall
  _                                                                 _
 // David Weinehall <tao@acc.umu.se> /> Northern lights wander      \\
//  Maintainer of the v2.0 kernel   //  Dance across the winter sky //
\>  http://www.acc.umu.se/~tao/    </   Full colour fire           </

Re: Linux 2.4.17-rc1

[J Sloan]

David Relson wrote:

> IMHO, 2.4.17-rc1 seems to be ready to be promoted to 2.4.17. It's passed a
> suitable "release candidate" test - available for a couple of days and
> nobody has found any major problems.

er - the loopback hangs?

cu

jjs

>
>
> David
>
> At 11:59 PM 12/15/01, Daniel Phillips wrote:
> >On December 13, 2001 09:44 pm, Marcelo Tosatti wrote:
> > > rc1:
> > >
> > > - Finish MODULE_LICENSE fixups for fs/nls     (Mark Hymers)
> > > - Console race fix                            (Andrew Morton/Robert Love)
> > > - Configure.help update                               (Eric S. Raymond)
> > > - Correctly fix Direct IO bug                 (Linus Benedict Torvalds)
> > > - Turn off aacraid debugging                  (Alan Cox)
> > > - Added missing spinlocking in do_loopback()  (Alexander Viro)
> > > - Added missing __devexit_p() in i82092
> > >   pcmcia driver                                       (Keith Owens)
> > > - ns83820 zerocopy bugfix                     (Benjamin LaHaise)
> > > - Fix VM problems where cache/buffers didn't get
> > >   freed                                               (me)
> >
> >Will there be a rc2?
> >
> >--
> >Daniel
> >-
> >To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> >the body of a message to majordomo@vger.kernel.org
> >More majordomo info at  http://vger.kernel.org/majordomo-info.html
> >Please read the FAQ at  http://www.tux.org/lkml/
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/

Re: Linux 2.4.17-rc1

[Erik Andersen]

On Thu Dec 13, 2001 at 06:44:16PM -0200, Marcelo Tosatti wrote:
> 
> Hi, 
> 
> I've just copied 2.4.17-rc1 to ftp.kernel.org... Its mirroring yet,
> probably.
> 
> Well, I want people with the "unfreeable" buffer/cache problem to confirm
> with me that 2.4.17-rc1 is working ok.
> 
> The same change which should fix that problem also should make 2.4 a bit
> less "swap happy".
> 
[---------snip-----------]
> 
> pre6:
> 
[---------snip-----------]
> - Fix lots of core NCR5380 bugs			(Alan Cox)

This fix from -pre6 broke NCR5380 so that it does not compile
when linked into the kernel (i.e.  not as a module).  This patch
fixes it.  Please apply for 2.4.17-rc2, 

 -Erik

--
Erik B. Andersen             http://codepoet-consulting.com/
--This message was written using 73% post-consumer electrons--


--- drivers/scsi.orig/g_NCR5380.c	Mon Dec 17 18:08:08 2001
+++ drivers/scsi/g_NCR5380.c	Mon Dec 17 18:22:34 2001
@@ -911,41 +911,7 @@
 MODULE_DEVICE_TABLE(isapnp, id_table);
 
 
-#ifndef MODULE
-
-static int __init do_NCR53C400_setup(char *str)
-{
-	int ints[10];
-
-	get_options(str, sizeof(ints) / sizeof(int), ints);
-	generic_NCR53C400_setup(str, ints);
-
-	return 1;
-}
-
-static int __init do_NCR53C400A_setup(char *str)
-{
-	int ints[10];
-
-	get_options(str, sizeof(ints) / sizeof(int), ints);
-	generic_NCR53C400A_setup(str, ints);
-
-	return 1;
-}
-
-static int __init do_DTC3181E_setup(char *str)
-{
-	int ints[10];
-
-	get_options(str, sizeof(ints) / sizeof(int), ints);
-	generic_DTC3181E_setup(str, ints);
-
-	return 1;
-}
-
 __setup("ncr5380=", do_NCR5380_setup);
 __setup("ncr53c400=", do_NCR53C400_setup);
 __setup("ncr53c400a=", do_NCR53C400A_setup);
 __setup("dtc3181e=", do_DTC3181E_setup);
-
-#endif

Re: Linux 2.4.17-rc1

[Dave Jones]

On Mon, 17 Dec 2001, Erik Andersen wrote:


> This fix from -pre6 broke NCR5380 so that it does not compile
> when linked into the kernel (i.e.  not as a module).  This patch
> fixes it.  Please apply for 2.4.17-rc2,

This doesn't look right..

> -static int __init do_NCR53C400_setup(char *str)
> -static int __init do_NCR53C400A_setup(char *str)
> -static int __init do_DTC3181E_setup(char *str)

You nuked the functions..

>  __setup("ncr5380=", do_NCR5380_setup);
>  __setup("ncr53c400=", do_NCR53C400_setup);
>  __setup("ncr53c400a=", do_NCR53C400A_setup);
>  __setup("dtc3181e=", do_DTC3181E_setup);

But not the references to them. What error are you seeing ?

Dave.

-- 
| Dave Jones.        http://www.codemonkey.org.uk
| SuSE Labs

Re: Linux 2.4.17-rc1

[Erik Andersen]

On Tue Dec 18, 2001 at 02:38:48AM +0100, Dave Jones wrote:
> On Mon, 17 Dec 2001, Erik Andersen wrote:
> 
> 
> > This fix from -pre6 broke NCR5380 so that it does not compile
> > when linked into the kernel (i.e.  not as a module).  This patch
> > fixes it.  Please apply for 2.4.17-rc2,
> 
> This doesn't look right..

Sure it does, look closer.  :-)

> > -static int __init do_NCR53C400_setup(char *str)
> > -static int __init do_NCR53C400A_setup(char *str)
> > -static int __init do_DTC3181E_setup(char *str)
> 
> You nuked the functions..

Exactly.  Because there were two copies of that code, but one
copy was wrapped inside an '#ifndef MODULE' so when compiling as
a module, everything was cool.  But if you link the driver into
the kernel you would get two copies of those init funcs....

> >  __setup("ncr5380=", do_NCR5380_setup);
> >  __setup("ncr53c400=", do_NCR53C400_setup);
> >  __setup("ncr53c400a=", do_NCR53C400A_setup);
> >  __setup("dtc3181e=", do_DTC3181E_setup);
> 
> But not the references to them. What error are you seeing ?

make[3]: Entering directory `/usr/src/linux/drivers/scsi'
ld -m elf_i386 -r -o scsi_mod.o scsi.o hosts.o scsi_ioctl.o constants.o scsicam.o scsi_proc.o scsi_error.o scsi_obsolete.o scsi_queue.o scsi_lib.o scsi_merge.o scsi_dma.o scsi_scan.o scsi_syms.o
gcc -D__KERNEL__ -I/usr/src/linux/include -Wall -Wstrict-prototypes -Wno-trigraphs -O2 -fomit-frame-pointer -fno-strict-aliasing -fno-common -pipe -mpreferred-stack-boundary=2 -march=i686 -malign-functions=4     -c -o g_NCR5380.o g_NCR5380.c
g_NCR5380.c:917: redefinition of `do_NCR53C400_setup'
g_NCR5380.c:230: `do_NCR53C400_setup' previously defined here
g_NCR5380.c: In function `do_NCR53C400_setup':
g_NCR5380.c:921: warning: implicit declaration of function `generic_NCR53C400_setup'
g_NCR5380.c: At top level:
g_NCR5380.c:927: redefinition of `do_NCR53C400A_setup'
g_NCR5380.c:248: `do_NCR53C400A_setup' previously defined here
g_NCR5380.c: In function `do_NCR53C400A_setup':
g_NCR5380.c:931: warning: implicit declaration of function `generic_NCR53C400A_setup'
g_NCR5380.c: At top level:
g_NCR5380.c:937: redefinition of `do_DTC3181E_setup'
g_NCR5380.c:266: `do_DTC3181E_setup' previously defined here
g_NCR5380.c: In function `do_DTC3181E_setup':
g_NCR5380.c:941: warning: implicit declaration of function `generic_DTC3181E_setup'
g_NCR5380.c: At top level:
NCR5380.c:458: warning: `NCR5380_print_phase' defined but not used
NCR5380.c:402: warning: `NCR5380_print' defined but not used
{standard input}: Assembler messages:
{standard input}:4585: Error: symbol `do_NCR53C400_setup' is already defined
{standard input}:4608: Error: symbol `do_NCR53C400A_setup' is already defined
{standard input}:4631: Error: symbol `do_DTC3181E_setup' is already defined
make[3]: *** [g_NCR5380.o] Error 1
make[3]: Leaving directory `/usr/src/linux/drivers/scsi'
make[2]: *** [first_rule] Error 2
make[2]: Leaving directory `/usr/src/linux/drivers/scsi'
make[1]: *** [_subdir_scsi] Error 2
make[1]: Leaving directory `/usr/src/linux/drivers'
make: *** [_dir_drivers] Error 2

 -Erik

--
Erik B. Andersen             http://codepoet-consulting.com/
--This message was written using 73% post-consumer electrons--

Re: Linux 2.4.17-rc1

[Dave Jones]

On Mon, 17 Dec 2001, Erik Andersen wrote:

> > This doesn't look right..
> Sure it does, look closer.  :-)

Ah yes, missed the context there (and was too lazy to crossref
with the source).. apologies.

Dave.

-- 
| Dave Jones.        http://www.codemonkey.org.uk
| SuSE Labs

Re: Linux 2.4.17-rc1

[Dieter Nützel]

What about the "new" eepro100.c driver from SUN (Cobalt)?
Early 2.4.18 stuff?

Thanks,
	Dieter
-- 
Dieter Nützel
Graduate Student, Computer Science

University of Hamburg
Department of Computer Science
@home: Dieter.Nuetzel@hamburg.de

Re: Linux 2.4.17-rc1

[Mario Vanoni]

IMHO integrate _please_ Andrea's patches 100%.

Dual SMP PIII550 with 1024MB memory CAS 2-2-2.

Reason: latency, <30% with Andrea.

Test: time qsbench -p 9 (Lorenzo Allegrucci):
      rc1aa1 resolves the latency in
      <30% the time with 120MB swap,
      <20% _without_ swap space present!

      time qsbench reaches 5m44.204s with rc1,
      <= 2.44.025s with rc1aa1,
      test a login in that time ..., wait Godot!

      Without swap the times are:
      <= 3m41.720s with rc1,
      <= 2m43.090s with rc1aa1.

What is better if e.g. a "qsbench" occusr,
wait 5 or 3 minutes for a response?

I run my 3xUP and 1xSMP normally _without_ swap,
all with lots of memory.

Note: this with or without 2 (two) setiathome
      24/7 on the SMP + all abitual work.
      See setiathome country Switzerland under
      Mario Vanoni, about place 87.

Kind regards

Mario, _not_ in lkml.

Re: Linux 2.4.17-rc1

[Martin Knoblauch]

> Linux 2.4.17-rc1
> 
> 
> Hi,
> 
> I've just copied 2.4.17-rc1 to ftp.kernel.org... Its mirroring yet,
> probably.
> 
> Well, I want people with the "unfreeable" buffer/cache problem to confirm
> with me that 2.4.17-rc1 is working ok.
> 
> The same change which should fix that problem also should make 2.4 a bit
> less "swap happy".
> 
Marcello,

 the system feels a bit better. This on a 320MB Toshiba notebook with
64MB swap. I have forced a complete run of "updatedb", while running
"vmware" and NetScape.

 Still swapping out, but the behaviour is [much] smoother. Not so many
"hangs" when the local disk is busy.

Martin
-- 
------------------------------------------------------------------
Martin Knoblauch         |    email:  Martin.Knoblauch@TeraPort.de
TeraPort GmbH            |    Phone:  +49-89-510857-309
C+ITS                    |    Fax:    +49-89-510857-111
http://www.teraport.de   |    Mobile: +49-170-4904759