Skip to contentSkip to conversation

2.6.29-rc3-git6: Reported regressions from 2.6.28

PATCH In review Rafael J. Wysocki 33 participants 156 messages Feb 4, 2009 – Feb 23, 2009

Subsystems (10)
lore
Share & tools
lore
#0 lore raw

[NOTES:
* Sorry for the delay, I had to take care of my own bugs first.
* Some of the bugs below are fixed, but I wasn't able to track the fixes down.
* There probably are some duplicates (mostly in the DRI area), but I wasn't 100% sure.]

This message contains a list of some regressions from 2.6.28, for which there are no fixes in the mainline I know of. If any of them have been fixed already, please let me know.

If you know of any other unresolved regressions from 2.6.28, please let me know either and I'll add them to the list. Also, please let me know if any of the entries below are invalid.

Each entry from the list will be sent additionally in an automatic reply to this message with CCs to the people involved in reporting and handling the issue.

Listed regressions statistics:

Date Total Pending Unresolved
----------------------------------------

  2009-02-04       66       51          39
  2009-01-20       38       35          27
  2009-01-11       13       13          10

Unresolved regressions
----------------------

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12618
Subject : hackbench [pthread mode] regression with 2.6.29-rc3
Submitter : Zhang, Yanmin <[email protected]>
Date : 2009-02-01 7:30 (4 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=490dea45d00f01847ebebd007685d564aaf2cd98
References : http://marc.info/?l=linux-kernel&m=123347347726527&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12617
Subject : unable to compile e100 firmware into kernel
Submitter : Andrey Borzenkov <[email protected]>
Date : 2009-01-31 15:59 (5 days old)
References : http://marc.info/?l=linux-kernel&m=123341764915181&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12616
Subject : boot hang: async vs. kexec
Submitter : Randy Dunlap <[email protected]>
Date : 2009-01-29 13:15 (7 days old)
References : http://lkml.org/lkml/2009/1/29/359

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12615
Subject : boot hangs while bringing up gianfar ethernet
Submitter : Ira Snyder <[email protected]>
Date : 2009-01-29 19:41 (7 days old)
References : http://marc.info/?l=linux-kernel&m=123325817201665&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12613
Subject : [Suspend regression][DRM, RADEON]
Submitter : etienne <[email protected]>
Date : 2009-01-28 22:00 (8 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=a9d51a5ad1154b5b20add1e8d30a5564f8aabbe9
References : http://marc.info/?l=linux-kernel&m=123318030419558&w=4

		  http://marc.info/?l=linux-kernel&m=123334865404574&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12611
Subject : kernel BUG at kernel/cgroup.c:398!
Submitter : Bharata B Rao <[email protected]>
Date : 2009-01-28 15:49 (8 days old)
References : http://marc.info/?l=linux-kernel&m=123315721707796&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12610
Subject : sync-Regression in 2.6.28.2?
Submitter : Ralf Hildebrandt <[email protected]>
Date : 2009-01-27 9:35 (9 days old)
References : http://marc.info/?l=linux-kernel&m=123304977706620&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12608
Subject : 2.6.29-rc powerpc G5 Xorg legacy_mem regression
Submitter : Hugh Dickins <[email protected]>
Date : 2009-01-21 21:12 (15 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=d3a54014e2a94bd37b7dee5e76e03f7bc4fab49a
References : http://marc.info/?l=linux-kernel&m=123257250431870&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12606
Subject : fb_mmap: circular locking dependency on hibernation
Submitter : Andrey Borzenkov <[email protected]>
Date : 2009-01-27 18:37 (9 days old)
References : http://marc.info/?l=linux-kernel&m=123308162731408&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12604
Subject : Commit 31a12666d8f0c22235297e1c1575f82061480029 slows down Berkeley DB
Submitter : Jan Kara <[email protected]>
Date : 2009-01-30 1:23 (6 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=31a12666d8f0c22235297e1c1575f82061480029
References : http://marc.info/?l=linux-kernel&m=123327862901800&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12602
Subject : CRED changes causing setuid failures
Submitter : David Smith <[email protected]>
Date : 2009-01-29 21:35 (7 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=d84f4f992cbd76e8f39c488cf0c5d123843923b1
References : http://marc.info/?l=linux-kernel&m=123326501813668&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12601
Subject : virt-manager broken on 2.6.29-rc2
Submitter : Stephen Hemminger <[email protected]>
Date : 2009-01-29 5:21 (7 days old)
References : http://marc.info/?l=linux-kernel&m=123320655819589&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12600
Subject : i915 lockdep warning
Submitter : Brandeburg, Jesse <[email protected]>
Date : 2009-01-13 23:17 (23 days old)
References : http://marc.info/?l=linux-kernel&m=123188898423532&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12599
Subject : dri /dev node disappeared with 2.6.29-rc1
Submitter : Norbert Preining <[email protected]>
Date : 2009-01-15 13:42 (21 days old)
References : http://marc.info/?l=linux-kernel&m=123202701026647&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12591
Subject : NULL pointer dereference in blk_queue_io_stat
Submitter : Petr Vandrovec <[email protected]>
Date : 2009-01-31 23:58 (5 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=bc58ba9468d94d62c56ab9b47173583ec140b165

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12574
Subject : possible circular locking dependency detected
Submitter : Michael S. Tsirkin <[email protected]>
Date : 2009-01-29 11:35 (7 days old)

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12571
Subject : Suspend-resume on Dell Latitude D410 newly broken in 2.6.29-rc*
Submitter : Ross Boswell <[email protected]>
Date : 2009-01-29 03:46 (7 days old)

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12551
Subject : end_request: I/O error, dev cciss/c0d0, sector 87435720
Submitter : Ralf Hildebrandt <[email protected]>
Date : 2009-01-27 06:51 (9 days old)

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12518
Subject : BUG: using smp_processor_id() in preemptible [00000000] code: dellWirelessCtl/...
Submitter : Alex Riesen <[email protected]>
Date : 2009-01-21 14:21 (15 days old)

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12511
Subject : WARNING: at drivers/dma/dmaengine.c:352
Submitter : Ingo Molnar <[email protected]>
Date : 2009-01-19 21:31 (17 days old)
References : http://marc.info/?l=linux-kernel&m=123240070614443&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12510
Subject : 2.6.29-rc2 dies on startup
Submitter : Ferenc Wagner <[email protected]>
Date : 2009-01-19 12:53 (17 days old)
References : http://marc.info/?l=linux-kernel&m=123236969321703&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12508
Subject : "powerpc/pci: Reserve legacy regions on PCI" broke my G3
Submitter : Mikael Pettersson <[email protected]>
Date : 2009-01-17 22:46 (19 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=c1f343028d35ba4e88cd4a3c44e0d8b8a84264ee
References : http://marc.info/?l=linux-kernel&m=123223247325452&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12506
Subject : Undefined symbols when CONFIG_MFD_PCF50633 is enabled
Submitter : Ozan Çağlayan <[email protected]>
Date : 2009-01-17 10:56 (19 days old)
References : http://marc.info/?l=linux-kernel&m=123218991104592&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12505
Subject : 2.6.29-rc1 Firefox crashing on page load
Submitter : Justin Madru <[email protected]>
Date : 2009-01-16 20:56 (20 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=4217458dafaa57d8e26a46f5d05ab8c53cf64191
References : http://marc.info/?l=linux-kernel&m=123213941914274&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12503
Subject : [slab corruption] BUG key_jar: Poison overwritten
Submitter : Ingo Molnar <[email protected]>
Date : 2009-01-15 18:16 (21 days old)
References : http://marc.info/?l=linux-kernel&m=123204353425825&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12502
Subject : pipe_read oops on sh
Submitter : Adrian McMenamin <[email protected]>
Date : 2009-01-15 9:48 (21 days old)
References : http://marc.info/?l=linux-kernel&m=123201298005600&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12501
Subject : build bug in eeepc-laptop.c
Submitter : Ingo Molnar <[email protected]>
Date : 2009-01-14 17:25 (22 days old)
References : http://lkml.org/lkml/2009/1/14/315

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12499
Subject : Problem with using bluetooth adaper connected to usb port
Submitter : Maciej Rutecki <[email protected]>
Date : 2009-01-13 18:34 (23 days old)
References : http://marc.info/?l=linux-kernel&m=123187185426236&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12497
Subject : new barrier warnings in 2.6.29-rc1
Submitter : Christoph Hellwig <[email protected]>
Date : 2009-01-12 15:46 (24 days old)
References : http://marc.info/?l=linux-kernel&m=123177528217154&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12494
Subject : Sony backlight regression from 2.6.28 to 29-rc
Submitter : Norbert Preining <[email protected]>
Date : 2009-01-19 8:14 (17 days old)
References : http://marc.info/?l=linux-acpi&m=123235286829512&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12491
Subject : i915 lockdep warning
Submitter : Brandeburg, Jesse <[email protected]>
Date : 2009-01-13 23:17 (23 days old)
References : http://marc.info/?l=linux-kernel&m=123188898423532&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12490
Subject : ath5k related kernel panic in 2.6.29-rc1
Submitter : Sergey S. Kostyliov <[email protected]>
Date : 2009-01-12 7:38 (24 days old)
References : http://marc.info/?l=linux-kernel&m=123174591509586&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12444
Subject : X hangs following switch from radeonfb console - Bisected
Submitter : Graham Murray <[email protected]>
Date : 2009-01-13 14:03 (23 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=7c1c2871a6a3a114853ec6836e9035ac1c0c7f7a

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12441
Subject : Xorg can't use dri on radeon X1950 AGP
Submitter : Daniel Vetter <[email protected]>
Date : 2009-01-13 05:20 (23 days old)

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12419
Subject : possible circular locking dependency on i915 dma
Submitter : Wang Chen <[email protected]>
Date : 2009-01-08 14:11 (28 days old)
References : http://marc.info/?l=linux-kernel&m=123142399720125&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12418
Subject : Repeated ioctl(4, 0x40046445, ..) loop in glxgears
Submitter : Pallipadi, Venkatesh <[email protected]>
Date : 2009-01-07 22:43 (29 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=7c1c2871a6a3a114853ec6836e9035ac1c0c7f7a
References : http://marc.info/?l=linux-kernel&m=123136836213319&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12417
Subject : glx performance drop with: "x86: PAT: implement track/untrack of pfnmap regions for x86 - v3"
Submitter : Alexey Fisher <[email protected]>
Date : 2009-01-06 18:46 (30 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=5899329b19100c0b82dc78e9b21ed8b920c9ffb3
References : http://marc.info/?l=linux-kernel&m=123126782822696&w=4

Ingo Molnar <[email protected]>

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12415
Subject : WARNING: at drivers/net/wireless/iwlwifi/iwl-sta.c:689
Submitter : Christian Borntraeger <[email protected]>
Date : 2009-01-05 10:36 (31 days old)
References : http://marc.info/?l=linux-wireless&m=123115178019082&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12414
Subject : iwl4965 cannot use "ap auto" on latest 2.6.28/29?
Submitter : Jeff Chua <[email protected]>
Date : 2009-01-05 4:13 (31 days old)
References : http://marc.info/?l=linux-kernel&m=123112882127823&w=4

Regressions with patches
------------------------

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12621
Subject : Resume broken on iBook
Submitter : Andreas Schwab <[email protected]>
Date : 2009-02-01 16:54 (4 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=aa8c6c93747f7b55fa11e1624fec8ca33763a805

Benjamin Herrenschmidt <[email protected]>
Patch : http://bugzilla.kernel.org/attachment.cgi?id=20085&action=view

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12609
Subject : v2.6.29-rc2 libata sff 32bit PIO regression
Submitter : Larry Finger <[email protected]>
Date : 2009-01-23 23:52 (13 days old)
References : http://marc.info/?l=linux-kernel&m=123275478111406&w=4

		  http://marc.info/?l=linux-kernel&m=123254501314058&w=4

Hugh Dickins <[email protected]>
Patch : http://marc.info/?l=linux-kernel&m=123254501314058&w=2

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12605
Subject : Suspend regression, hang after matroxfb
Submitter : Eric Sesterhenn <[email protected]>
Date : 2009-01-29 12:58 (7 days old)
References : http://thread.gmane.org/gmane.linux.power-management.general/12640

Patch : http://marc.info/?l=linux-kernel&m=123335902122029&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12598
Subject : tg3 dead after resume
Submitter : Parag Warudkar <[email protected]>
Date : 2009-01-29 0:14 (7 days old)
References : http://marc.info/?l=linux-kernel&m=123318968501524&w=4

Rafael J. Wysocki <[email protected]>
Patch : http://marc.info/?l=linux-kernel&m=123365342301877&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12538
Subject : xfs_fsr fails on 2.6.29-rc kernels
Submitter : Paul Martin <[email protected]>
Date : 2009-01-25 13:02 (11 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=743bb4650da9e2595d6cedd01c680b5b9398c74a

Patch : http://bugzilla.kernel.org/attachment.cgi?id=19986&action=view

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12509
Subject : lockdep report. fb_mmap vs sys_mmap2
Submitter : Dave Jones <[email protected]>
Date : 2009-01-17 18:19 (19 days old)
References : http://lkml.org/lkml/2009/1/17/176

Patch : http://marc.info/?l=linux-kernel&m=123231032115956&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12496
Subject : swsusp cannot find resume device (sometimes)
Submitter : Rafael J. Wysocki <[email protected]>
Date : 2009-01-09 12:24 (27 days old)
References : http://marc.info/?l=linux-kernel&m=123150395731165&w=4

Patch : http://marc.info/?l=linux-kernel&m=123156441218358&w=4

		  http://marc.info/?t=123156453100002&r=1&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12495
Subject : thinkpad problems during resume
Submitter : Christian Borntraeger <[email protected]>
Date : 2009-01-17 15:48 (19 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=e39ad415ac15116df213dfa2aa2a4f1b0857af9c
References : http://marc.info/?l=linux-kernel&m=123220746125479&w=4

		  http://marc.info/?l=linux-kernel&m=123170361827197&w=4

Patch : http://lkml.org/lkml/2009/1/16/375

		  http://lkml.org/lkml/2009/1/16/376
		  http://lkml.org/lkml/2009/1/16/378

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12493
Subject : ACPI related kernel panic when booting 2.6.29-rc2
Submitter : Peter Klotz <[email protected]>
Date : 2009-01-17 20:47 (19 days old)
References : http://marc.info/?l=linux-acpi&m=123222565017047&w=4

Tero Roponen <[email protected]>
Patch : http://marc.info/?l=linux-acpi&m=123223261225680&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12427
Subject : cpumask change causes sparc build bustage
Submitter : David Miller <[email protected]>
Date : 2009-01-11 8:31 (25 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=588235bb53f2c215f0d4b08fd30b461fedc3338e
References : http://marc.info/?l=linux-kernel&m=123166268013838&w=4

Patch : http://marc.info/?l=linux-kernel&m=123167571525706&w=4

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12416
Subject : Recent change to kernel spikes out ccache/distcc
Submitter : Theodore Ts'o <[email protected]>
Date : 2009-01-06 15:15 (30 days old)
References : http://marc.info/?l=linux-kernel&m=123125495000603&w=4
Patch : http://lkml.org/lkml/2009/1/20/78

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12400
Subject : git-latest: kernel oops in IOMMU setup
Submitter : Dirk Hohndel <[email protected]>
Date : 2009-01-08 20:05 (28 days old)
References : http://marc.info/?l=linux-kernel&m=123144520124853&w=4

Patch : http://marc.info/?l=linux-kernel&m=123152033725557&w=4

For details, please visit the bug entries and follow the links given in references.

As you can see, there is a Bugzilla entry for each of the listed regressions. There also is a Bugzilla entry used for tracking the regressions from 2.6.28, unresolved as well as resolved, at:

http://bugzilla.kernel.org/show_bug.cgi?id=12398

Please let me know if there are any Bugzilla entries that should be added to the list in there.

Thanks,
Rafael

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

#1 lore raw

[Bug #12400] git-latest: kernel oops in IOMMU setup

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12400
Subject : git-latest: kernel oops in IOMMU setup
Submitter : Dirk Hohndel <[email protected]>
Date : 2009-01-08 20:05 (28 days old)
References : http://marc.info/?l=linux-kernel&m=123144520124853&w=4

Patch : http://marc.info/?l=linux-kernel&m=123152033725557&w=4

#2 lore raw

[Bug #12416] Recent change to kernel spikes out ccache/distcc

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12416
Subject : Recent change to kernel spikes out ccache/distcc
Submitter : Theodore Ts'o <[email protected]>
Date : 2009-01-06 15:15 (30 days old)
References : http://marc.info/?l=linux-kernel&m=123125495000603&w=4
Patch : http://lkml.org/lkml/2009/1/20/78

#3 lore raw

Re: [Bug #12416] Recent change to kernel spikes out ccache/distcc

Quoted 8 lines · Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 6 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12416
Subject : Recent change to kernel spikes out ccache/distcc
Submitter : Theodore Ts'o <[email protected]>
Date : 2009-01-06 15:15 (30 days old)
References : http://marc.info/?l=linux-kernel&m=123125495000603&w=4
Patch : http://lkml.org/lkml/2009/1/20/78

This can be closed; Sam reverted the necessary patch the relevent commits in 2ea03891, which is in Linus's tree.

Rafael, I've closed the bug in the Kernel Bugzilla for your convenience.

Thanks,

- Ted

#4 lore raw

Re: [Bug #12416] Recent change to kernel spikes out ccache/distcc

Quoted 16 lines · Theodore Tso
Quoted 8 lines · Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 6 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12416
Subject : Recent change to kernel spikes out ccache/distcc
Submitter : Theodore Ts'o <[email protected]>
Date : 2009-01-06 15:15 (30 days old)
References : http://marc.info/?l=linux-kernel&m=123125495000603&w=4
Patch : http://lkml.org/lkml/2009/1/20/78

Quoted 4 lines

This can be closed; Sam reverted the necessary patch the relevent commits in 2ea03891, which is in Linus's tree.

Rafael, I've closed the bug in the Kernel Bugzilla for your convenience.

Great, thanks a lot!

Best,
Rafael

#5 lore raw

[Bug #12415] WARNING: at drivers/net/wireless/iwlwifi/iwl-sta.c:689

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12415
Subject : WARNING: at drivers/net/wireless/iwlwifi/iwl-sta.c:689
Submitter : Christian Borntraeger <[email protected]>
Date : 2009-01-05 10:36 (31 days old)
References : http://marc.info/?l=linux-wireless&m=123115178019082&w=4

#6 lore raw

Re: [Bug #12415] WARNING: at drivers/net/wireless/iwlwifi/iwl-sta.c:689

Am Wednesday 04 February 2009 11:23:55 schrieb Rafael J. Wysocki:

Quoted 8 lines

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 5 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12415
Subject : WARNING: at drivers/net/wireless/iwlwifi/iwl-sta.c:689
Submitter : Christian Borntraeger <[email protected]>
Date : 2009-01-05 10:36 (31 days old)
References : http://marc.info/?l=linux-wireless&m=123115178019082&w=4

Patch is available but not yet upstream:
http://marc.info/?l=linux-wireless&m=123316414222201&w=2

#7 lore raw

Re: [Bug #12415] WARNING: at drivers/net/wireless/iwlwifi/iwl-sta.c:689

Quoted 9 lines · Christian Borntraeger

Am Wednesday 04 February 2009 11:23:55 schrieb Rafael J. Wysocki:

Quoted 8 lines

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 8 lines
Quoted 5 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12415
Subject : WARNING: at drivers/net/wireless/iwlwifi/iwl-sta.c:689
Submitter : Christian Borntraeger <[email protected]>
Date : 2009-01-05 10:36 (31 days old)
References : http://marc.info/?l=linux-wireless&m=123115178019082&w=4

Patch is available but not yet upstream:
http://marc.info/?l=linux-wireless&m=123316414222201&w=2

Thanks, I've updated the bug entry.

Rafael

#8 lore raw

[Bug #12414] iwl4965 cannot use "ap auto" on latest 2.6.28/29?

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12414
Subject : iwl4965 cannot use "ap auto" on latest 2.6.28/29?
Submitter : Jeff Chua <[email protected]>
Date : 2009-01-05 4:13 (31 days old)
References : http://marc.info/?l=linux-kernel&m=123112882127823&w=4

#9 lore raw

Re: [Bug #12414] iwl4965 cannot use "ap auto" on latest 2.6.28/29?

Quoted 5 lines · Rafael J. Wysocki

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12414
Subject : iwl4965 cannot use "ap auto" on latest 2.6.28/29?
Submitter : Jeff Chua <[email protected]>
Date : 2009-01-05 4:13 (31 days old)
References : http://marc.info/?l=linux-kernel&m=123112882127823&w=4

Latest linux git pull seems to behave slightly better. Restarting the interface a few times fixes the problem 70% of the time. The rest of the time, it can't associate with the AP.

Jeff.

#10 lore raw

Re: [Bug #12414] iwl4965 cannot use "ap auto" on latest 2.6.28/29?

Quoted 10 lines · Jeff Chua
Quoted 5 lines · Rafael J. Wysocki

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12414
Subject : iwl4965 cannot use "ap auto" on latest 2.6.28/29?
Submitter : Jeff Chua <[email protected]>
Date : 2009-01-05 4:13 (31 days old)
References : http://marc.info/?l=linux-kernel&m=123112882127823&w=4

Latest linux git pull seems to behave slightly better. Restarting the interface a few times fixes the problem 70% of the time. The rest of the time, it can't associate with the AP.

Thanks for the update, there is a patch for you to test at
http://bugzilla.kernel.org/show_bug.cgi?id=12414#c3

Rafael

#11 lore raw

Re: [Bug #12414] iwl4965 cannot use "ap auto" on latest 2.6.28/29?

Quoted 8 lines · Rafael J. Wysocki
Quoted 7 lines · Jeff Chua
Quoted 5 lines · Rafael J. Wysocki

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12414
Subject : iwl4965 cannot use "ap auto" on latest 2.6.28/29?
Submitter : Jeff Chua <[email protected]>
Date : 2009-01-05 4:13 (31 days old)
References : http://marc.info/?l=linux-kernel&m=123112882127823&w=4

Quoted 6 lines
Quoted 3 lines

Latest linux git pull seems to behave slightly better. Restarting the interface a few times fixes the problem 70% of the time. The rest of the time, it can't associate with the AP.

Thanks for the update, there is a patch for you to test at
http://bugzilla.kernel.org/show_bug.cgi?id=12414#c3

Ok, will try that patch next weekend when I get home.

Thanks for the patch.

Jeff.

#12 lore raw

Re: [Bug #12414] iwl4965 cannot use "ap auto" on latest 2.6.28/29?

In reply to Rafael J. Wysocki
Quoted 5 lines · Rafael J. Wysocki

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12414
Subject : iwl4965 cannot use "ap auto" on latest 2.6.28/29?
Submitter : Jeff Chua <[email protected]>
Date : 2009-01-05 4:13 (31 days old)
References : http://marc.info/?l=linux-kernel&m=123112882127823&w=4

I applied the patch
http://marc.info/?l=linux-wireless&m=123320213115108&w=2 and it seems better now. One instant of not association out of 20 tries, but reloading the modules seems to make it work again. So, let's just close the case?

Here's part of the dmesg. Kernel is 2.6.29-rc5 with patch above applied.

wmaster0 (iwlagn): not using net_device_ops yet
phy0: Selected rate control algorithm 'iwl-agn-rs'
wlan0 (iwlagn): not using net_device_ops yet
iwlagn 0000:03:00.0: PCI INT A -> GSI 17 (level, low) -> IRQ 17 iwlagn 0000:03:00.0: irq 29 for MSI/MSI-X
iwlagn 0000:03:00.0: firmware: requesting iwlwifi-4965-2.ucode iwlagn loaded firmware version 228.57.2.23
Registered led device: iwl-phy0:radio
Registered led device: iwl-phy0:assoc
Registered led device: iwl-phy0:RX
Registered led device: iwl-phy0:TX
iwlagn: TX Power requested while scanning!
iwlagn: Error sending TX power (-11)
wlan0: authenticate with AP xxxxxxxxxxxxx
wlan0: authenticated
wlan0: associate with AP xxxxxxxxxxxxxx
wlan0: RX AssocResp from xxxxxxxxxxxxxx (capab=0x471 status=0 aid=24)
wlan0: associated

Thanks,
Jeff.

#13 lore raw

Re: [Bug #12414] iwl4965 cannot use "ap auto" on latest 2.6.28/29?

Quoted 7 lines · Jeff Chua
Quoted 5 lines · Rafael J. Wysocki

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12414
Subject : iwl4965 cannot use "ap auto" on latest 2.6.28/29?
Submitter : Jeff Chua <[email protected]>
Date : 2009-01-05 4:13 (31 days old)
References : http://marc.info/?l=linux-kernel&m=123112882127823&w=4

Quoted 5 lines

I applied the patch
http://marc.info/?l=linux-wireless&m=123320213115108&w=2 and it seems better now. One instant of not association out of 20 tries, but reloading the modules seems to make it work again. So, let's just close the case?

I tested on another AP (WAG200G ), and this one, it's failing to associate more often. And I noticed that when it does associate, I don't get these two lines ...

iwlagn: TX Power requested while scanning!
iwlagn: Error sending TX power (-11)

Here's a good run ...

iwlagn: Intel(R) Wireless WiFi Link AGN driver for Linux, 1.3.27ks
iwlagn: Copyright(c) 2003-2008 Intel Corporation
iwlagn 0000:03:00.0: PCI INT A -> GSI 17 (level, low) -> IRQ 17 iwlagn 0000:03:00.0: setting latency timer to 64
iwlagn: Detected Intel Wireless WiFi Link 4965AGN REV=0x4
iwlagn: Tunable channels: 11 802.11bg, 13 802.11a channels wmaster0 (iwlagn): not using net_device_ops yet
phy0: Selected rate control algorithm 'iwl-agn-rs' wlan0 (iwlagn): not using net_device_ops yet
iwlagn 0000:03:00.0: PCI INT A -> GSI 17 (level, low) -> IRQ 17 iwlagn 0000:03:00.0: irq 28 for MSI/MSI-X
iwlagn 0000:03:00.0: firmware: requesting iwlwifi-4965-2.ucode iwlagn loaded firmware version 228.57.2.23
Registered led device: iwl-phy0:radio
Registered led device: iwl-phy0:assoc
Registered led device: iwl-phy0:RX
Registered led device: iwl-phy0:TX
wlan0: authenticate with AP xxxxxxxxxxxxx
wlan0: authenticate with AP xxxxxxxxxxxxx
wlan0: authenticate with AP xxxxxxxxxxxxx
wlan0: authenticated
wlan0: associate with AP xxxxxxxxxxxx
wlan0: RX AssocResp from xxxxxxxxxxxx (capab=0x471 status=0 aid=1)
wlan0: associated

Thanks,
Jeff.

#14 lore raw

Re: [Bug #12414] iwl4965 cannot use "ap auto" on latest 2.6.28/29?

Quoted 41 lines · Jeff Chua
Quoted 7 lines · Jeff Chua
Quoted 5 lines · Rafael J. Wysocki

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12414
Subject : iwl4965 cannot use "ap auto" on latest 2.6.28/29?
Submitter : Jeff Chua <[email protected]>
Date : 2009-01-05 4:13 (31 days old)
References : http://marc.info/?l=linux-kernel&m=123112882127823&w=4

Quoted 5 lines

I applied the patch
http://marc.info/?l=linux-wireless&m=123320213115108&w=2 and it seems better now. One instant of not association out of 20 tries, but reloading the modules seems to make it work again. So, let's just close the case?

I tested on another AP (WAG200G ), and this one, it's failing to associate more often. And I noticed that when it does associate, I don't get these two lines ...

iwlagn: TX Power requested while scanning!
iwlagn: Error sending TX power (-11)

Here's a good run ...

iwlagn: Intel(R) Wireless WiFi Link AGN driver for Linux, 1.3.27ks
iwlagn: Copyright(c) 2003-2008 Intel Corporation
iwlagn 0000:03:00.0: PCI INT A -> GSI 17 (level, low) -> IRQ 17 iwlagn 0000:03:00.0: setting latency timer to 64
iwlagn: Detected Intel Wireless WiFi Link 4965AGN REV=0x4
iwlagn: Tunable channels: 11 802.11bg, 13 802.11a channels wmaster0 (iwlagn): not using net_device_ops yet
phy0: Selected rate control algorithm 'iwl-agn-rs' wlan0 (iwlagn): not using net_device_ops yet
iwlagn 0000:03:00.0: PCI INT A -> GSI 17 (level, low) -> IRQ 17 iwlagn 0000:03:00.0: irq 28 for MSI/MSI-X
iwlagn 0000:03:00.0: firmware: requesting iwlwifi-4965-2.ucode iwlagn loaded firmware version 228.57.2.23
Registered led device: iwl-phy0:radio
Registered led device: iwl-phy0:assoc
Registered led device: iwl-phy0:RX

Quoted 8 lines

Registered led device: iwl-phy0:TX
wlan0: authenticate with AP xxxxxxxxxxxxx
wlan0: authenticate with AP xxxxxxxxxxxxx
wlan0: authenticate with AP xxxxxxxxxxxxx
wlan0: authenticated
wlan0: associate with AP xxxxxxxxxxxx
wlan0: RX AssocResp from xxxxxxxxxxxx (capab=0x471 status=0 aid=1)
wlan0: associated

Thanks for the updates.

Rafael

#15 lore raw

[Bug #12427] cpumask change causes sparc build bustage

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12427
Subject : cpumask change causes sparc build bustage
Submitter : David Miller <davem-fT/PcQaiUtIeIZ0/[email protected]>
Date : 2009-01-11 8:31 (25 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=588235bb53f2c215f0d4b08fd30b461fedc3338e
References : http://marc.info/?l=linux-kernel&m=123166268013838&w=4

Patch : http://marc.info/?l=linux-kernel&m=123167571525706&w=4

#16 lore raw

[Bug #12419] possible circular locking dependency on i915 dma

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12419
Subject : possible circular locking dependency on i915 dma
Submitter : Wang Chen <[email protected]>
Date : 2009-01-08 14:11 (28 days old)
References : http://marc.info/?l=linux-kernel&m=123142399720125&w=4

#17 lore raw

[Bug #12417] glx performance drop with: "x86: PAT: implement track/untrack of pfnmap regions for x86 - v3"

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12417
Subject : glx performance drop with: "x86: PAT: implement track/untrack of pfnmap regions for x86 - v3"
Submitter : Alexey Fisher <[email protected]>
Date : 2009-01-06 18:46 (30 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=5899329b19100c0b82dc78e9b21ed8b920c9ffb3
References : http://marc.info/?l=linux-kernel&m=123126782822696&w=4

Ingo Molnar <[email protected]>

#18 lore raw

Re: [Bug #12417] glx performance drop with: "x86: PAT: implement track/untrack of pfnmap regions for x86 - v3"

Quoted 8 lines · Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 9 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12417
Subject : glx performance drop with: "x86: PAT: implement track/untrack of pfnmap regions for x86 - v3"
Submitter : Alexey Fisher <[email protected]>
Date : 2009-01-06 18:46 (30 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=5899329b19100c0b82dc78e9b21ed8b920c9ffb3
References : http://marc.info/?l=linux-kernel&m=123126782822696&w=4

Ingo Molnar <[email protected]>

this should be fixed in latest -git. Alexey, can you confirm?

Ingo

#19 lore raw

Re: [Bug #12417] glx performance drop with: "x86: PAT: implement track/untrack of pfnmap regions for x86 - v3"

Quoted 20 lines · Ingo Molnar
Quoted 8 lines · Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 9 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12417
Subject : glx performance drop with: "x86: PAT: implement track/untrack of pfnmap regions for x86 - v3"
Submitter : Alexey Fisher <[email protected]>
Date : 2009-01-06 18:46 (30 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=5899329b19100c0b82dc78e9b21ed8b920c9ffb3
References : http://marc.info/?l=linux-kernel&m=123126782822696&w=4

Ingo Molnar <[email protected]>

Quoted 3 lines

this should be fixed in latest -git. Alexey, can you confirm?

Ingo

I confirm it. It working great with latest git 2.6.29-rc3-00495-gdcf6a79.
Alexey

#20 lore raw

Re: [Bug #12417] glx performance drop with: "x86: PAT: implement track/untrack of pfnmap regions for x86 - v3"

Quoted 21 lines · Alexey Fisher
Quoted 20 lines · Ingo Molnar
Quoted 17 lines · Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12417
Subject : glx performance drop with: "x86: PAT: implement track/untrack of pfnmap regions for x86 - v3"
Submitter : Alexey Fisher <[email protected]>
Date : 2009-01-06 18:46 (30 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=5899329b19100c0b82dc78e9b21ed8b920c9ffb3
References : http://marc.info/?l=linux-kernel&m=123126782822696&w=4

Ingo Molnar <[email protected]>

Quoted 5 lines
Quoted 3 lines

this should be fixed in latest -git. Alexey, can you confirm?

Ingo

I confirm it. It working great with latest git 2.6.29-rc3-00495-gdcf6a79.

Thanks for the testing!

Ingo

#21 lore raw

Re: [Bug #12417] glx performance drop with: "x86: PAT: implement track/untrack of pfnmap regions for x86 - v3"

Quoted 24 lines · Ingo Molnar
Quoted 21 lines · Alexey Fisher
Quoted 20 lines · Ingo Molnar
Quoted 17 lines · Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12417
Subject : glx performance drop with: "x86: PAT: implement track/untrack of pfnmap regions for x86 - v3"
Submitter : Alexey Fisher <[email protected]>
Date : 2009-01-06 18:46 (30 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=5899329b19100c0b82dc78e9b21ed8b920c9ffb3
References : http://marc.info/?l=linux-kernel&m=123126782822696&w=4

Ingo Molnar <[email protected]>

Quoted 7 lines
Quoted 5 lines
Quoted 3 lines

this should be fixed in latest -git. Alexey, can you confirm?

Ingo

I confirm it. It working great with latest git 2.6.29-rc3-00495-gdcf6a79.

Thanks for the testing!

Closed.

Thanks,
Rafael

#22 lore raw

[Bug #12418] Repeated ioctl(4, 0x40046445, ..) loop in glxgears

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12418
Subject : Repeated ioctl(4, 0x40046445, ..) loop in glxgears
Submitter : Pallipadi, Venkatesh <[email protected]>
Date : 2009-01-07 22:43 (29 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=7c1c2871a6a3a114853ec6836e9035ac1c0c7f7a
References : http://marc.info/?l=linux-kernel&m=123136836213319&w=4

#23 lore raw

[Bug #12491] i915 lockdep warning

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12491
Subject : i915 lockdep warning
Submitter : Brandeburg, Jesse <[email protected]>
Date : 2009-01-13 23:17 (23 days old)
References : http://marc.info/?l=linux-kernel&m=123188898423532&w=4

#24 lore raw

Re: [Bug #12491] i915 lockdep warning

Quoted 5 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12491
Subject : i915 lockdep warning
Submitter : Brandeburg, Jesse <[email protected]>
Date : 2009-01-13 23:17 (23 days old)
References : http://marc.info/?l=linux-kernel&m=123188898423532&w=4

Looking at the code, it seems that the issue is that the DRM struct_mutex must be taken inside mmap_sem (because struct_mutex is taken in drm_vm_open(), which is called with mmap_sem already held), but i915_gem_execbuffer() does a copy_to_user() while holding struct_mutex, and if this copy faults, then the VM tries to acquire mmap_sem -- ie lockdep identifies correctly a potential AB/BA deadlock.

I don't pretend to fully understand the DRM or GEM, but a possible fix is below -- would be worth it to test and review, and get into 2.6.29 if it is a correct fix:

---
i915: Fix potential AB-BA deadlock in i915_gem_execbuffer()

Lockdep warns that i915_gem_execbuffer() can trigger a page fault (which takes mmap_sem) while holding dev->struct_mutex, while drm_vm_open() (which is called with mmap_sem already held) takes dev->struct_mutex. So this is a potential AB-BA deadlock.

The way that i915_gem_execbuffer() triggers a page fault is by doing copy_to_user() when returning new buffer offsets back to userspace; however there is no reason to hold the struct_mutex when doing this copy, since what is being copied is a private array anyway. So we can fix the potential deadlock (and get rid of the lockdep warning) by simply moving the copy_to_user() outside of where struct_mutex is held.

This fixes <http://bugzilla.kernel.org/show_bug.cgi?id=12491>.

---
 drivers/gpu/drm/i915/i915_gem.c |   21 ++++++++++++---------
 1 files changed, 12 insertions(+), 9 deletions(-)
drivers/gpu/drm/i915/i915_gem.c +12 -9
index debad5c..23aad8c 100644
@@ -2610,15 +2610,6 @@ i915_gem_execbuffer(struct drm_device *dev, void *data,
 
 	i915_verify_inactive(dev, __FILE__, __LINE__);
 
-	/* Copy the new buffer offsets back to the user's exec list. */
-	ret = copy_to_user((struct drm_i915_relocation_entry __user *)
-			   (uintptr_t) args->buffers_ptr,
-			   exec_list,
-			   sizeof(*exec_list) * args->buffer_count);
-	if (ret)
-		DRM_ERROR("failed to copy %d exec entries "
-			  "back to user (%d)\n",
-			   args->buffer_count, ret);
 err:
 	for (i = 0; i < pinned; i++)
 		i915_gem_object_unpin(object_list[i]);
@@ -2628,6 +2619,18 @@ err:
 
 	mutex_unlock(&dev->struct_mutex);
 
+	if (!ret) {
+		/* Copy the new buffer offsets back to the user's exec list. */
+		ret = copy_to_user((struct drm_i915_relocation_entry __user *)
+				   (uintptr_t) args->buffers_ptr,
+				   exec_list,
+				   sizeof(*exec_list) * args->buffer_count);
+		if (ret)
+			DRM_ERROR("failed to copy %d exec entries "
+				  "back to user (%d)\n",
+				  args->buffer_count, ret);
+	}
+
 pre_mutex_err:
 	drm_free(object_list, sizeof(*object_list) * args->buffer_count,
 		 DRM_MEM_DRIVER);
#25 lore raw

Re: [Bug #12491] i915 lockdep warning

[Adding CCs.]

Quoted 80 lines · Roland Dreier
Quoted 5 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12491
Subject : i915 lockdep warning
Submitter : Brandeburg, Jesse <[email protected]>
Date : 2009-01-13 23:17 (23 days old)
References : http://marc.info/?l=linux-kernel&m=123188898423532&w=4

Looking at the code, it seems that the issue is that the DRM struct_mutex must be taken inside mmap_sem (because struct_mutex is taken in drm_vm_open(), which is called with mmap_sem already held), but i915_gem_execbuffer() does a copy_to_user() while holding struct_mutex, and if this copy faults, then the VM tries to acquire mmap_sem -- ie lockdep identifies correctly a potential AB/BA deadlock.

I don't pretend to fully understand the DRM or GEM, but a possible fix is below -- would be worth it to test and review, and get into 2.6.29 if it is a correct fix:

---
i915: Fix potential AB-BA deadlock in i915_gem_execbuffer()

Lockdep warns that i915_gem_execbuffer() can trigger a page fault (which takes mmap_sem) while holding dev->struct_mutex, while drm_vm_open() (which is called with mmap_sem already held) takes dev->struct_mutex. So this is a potential AB-BA deadlock.

The way that i915_gem_execbuffer() triggers a page fault is by doing copy_to_user() when returning new buffer offsets back to userspace; however there is no reason to hold the struct_mutex when doing this copy, since what is being copied is a private array anyway. So we can fix the potential deadlock (and get rid of the lockdep warning) by simply moving the copy_to_user() outside of where struct_mutex is held.

This fixes <http://bugzilla.kernel.org/show_bug.cgi?id=12491>.

---
 drivers/gpu/drm/i915/i915_gem.c |   21 ++++++++++++---------
 1 files changed, 12 insertions(+), 9 deletions(-)
drivers/gpu/drm/i915/i915_gem.c +12 -10
index debad5c..23aad8c 100644
@@ -2610,15 +2610,6 @@ i915_gem_execbuffer(struct drm_device *dev, void *data,
 
 	i915_verify_inactive(dev, __FILE__, __LINE__);
 
-	/* Copy the new buffer offsets back to the user's exec list. */
-	ret = copy_to_user((struct drm_i915_relocation_entry __user *)
-			   (uintptr_t) args->buffers_ptr,
-			   exec_list,
-			   sizeof(*exec_list) * args->buffer_count);
-	if (ret)
-		DRM_ERROR("failed to copy %d exec entries "
-			  "back to user (%d)\n",
-			   args->buffer_count, ret);
 err:
 	for (i = 0; i < pinned; i++)
 		i915_gem_object_unpin(object_list[i]);
@@ -2628,6 +2619,18 @@ err:
 
 	mutex_unlock(&dev->struct_mutex);
 
+	if (!ret) {
+		/* Copy the new buffer offsets back to the user's exec list. */
+		ret = copy_to_user((struct drm_i915_relocation_entry __user *)
+				   (uintptr_t) args->buffers_ptr,
+				   exec_list,
+				   sizeof(*exec_list) * args->buffer_count);
+		if (ret)
+			DRM_ERROR("failed to copy %d exec entries "
+				  "back to user (%d)\n",
+				  args->buffer_count, ret);
+	}
+
 pre_mutex_err:
 	drm_free(object_list, sizeof(*object_list) * args->buffer_count,
 		 DRM_MEM_DRIVER);
--
#27 lore raw

Re: [Bug #12491] i915 lockdep warning

Quoted 5 lines · Roland Dreier
Quoted 1 line

i915: Fix potential AB-BA deadlock in i915_gem_execbuffer()

Jesse, Eric, or anybody -- does this look correct to you? This seems to be triggering for several people, eg
https://bugzilla.redhat.com/show_bug.cgi?id=481687

Eric's been looking at this issue, and there's some discussion about it on the intel-gfx list, but I'm not sure what the latest is (Eric is travelling atm).

Signature · 1 line
Jesse Barnes, Intel Open Source Technology Center
#28 lore raw

RE: [Bug #12491] i915 lockdep warning

In reply to Roland Dreier
Quoted 7 lines · Roland Dreier
Quoted 1 line

i915: Fix potential AB-BA deadlock in i915_gem_execbuffer()

Jesse, Eric, or anybody -- does this look correct to you? This seems to be triggering for several people, eg
https://bugzilla.redhat.com/show_bug.cgi?id=481687

- R.

I tested this on my 965 based system that showed the issue and it seemed resolved, at least no warnings in dmesg. Lockdep was confirmed to still be enabled.

I can mark the bug fixed in korg bugzilla too.

Thank you Roland for looking into this, good work!

Tested-by Jesse Brandeburg
#29 lore raw

Re: [Bug #12491] i915 lockdep warning

[intel-gfx CC added]

Quoted 3 lines

I tested this on my 965 based system that showed the issue and it seemed resolved, at least no warnings in dmesg. Lockdep was confirmed to still be enabled.

I can mark the bug fixed in korg bugzilla too.

Probably we just want to add a comment saying the patch works -- the bug should stay open until we get the fix upstream I guess.

Quoted 1 line

Thank you Roland for looking into this, good work!

Thanks for testing! Patch with updated changelog (including your tested-by) below. Jesse [Barnes] and/or Eric please review and apply:

---

i915: Fix potential AB-BA deadlock in i915_gem_execbuffer()

Lockdep warns that i915_gem_execbuffer() can trigger a page fault (which takes mmap_sem) while holding dev->struct_mutex, while drm_vm_open() (which is called with mmap_sem already held) takes dev->struct_mutex. So this is a potential AB-BA deadlock.

The way that i915_gem_execbuffer() triggers a page fault is by doing copy_to_user() when returning new buffer offsets back to userspace; however there is no reason to hold the struct_mutex when doing this copy, since what is being copied is the contents of an array private to i915_gem_execbuffer() anyway. So we can fix the potential deadlock (and get rid of the lockdep warning) by simply moving the copy_to_user() outside of where struct_mutex is held.

This fixes <http://bugzilla.kernel.org/show_bug.cgi?id=12491>.

---
 drivers/gpu/drm/i915/i915_gem.c |   21 ++++++++++++---------
 1 files changed, 12 insertions(+), 9 deletions(-)
drivers/gpu/drm/i915/i915_gem.c +12 -9
index debad5c..23aad8c 100644
@@ -2610,15 +2610,6 @@ i915_gem_execbuffer(struct drm_device *dev, void *data,
 
 	i915_verify_inactive(dev, __FILE__, __LINE__);
 
-	/* Copy the new buffer offsets back to the user's exec list. */
-	ret = copy_to_user((struct drm_i915_relocation_entry __user *)
-			   (uintptr_t) args->buffers_ptr,
-			   exec_list,
-			   sizeof(*exec_list) * args->buffer_count);
-	if (ret)
-		DRM_ERROR("failed to copy %d exec entries "
-			  "back to user (%d)\n",
-			   args->buffer_count, ret);
 err:
 	for (i = 0; i < pinned; i++)
 		i915_gem_object_unpin(object_list[i]);
@@ -2628,6 +2619,18 @@ err:
 
 	mutex_unlock(&dev->struct_mutex);
 
+	if (!ret) {
+		/* Copy the new buffer offsets back to the user's exec list. */
+		ret = copy_to_user((struct drm_i915_relocation_entry __user *)
+				   (uintptr_t) args->buffers_ptr,
+				   exec_list,
+				   sizeof(*exec_list) * args->buffer_count);
+		if (ret)
+			DRM_ERROR("failed to copy %d exec entries "
+				  "back to user (%d)\n",
+				  args->buffer_count, ret);
+	}
+
 pre_mutex_err:
 	drm_free(object_list, sizeof(*object_list) * args->buffer_count,
 		 DRM_MEM_DRIVER);
#30 lore raw

Re: [Bug #12491] i915 lockdep warning

Quoted 7 lines · Roland Dreier

[intel-gfx CC added]

Quoted 3 lines

I tested this on my 965 based system that showed the issue and it seemed resolved, at least no warnings in dmesg. Lockdep was confirmed to still be enabled.

I can mark the bug fixed in korg bugzilla too.

Probably we just want to add a comment saying the patch works

Yes, I have already added this information to the bug entry.

Quoted 1 line

-- the bug should stay open until we get the fix upstream I guess.

Yes, the bug will be closed as soon as it's fixed upstream.

Thanks,
Rafael

#31 lore raw

Re: [Bug #12491] i915 lockdep warning

In reply to Roland Dreier
Quoted 31 lines · Roland Dreier

[intel-gfx CC added]

Quoted 3 lines

I tested this on my 965 based system that showed the issue and it seemed resolved, at least no warnings in dmesg. Lockdep was confirmed to still be enabled.

I can mark the bug fixed in korg bugzilla too.

Probably we just want to add a comment saying the patch works -- the bug should stay open until we get the fix upstream I guess.

Quoted 1 line

Thank you Roland for looking into this, good work!

Thanks for testing! Patch with updated changelog (including your tested-by) below. Jesse [Barnes] and/or Eric please review and apply:

---

i915: Fix potential AB-BA deadlock in i915_gem_execbuffer()

Lockdep warns that i915_gem_execbuffer() can trigger a page fault (which takes mmap_sem) while holding dev->struct_mutex, while drm_vm_open() (which is called with mmap_sem already held) takes dev->struct_mutex. So this is a potential AB-BA deadlock.

The way that i915_gem_execbuffer() triggers a page fault is by doing copy_to_user() when returning new buffer offsets back to userspace; however there is no reason to hold the struct_mutex when doing this copy, since what is being copied is the contents of an array private to i915_gem_execbuffer() anyway. So we can fix the potential deadlock (and get rid of the lockdep warning) by simply moving the copy_to_user() outside of where struct_mutex is held.

Quoted 5 lines

This fixes <http://bugzilla.kernel.org/show_bug.cgi?id=12491>.

This only fixes 1 of the about 4 of this type of deadlock, but it looks like a fine fix for that and I've put it in my tree. Thanks!

(I'd been holding off on hitting the easy ones to figure out if we could get the lock order the other way around. It would be really nice if we could have struct_mutex -> mmap_sem, but it looks like that's just not going to happen. pread and pwrite are going to be a mess.)

Signature · 2 lines
Eric Anholt
[email protected]                         [email protected]
#32 lore raw

Re: [Bug #12491] i915 lockdep warning

Quoted 2 lines

This only fixes 1 of the about 4 of this type of deadlock, but it looks like a fine fix for that and I've put it in my tree. Thanks!

Heh, just fixing what I saw reported. probably once this gets in the harder reports will pop up because lockdep won't disable itself quite so soon...

- R.

#33 lore raw

[Bug #12441] Xorg can't use dri on radeon X1950 AGP

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12441
Subject : Xorg can't use dri on radeon X1950 AGP
Submitter : Daniel Vetter <daniel-/[email protected]>
Date : 2009-01-13 05:20 (23 days old)

#34 lore raw

[Bug #12444] X hangs following switch from radeonfb console - Bisected

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12444
Subject : X hangs following switch from radeonfb console - Bisected
Submitter : Graham Murray <[email protected]>
Date : 2009-01-13 14:03 (23 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=7c1c2871a6a3a114853ec6836e9035ac1c0c7f7a

#35 lore raw

[Bug #12490] ath5k related kernel panic in 2.6.29-rc1

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12490
Subject : ath5k related kernel panic in 2.6.29-rc1
Submitter : Sergey S. Kostyliov <[email protected]>
Date : 2009-01-12 7:38 (24 days old)
References : http://marc.info/?l=linux-kernel&m=123174591509586&w=4

#36 lore raw

Re: [Bug #12490] ath5k related kernel panic in 2.6.29-rc1

On Wed, 4 Feb 2009 11:23:57 +0100 (CET), Rafael J. Wysocki wrote

Quoted 7 lines

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 6 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12490
Subject : ath5k related kernel panic in 2.6.29-rc1
Submitter : Sergey S. Kostyliov <[email protected]>
Date : 2009-01-12 7:38 (24 days old)
References : http://marc.info/?l=linux-kernel&m=123174591509586&w=4

Yes, this should still be listed. Seems there is corruption of the rate indexes but I haven't found it yet. Felix Fietkau added to CC in case he has any thoughts.

Signature · 1 line
Bob Copeland %% www.bobcopeland.com
#37 lore raw

Re: [Bug #12490] ath5k related kernel panic in 2.6.29-rc1

Quoted 15 lines · Bob Copeland

On Wed, 4 Feb 2009 11:23:57 +0100 (CET), Rafael J. Wysocki wrote

Quoted 7 lines

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 6 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12490
Subject : ath5k related kernel panic in 2.6.29-rc1
Submitter : Sergey S. Kostyliov <[email protected]>
Date : 2009-01-12 7:38 (24 days old)
References : http://marc.info/?l=linux-kernel&m=123174591509586&w=4

Quoted 3 lines

Yes, this should still be listed. Seems there is corruption of the rate indexes but I haven't found it yet. Felix Fietkau added to CC in case he has any thoughts.

Thanks for the update.

Rafael

#38 lore raw

[Bug #12494] Sony backlight regression from 2.6.28 to 29-rc

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12494
Subject : Sony backlight regression from 2.6.28 to 29-rc
Submitter : Norbert Preining <[email protected]>
Date : 2009-01-19 8:14 (17 days old)
References : http://marc.info/?l=linux-acpi&m=123235286829512&w=4

#39 lore raw

[Bug #12493] ACPI related kernel panic when booting 2.6.29-rc2

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12493
Subject : ACPI related kernel panic when booting 2.6.29-rc2
Submitter : Peter Klotz <[email protected]>
Date : 2009-01-17 20:47 (19 days old)
References : http://marc.info/?l=linux-acpi&m=123222565017047&w=4

Tero Roponen <[email protected]>
Patch : http://marc.info/?l=linux-acpi&m=123223261225680&w=4

#40 lore raw

[Bug #12495] thinkpad problems during resume

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12495
Subject : thinkpad problems during resume
Submitter : Christian Borntraeger <[email protected]>
Date : 2009-01-17 15:48 (19 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=e39ad415ac15116df213dfa2aa2a4f1b0857af9c
References : http://marc.info/?l=linux-kernel&m=123220746125479&w=4

		  http://marc.info/?l=linux-kernel&m=123170361827197&w=4

Patch : http://lkml.org/lkml/2009/1/16/375

		  http://lkml.org/lkml/2009/1/16/376
		  http://lkml.org/lkml/2009/1/16/378
#41 lore raw

Re: [Bug #12495] thinkpad problems during resume

Am Wednesday 04 February 2009 11:23:58 schrieb Rafael J. Wysocki:

Quoted 17 lines

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12495
Subject : thinkpad problems during resume
Submitter : Christian Borntraeger <[email protected]>
Date : 2009-01-17 15:48 (19 days old)
First-Bad-Commit:
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h
=e39ad415ac15116df213dfa2aa2a4f1b0857af9c References :
http://marc.info/?l=linux-kernel&m=123220746125479&w=4
http://marc.info/?l=linux-kernel&m=123170361827197&w=4

Quoted 4 lines

Patch : http://lkml.org/lkml/2009/1/16/375

		  http://lkml.org/lkml/2009/1/16/376
		  http://lkml.org/lkml/2009/1/16/378

Fixed upstream.

#42 lore raw

Re: [Bug #12495] thinkpad problems during resume

Quoted 23 lines · Christian Borntraeger

Am Wednesday 04 February 2009 11:23:58 schrieb Rafael J. Wysocki:

Quoted 17 lines

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12495
Subject : thinkpad problems during resume
Submitter : Christian Borntraeger <[email protected]>
Date : 2009-01-17 15:48 (19 days old)
First-Bad-Commit:
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h
=e39ad415ac15116df213dfa2aa2a4f1b0857af9c References :
http://marc.info/?l=linux-kernel&m=123220746125479&w=4
http://marc.info/?l=linux-kernel&m=123170361827197&w=4

Quoted 4 lines

Patch : http://lkml.org/lkml/2009/1/16/375

		  http://lkml.org/lkml/2009/1/16/376
		  http://lkml.org/lkml/2009/1/16/378
Quoted 1 line

Fixed upstream.

Thanks, closed.

Rafael

#43 lore raw

[Bug #12496] swsusp cannot find resume device (sometimes)

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12496
Subject : swsusp cannot find resume device (sometimes)
Submitter : Rafael J. Wysocki <[email protected]>
Date : 2009-01-09 12:24 (27 days old)
References : http://marc.info/?l=linux-kernel&m=123150395731165&w=4

Patch : http://marc.info/?l=linux-kernel&m=123156441218358&w=4

		  http://marc.info/?t=123156453100002&r=1&w=4
#44 lore raw

[Bug #12499] Problem with using bluetooth adaper connected to usb port

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12499
Subject : Problem with using bluetooth adaper connected to usb port
Submitter : Maciej Rutecki <[email protected]>
Date : 2009-01-13 18:34 (23 days old)
References : http://marc.info/?l=linux-kernel&m=123187185426236&w=4

#45 lore raw

[Bug #12501] build bug in eeepc-laptop.c

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12501
Subject : build bug in eeepc-laptop.c
Submitter : Ingo Molnar <[email protected]>
Date : 2009-01-14 17:25 (22 days old)
References : http://lkml.org/lkml/2009/1/14/315

#46 lore raw

[Bug #12497] new barrier warnings in 2.6.29-rc1

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12497
Subject : new barrier warnings in 2.6.29-rc1
Submitter : Christoph Hellwig <[email protected]>
Date : 2009-01-12 15:46 (24 days old)
References : http://marc.info/?l=linux-kernel&m=123177528217154&w=4

#47 lore raw

Re: [Bug #12497] new barrier warnings in 2.6.29-rc1

Quoted 6 lines · Rafael J. Wysocki

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12497
Subject : new barrier warnings in 2.6.29-rc1
Submitter : Christoph Hellwig <[email protected]>
Date : 2009-01-12 15:46 (24 days old)
References : http://marc.info/?l=linux-kernel&m=123177528217154&w=4

We're now done to a single warning per mount instead of the previous two:

[  485.307225] end_request: I/O error, dev vdb, sector 0
[  485.604639] Filesystem "vdb": Disabling barriers, trial barrier write failed
[  485.627352] XFS mounting filesystem vdb

but it's still around.

#48 lore raw

[Bug #12502] pipe_read oops on sh

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12502
Subject : pipe_read oops on sh
Submitter : Adrian McMenamin <[email protected]>
Date : 2009-01-15 9:48 (21 days old)
References : http://marc.info/?l=linux-kernel&m=123201298005600&w=4

#49 lore raw

[Bug #12503] [slab corruption] BUG key_jar: Poison overwritten

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12503
Subject : [slab corruption] BUG key_jar: Poison overwritten
Submitter : Ingo Molnar <[email protected]>
Date : 2009-01-15 18:16 (21 days old)
References : http://marc.info/?l=linux-kernel&m=123204353425825&w=4

#50 lore raw

[Bug #12505] 2.6.29-rc1 Firefox crashing on page load

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12505
Subject : 2.6.29-rc1 Firefox crashing on page load
Submitter : Justin Madru <[email protected]>
Date : 2009-01-16 20:56 (20 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=4217458dafaa57d8e26a46f5d05ab8c53cf64191
References : http://marc.info/?l=linux-kernel&m=123213941914274&w=4

#51 lore raw

Re: [Bug #12505] 2.6.29-rc1 Firefox crashing on page load

Quoted 8 lines · Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 10 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12505
Subject : 2.6.29-rc1 Firefox crashing on page load
Submitter : Justin Madru <[email protected]>
Date : 2009-01-16 20:56 (20 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=4217458dafaa57d8e26a46f5d05ab8c53cf64191
References : http://marc.info/?l=linux-kernel&m=123213941914274&w=4

Well the hardest for this, was to reproduce the error.(never did). I guess it can be taken of if there was a patch for the fix, and the patch made it into the main tree. but if this is a firefox issue, then a report needs to be filed with them. so overall my answer is "I don't know."
:^)

regards;

Signature · 1 line
Justin P. Mattock
#52 lore raw

Re: [Bug #12505] 2.6.29-rc1 Firefox crashing on page load

Quoted 20 lines · Justin Mattock
Quoted 8 lines · Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 10 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12505
Subject : 2.6.29-rc1 Firefox crashing on page load
Submitter : Justin Madru <[email protected]>
Date : 2009-01-16 20:56 (20 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=4217458dafaa57d8e26a46f5d05ab8c53cf64191
References : http://marc.info/?l=linux-kernel&m=123213941914274&w=4

Quoted 7 lines

Well the hardest for this, was to reproduce the error.(never did). I guess it can be taken of if there was a patch for the fix, and the patch made it into the main tree. but if this is a firefox issue, then a report needs to be filed with them. so overall my answer is "I don't know."
:^)

the error is spurious because gcc randomly stomps on user register state so the effects can be rather random and unpredictable.

The fix:

552b8aa: Revert "x86: signal: change type of paramter for sys_rt_sigreturn()"

was included in -rc2 already, so please test latest -git whether you can see any such troubles.

Ingo

#53 lore raw

Re: [Bug #12505] 2.6.29-rc1 Firefox crashing on page load

Quoted 34 lines · Ingo Molnar
Quoted 20 lines · Justin Mattock
Quoted 8 lines · Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 10 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12505
Subject : 2.6.29-rc1 Firefox crashing on page load
Submitter : Justin Madru <[email protected]>
Date : 2009-01-16 20:56 (20 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=4217458dafaa57d8e26a46f5d05ab8c53cf64191
References : http://marc.info/?l=linux-kernel&m=123213941914274&w=4

Quoted 7 lines

Well the hardest for this, was to reproduce the error.(never did). I guess it can be taken of if there was a patch for the fix, and the patch made it into the main tree. but if this is a firefox issue, then a report needs to be filed with them. so overall my answer is "I don't know."
:^)

the error is spurious because gcc randomly stomps on user register state so the effects can be rather random and unpredictable.

Quoted 9 lines

The fix:

552b8aa: Revert "x86: signal: change type of paramter for sys_rt_sigreturn()"

was included in -rc2 already, so please test latest -git whether you can see any such troubles.

Ingo

Sure I'll test it.
I'm just in the process of compiling libc.
which looks like I'll be here for a while.

Signature · 1 line
Justin P. Mattock
#54 lore raw

Re: [Bug #12505] 2.6.29-rc1 Firefox crashing on page load

Quoted 42 lines · Justin Mattock
Quoted 41 lines · Ingo Molnar
Quoted 20 lines · Justin Mattock
Quoted 8 lines · Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 10 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12505
Subject : 2.6.29-rc1 Firefox crashing on page load
Submitter : Justin Madru <[email protected]>
Date : 2009-01-16 20:56 (20 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=4217458dafaa57d8e26a46f5d05ab8c53cf64191
References : http://marc.info/?l=linux-kernel&m=123213941914274&w=4

Quoted 7 lines

Well the hardest for this, was to reproduce the error.(never did). I guess it can be taken of if there was a patch for the fix, and the patch made it into the main tree. but if this is a firefox issue, then a report needs to be filed with them. so overall my answer is "I don't know."
:^)

the error is spurious because gcc randomly stomps on user register state so the effects can be rather random and unpredictable.

The fix:

552b8aa: Revert "x86: signal: change type of paramter for sys_rt_sigreturn()"

was included in -rc2 already, so please test latest -git whether you can see any such troubles.

Quoted 6 lines
Quoted 2 lines

Ingo

Sure I'll test it.
I'm just in the process of compiling libc.
which looks like I'll be here for a while.

I've closed the bug as hopefully fixed.

Thanks,
Rafael

#55 lore raw

Re: [Bug #12505] 2.6.29-rc1 Firefox crashing on page load

Quoted 46 lines · Rafael J. Wysocki
Quoted 45 lines · Justin Mattock
Quoted 43 lines · Ingo Molnar
Quoted 20 lines · Justin Mattock
Quoted 8 lines · Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 10 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12505
Subject : 2.6.29-rc1 Firefox crashing on page load
Submitter : Justin Madru <[email protected]>
Date : 2009-01-16 20:56 (20 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=4217458dafaa57d8e26a46f5d05ab8c53cf64191
References : http://marc.info/?l=linux-kernel&m=123213941914274&w=4

Quoted 7 lines

Well the hardest for this, was to reproduce the error.(never did). I guess it can be taken of if there was a patch for the fix, and the patch made it into the main tree. but if this is a firefox issue, then a report needs to be filed with them. so overall my answer is "I don't know."
:^)

the error is spurious because gcc randomly stomps on user register state so the effects can be rather random and unpredictable.

The fix:

552b8aa: Revert "x86: signal: change type of paramter for sys_rt_sigreturn()"

was included in -rc2 already, so please test latest -git whether you can see any such troubles.

Ingo

Quoted 9 lines
Quoted 3 lines

Sure I'll test it.
I'm just in the process of compiling libc.
which looks like I'll be here for a while.

I've closed the bug as hopefully fixed.

Thanks,
Rafael

Cool,

Signature · 1 line
Justin P. Mattock
#56 lore raw

[Bug #12510] 2.6.29-rc2 dies on startup

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12510
Subject : 2.6.29-rc2 dies on startup
Submitter : Ferenc Wagner <[email protected]>
Date : 2009-01-19 12:53 (17 days old)
References : http://marc.info/?l=linux-kernel&m=123236969321703&w=4

#57 lore raw

[Bug #12508] "powerpc/pci: Reserve legacy regions on PCI" broke my G3

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12508
Subject : "powerpc/pci: Reserve legacy regions on PCI" broke my G3
Submitter : Mikael Pettersson <[email protected]>
Date : 2009-01-17 22:46 (19 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=c1f343028d35ba4e88cd4a3c44e0d8b8a84264ee
References : http://marc.info/?l=linux-kernel&m=123223247325452&w=4

#58 lore raw

Re: [Bug #12508] "powerpc/pci: Reserve legacy regions on PCI" broke my G3

Quoted 6 lines · Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Can you check if it's still broken ?

59b608c2c33feacc8be281ec3ba9ca2a3a5cb9a7 might have fixed it...

Cheers,
Ben.

Quoted 9 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12508
Subject : "powerpc/pci: Reserve legacy regions on PCI" broke my G3
Submitter : Mikael Pettersson <[email protected]>
Date : 2009-01-17 22:46 (19 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=c1f343028d35ba4e88cd4a3c44e0d8b8a84264ee
References : http://marc.info/?l=linux-kernel&m=123223247325452&w=4

#59 lore raw

Re: [Bug #12508] "powerpc/pci: Reserve legacy regions on PCI" broke my G3

Quoted 8 lines · Benjamin Herrenschmidt
Quoted 6 lines · Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 6 lines

Can you check if it's still broken ?

59b608c2c33feacc8be281ec3ba9ca2a3a5cb9a7 might have fixed it...

Cheers,
Ben.

Confirmed. Adding 59b608c2c33feacc8be281ec3ba9ca2a3a5cb9a7 to 2.6.29-rc3 makes it boot fine on my G3.

Thanks,

/Mikael

Quoted 7 lines
Quoted 7 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12508
Subject : "powerpc/pci: Reserve legacy regions on PCI" broke my G3
Submitter : Mikael Pettersson <[email protected]>
Date : 2009-01-17 22:46 (19 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=c1f343028d35ba4e88cd4a3c44e0d8b8a84264ee
References : http://marc.info/?l=linux-kernel&m=123223247325452&w=4

#60 lore raw

[Bug #12506] Undefined symbols when CONFIG_MFD_PCF50633 is enabled

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12506
Subject : Undefined symbols when CONFIG_MFD_PCF50633 is enabled
Submitter : Ozan Çağlayan <[email protected]>
Date : 2009-01-17 10:56 (19 days old)
References : http://marc.info/?l=linux-kernel&m=123218991104592&w=4

#61 lore raw

[Bug #12509] lockdep report. fb_mmap vs sys_mmap2

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12509
Subject : lockdep report. fb_mmap vs sys_mmap2
Submitter : Dave Jones <[email protected]>
Date : 2009-01-17 18:19 (19 days old)
References : http://lkml.org/lkml/2009/1/17/176

Patch : http://marc.info/?l=linux-kernel&m=123231032115956&w=4

#62 lore raw

[Bug #12538] xfs_fsr fails on 2.6.29-rc kernels

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12538
Subject : xfs_fsr fails on 2.6.29-rc kernels
Submitter : Paul Martin <[email protected]>
Date : 2009-01-25 13:02 (11 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=743bb4650da9e2595d6cedd01c680b5b9398c74a

Patch : http://bugzilla.kernel.org/attachment.cgi?id=19986&action=view

#63 lore raw

[Bug #12518] BUG: using smp_processor_id() in preemptible [00000000] code: dellWirelessCtl/...

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12518
Subject : BUG: using smp_processor_id() in preemptible [00000000] code: dellWirelessCtl/...
Submitter : Alex Riesen <[email protected]>
Date : 2009-01-21 14:21 (15 days old)

#64 lore raw

Re: [Bug #12518] BUG: using smp_processor_id() in preemptible [00000000] code: dellWirelessCtl/...

2009/2/4 Rafael J. Wysocki <[email protected]>:

Quoted 7 lines

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

I confirm the problem fixed with the patch in the bugs discussion.

Quoted 6 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12518
Subject : BUG: using smp_processor_id() in preemptible [00000000] code: dellWirelessCtl/...
Submitter : Alex Riesen <[email protected]>
Date : 2009-01-21 14:21 (15 days old)

#65 lore raw

[Bug #12511] WARNING: at drivers/dma/dmaengine.c:352

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12511
Subject : WARNING: at drivers/dma/dmaengine.c:352
Submitter : Ingo Molnar <[email protected]>
Date : 2009-01-19 21:31 (17 days old)
References : http://marc.info/?l=linux-kernel&m=123240070614443&w=4

#66 lore raw

Re: [Bug #12511] WARNING: at drivers/dma/dmaengine.c:352

Quoted 7 lines · Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

This should be resolved as of commit 83436a05 "dmaengine: kill some dubious WARN_ONCEs"

Thanks,
Dan

#67 lore raw

Re: [Bug #12511] WARNING: at drivers/dma/dmaengine.c:352

Quoted 11 lines · Dan Williams
Quoted 7 lines · Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

This should be resolved as of commit 83436a05 "dmaengine: kill some dubious WARN_ONCEs"

Thanks, closed.

Rafael

#68 lore raw

[Bug #12551] end_request: I/O error, dev cciss/c0d0, sector 87435720

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12551
Subject : end_request: I/O error, dev cciss/c0d0, sector 87435720
Submitter : Ralf Hildebrandt <[email protected]>
Date : 2009-01-27 06:51 (9 days old)

#69 lore raw

[Bug #12571] Suspend-resume on Dell Latitude D410 newly broken in 2.6.29-rc*

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12571
Subject : Suspend-resume on Dell Latitude D410 newly broken in 2.6.29-rc*
Submitter : Ross Boswell <[email protected]>
Date : 2009-01-29 03:46 (7 days old)

#70 lore raw

[Bug #12574] possible circular locking dependency detected

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12574
Subject : possible circular locking dependency detected
Submitter : Michael S. Tsirkin <[email protected]>
Date : 2009-01-29 11:35 (7 days old)

#71 lore raw

[Bug #12591] NULL pointer dereference in blk_queue_io_stat

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12591
Subject : NULL pointer dereference in blk_queue_io_stat
Submitter : Petr Vandrovec <[email protected]>
Date : 2009-01-31 23:58 (5 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=bc58ba9468d94d62c56ab9b47173583ec140b165

#72 lore raw

Re: [Bug #12591] NULL pointer dereference in blk_queue_io_stat

Quoted 8 lines · Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 6 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12591
Subject : NULL pointer dereference in blk_queue_io_stat
Submitter : Petr Vandrovec <[email protected]>
Date : 2009-01-31 23:58 (5 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=bc58ba9468d94d62c56ab9b47173583ec140b165

This one is already fixed in -git by commit
fb8ec18c316d869271137c97320dbfd2def56569.

Signature · 1 line
Jens Axboe
#73 lore raw

Re: [Bug #12591] NULL pointer dereference in blk_queue_io_stat

Quoted 16 lines · Jens Axboe
Quoted 8 lines · Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 6 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12591
Subject : NULL pointer dereference in blk_queue_io_stat
Submitter : Petr Vandrovec <[email protected]>
Date : 2009-01-31 23:58 (5 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=bc58ba9468d94d62c56ab9b47173583ec140b165

Quoted 2 lines

This one is already fixed in -git by commit
fb8ec18c316d869271137c97320dbfd2def56569.

Thanks, closed.

Rafael

#74 lore raw

[Bug #12598] tg3 dead after resume

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12598
Subject : tg3 dead after resume
Submitter : Parag Warudkar <[email protected]>
Date : 2009-01-29 0:14 (7 days old)
References : http://marc.info/?l=linux-kernel&m=123318968501524&w=4

Rafael J. Wysocki <[email protected]>
Patch : http://marc.info/?l=linux-kernel&m=123365342301877&w=4

#75 lore raw

Re: [Bug #12598] tg3 dead after resume

Quoted 8 lines · Rafael J. Wysocki

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12598
Subject : tg3 dead after resume
Submitter : Parag Warudkar <[email protected]>
Date : 2009-01-29 0:14 (7 days old)
References : http://marc.info/?l=linux-kernel&m=123318968501524&w=4

Rafael J. Wysocki <[email protected]>
Patch : http://marc.info/?l=linux-kernel&m=123365342301877&w=4

OK to close - haven't checked if the patch made it to Linus' tree, assuming yes.

Parag

#76 lore raw

Re: [Bug #12598] tg3 dead after resume

Quoted 10 lines · Parag Warudkar
Quoted 8 lines · Rafael J. Wysocki

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12598
Subject : tg3 dead after resume
Submitter : Parag Warudkar <[email protected]>
Date : 2009-01-29 0:14 (7 days old)
References : http://marc.info/?l=linux-kernel&m=123318968501524&w=4

Rafael J. Wysocki <[email protected]>
Patch : http://marc.info/?l=linux-kernel&m=123365342301877&w=4

Quoted 2 lines

OK to close - haven't checked if the patch made it to Linus' tree, assuming yes.

It's on its way. Just applied it to my for-linus branch.

Signature · 1 line
Jesse Barnes, Intel Open Source Technology Center
#77 lore raw

[Bug #12599] dri /dev node disappeared with 2.6.29-rc1

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12599
Subject : dri /dev node disappeared with 2.6.29-rc1
Submitter : Norbert Preining <[email protected]>
Date : 2009-01-15 13:42 (21 days old)
References : http://marc.info/?l=linux-kernel&m=123202701026647&w=4

#78 lore raw

[Bug #12601] virt-manager broken on 2.6.29-rc2

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12601
Subject : virt-manager broken on 2.6.29-rc2
Submitter : Stephen Hemminger <[email protected]>
Date : 2009-01-29 5:21 (7 days old)
References : http://marc.info/?l=linux-kernel&m=123320655819589&w=4

#79 lore raw

Re: [Bug #12601] virt-manager broken on 2.6.29-rc2

Quoted 8 lines · On Wed, 4 Feb 2009 11:24:04 +0100 (CET) "Rafael J. Wysocki"…

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 7 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12601
Subject : virt-manager broken on 2.6.29-rc2
Submitter : Stephen Hemminger <[email protected]>
Date : 2009-01-29 5:21 (7 days old)
References : http://marc.info/?l=linux-kernel&m=123320655819589&w=4

This was in linux-next and already fixed.

#80 lore raw

Re: [Bug #12601] virt-manager broken on 2.6.29-rc2

Quoted 19 lines · Stephen Hemminger
Quoted 8 lines · On Wed, 4 Feb 2009 11:24:04 +0100 (CET) "Rafael J. Wysocki"…

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 7 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12601
Subject : virt-manager broken on 2.6.29-rc2
Submitter : Stephen Hemminger <[email protected]>
Date : 2009-01-29 5:21 (7 days old)
References : http://marc.info/?l=linux-kernel&m=123320655819589&w=4

Quoted 1 line

This was in linux-next and already fixed.

Thanks, closed.

Rafael

#81 lore raw

[Bug #12600] i915 lockdep warning

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12600
Subject : i915 lockdep warning
Submitter : Brandeburg, Jesse <[email protected]>
Date : 2009-01-13 23:17 (23 days old)
References : http://marc.info/?l=linux-kernel&m=123188898423532&w=4

#82 lore raw

[Bug #12602] CRED changes causing setuid failures

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12602
Subject : CRED changes causing setuid failures
Submitter : David Smith <[email protected]>
Date : 2009-01-29 21:35 (7 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=d84f4f992cbd76e8f39c488cf0c5d123843923b1
References : http://marc.info/?l=linux-kernel&m=123326501813668&w=4

#83 lore raw

Re: [Bug #12602] CRED changes causing setuid failures

Quoted 7 lines · Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 6 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12602
Subject : CRED changes causing setuid failures
Submitter : David Smith <[email protected]>
Date : 2009-01-29 21:35 (7 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=d84f4f992cbd76e8f39c488cf0c5d123843923b1
References : http://marc.info/?l=linux-kernel&m=123326501813668&w=4

I've verified this still exists in 2.6.29-rc3.

(git describe says I'm at v2.6.29-rc3-495-gdcf6a79)

Signature · 6 lines
David Smith
[email protected]
Red Hat
http://www.redhat.com
256.217.0141 (direct)
256.837.0057 (fax)
#84 lore raw

Re: [Bug #12602] CRED changes causing setuid failures

Quoted 15 lines · David Smith
Quoted 7 lines · Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 6 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12602
Subject : CRED changes causing setuid failures
Submitter : David Smith <[email protected]>
Date : 2009-01-29 21:35 (7 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=d84f4f992cbd76e8f39c488cf0c5d123843923b1
References : http://marc.info/?l=linux-kernel&m=123326501813668&w=4

Quoted 3 lines

I've verified this still exists in 2.6.29-rc3.

(git describe says I'm at v2.6.29-rc3-495-gdcf6a79)

Thanks for the update.

Rafael

#85 lore raw

[Bug #12605] Suspend regression, hang after matroxfb

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12605
Subject : Suspend regression, hang after matroxfb
Submitter : Eric Sesterhenn <[email protected]>
Date : 2009-01-29 12:58 (7 days old)
References : http://thread.gmane.org/gmane.linux.power-management.general/12640

Patch : http://marc.info/?l=linux-kernel&m=123335902122029&w=4

#86 lore raw

Re: [Bug #12605] Suspend regression, hang after matroxfb

Quoted 7 lines · * Rafael J. Wysocki ([email protected]) wrote:

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 7 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12605
Subject : Suspend regression, hang after matroxfb
Submitter : Eric Sesterhenn <[email protected]>
Date : 2009-01-29 12:58 (7 days old)
References : http://thread.gmane.org/gmane.linux.power-management.general/12640

Patch : http://marc.info/?l=linux-kernel&m=123335902122029&w=4

with todays -git I still get the same error, I have seen that some of your PM patches got applied, seems the one I need wasnt one of them.

Greetings, Eric

#87 lore raw

Re: [Bug #12605] Suspend regression, hang after matroxfb

Quoted 16 lines · Eric Sesterhenn
Quoted 7 lines · * Rafael J. Wysocki ([email protected]) wrote:

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 7 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12605
Subject : Suspend regression, hang after matroxfb
Submitter : Eric Sesterhenn <[email protected]>
Date : 2009-01-29 12:58 (7 days old)
References : http://thread.gmane.org/gmane.linux.power-management.general/12640

Patch : http://marc.info/?l=linux-kernel&m=123335902122029&w=4

Quoted 2 lines

with todays -git I still get the same error, I have seen that some of your PM patches got applied, seems the one I need wasnt one of them.

No, it's not been merged yet.

Thanks,
Rafael

#88 lore raw

Re: [Bug #12605] Suspend regression, hang after matroxfb

Quoted 17 lines · * Rafael J. Wysocki ([email protected]) wrote:
Quoted 16 lines · Eric Sesterhenn
Quoted 14 lines · * Rafael J. Wysocki ([email protected]) wrote:

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12605
Subject : Suspend regression, hang after matroxfb
Submitter : Eric Sesterhenn <[email protected]>
Date : 2009-01-29 12:58 (7 days old)
References : http://thread.gmane.org/gmane.linux.power-management.general/12640

Patch : http://marc.info/?l=linux-kernel&m=123335902122029&w=4

Quoted 4 lines
Quoted 2 lines

with todays -git I still get the same error, I have seen that some of your PM patches got applied, seems the one I need wasnt one of them.

No, it's not been merged yet.

This is fixed with todays -git. Thanks
A CC on the newer patches would have been nice, since I read the list only via web archives, but i saw them floating by.

Greetings, Eric

#89 lore raw

[Bug #12604] Commit 31a12666d8f0c22235297e1c1575f82061480029 slows down Berkeley DB

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12604
Subject : Commit 31a12666d8f0c22235297e1c1575f82061480029 slows down Berkeley DB
Submitter : Jan Kara <[email protected]>
Date : 2009-01-30 1:23 (6 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=31a12666d8f0c22235297e1c1575f82061480029
References : http://marc.info/?l=linux-kernel&m=123327862901800&w=4

#90 lore raw

[Bug #12606] fb_mmap: circular locking dependency on hibernation

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12606
Subject : fb_mmap: circular locking dependency on hibernation
Submitter : Andrey Borzenkov <arvidjaar-JGs/[email protected]>
Date : 2009-01-27 18:37 (9 days old)
References : http://marc.info/?l=linux-kernel&m=123308162731408&w=4

#91 lore raw

Re: [Bug #12606] fb_mmap: circular locking dependency on hibernation

Quoted 8 lines · On 4 of February 2009 13:24:06 Rafael J. Wysocki wrote:

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 5 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12606
Subject : fb_mmap: circular locking dependency on hibernation
Submitter : Andrey Borzenkov <[email protected]>
Date : 2009-01-27 18:37 (9 days old)
References : http://marc.info/?l=linux-kernel&m=123308162731408&w=4

for me it is fixed with
http://marc.info/?l=linux-kernel&m=123365581406194&w=2

#92 lore raw

Re: [Bug #12606] fb_mmap: circular locking dependency on hibernation

Quoted 9 lines · Andrey Borzenkov
Quoted 8 lines · On 4 of February 2009 13:24:06 Rafael J. Wysocki wrote:

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 8 lines
Quoted 5 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12606
Subject : fb_mmap: circular locking dependency on hibernation
Submitter : Andrey Borzenkov <arvidjaar-JGs/[email protected]>
Date : 2009-01-27 18:37 (9 days old)
References : http://marc.info/?l=linux-kernel&m=123308162731408&w=4

for me it is fixed with
http://marc.info/?l=linux-kernel&m=123365581406194&w=2

Thanks, I've updated the bug entry.

Rafael

#93 lore raw

[Bug #12611] kernel BUG at kernel/cgroup.c:398!

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12611
Subject : kernel BUG at kernel/cgroup.c:398!
Submitter : Bharata B Rao <[email protected]>
Date : 2009-01-28 15:49 (8 days old)
References : http://marc.info/?l=linux-kernel&m=123315721707796&w=4

#94 lore raw

Re: [Bug #12611] kernel BUG at kernel/cgroup.c:398!

Quoted 8 lines · Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 6 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12611
Subject : kernel BUG at kernel/cgroup.c:398!
Submitter : Bharata B Rao <[email protected]>
Date : 2009-01-28 15:49 (8 days old)
References : http://marc.info/?l=linux-kernel&m=123315721707796&w=4

It has been fixed by the following commit:

http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=839ec5452ebfd5905b9c69b20ceb640903a8ea1a

#95 lore raw

Re: [Bug #12611] kernel BUG at kernel/cgroup.c:398!

Quoted 16 lines · Li Zefan
Quoted 8 lines · Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 6 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12611
Subject : kernel BUG at kernel/cgroup.c:398!
Submitter : Bharata B Rao <[email protected]>
Date : 2009-01-28 15:49 (8 days old)
References : http://marc.info/?l=linux-kernel&m=123315721707796&w=4

Quoted 3 lines

It has been fixed by the following commit:

http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=839ec5452ebfd5905b9c69b20ceb640903a8ea1a

Already closed.

Thanks,
Rafael

#96 lore raw

[Bug #12610] sync-Regression in 2.6.28.2?

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12610
Subject : sync-Regression in 2.6.28.2?
Submitter : Ralf Hildebrandt <[email protected]>
Date : 2009-01-27 9:35 (9 days old)
References : http://marc.info/?l=linux-kernel&m=123304977706620&w=4

#97 lore raw

[Bug #12609] v2.6.29-rc2 libata sff 32bit PIO regression

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12609
Subject : v2.6.29-rc2 libata sff 32bit PIO regression
Submitter : Larry Finger <[email protected]>
Date : 2009-01-23 23:52 (13 days old)
References : http://marc.info/?l=linux-kernel&m=123275478111406&w=4

		  http://marc.info/?l=linux-kernel&m=123254501314058&w=4

Hugh Dickins <[email protected]>
Patch : http://marc.info/?l=linux-kernel&m=123254501314058&w=2

#98 lore raw

Re: [Bug #12609] v2.6.29-rc2 libata sff 32bit PIO regression

Quoted 10 lines · Rafael J. Wysocki

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12609
Subject : v2.6.29-rc2 libata sff 32bit PIO regression
Submitter : Larry Finger <[email protected]>
Date : 2009-01-23 23:52 (13 days old)
References : http://marc.info/?l=linux-kernel&m=123275478111406&w=4

		  http://marc.info/?l=linux-kernel&m=123254501314058&w=4

Hugh Dickins <[email protected]>
Patch : http://marc.info/?l=linux-kernel&m=123254501314058&w=2

Yes, this does still need to be listed. My initial patch for it was not enough, but I think the three patches necessary have all now been posted (though Sergei may rewrite mine). I'm expecting Alan to gather them together and send them in due course, but I believe he's occupied with other stuff just at the moment.

Hugh

#99 lore raw

Re: [Bug #12609] v2.6.29-rc2 libata sff 32bit PIO regression

Quoted 12 lines · Hugh Dickins
Quoted 10 lines · Rafael J. Wysocki

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12609
Subject : v2.6.29-rc2 libata sff 32bit PIO regression
Submitter : Larry Finger <[email protected]>
Date : 2009-01-23 23:52 (13 days old)
References : http://marc.info/?l=linux-kernel&m=123275478111406&w=4

		  http://marc.info/?l=linux-kernel&m=123254501314058&w=4

Hugh Dickins <[email protected]>
Patch : http://marc.info/?l=linux-kernel&m=123254501314058&w=2

Quoted 5 lines

Yes, this does still need to be listed. My initial patch for it was not enough, but I think the three patches necessary have all now been posted (though Sergei may rewrite mine). I'm expecting Alan to gather them together and send them in due course, but I believe he's occupied with other stuff just at the moment.

Thanks for the update.

Rafael

#100 lore raw

Re: [Bug #12609] v2.6.29-rc2 libata sff 32bit PIO regression

Quoted 11 lines · Rafael J. Wysocki
Quoted 10 lines · Hugh Dickins
Quoted 9 lines · Rafael J. Wysocki

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12609
Subject : v2.6.29-rc2 libata sff 32bit PIO regression
Submitter : Larry Finger <[email protected]>
Date : 2009-01-23 23:52 (13 days old)
References : http://marc.info/?l=linux-kernel&m=123275478111406&w=4

		  http://marc.info/?l=linux-kernel&m=123254501314058&w=4

Hugh Dickins <[email protected]>
Patch : http://marc.info/?l=linux-kernel&m=123254501314058&w=2

Quoted 5 lines
Quoted 5 lines

Yes, this does still need to be listed. My initial patch for it was not enough, but I think the three patches necessary have all now been posted (though Sergei may rewrite mine). I'm expecting Alan to gather them together and send them in due course, but I believe he's occupied with other stuff just at the moment.

I still have the problem as of 2.6.29-rc3-00634-g9be260a

Larry

#101 lore raw

Re: [Bug #12609] v2.6.29-rc2 libata sff 32bit PIO regression

Quoted 12 lines · Larry Finger
Quoted 11 lines · Rafael J. Wysocki
Quoted 10 lines · Hugh Dickins
Quoted 9 lines · Rafael J. Wysocki

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12609
Subject : v2.6.29-rc2 libata sff 32bit PIO regression
Submitter : Larry Finger <[email protected]>
Date : 2009-01-23 23:52 (13 days old)
References : http://marc.info/?l=linux-kernel&m=123275478111406&w=4

		  http://marc.info/?l=linux-kernel&m=123254501314058&w=4

Hugh Dickins <[email protected]>
Patch : http://marc.info/?l=linux-kernel&m=123254501314058&w=2

Quoted 7 lines
Quoted 5 lines
Quoted 5 lines

Yes, this does still need to be listed. My initial patch for it was not enough, but I think the three patches necessary have all now been posted (though Sergei may rewrite mine). I'm expecting Alan to gather them together and send them in due course, but I believe he's occupied with other stuff just at the moment.

I still have the problem as of 2.6.29-rc3-00634-g9be260a

Thanks for the updates.

Rafael

#102 lore raw

Re: [Bug #12609] v2.6.29-rc2 libata sff 32bit PIO regression

Quoted 13 lines · Rafael J. Wysocki
Quoted 12 lines · Larry Finger
Quoted 11 lines · Rafael J. Wysocki
Quoted 10 lines · Hugh Dickins
Quoted 9 lines · Rafael J. Wysocki

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12609
Subject : v2.6.29-rc2 libata sff 32bit PIO regression
Submitter : Larry Finger <[email protected]>
Date : 2009-01-23 23:52 (13 days old)
References : http://marc.info/?l=linux-kernel&m=123275478111406&w=4

		  http://marc.info/?l=linux-kernel&m=123254501314058&w=4

Hugh Dickins <[email protected]>
Patch : http://marc.info/?l=linux-kernel&m=123254501314058&w=2

Quoted 7 lines
Quoted 6 lines
Quoted 5 lines
Quoted 5 lines

Yes, this does still need to be listed. My initial patch for it was not enough, but I think the three patches necessary have all now been posted (though Sergei may rewrite mine). I'm expecting Alan to gather them together and send them in due course, but I believe he's occupied with other stuff just at the moment.

I still have the problem as of 2.6.29-rc3-00634-g9be260a

The problem was fixed as of 2.6.29-rc5-00168-gba95fd4 with commit d1b3525. At least for my system, the regression is cleared.

Larry

#103 lore raw

Re: [Bug #12609] v2.6.29-rc2 libata sff 32bit PIO regression

Quoted 19 lines · Larry Finger
Quoted 18 lines · Rafael J. Wysocki
Quoted 17 lines · Larry Finger
Quoted 16 lines · Rafael J. Wysocki
Quoted 15 lines · Hugh Dickins
Quoted 9 lines · Rafael J. Wysocki

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12609
Subject : v2.6.29-rc2 libata sff 32bit PIO regression
Submitter : Larry Finger <[email protected]>
Date : 2009-01-23 23:52 (13 days old)
References : http://marc.info/?l=linux-kernel&m=123275478111406&w=4

		  http://marc.info/?l=linux-kernel&m=123254501314058&w=4

Hugh Dickins <[email protected]>
Patch : http://marc.info/?l=linux-kernel&m=123254501314058&w=2

Yes, this does still need to be listed. My initial patch for it was not enough, but I think the three patches necessary have all now been posted (though Sergei may rewrite mine). I'm expecting Alan to gather them together and send them in due course, but I believe he's occupied with other stuff just at the moment.

Quoted 5 lines
Quoted 2 lines
Quoted 1 line

I still have the problem as of 2.6.29-rc3-00634-g9be260a

The problem was fixed as of 2.6.29-rc5-00168-gba95fd4 with commit d1b3525. At least for my system, the regression is cleared.

It's already been closed, thanks for the confirmation.

Rafael

#104 lore raw

[Bug #12608] 2.6.29-rc powerpc G5 Xorg legacy_mem regression

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12608
Subject : 2.6.29-rc powerpc G5 Xorg legacy_mem regression
Submitter : Hugh Dickins <[email protected]>
Date : 2009-01-21 21:12 (15 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=d3a54014e2a94bd37b7dee5e76e03f7bc4fab49a
References : http://marc.info/?l=linux-kernel&m=123257250431870&w=4

#105 lore raw

Re: [Bug #12608] 2.6.29-rc powerpc G5 Xorg legacy_mem regression

Quoted 6 lines · Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

I still don't know what's the best way to handle that one... the bug is in X and I don't see a way to work around it without removing support for legacy memory access from the kernel :-( Or doing it in a way that doesn't allow userspace to differenciate between the kernel not supporting it vs. the HW not supporting it, causing X to fallback to even more broken crap.

I'll try to find out the extent of the X problem and whether that's fixable in a way that can hit distros.

Quoted 9 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12608
Subject : 2.6.29-rc powerpc G5 Xorg legacy_mem regression
Submitter : Hugh Dickins <[email protected]>
Date : 2009-01-21 21:12 (15 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=d3a54014e2a94bd37b7dee5e76e03f7bc4fab49a
References : http://marc.info/?l=linux-kernel&m=123257250431870&w=4

#106 lore raw

Re: [Bug #12608] 2.6.29-rc powerpc G5 Xorg legacy_mem regression

Quoted 8 lines · Benjamin Herrenschmidt
Quoted 6 lines · Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 6 lines

I still don't know what's the best way to handle that one... the bug is in X and I don't see a way to work around it without removing support for legacy memory access from the kernel :-( Or doing it in a way that doesn't allow userspace to differenciate between the kernel not supporting it vs. the HW not supporting it, causing X to fallback to even more broken crap.

Is it a really a bug in X, or a misunderstanding between X and the kernel as to what existence of the legacy_mem file implies?

I may have got this quite wrong, but to me it appears that X assumes that existence of the legacy_mem file implies that it will be useful; whereas the kernel thinks it can make the legacy_mem file available, even if it cannot be used for mmapping mem - which is its sole purpose?

What if pci_create_legacy_files() were to call some new verification routine, and only create the legacy_mem file if it would be usable? (But perhaps that cannot be known at the time it needs to be created.)

Hugh

Quoted 12 lines

I'll try to find out the extent of the X problem and whether that's fixable in a way that can hit distros.

Quoted 8 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12608
Subject : 2.6.29-rc powerpc G5 Xorg legacy_mem regression
Submitter : Hugh Dickins <[email protected]>
Date : 2009-01-21 21:12 (15 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=d3a54014e2a94bd37b7dee5e76e03f7bc4fab49a
References : http://marc.info/?l=linux-kernel&m=123257250431870&w=4

#107 lore raw

Re: [Bug #12608] 2.6.29-rc powerpc G5 Xorg legacy_mem regression

Quoted 4 lines

Is it a really a bug in X, or a misunderstanding between X and the kernel as to what existence of the legacy_mem file implies?

Quoted 8 lines

I may have got this quite wrong, but to me it appears that X assumes that existence of the legacy_mem file implies that it will be useful; whereas the kernel thinks it can make the legacy_mem file available, even if it cannot be used for mmapping mem - which is its sole purpose?

What if pci_create_legacy_files() were to call some new verification routine, and only create the legacy_mem file if it would be usable? (But perhaps that cannot be known at the time it needs to be created.)

Well, first X should certainly not -fail- to launch if it fails to map legacy memory, which is generally not useful anyway. That's where the bug is. Jesse, did you have a chance to fix that yet or should I give it a go ?

The second problem is that if I just don't expose the legacy_mem file, then X has no way to know whether the kernel doesn't support the interface or whether the HW doesn't support legacy memory access. So X will fallback to whacking /dev/mem which is even more bogus. At least that's what I remember from last I looked at that part of X code.

It should be a trivial fix on X side tho.

Cheers,
Ben.

#108 lore raw

Re: [Bug #12608] 2.6.29-rc powerpc G5 Xorg legacy_mem regression

Quoted 2 lines · Benjamin Herrenschmidt

It should be a trivial fix on X side tho.

It will be unusual to require users of a new kernel to install a new X.

Hugh

maintainer
#109 lore raw

Re: [Bug #12608] 2.6.29-rc powerpc G5 Xorg legacy_mem regression

In reply to Benjamin Herrenschmidt
Quoted 18 lines · Benjamin Herrenschmidt
Quoted 4 lines

Is it a really a bug in X, or a misunderstanding between X and the kernel as to what existence of the legacy_mem file implies?

Quoted 8 lines

I may have got this quite wrong, but to me it appears that X assumes that existence of the legacy_mem file implies that it will be useful; whereas the kernel thinks it can make the legacy_mem file available, even if it cannot be used for mmapping mem - which is its sole purpose?

What if pci_create_legacy_files() were to call some new verification routine, and only create the legacy_mem file if it would be usable? (But perhaps that cannot be known at the time it needs to be created.)

Well, first X should certainly not -fail- to launch if it fails to map legacy memory, which is generally not useful anyway. That's where the bug is. Jesse, did you have a chance to fix that yet or should I give it a go ?

Quoted 7 lines

The second problem is that if I just don't expose the legacy_mem file, then X has no way to know whether the kernel doesn't support the interface or whether the HW doesn't support legacy memory access. So X will fallback to whacking /dev/mem which is even more bogus. At least that's what I remember from last I looked at that part of X code.

It should be a trivial fix on X side tho.

I think the correct answer is the ugly one, try again.

Add a new legacy_mem interface that works cleanly, update X to use it, leave the old
broken one broken as it for older X to use.

Dave.

Quoted 11 lines

Cheers,
Ben.

--
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/

#110 lore raw

Re: [Bug #12608] 2.6.29-rc powerpc G5 Xorg legacy_mem regression

Quoted 2 lines

I think the correct answer is the ugly one, try again.

Well... it's a fairly recent damage in X which is why I was somewhat hoping that it wasn't too widespread it couldn't be contained before the kernel got butchered :-)

But I'll just stick anonymous memory in there instead, it should be trivial enough.

Quoted 3 lines

Add a new legacy_mem interface that works cleanly, update X to use it, leave the old
broken one broken as it for older X to use.

:-)

Ben.

#111 lore raw

Re: [Bug #12608] 2.6.29-rc powerpc G5 Xorg legacy_mem regression

In reply to Benjamin Herrenschmidt
Quoted 8 lines · Benjamin Herrenschmidt
Quoted 8 lines

Is it a really a bug in X, or a misunderstanding between X and the kernel as to what existence of the legacy_mem file implies?

I may have got this quite wrong, but to me it appears that X assumes that existence of the legacy_mem file implies that it will be useful; whereas the kernel thinks it can make the legacy_mem file available, even if it cannot be used for mmapping mem - which is its sole purpose?

Quoted 8 lines
Quoted 3 lines

What if pci_create_legacy_files() were to call some new verification routine, and only create the legacy_mem file if it would be usable? (But perhaps that cannot be known at the time it needs to be created.)

Well, first X should certainly not -fail- to launch if it fails to map legacy memory, which is generally not useful anyway. That's where the bug is. Jesse, did you have a chance to fix that yet or should I give it a go ?

No, sorry, but I just took a look and as long as the various callers can handle it (haven't checked), this patch would work.

Quoted 7 lines

The second problem is that if I just don't expose the legacy_mem file, then X has no way to know whether the kernel doesn't support the interface or whether the HW doesn't support legacy memory access. So X will fallback to whacking /dev/mem which is even more bogus. At least that's what I remember from last I looked at that part of X code.

It should be a trivial fix on X side tho.

One option there would be to provide the file but just use anonymous memory to back it. X will happily think it's messing with legacy VGA memory, but it shouldn't matter that it's not actually affecting hw.

(quoted diff) +0 -0

b/hw/xfree86/os-support/bus/li
index 263fd8f..5d2da32 100644

hw/xfree86/os-support/bus/linuxPci.c +0 -0
@@ -484,8 +484,9 @@ xf86MapDomainMemory(int ScreenNum, int Flags, struct 

pci_dev

     if (fd >= 0)
        close(fd);
     if (addr == NULL || addr == MAP_FAILED) {
-       perror("mmap failure");
-       FatalError("xf86MapDomainMem():  mmap() failure\n");
+       xf86Msg(X_WARNING, "xf86MapDomainMem():  mmap() failure: %s\n",
+               strerror(errno));
+       return NULL;
     }
     return addr;

}

Signature · 1 line
Jesse Barnes, Intel Open Source Technology Center
#112 lore raw

Re: [Bug #12608] 2.6.29-rc powerpc G5 Xorg legacy_mem regression

Quoted 10 lines

No, sorry, but I just took a look and as long as the various callers can handle it (haven't checked), this patch would work.

Quoted 6 lines

The second problem is that if I just don't expose the legacy_mem file, then X has no way to know whether the kernel doesn't support the interface or whether the HW doesn't support legacy memory access. So X will fallback to whacking /dev/mem which is even more bogus. At least that's what I remember from last I looked at that part of X code.

Quoted 5 lines
Quoted 1 line

It should be a trivial fix on X side tho.

One option there would be to provide the file but just use anonymous memory to back it. X will happily think it's messing with legacy VGA memory, but it shouldn't matter that it's not actually affecting hw.

That might be the best approach ... I'll have a look.

Ben.

Quoted 20 lines
hw/xfree86/os-support/bus/linuxPci.c +0 -0
index 263fd8f..5d2da32 100644
@@ -484,8 +484,9 @@ xf86MapDomainMemory(int ScreenNum, int Flags, struct 

pci_dev

     if (fd >= 0)
        close(fd);
     if (addr == NULL || addr == MAP_FAILED) {
hw/xfree86/os-support/bus/linuxPci.c +3 -2
-       perror("mmap failure");
-       FatalError("xf86MapDomainMem():  mmap() failure\n");
+       xf86Msg(X_WARNING, "xf86MapDomainMem():  mmap() failure: %s\n",
+               strerror(errno));
+       return NULL;
     }
     return addr;
 }
#113 lore raw

Re: [Bug #12608] 2.6.29-rc powerpc G5 Xorg legacy_mem regression

In reply to Jesse Barnes
Quoted 3 lines · Jesse Barnes

One option there would be to provide the file but just use anonymous memory to back it. X will happily think it's messing with legacy VGA memory, but it shouldn't matter that it's not actually affecting hw.

What about this (untested) patch ?

powerpc/pci: mmap anonymous memory when legacy_mem doesn't exist

The new legacy_mem file in sysfs is causing problems with X on machines that don't support legacy memory access. The way I initially implemented it, we would fail with -ENXIO when trying to mmap it, thus exposing to X that we do support the API but there is no legacy memory.

Unfortunately, X poor error handling is causing it to fail to start when it gets this error.

This implements a workaround hack that instead maps anonymous memory instead (using shmem if VM_SHARED is set, just like /dev/zero does).

---

arch/powerpc/kernel/pci-common.c +15 -2
Index: linux-work/arch/powerpc/kernel/pci-common.c
===================================================================
@@ -561,8 +561,21 @@ int pci_mmap_legacy_page_range(struct pc
 		 (unsigned long long)(offset + size - 1));
 
 	if (mmap_state == pci_mmap_mem) {
-		if ((offset + size) > hose->isa_mem_size)
-			return -ENXIO;
+		/* Hack alert !
+		 *
+		 * Because X is lame and can fail starting if it gets an error trying
+		 * to mmap legacy_mem (instead of just moving on without legacy memory
+		 * access) we fake it here by giving it anonymous memory, effectively
+		 * behaving just like /dev/zero
+		 */
+		if ((offset + size) > hose->isa_mem_size) {
+			printk(KERN_DEBUG
+			       "Process %s (pid:%d) mapped non-existing PCI legacy memory for 0%04x:%02x\n",
+			       current->comm, current->pid, pci_domain_nr(bus), bus->number);
+			if (vma->vm_flags & VM_SHARED)
+				return shmem_zero_setup(vma);
+			return 0;
+		}
 		offset += hose->isa_mem_phys;
 	} else {
 		unsigned long io_offset = (unsigned long)hose->io_base_virt - _IO_BASE;
#114 lore raw

Re: [Bug #12608] 2.6.29-rc powerpc G5 Xorg legacy_mem regression

Quoted 15 lines · Benjamin Herrenschmidt
Quoted 3 lines · Jesse Barnes

One option there would be to provide the file but just use anonymous memory to back it. X will happily think it's messing with legacy VGA memory, but it shouldn't matter that it's not actually affecting hw.

What about this (untested) patch ?

powerpc/pci: mmap anonymous memory when legacy_mem doesn't exist

The new legacy_mem file in sysfs is causing problems with X on machines that don't support legacy memory access. The way I initially implemented it, we would fail with -ENXIO when trying to mmap it, thus exposing to X that we do support the API but there is no legacy memory.

Quoted 8 lines

Unfortunately, X poor error handling is causing it to fail to start when it gets this error.

This implements a workaround hack that instead maps anonymous memory instead (using shmem if VM_SHARED is set, just like /dev/zero does).

---

Beautiful idea, works very nicely for me. Put a dummy in the baby's mouth to stop it crying.

I still don't understand the point of this legacy_mem file at all, and why X should want to mmap it, if it works so well when it's thus divorced from reality. But I won't worry my pretty little head over it any further - please don't waste your time trying to explain it to me!

Thanks,
Hugh

Quoted 29 lines

Index: linux-work/arch/powerpc/kernel/pci-common.c
===================================================================

arch/powerpc/kernel/pci-common.c +15 -2
@@ -561,8 +561,21 @@ int pci_mmap_legacy_page_range(struct pc
 		 (unsigned long long)(offset + size - 1));
 
 	if (mmap_state == pci_mmap_mem) {
-		if ((offset + size) > hose->isa_mem_size)
-			return -ENXIO;
+		/* Hack alert !
+		 *
+		 * Because X is lame and can fail starting if it gets an error trying
+		 * to mmap legacy_mem (instead of just moving on without legacy memory
+		 * access) we fake it here by giving it anonymous memory, effectively
+		 * behaving just like /dev/zero
+		 */
+		if ((offset + size) > hose->isa_mem_size) {
+			printk(KERN_DEBUG
+			       "Process %s (pid:%d) mapped non-existing PCI legacy memory for 0%04x:%02x\n",
+			       current->comm, current->pid, pci_domain_nr(bus), bus->number);
+			if (vma->vm_flags & VM_SHARED)
+				return shmem_zero_setup(vma);
+			return 0;
+		}
 		offset += hose->isa_mem_phys;
 	} else {
 		unsigned long io_offset = (unsigned long)hose->io_base_virt - _IO_BASE;
#115 lore raw

Re: [Bug #12608] 2.6.29-rc powerpc G5 Xorg legacy_mem regression

Quoted 24 lines · Hugh Dickins
Quoted 14 lines · Benjamin Herrenschmidt
Quoted 3 lines · Jesse Barnes

One option there would be to provide the file but just use anonymous memory to back it. X will happily think it's messing with legacy VGA memory, but it shouldn't matter that it's not actually affecting hw.

What about this (untested) patch ?

powerpc/pci: mmap anonymous memory when legacy_mem doesn't exist

The new legacy_mem file in sysfs is causing problems with X on machines that don't support legacy memory access. The way I initially implemented it, we would fail with -ENXIO when trying to mmap it, thus exposing to X that we do support the API but there is no legacy memory.

Quoted 8 lines

Unfortunately, X poor error handling is causing it to fail to start when it gets this error.

This implements a workaround hack that instead maps anonymous memory instead (using shmem if VM_SHARED is set, just like /dev/zero does).

---

Quoted 8 lines

Beautiful idea, works very nicely for me. Put a dummy in the baby's mouth to stop it crying.

I still don't understand the point of this legacy_mem file at all, and why X should want to mmap it, if it works so well when it's thus divorced from reality. But I won't worry my pretty little head over it any further - please don't waste your time trying to explain it to me!

Hugh, did you get a chance to try my X patch (w/o Ben's patch applied)? If it also works, we should apply it to X too, if only to make porting to future platforms a little easier.

Since you probably don't want to know, you can skip this paragraph as it describes the motivation behind the legacy_mem file. :) The motivation for it was to support machines with multihead video card configurations, but which could also support multiple legacy port I/O and memory spaces (e.g. on a per PCI domain basis). It allows X to run option ROMs on multiple devices simultaneously, and can also help with hardware that can't route legacy space from a given host bus to arbitrary system addresses. When we put this in initially, Ben knew there were some platforms where legacy memory simply wasn't available, but we figured that shouldn't be a problem, since X didn't really *need* that space to function (most ROMs just put splash screens there). Unfortunately in the interim, X made the lack of an mmap'able legacy_mem file into a fatal error.

In the past, doing Linux bringup on a platform with a complex I/O topology (sparc, ia64, alpha, etc.) usually meant doing X bring up as well, often by adding new interfaces or hacks to X. With the legacy_mem and io files, there's a chance that X won't have to be modified at bring up time, so in that sense the APIs are a step forward. However that also means your platform has to support the APIs in a way consistent with how X uses them, as you discovered. :)

Signature · 1 line
Jesse Barnes, Intel Open Source Technology Center
#116 lore raw

Re: [Bug #12608] 2.6.29-rc powerpc G5 Xorg legacy_mem regression

Quoted 4 lines · Jesse Barnes

Hugh, did you get a chance to try my X patch (w/o Ben's patch applied)? If it also works, we should apply it to X too, if only to make porting to future platforms a little easier.

I have tried it now: I'm sorry to say it does not work, fails with (WW) xf86MapDomainMem(): mmap() failure: No such device or address

Fatal server error:
AddScreen/ScreenInit failed for driver 0

The "mmap() failure" line is of course the one you're intentionally showing, but the AddScreen/ScreenInit one is not what you want. Presumably it demands something other than a NULL addr returned, but I haven't followed that up at all.

I have tried the obvious patch below, which reverts to the original code if the mmap fails: this works, but presumably negates much of what you intended to achieve with legacy_mem.

Easy for me to try something else, fire away.

Hugh

hw/xfree86/os-support/bus/linuxPci.c +4 -2
@@ -501,8 +501,10 @@ xf86MapDomainMemory(int ScreenNum, int F
     if (fd >= 0)
 	close(fd);
     if (addr == NULL || addr == MAP_FAILED) {
-	perror("mmap failure");
-	FatalError("xf86MapDomainMem():  mmap() failure\n");
+	xf86Msg(X_WARNING, "xf86MapDomainMem():  mmap() failure: %s\n",
+		strerror(errno));
+	return linuxMapPci(ScreenNum, Flags, dev, Base, Size,
+			   PCIIOC_MMAP_IS_MEM);
     }
     return addr;
 }
#117 lore raw

Re: [Bug #12608] 2.6.29-rc powerpc G5 Xorg legacy_mem regression

Quoted 29 lines · Hugh Dickins
Quoted 3 lines · Jesse Barnes

Hugh, did you get a chance to try my X patch (w/o Ben's patch applied)? If it also works, we should apply it to X too, if only to make porting to future platforms a little easier.

I have tried it now: I'm sorry to say it does not work, fails with (WW) xf86MapDomainMem(): mmap() failure: No such device or address

Fatal server error:
AddScreen/ScreenInit failed for driver 0

The "mmap() failure" line is of course the one you're intentionally showing, but the AddScreen/ScreenInit one is not what you want. Presumably it demands something other than a NULL addr returned, but I haven't followed that up at all.

I have tried the obvious patch below, which reverts to the original code if the mmap fails: this works, but presumably negates much of what you intended to achieve with legacy_mem.

Easy for me to try something else, fire away.

Hugh

hw/xfree86/os-support/bus/linuxPci.c +0 -0
@@ -501,8 +501,10 @@ xf86MapDomainMemory(int ScreenNum, int F
     if (fd >= 0)
 	close(fd);

hw/xfree86/os-support/bus/linuxPci.c xf86MapDomainMemory(int ScreenNum, int F · line 510

Quoted 10 lines
hw/xfree86/os-support/bus/linuxPci.c +4 -2
     if (addr == NULL || addr == MAP_FAILED) {
-	perror("mmap failure");
-	FatalError("xf86MapDomainMem():  mmap() failure\n");
+	xf86Msg(X_WARNING, "xf86MapDomainMem():  mmap() failure: %s\n",
+		strerror(errno));
+	return linuxMapPci(ScreenNum, Flags, dev, Base, Size,
+			   PCIIOC_MMAP_IS_MEM);
     }
     return addr;
 }

Yeah that should work. I always hated this code (well most of X actually); we could probably clean up the logic a little. If this works I'll send it out to the Xorg list for review.

Signature · 1 line
Jesse Barnes, Intel Open Source Technology Center
hw/xfree86/os-support/bus/linuxPci.c +11 -16
index 263fd8f..fa0fc0a 100644
@@ -471,23 +471,18 @@ xf86MapDomainMemory(int ScreenNum, int Flags, struct pci_device *dev,
     int fd = -1;
     pointer addr;
 
-    /*
-     * We use /proc/bus/pci on non-legacy addresses or if the Linux sysfs
-     * legacy_mem interface is unavailable.
-     */
-    if ((Base > 1024*1024) || ((fd = linuxOpenLegacy(dev, "legacy_mem")) < 0))
-	return linuxMapPci(ScreenNum, Flags, dev, Base, Size,
-			   PCIIOC_MMAP_IS_MEM);
-    else
-	addr = mmap(NULL, Size, PROT_READ|PROT_WRITE, MAP_SHARED, fd, Base);
-
-    if (fd >= 0)
-	close(fd);
-    if (addr == NULL || addr == MAP_FAILED) {
-	perror("mmap failure");
-	FatalError("xf86MapDomainMem():  mmap() failure\n");
+    if (Base <= 1024*1024) {
+        /* Try legacy_mem (may not be available or implemented) */
+        if ((fd = linuxOpenLegacy(dev, "legacy_mem")) < 0) {
+            addr = mmap(NULL, Size, PROT_READ|PROT_WRITE, MAP_SHARED, fd, Base);
+            close(fd);
+            if (addr && addr != MAP_FAILED)
+                return addr;
+        }
     }
-    return addr;
+
+    /* Fall back to old method if legacy_mem fails or Base >= 1M */
+    return linuxMapPci(ScreenNum, Flags, dev, Base, Size, PCIIOC_MMAP_IS_MEM);
 }
 
 /**
#118 lore raw

Re: [Bug #12608] 2.6.29-rc powerpc G5 Xorg legacy_mem regression

Quoted 4 lines · Jesse Barnes

Yeah that should work. I always hated this code (well most of X actually); we could probably clean up the logic a little. If this works I'll send it out to the Xorg list for review.

Yes, thanks for restructuring it, I was rather itching to do so.

With the one small change below which I think you'll approve, your patch works for me on the G5 - but I've not tested it on other (x86) machines which were already working.

Hugh

Quoted 20 lines
Signature · 2 lines
-- 
Jesse Barnes, Intel Open Source Technology Center
hw/xfree86/os-support/bus/linuxPci.c +0 -8
index 263fd8f..fa0fc0a 100644
@@ -471,23 +471,18 @@ xf86MapDomainMemory(int ScreenNum, int Flags, struct pci_device *dev,
     int fd = -1;
     pointer addr;
 
-    /*
-     * We use /proc/bus/pci on non-legacy addresses or if the Linux sysfs
-     * legacy_mem interface is unavailable.
-     */
-    if ((Base > 1024*1024) || ((fd = linuxOpenLegacy(dev, "legacy_mem")) < 0))
-	return linuxMapPci(ScreenNum, Flags, dev, Base, Size,
-			   PCIIOC_MMAP_IS_MEM);
-    else

hw/xfree86/os-support/bus/linuxPci.c xf86MapDomainMemory(int ScreenNum, int Flags, struct pci_device *dev, · line 476

Quoted 10 lines
hw/xfree86/os-support/bus/linuxPci.c +3 -7
-	addr = mmap(NULL, Size, PROT_READ|PROT_WRITE, MAP_SHARED, fd, Base);
-
-    if (fd >= 0)
-	close(fd);
-    if (addr == NULL || addr == MAP_FAILED) {
-	perror("mmap failure");
-	FatalError("xf86MapDomainMem():  mmap() failure\n");
+    if (Base <= 1024*1024) {
+        /* Try legacy_mem (may not be available or implemented) */
+        if ((fd = linuxOpenLegacy(dev, "legacy_mem")) < 0) {
                                                         >=
Quoted 13 lines
hw/xfree86/os-support/bus/linuxPci.c +8 -1
+            addr = mmap(NULL, Size, PROT_READ|PROT_WRITE, MAP_SHARED, fd, Base);
+            close(fd);
+            if (addr && addr != MAP_FAILED)
+                return addr;
+        }
     }
-    return addr;
+
+    /* Fall back to old method if legacy_mem fails or Base >= 1M */
+    return linuxMapPci(ScreenNum, Flags, dev, Base, Size, PCIIOC_MMAP_IS_MEM);
 }
 
 /**
#119 lore raw

Re: [Bug #12608] 2.6.29-rc powerpc G5 Xorg legacy_mem regression

Quoted 10 lines · Hugh Dickins
Quoted 3 lines · Jesse Barnes

Yeah that should work. I always hated this code (well most of X actually); we could probably clean up the logic a little. If this works I'll send it out to the Xorg list for review.

Yes, thanks for restructuring it, I was rather itching to do so.

With the one small change below which I think you'll approve, your patch works for me on the G5 - but I've not tested it on other (x86) machines which were already working.

Heh, yeah thanks. Ok I'll get it into X.

Signature · 1 line
Jesse Barnes, Intel Open Source Technology Center
#120 lore raw

Re: [Bug #12608] 2.6.29-rc powerpc G5 Xorg legacy_mem regression

In reply to Jesse Barnes
Quoted 4 lines · Jesse Barnes
hw/xfree86/os-support/bus/linuxPci.c +4 -0
+    if (Base <= 1024*1024) {
+        /* Try legacy_mem (may not be available or implemented) */
+        if ((fd = linuxOpenLegacy(dev, "legacy_mem")) < 0) {
+            addr = mmap(NULL, Size, PROT_READ|PROT_WRITE, MAP_SHARED, fd, Base);

hw/xfree86/os-support/bus/linuxPci.c xf86MapDomainMemory(int ScreenNum, int Flags, struct pci_device *dev, · line 486

Quoted 10 lines
hw/xfree86/os-support/bus/linuxPci.c +7 -1
+            close(fd);
+            if (addr && addr != MAP_FAILED)
+                return addr;
+        }
     }
-    return addr;
+
+    /* Fall back to old method if legacy_mem fails or Base >= 1M */
+    return linuxMapPci(ScreenNum, Flags, dev, Base, Size, PCIIOC_MMAP_IS_MEM);
 }

I don't like the fallback if legacy_mem exists and returns an error, that's an indication that legacy memory is -not- available and thus whatever 'fallback' X will try (supposedly using /dev/mem) will be horribly broken and will probably end up scribbling all over system memory :-)

Ben.

#121 lore raw

Re: [Bug #12608] 2.6.29-rc powerpc G5 Xorg legacy_mem regression

Quoted 16 lines · Benjamin Herrenschmidt
Quoted 13 lines · Jesse Barnes
hw/xfree86/os-support/bus/linuxPci.c +4 -0
+    if (Base <= 1024*1024) {
+        /* Try legacy_mem (may not be available or implemented) */
+        if ((fd = linuxOpenLegacy(dev, "legacy_mem")) < 0) {
+            addr = mmap(NULL, Size, PROT_READ|PROT_WRITE, MAP_SHARED,

fd, Base); + close(fd);

hw/xfree86/os-support/bus/linuxPci.c +6 -1
+            if (addr && addr != MAP_FAILED)
+                return addr;
+        }
     }
-    return addr;
+
+    /* Fall back to old method if legacy_mem fails or Base >= 1M */
+    return linuxMapPci(ScreenNum, Flags, dev, Base, Size,
Quoted 1 line

PCIIOC_MMAP_IS_MEM); }

Quoted 5 lines

I don't like the fallback if legacy_mem exists and returns an error, that's an indication that legacy memory is -not- available and thus whatever 'fallback' X will try (supposedly using /dev/mem) will be horribly broken and will probably end up scribbling all over system memory :-)

Yeah, but unless we fix all the callers (and possibly their callers), not falling back will keep X from starting...

Signature · 1 line
Jesse Barnes, Intel Open Source Technology Center
#122 lore raw

Re: [Bug #12608] 2.6.29-rc powerpc G5 Xorg legacy_mem regression

In reply to Hugh Dickins
Quoted 6 lines

I still don't understand the point of this legacy_mem file at all, and why X should want to mmap it, if it works so well when it's thus divorced from reality. But I won't worry my pretty little head over it any further - please don't waste your time trying to explain it to me!

VGA memory hole :-) ie, text mode & fonts

Ben.

Quoted 32 lines

Thanks,
Hugh

Quoted 29 lines

Index: linux-work/arch/powerpc/kernel/pci-common.c
===================================================================

arch/powerpc/kernel/pci-common.c +15 -2
@@ -561,8 +561,21 @@ int pci_mmap_legacy_page_range(struct pc
 		 (unsigned long long)(offset + size - 1));
 
 	if (mmap_state == pci_mmap_mem) {
-		if ((offset + size) > hose->isa_mem_size)
-			return -ENXIO;
+		/* Hack alert !
+		 *
+		 * Because X is lame and can fail starting if it gets an error trying
+		 * to mmap legacy_mem (instead of just moving on without legacy memory
+		 * access) we fake it here by giving it anonymous memory, effectively
+		 * behaving just like /dev/zero
+		 */
+		if ((offset + size) > hose->isa_mem_size) {
+			printk(KERN_DEBUG
+			       "Process %s (pid:%d) mapped non-existing PCI legacy memory for 0%04x:%02x\n",
+			       current->comm, current->pid, pci_domain_nr(bus), bus->number);
+			if (vma->vm_flags & VM_SHARED)
+				return shmem_zero_setup(vma);
+			return 0;
+		}
 		offset += hose->isa_mem_phys;
 	} else {
 		unsigned long io_offset = (unsigned long)hose->io_base_virt - _IO_BASE;
#123 lore raw

[Bug #12615] boot hangs while bringing up gianfar ethernet

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12615
Subject : boot hangs while bringing up gianfar ethernet
Submitter : Ira Snyder <[email protected]>
Date : 2009-01-29 19:41 (7 days old)
References : http://marc.info/?l=linux-kernel&m=123325817201665&w=4

#124 lore raw

Re: [Bug #12615] boot hangs while bringing up gianfar ethernet

Quoted 8 lines · Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 7 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12615
Subject : boot hangs while bringing up gianfar ethernet
Submitter : Ira Snyder <[email protected]>
Date : 2009-01-29 19:41 (7 days old)
References : http://marc.info/?l=linux-kernel&m=123325817201665&w=4

Tested this morning with Linux version 2.6.29-rc3-00563-g37f752c, and I still see the same behavior.

Thanks,
Ira

#125 lore raw

Re: [Bug #12615] boot hangs while bringing up gianfar ethernet

Quoted 17 lines · Ira Snyder
Quoted 8 lines · Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 7 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12615
Subject : boot hangs while bringing up gianfar ethernet
Submitter : Ira Snyder <[email protected]>
Date : 2009-01-29 19:41 (7 days old)
References : http://marc.info/?l=linux-kernel&m=123325817201665&w=4

Quoted 2 lines

Tested this morning with Linux version 2.6.29-rc3-00563-g37f752c, and I still see the same behavior.

Thanks for the update.

Rafael

#126 lore raw

[Bug #12617] unable to compile e100 firmware into kernel

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12617
Subject : unable to compile e100 firmware into kernel
Submitter : Andrey Borzenkov <arvidjaar-JGs/[email protected]>
Date : 2009-01-31 15:59 (5 days old)
References : http://marc.info/?l=linux-kernel&m=123341764915181&w=4

#127 lore raw

[Bug #12613] [Suspend regression][DRM, RADEON]

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12613
Subject : [Suspend regression][DRM, RADEON]
Submitter : etienne <etienne.basset-Bf/[email protected]>
Date : 2009-01-28 22:00 (8 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=a9d51a5ad1154b5b20add1e8d30a5564f8aabbe9
References : http://marc.info/?l=linux-kernel&m=123318030419558&w=4

		  http://marc.info/?l=linux-kernel&m=123334865404574&w=4
#128 lore raw

Re: [Bug #12613] [Suspend regression][DRM, RADEON]

Quoted 8 lines · Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 10 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12613
Subject : [Suspend regression][DRM, RADEON]
Submitter : etienne <[email protected]>
Date : 2009-01-28 22:00 (8 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=a9d51a5ad1154b5b20add1e8d30a5564f8aabbe9
References : http://marc.info/?l=linux-kernel&m=123318030419558&w=4

		  http://marc.info/?l=linux-kernel&m=123334865404574&w=4

hello,

yes it is still here with today's git
regards
Etienne

#129 lore raw

Re: [Bug #12613] [Suspend regression][DRM, RADEON]

Quoted 19 lines · etienne
Quoted 8 lines · Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 10 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12613
Subject : [Suspend regression][DRM, RADEON]
Submitter : etienne <etienne.basset-Bf/[email protected]>
Date : 2009-01-28 22:00 (8 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=a9d51a5ad1154b5b20add1e8d30a5564f8aabbe9
References : http://marc.info/?l=linux-kernel&m=123318030419558&w=4

		  http://marc.info/?l=linux-kernel&m=123334865404574&w=4
Quoted 3 lines

hello,

yes it is still here with today's git

Thanks for the update.

Rafael

#130 lore raw

[Bug #12616] boot hang: async vs. kexec

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12616
Subject : boot hang: async vs. kexec
Submitter : Randy Dunlap <[email protected]>
Date : 2009-01-29 13:15 (7 days old)
References : http://lkml.org/lkml/2009/1/29/359

#131 lore raw

Re: [Bug #12616] boot hang: async vs. kexec

Quoted 8 lines · Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 5 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12616
Subject : boot hang: async vs. kexec
Submitter : Randy Dunlap <[email protected]>
Date : 2009-01-29 13:15 (7 days old)
References : http://lkml.org/lkml/2009/1/29/359

Please reject this for now. I'll keep debugging/working on it and see if any fixes are needed.

Thanks,

Signature · 1 line
~Randy
#132 lore raw

Re: [Bug #12616] boot hang: async vs. kexec

Quoted 9 lines · Randy Dunlap
Quoted 8 lines · Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Quoted 8 lines
Quoted 5 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12616
Subject : boot hang: async vs. kexec
Submitter : Randy Dunlap <[email protected]>
Date : 2009-01-29 13:15 (7 days old)
References : http://lkml.org/lkml/2009/1/29/359

Please reject this for now. I'll keep debugging/working on it and see if any fixes are needed.

Done.

Thanks,
Rafael

#133 lore raw

[Bug #12621] Resume broken on iBook

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12621
Subject : Resume broken on iBook
Submitter : Andreas Schwab <[email protected]>
Date : 2009-02-01 16:54 (4 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=aa8c6c93747f7b55fa11e1624fec8ca33763a805

Benjamin Herrenschmidt <[email protected]>
Patch : http://bugzilla.kernel.org/attachment.cgi?id=20085&action=view

#134 lore raw

[Bug #12618] hackbench [pthread mode] regression with 2.6.29-rc3

In reply to Rafael J. Wysocki

This message has been generated automatically as a part of a report of recent regressions.

The following bug entry is on the current list of known regressions from 2.6.28. Please verify if it still should be listed and let me know (either way).

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12618
Subject : hackbench [pthread mode] regression with 2.6.29-rc3
Submitter : Zhang, Yanmin <yanmin_zhang-VuQAYsv1563Yd54FQh9/[email protected]>
Date : 2009-02-01 7:30 (4 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=490dea45d00f01847ebebd007685d564aaf2cd98
References : http://marc.info/?l=linux-kernel&m=123347347726527&w=4

#135 lore raw
In reply to Rafael J. Wysocki
Quoted 6 lines · Rafael J. Wysocki

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12599
Subject : dri /dev node disappeared with 2.6.29-rc1
Submitter : Norbert Preining <[email protected]>
Date : 2009-01-15 13:42 (21 days old)
References : http://marc.info/?l=linux-kernel&m=123202701026647&w=4

I think this one is simply that Intel DRI now requires CONFIG_FB.

So make sure that you say yes to the CONFIG_FB question (easiest fix: just edit the .config file and change the line "# CONFIG_FB is not set" to "CONFIG_FB=y" and do "make oldconfig" - I realize that people think that graphical config front-ends are simpler, but often just editing the file is the quickest way).

Quoted 6 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12591
Subject : NULL pointer dereference in blk_queue_io_stat
Submitter : Petr Vandrovec <[email protected]>
Date : 2009-01-31 23:58 (5 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=bc58ba9468d94d62c56ab9b47173583ec140b165

I think this one is fixed by commit fb8ec18c316d8.

Quoted 6 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12511
Subject : WARNING: at drivers/dma/dmaengine.c:352
Submitter : Ingo Molnar <[email protected]>
Date : 2009-01-19 21:31 (17 days old)
References : http://marc.info/?l=linux-kernel&m=123240070614443&w=4

This warning just got removed as bogus.

Commit 83436a0560e9ef8af2f0796264dde4bed1415359

Quoted 5 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12506
Subject : Undefined symbols when CONFIG_MFD_PCF50633 is enabled
Submitter : Ozan Çağlayan <[email protected]>
Date : 2009-01-17 10:56 (19 days old)
References : http://marc.info/?l=linux-kernel&m=123218991104592&w=4

Commit 9e6f8ed7c3a303d37eb119847dd3029701e37e28

Quoted 7 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12505
Subject : 2.6.29-rc1 Firefox crashing on page load
Submitter : Justin Madru <[email protected]>
Date : 2009-01-16 20:56 (20 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=4217458dafaa57d8e26a46f5d05ab8c53cf64191
References : http://marc.info/?l=linux-kernel&m=123213941914274&w=4

Heh. That commit you point to is very innocuous.

Commit 4217458dafaa57d8e26a46f5d05ab8c53cf64191 is a pure cleanup, and it _does_ make the code look nicer, but while it looks lik a total no-op, it actually has a very subtle behavioural change: it makes gcc think that it owns the whole argument stack. Which is _not_ true of 'asmlinkage' functions, but we've never been able to tell gcc to keep its grubby hands off our stack.

So I suspect that gcc inlines the function and then creates a function call that re-uses the "struct pt_regs" on the stack as the argument area too. Which corrupts the "pt_regs", and then we return to user space with some random register contents.

So I think we need to just revert it. Ingo, Hiroshi?

Linus

Tested-by Justin Madru
#136 lore raw
Quoted 17 lines · Linus Torvalds
Quoted 7 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12505
Subject : 2.6.29-rc1 Firefox crashing on page load
Submitter : Justin Madru <[email protected]>
Date : 2009-01-16 20:56 (20 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=4217458dafaa57d8e26a46f5d05ab8c53cf64191
References : http://marc.info/?l=linux-kernel&m=123213941914274&w=4

Heh. That commit you point to is very innocuous.

Commit 4217458dafaa57d8e26a46f5d05ab8c53cf64191 is a pure cleanup, and it _does_ make the code look nicer, but while it looks lik a total no-op, it actually has a very subtle behavioural change: it makes gcc think that it owns the whole argument stack. Which is _not_ true of 'asmlinkage' functions, but we've never been able to tell gcc to keep its grubby hands off our stack.

Quoted 4 lines

So I suspect that gcc inlines the function and then creates a function call that re-uses the "struct pt_regs" on the stack as the argument area too. Which corrupts the "pt_regs", and then we return to user space with some random register contents.

yes - i tracked it down back then, just the regression entry didnt get updated.

Quoted 1 line

So I think we need to just revert it. Ingo, Hiroshi?

should be fixed upstream by:

552b8aa: Revert "x86: signal: change type of paramter for sys_rt_sigreturn()"

the revert commit is below with explaination - also with a Tested-by tag. The revert is not a pure revert - it also adds a comment to preserve this circumstance cleanly.

We first had a few fruitless attempts to annotate it in some clear fashion (the existing tailcall-avoidance macros we have didnt suit this case) - then went for the revert because none of the variants that told gcc not to screw up actually made the code better than it was originally.

Ingo

-------------->

Quoted 1 line

From 552b8aa4d1edcc1c764ff6f61a7686347a2d1827 Mon Sep 17 00:00:00 2001

From: Ingo Molnar <[email protected]>
Date: Tue, 20 Jan 2009 09:31:49 +0100
Subject: [PATCH] Revert "x86: signal: change type of paramter for sys_rt_sigreturn()"

This reverts commit 4217458dafaa57d8e26a46f5d05ab8c53cf64191.

Justin Madru bisected this commit, it was causing weird Firefox crashes.

The reason is that GCC mis-optimizes (re-uses) the on-stack parameters of the calling frame, which corrupts the syscall return pt_regs state and thus corrupts user-space register state.

So we go back to the slightly less clean but more optimization-safe method of getting to pt_regs. Also add a comment to explain this.

Resolves: http://bugzilla.kernel.org/show_bug.cgi?id=12505

---
 arch/x86/include/asm/syscalls.h |    2 +-
 arch/x86/kernel/signal.c        |   11 +++++++++--
 2 files changed, 10 insertions(+), 3 deletions(-)
arch/x86/include/asm/syscalls.h +1 -1
index 9c6797c..c0b0bda 100644
@@ -40,7 +40,7 @@ asmlinkage int sys_sigaction(int, const struct old_sigaction __user *,
 			     struct old_sigaction __user *);
 asmlinkage int sys_sigaltstack(unsigned long);
 asmlinkage unsigned long sys_sigreturn(unsigned long);
-asmlinkage int sys_rt_sigreturn(struct pt_regs);
+asmlinkage int sys_rt_sigreturn(unsigned long);
 
 /* kernel/ioport.c */
 asmlinkage long sys_iopl(unsigned long);
arch/x86/kernel/signal.c +9 -2
index 89bb766..df0587f 100644
@@ -632,9 +632,16 @@ badframe:
 }
 
 #ifdef CONFIG_X86_32
-asmlinkage int sys_rt_sigreturn(struct pt_regs regs)
+/*
+ * Note: do not pass in pt_regs directly as with tail-call optimization
+ * GCC will incorrectly stomp on the caller's frame and corrupt user-space
+ * register state:
+ */
+asmlinkage int sys_rt_sigreturn(unsigned long __unused)
 {
-	return do_rt_sigreturn(&regs);
+	struct pt_regs *regs = (struct pt_regs *)&__unused;
+
+	return do_rt_sigreturn(regs);
 }
 #else /* !CONFIG_X86_32 */
 asmlinkage long sys_rt_sigreturn(struct pt_regs *regs)
#137 lore raw
In reply to Linus Torvalds
Quoted 7 lines · Linus Torvalds
Quoted 5 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12599
Subject : dri /dev node disappeared with 2.6.29-rc1
Submitter : Norbert Preining <[email protected]>
Date : 2009-01-15 13:42 (21 days old)
References : http://marc.info/?l=linux-kernel&m=123202701026647&w=4

I think this one is simply that Intel DRI now requires CONFIG_FB.

Right. I mentioned that in another bug report:
http://bugzilla.kernel.org/show_bug.cgi?id=12494

Quoted 5 lines

So make sure that you say yes to the CONFIG_FB question (easiest fix: just edit the .config file and change the line "# CONFIG_FB is not set" to "CONFIG_FB=y" and do "make oldconfig" - I realize that people think that graphical config front-ends are simpler, but often just editing the file is the quickest way).

The problem is that if you have a configuration under 2.6.28 without CONFIG_FB and just call make oldconfig, or even make config and don't know that you loose the DRM. And I was using make oldconfig (there is a graphical config?? ;-))

There is that other bug 12494 mentioned, namely due to the disappearance of DRM the brightness adjustment on my vaio z11 stopped working.

So maybe there should be a select FB or so under the CONFIG_DRM option (not that I actually can program Kconfig).

Best wishes

Norbert

-------------------------------------------------------------------------------
Dr. Norbert Preining <[email protected]> Vienna University of Technology
Debian Developer <[email protected]> Debian TeX Group
gpg DSA: 0x09C5B094 fp: 14DF 2E6C 0307 BE6D AD76 A9C0 D2BF 4AA3 09C5 B094
-------------------------------------------------------------------------------
PITLOCHRY (n.)
The background gurgling noise heard in Wimby Bars caused by people trying to get the last bubbles out of their milkshakes by slurping loudly through their straws.
--- Douglas Adams, The Meaning of Liff

#138 lore raw
Quoted 5 lines · Norbert Preining

The problem is that if you have a configuration under 2.6.28 without CONFIG_FB and just call make oldconfig, or even make config and don't know that you loose the DRM. And I was using make oldconfig (there is a graphical config?? ;-))

Sure. It's inconvenient, no question about that. I asked the i915 people to look into not requiring CONFIG_FB, and I hope they will, but my point is that I don't think we can consider "small one-time inconvenience" to be a "regression".

Quoted 2 lines

So maybe there should be a select FB or so under the CONFIG_DRM option (not that I actually can program Kconfig).

CONFIG_FB is one of the things that _could_ be selected for (it doesn't depend on anything else that I know about), so just changing the "depends on FB" into a "select FB" may get rid of the issue - but it would not actually fix your problem, since your CONFIG_DRM_I915 has already been set to 'n' by now - so you'd _still_ have to enable it by hand again.

So the upside is not very big - people who have custom configs are the often the same people who try -rc1's, so they've already hit this.

Linus

#139 lore raw
Quoted 2 lines · Linus Torvalds

is that I don't think we can consider "small one-time inconvenience" to be a "regression".

Agreed.

Best wishes

Norbert

-------------------------------------------------------------------------------
Dr. Norbert Preining <[email protected]> Vienna University of Technology
Debian Developer <[email protected]> Debian TeX Group
gpg DSA: 0x09C5B094 fp: 14DF 2E6C 0307 BE6D AD76 A9C0 D2BF 4AA3 09C5 B094
-------------------------------------------------------------------------------
EAST WITTERING (n.)
The same as west wittering (q.v.) only it's you they've trying to get away from.
--- Douglas Adams, The Meaning of Liff

#140 lore raw
In reply to Linus Torvalds
Quoted 11 lines · Linus Torvalds
Quoted 5 lines · Norbert Preining

The problem is that if you have a configuration under 2.6.28 without CONFIG_FB and just call make oldconfig, or even make config and don't know that you loose the DRM. And I was using make oldconfig (there is a graphical config?? ;-))

Sure. It's inconvenient, no question about that. I asked the i915 people to look into not requiring CONFIG_FB, and I hope they will, but my point is that I don't think we can consider "small one-time inconvenience" to be a "regression".

if you mean that as a general principle, there's four very real downsides in my opinion.

Firstly, we could have done better (and still can do better), via various easy and non-intrusive measures:

- We could add a runtime warning:

for example a WARN_ONCE("please enable CONFIG_DRM_I915 and CONFIG_FB") that there's no DRM because CONFIG_FB is not selected and oldconfig loses the I915 setting silently - placed in a key DRM ioctl, would have gone a long way addressing the issue. Testers do notice kernel warnings that pop up when their X gets slow. (This approach might also have the added bonus of warning folks who enable the wrong driver for the hardware.)

- Or we could add a more thoughtful Kconfig migration:

Rename DRM_I915 to DRM_I915_FB [which it really is now], and keep DRM_I915 as a non-interactive migration helper: if set, it auto-selects both FB and DRM_I915_FB.

While CONFIG_FB is an interactive Kconfig option so a select can be dangerous to a correct dependency tree, it seems safe to do in this specific case because it seems to be a rather leaf entry with no dependencies.

Sure, upgrading systems and following development is never easy and well structured brains are in heavy demand in any case - but we should really engineer things so that people spend time on reporting _real_ regressions - and not waste their time and energy on self-inflicted wounds.

The four very real downsides are:

1) I dont think it's honest to make an artificial distinction between a user-visible bug that can be fixed via a one-time tweak to a .config, and a user-visible bug that can be fixed via a one-time upgrade to a fixed kernel.

As far as the tester is involved the two are largely the same, and both were caused by this shiny new 'upstream kernel' thing they just tested.

Yes, to us developers there's a fundamental difference because one is a fix in the code the other is a change in the environment - but the user really doesnt make that distinction - both are largely black boxes to them.

Also, it can actually get _worse_ than a regression: if it gets declared a feature/annoyance and no action is taken - then it's an unfixed bug and that's far worse than even the painful memory of a fixed bug.

2) Furthermore, many testers dont touch early -rcs because they know they have rough edges. So it's a linearly ongoing cost beyond -rc1.

3) It's not the individual small annoyances that matter but the sheer sum of of such annoyances when migrating from one kernel to another. I still struggle with some of such issues myself when testing back to some old kernel and keep wasting time on it - again and again. Keeping oldconfig compatibility is really a prime-type quality of Linux kernel testing i think.

Especially a casual tester will be discouraged by such an experience:
and more so if it's clearly avoidable via a few trivial tweaks - and the tester senses this 'ease of fix' intuitively and gets a "not a bug" pushback.

4) Also, some maintainers are using every excuse they can find to not have to keep 'make oldconfig' compatibility. That is not because they are lazy, embarrased and defensive or evil, but it is a natural reaction:
they only see the small trivial annoyance they intruduce themselves - which is in a code area and usecase they are prominently familiar with, so they cannot personally relate to the trouble that users go through if they hit such issues.

But the thing is, we've got dozens and dozens of critical subsystems, and if each one does just a single such stupid looking tiny thing in a kernel cycle it mounts up to a real force. The sheer size of Linux is a real multiplicator factor here IMO. The moment Linus condones it the precedent is set and everyone will just say "others do it too every now and then and Linus doesnt mind so buzz off".

In my experience the sum of such small stupid careless migration annoyances during upgrades snowballs way too much in practice, and it has become the main vector along which we lose good testers - so i am never shy to flame folks if they introduce such issues ;-) YMMV.

</soapbox ;) >

Ingo

#141 lore raw
Quoted 5 lines · Ingo Molnar

[...] it is a natural reaction:
they only see the small trivial annoyance they intruduce themselves - which is in a code area and usecase they are prominently familiar with, so they cannot personally relate to the trouble that users go through if they hit such issues.

Amen. Preach it. I spent quite a while just a week ago arguing that every semi-loaded machine out there using epoll should not require the admin to discover that their previously happy software stack was suddenly hitting an artificially tiny per-user instances count.

Luckily I was able to find multiple blog posts and mailing list archives with people who had literally spent _days_ tracking down why things had broken for them when they upgraded to a new -stable kernel.

You really do have to assume that your users don't have time for this shit. Anything that really can't DTRT automatically needs to be covered with plenty of easy to follow instructions on how to fix the problem - because for someone unfamiliar with that area of the system it really does take enormous effort to track down what's changed.

Bron.

#142 lore raw

(Cc:-ed Davide)

Quoted 16 lines · Bron Gondwana
Quoted 5 lines · Ingo Molnar

[...] it is a natural reaction:
they only see the small trivial annoyance they intruduce themselves - which is in a code area and usecase they are prominently familiar with, so they cannot personally relate to the trouble that users go through if they hit such issues.

Amen. Preach it. I spent quite a while just a week ago arguing that every semi-loaded machine out there using epoll should not require the admin to discover that their previously happy software stack was suddenly hitting an artificially tiny per-user instances count.

Luckily I was able to find multiple blog posts and mailing list archives with people who had literally spent _days_ tracking down why things had broken for them when they upgraded to a new -stable kernel.

Quoted 5 lines

You really do have to assume that your users don't have time for this shit. Anything that really can't DTRT automatically needs to be covered with plenty of easy to follow instructions on how to fix the problem - because for someone unfamiliar with that area of the system it really does take enormous effort to track down what's changed.

do you know which commit that was (or which exact tunable default value it is about) and whether we could restore the old default safely, and whether there's some reasonable minium must-have value that still works well in practice?

With Moore's law still alive and kicking there's normally no reason to narrow defaults - if then they get increased or get changed to some auto-size-to-hw-capabilities dynamic method.

Upstream defaults usually get narrowed only for really good reasons and often the reason is DoS and security and a specific testcase that kills a default box with a too large default. Sometimes they get narrowed spuriously and then we can fix things reasonably.

Ingo

#143 lore raw
Quoted 2 lines · Ingo Molnar

(Cc:-ed Davide)

eep

Quoted 24 lines · Bron Gondwana
Quoted 17 lines · Ingo Molnar
Quoted 5 lines

[...] it is a natural reaction:
they only see the small trivial annoyance they intruduce themselves - which is in a code area and usecase they are prominently familiar with, so they cannot personally relate to the trouble that users go through if they hit such issues.

Amen. Preach it. I spent quite a while just a week ago arguing that every semi-loaded machine out there using epoll should not require the admin to discover that their previously happy software stack was suddenly hitting an artificially tiny per-user instances count.

Luckily I was able to find multiple blog posts and mailing list archives with people who had literally spent _days_ tracking down why things had broken for them when they upgraded to a new -stable kernel.

Quoted 5 lines

You really do have to assume that your users don't have time for this shit. Anything that really can't DTRT automatically needs to be covered with plenty of easy to follow instructions on how to fix the problem - because for someone unfamiliar with that area of the system it really does take enormous effort to track down what's changed.

Quoted 4 lines

do you know which commit that was (or which exact tunable default value it is about) and whether we could restore the old default safely, and whether there's some reasonable minium must-have value that still works well in practice?

Oh, it got sorted out eventually, but not without a whole lot of debate about how it wasn't that hard (per individual). Let's not stir this one up again :) We've resolved it to everyone's satisfaction.

It's the more abstract "I understand the issue and it's easy for me to set a sane config for my environment" being extended to everyone having to understand yet another bloody tunable.

And I'm somewhat guilty of it myself with Cyrus. You run into the thorny issue of: there's (a) the sane choice, (b) the backwards compatible choice. Every new site should be running (a), but you don't want to ship a new stable version with (a) as the default because it will break everything and people will need to figure out what your stupid little change was and why it broke them. Especially if the underlying issue doesn't actually bother their config.

I tend to side with defaulting to (b), and the basic config file on our of our imap servers just keeps getting longer.

Quoted 3 lines

With Moore's law still alive and kicking there's normally no reason to narrow defaults - if then they get increased or get changed to some auto-size-to-hw-capabilities dynamic method.

It was an N^2 issue. I appreciate the DoS risk it solved, just that the solution was a stab in the dark, and it wound up hitting a lot more people than expected (who knew that most epoll users create one watcher per process, but still create lots of processes as well, obviously not many people until it was shown to affect at least postfix, apache and java. Quite the collection!)

Quoted 4 lines

Upstream defaults usually get narrowed only for really good reasons and often the reason is DoS and security and a specific testcase that kills a default box with a too large default. Sometimes they get narrowed spuriously and then we can fix things reasonably.

Yeah. It's a pain - especially since more fine-grained tracking to distinguish between a DoS and a reasonable user comes with its own costs (see companies that expect you to track your time in 5 minute increments, and act all surprised when half of everyone's time comes back coded as "filling in the stupid timesheets")

Bron.

#144 lore raw
In reply to Ingo Molnar
Quoted 32 lines · Ingo Molnar
Quoted 11 lines · Linus Torvalds
Quoted 5 lines · Norbert Preining

The problem is that if you have a configuration under 2.6.28 without CONFIG_FB and just call make oldconfig, or even make config and don't know that you loose the DRM. And I was using make oldconfig (there is a graphical config?? ;-))

Sure. It's inconvenient, no question about that. I asked the i915 people to look into not requiring CONFIG_FB, and I hope they will, but my point is that I don't think we can consider "small one-time inconvenience" to be a "regression".

if you mean that as a general principle, there's four very real downsides in my opinion.

Firstly, we could have done better (and still can do better), via various easy and non-intrusive measures:

- We could add a runtime warning:

for example a WARN_ONCE("please enable CONFIG_DRM_I915 and CONFIG_FB") that there's no DRM because CONFIG_FB is not selected and oldconfig loses the I915 setting silently - placed in a key DRM ioctl, would have gone a long way addressing the issue. Testers do notice kernel warnings that pop up when their X gets slow. (This approach might also have the added bonus of warning folks who enable the wrong driver for the hardware.)

- Or we could add a more thoughtful Kconfig migration:

Quoted 8 lines

Rename DRM_I915 to DRM_I915_FB [which it really is now], and keep DRM_I915 as a non-interactive migration helper: if set, it auto-selects both FB and DRM_I915_FB.

While CONFIG_FB is an interactive Kconfig option so a select can be dangerous to a correct dependency tree, it seems safe to do in this specific case because it seems to be a rather leaf entry with no dependencies.

I tried select FB. It's the right thing to do. It doesn't work. I posted to the mailing list two weeks ago about the insane dependency chain that kbuild comes up with and fails on when we do this, and got silence.

Believe me, I hate this inconvenience to users even more than each of you do, because I get to deal with the reports. But I haven't had the time to sit down and figure out what drugs kbuild is on, or even how to work around it (despite IRC help from a few other kernel guys).

The alternative I can see is to ifdef the code for something that will be on by default and which stable userland will require in 6 months. That seems wrong.

Signature · 2 lines
Eric Anholt
[email protected]                         [email protected]
#145 lore raw
Quoted 4 lines · Eric Anholt

I tried select FB. It's the right thing to do. It doesn't work. I posted to the mailing list two weeks ago about the insane dependency chain that kbuild comes up with and fails on when we do this, and got silence.

Right, I agree that I have tried to do it myself, and it was impossible when:
- editing .config by hand removing any *FB* and *DRM* related stuff
- calling make oldconfig
in all cases DRM was disabled.

I managed only to activate it by hand-editing .config and there putting by hand
CONFIG_something=y
(don't remember which it was) and after that the make oldconfig did actually select everything I need.

So, that seems to be a Kconfig bug.

Best wishes

Norbert

-------------------------------------------------------------------------------
Dr. Norbert Preining <[email protected]> Vienna University of Technology
Debian Developer <[email protected]> Debian TeX Group
gpg DSA: 0x09C5B094 fp: 14DF 2E6C 0307 BE6D AD76 A9C0 D2BF 4AA3 09C5 B094
-------------------------------------------------------------------------------
KINGSTON BAGPUISE (n.)
A forty-year-old sixteen-stone man trying to commit suicide by jogging.
--- Douglas Adams, The Meaning of Liff

#146 lore raw

PATCH 1/1 drm, i915: select framebuffer support automatically

In reply to Eric Anholt
Quoted 42 lines · Eric Anholt
Quoted 32 lines · Ingo Molnar
Quoted 11 lines · Linus Torvalds
Quoted 5 lines · Norbert Preining

The problem is that if you have a configuration under 2.6.28 without CONFIG_FB and just call make oldconfig, or even make config and don't know that you loose the DRM. And I was using make oldconfig (there is a graphical config?? ;-))

Sure. It's inconvenient, no question about that. I asked the i915 people to look into not requiring CONFIG_FB, and I hope they will, but my point is that I don't think we can consider "small one-time inconvenience" to be a "regression".

if you mean that as a general principle, there's four very real downsides in my opinion.

Firstly, we could have done better (and still can do better), via various easy and non-intrusive measures:

- We could add a runtime warning:

for example a WARN_ONCE("please enable CONFIG_DRM_I915 and CONFIG_FB") that there's no DRM because CONFIG_FB is not selected and oldconfig loses the I915 setting silently - placed in a key DRM ioctl, would have gone a long way addressing the issue. Testers do notice kernel warnings that pop up when their X gets slow. (This approach might also have the added bonus of warning folks who enable the wrong driver for the hardware.)

- Or we could add a more thoughtful Kconfig migration:

Quoted 8 lines

Rename DRM_I915 to DRM_I915_FB [which it really is now], and keep DRM_I915 as a non-interactive migration helper: if set, it auto-selects both FB and DRM_I915_FB.

While CONFIG_FB is an interactive Kconfig option so a select can be dangerous to a correct dependency tree, it seems safe to do in this specific case because it seems to be a rather leaf entry with no dependencies.

Quoted 4 lines

I tried select FB. It's the right thing to do. It doesn't work. I posted to the mailing list two weeks ago about the insane dependency chain that kbuild comes up with and fails on when we do this, and got silence.

I mean the patch below.

I have tested it here it works fine and has no dependency problems, nor any build breakages on 32-bit or 64-bit x86. When you tried this you probably ran into the FB_I810 and FB_INTEL complication - this patch solves that too.

What do you think?

Ingo

--------------->

Quoted 1 line

From ca835567dcb7c513a26a4396d86f12848e62de8d Mon Sep 17 00:00:00 2001

From: Ingo Molnar <[email protected]>
Date: Thu, 5 Feb 2009 16:03:34 +0100
Subject: [PATCH] drm, i915: select framebuffer support automatically

Migration helper.

The i915 driver recently added a 'depends on FB' rule to its Kconfig entry - which silently turns off DRM_I915 if someone has a working config but no CONFIG_FB selected, and upgrades to the latest upstream kernel.

Norbert Preining reported this problem:

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12599
Subject : dri /dev node disappeared with 2.6.29-rc1

So change it to "select FB", which auto-selects framebuffer support. This way the driver keeps working, regardless of whether FB was enabled before or not.

Kconfig select's of interactive options can be problematic to dependencies and can cause build breakages - but in this case it's safe because it's a leaf entry with no dependencies of its own.

( There is some minor circular dependency fallout as FB_I810
and FB_INTEL also used 'depends on FB' constructs - update those to "select FB" too. )

---
 drivers/gpu/drm/Kconfig |    2 +-
 drivers/video/Kconfig   |    6 ++++--
 2 files changed, 5 insertions(+), 3 deletions(-)
drivers/gpu/drm/Kconfig +1 -1
index 5130b72..4be3acb 100644
@@ -70,7 +70,7 @@ config DRM_I915
 	select FB_CFB_FILLRECT
 	select FB_CFB_COPYAREA
 	select FB_CFB_IMAGEBLIT
-	depends on FB
+	select FB
 	tristate "i915 driver"
 	help
 	  Choose this option if you have a system that has Intel 830M, 845G,
drivers/video/Kconfig +4 -2
index f026770..bf0af66 100644
@@ -1054,9 +1054,10 @@ config FB_RIVA_BACKLIGHT
 
 config FB_I810
 	tristate "Intel 810/815 support (EXPERIMENTAL)"
-	depends on FB && EXPERIMENTAL && PCI && X86_32
+	depends on EXPERIMENTAL && PCI && X86_32
 	select AGP
 	select AGP_INTEL
+	select FB
 	select FB_MODE_HELPERS
 	select FB_CFB_FILLRECT
 	select FB_CFB_COPYAREA
@@ -1119,7 +1120,8 @@ config FB_CARILLO_RANCH
 
 config FB_INTEL
 	tristate "Intel 830M/845G/852GM/855GM/865G/915G/945G/945GM/965G/965GM support (EXPERIMENTAL)"
-	depends on FB && EXPERIMENTAL && PCI && X86
+	depends on EXPERIMENTAL && PCI && X86
+	select FB
 	select AGP
 	select AGP_INTEL
 	select FB_MODE_HELPERS
maintainer
#147 lore raw

Re: [PATCH] drm, i915: select framebuffer support automatically

Quoted 26 lines · Ingo Molnar
Quoted 23 lines · Eric Anholt
Quoted 22 lines · Ingo Molnar
Quoted 11 lines · Linus Torvalds
Quoted 5 lines · Norbert Preining

The problem is that if you have a configuration under 2.6.28 without CONFIG_FB and just call make oldconfig, or even make config and don't know that you loose the DRM. And I was using make oldconfig (there is a graphical config?? ;-))

Sure. It's inconvenient, no question about that. I asked the i915 people to look into not requiring CONFIG_FB, and I hope they will, but my point is that I don't think we can consider "small one-time inconvenience" to be a "regression".

if you mean that as a general principle, there's four very real downsides in my opinion.

Firstly, we could have done better (and still can do better), via various easy and non-intrusive measures:

- We could add a runtime warning:

Quoted 7 lines
Quoted 7 lines
Quoted 7 lines

for example a WARN_ONCE("please enable CONFIG_DRM_I915 and CONFIG_FB") that there's no DRM because CONFIG_FB is not selected and oldconfig loses the I915 setting silently - placed in a key DRM ioctl, would have gone a long way addressing the issue. Testers do notice kernel warnings that pop up when their X gets slow. (This approach might also have the added bonus of warning folks who enable the wrong driver for the hardware.)

We can't actually do this, if you don't have a DRM driver for the hw, you don't get any device node or ioctl
to play with. X just loads the module, if it doesn't exist X complains, nothing we could do about it in the kernel without adding some wierdass fake drm devices.

Quoted 11 lines
Quoted 11 lines
Quoted 11 lines

- Or we could add a more thoughtful Kconfig migration:

Rename DRM_I915 to DRM_I915_FB [which it really is now], and keep DRM_I915 as a non-interactive migration helper: if set, it auto-selects both FB and DRM_I915_FB.

While CONFIG_FB is an interactive Kconfig option so a select can be dangerous to a correct dependency tree, it seems safe to do in this specific case because it seems to be a rather leaf entry with no dependencies.

This is the correct answer, the only other option was to allow i915 to build without modesetting
if FB isn't selected but Intel didn't want to go down that road.

Quoted 43 lines
Quoted 5 lines

I tried select FB. It's the right thing to do. It doesn't work. I posted to the mailing list two weeks ago about the insane dependency chain that kbuild comes up with and fails on when we do this, and got silence.

I mean the patch below.

I have tested it here it works fine and has no dependency problems, nor any build breakages on 32-bit or 64-bit x86. When you tried this you probably ran into the FB_I810 and FB_INTEL complication - this patch solves that too.

What do you think?

Ingo

--------------->
From ca835567dcb7c513a26a4396d86f12848e62de8d Mon Sep 17 00:00:00 2001
From: Ingo Molnar <[email protected]>
Date: Thu, 5 Feb 2009 16:03:34 +0100
Subject: [PATCH] drm, i915: select framebuffer support automatically

Migration helper.

The i915 driver recently added a 'depends on FB' rule to its Kconfig entry - which silently turns off DRM_I915 if someone has a working config but no CONFIG_FB selected, and upgrades to the latest upstream kernel.

Norbert Preining reported this problem:

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12599
Subject : dri /dev node disappeared with 2.6.29-rc1

So change it to "select FB", which auto-selects framebuffer support. This way the driver keeps working, regardless of whether FB was enabled before or not.

Kconfig select's of interactive options can be problematic to dependencies and can cause build breakages - but in this case it's safe because it's a leaf entry with no dependencies of its own.

Quoted 6 lines

( There is some minor circular dependency fallout as FB_I810
and FB_INTEL also used 'depends on FB' constructs - update those to "select FB" too. )

This looks like the best option to me so I'll suck this patch into my tree.

Thanks Ingo for solving the wierdass depends chain.

Dave.

Quoted 50 lines
---
 drivers/gpu/drm/Kconfig |    2 +-
 drivers/video/Kconfig   |    6 ++++--
 2 files changed, 5 insertions(+), 3 deletions(-)
drivers/gpu/drm/Kconfig +1 -1
index 5130b72..4be3acb 100644
@@ -70,7 +70,7 @@ config DRM_I915
       select FB_CFB_FILLRECT
       select FB_CFB_COPYAREA
       select FB_CFB_IMAGEBLIT
-       depends on FB
+       select FB
       tristate "i915 driver"
       help
         Choose this option if you have a system that has Intel 830M, 845G,
drivers/video/Kconfig +4 -3
index f026770..bf0af66 100644
@@ -1054,9 +1054,10 @@ config FB_RIVA_BACKLIGHT
 
 config FB_I810
       tristate "Intel 810/815 support (EXPERIMENTAL)"
-       depends on FB && EXPERIMENTAL && PCI && X86_32
+       depends on EXPERIMENTAL && PCI && X86_32
       select AGP
       select AGP_INTEL
+       select FB
       select FB_MODE_HELPERS
       select FB_CFB_FILLRECT
       select FB_CFB_COPYAREA
@@ -1119,7 +1120,8 @@ config FB_CARILLO_RANCH
 
 config FB_INTEL
       tristate "Intel 830M/845G/852GM/855GM/865G/915G/945G/945GM/965G/965GM support (EXPERIMENTAL)"
-       depends on FB && EXPERIMENTAL && PCI && X86
+       depends on EXPERIMENTAL && PCI && X86
+       select FB
       select AGP
       select AGP_INTEL
       select FB_MODE_HELPERS
--

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/

#148 lore raw
In reply to Eric Anholt
Quoted 39 lines · Eric Anholt
Quoted 29 lines · Ingo Molnar
Quoted 9 lines · Linus Torvalds
Quoted 4 lines · Norbert Preining

The problem is that if you have a configuration under 2.6.28 without CONFIG_FB and just call make oldconfig, or even make config and don't know that you loose the DRM. And I was using make oldconfig (there is a graphical config?? ;-))

Sure. It's inconvenient, no question about that. I asked the i915 people to look into not requiring CONFIG_FB, and I hope they will, but my point is that I don't think we can consider "small one-time inconvenience" to be a "regression".

if you mean that as a general principle, there's four very real downsides in my opinion.

Firstly, we could have done better (and still can do better), via various easy and non-intrusive measures:

- We could add a runtime warning:

for example a WARN_ONCE("please enable CONFIG_DRM_I915 and CONFIG_FB") that there's no DRM because CONFIG_FB is not selected and oldconfig loses the I915 setting silently - placed in a key DRM ioctl, would have gone a long way addressing the issue. Testers do notice kernel warnings that pop up when their X gets slow. (This approach might also have the added bonus of warning folks who enable the wrong driver for the hardware.)

- Or we could add a more thoughtful Kconfig migration:

Quoted 8 lines

Rename DRM_I915 to DRM_I915_FB [which it really is now], and keep DRM_I915 as a non-interactive migration helper: if set, it auto-selects both FB and DRM_I915_FB.

While CONFIG_FB is an interactive Kconfig option so a select can be dangerous to a correct dependency tree, it seems safe to do in this specific case because it seems to be a rather leaf entry with no dependencies.

Quoted 4 lines

I tried select FB. It's the right thing to do. It doesn't work. I posted to the mailing list two weeks ago about the insane dependency chain that kbuild comes up with and fails on when we do this, and got silence.

I tried what you had described in that email (from 2 weeks ago), got the same results that you did, but kbuild does seem very confused (to me).

reference email from 2+ weeks ago:
http://marc.info/?l=linux-kernel&m=123197341316461&w=2

Adding Sam to cc.

Quoted 9 lines

Believe me, I hate this inconvenience to users even more than each of you do, because I get to deal with the reports. But I haven't had the time to sit down and figure out what drugs kbuild is on, or even how to work around it (despite IRC help from a few other kernel guys).

The alternative I can see is to ifdef the code for something that will be on by default and which stable userland will require in 6 months. That seems wrong.

Signature · 1 line
~Randy
#149 lore raw
Quoted 45 lines · Randy Dunlap
Quoted 39 lines · Eric Anholt
Quoted 29 lines · Ingo Molnar
Quoted 9 lines · Linus Torvalds
Quoted 4 lines · Norbert Preining

The problem is that if you have a configuration under 2.6.28 without CONFIG_FB and just call make oldconfig, or even make config and don't know that you loose the DRM. And I was using make oldconfig (there is a graphical config?? ;-))

Sure. It's inconvenient, no question about that. I asked the i915 people to look into not requiring CONFIG_FB, and I hope they will, but my point is that I don't think we can consider "small one-time inconvenience" to be a "regression".

if you mean that as a general principle, there's four very real downsides in my opinion.

Firstly, we could have done better (and still can do better), via various easy and non-intrusive measures:

- We could add a runtime warning:

for example a WARN_ONCE("please enable CONFIG_DRM_I915 and CONFIG_FB") that there's no DRM because CONFIG_FB is not selected and oldconfig loses the I915 setting silently - placed in a key DRM ioctl, would have gone a long way addressing the issue. Testers do notice kernel warnings that pop up when their X gets slow. (This approach might also have the added bonus of warning folks who enable the wrong driver for the hardware.)

- Or we could add a more thoughtful Kconfig migration:

Quoted 8 lines

Rename DRM_I915 to DRM_I915_FB [which it really is now], and keep DRM_I915 as a non-interactive migration helper: if set, it auto-selects both FB and DRM_I915_FB.

While CONFIG_FB is an interactive Kconfig option so a select can be dangerous to a correct dependency tree, it seems safe to do in this specific case because it seems to be a rather leaf entry with no dependencies.

Quoted 4 lines

I tried select FB. It's the right thing to do. It doesn't work. I posted to the mailing list two weeks ago about the insane dependency chain that kbuild comes up with and fails on when we do this, and got silence.

Quoted 7 lines

I tried what you had described in that email (from 2 weeks ago), got the same results that you did, but kbuild does seem very confused (to me).

reference email from 2+ weeks ago:
http://marc.info/?l=linux-kernel&m=123197341316461&w=2

Adding Sam to cc.

Check the patch i posted in this thread earlier today, it solves this problem.

Ingo

#150 lore raw
Quoted 48 lines · Ingo Molnar
Quoted 46 lines · Randy Dunlap
Quoted 34 lines · Eric Anholt
Quoted 33 lines · Ingo Molnar
Quoted 9 lines · Linus Torvalds
Quoted 4 lines · Norbert Preining

The problem is that if you have a configuration under 2.6.28 without CONFIG_FB and just call make oldconfig, or even make config and don't know that you loose the DRM. And I was using make oldconfig (there is a graphical config?? ;-))

Sure. It's inconvenient, no question about that. I asked the i915 people to look into not requiring CONFIG_FB, and I hope they will, but my point is that I don't think we can consider "small one-time inconvenience" to be a "regression".

if you mean that as a general principle, there's four very real downsides in my opinion.

Firstly, we could have done better (and still can do better), via various easy and non-intrusive measures:

- We could add a runtime warning:

for example a WARN_ONCE("please enable CONFIG_DRM_I915 and CONFIG_FB") that there's no DRM because CONFIG_FB is not selected and oldconfig loses the I915 setting silently - placed in a key DRM ioctl, would have gone a long way addressing the issue. Testers do notice kernel warnings that pop up when their X gets slow. (This approach might also have the added bonus of warning folks who enable the wrong driver for the hardware.)

- Or we could add a more thoughtful Kconfig migration:

Rename DRM_I915 to DRM_I915_FB [which it really is now], and keep DRM_I915 as a non-interactive migration helper: if set, it auto-selects both FB and DRM_I915_FB.

Quoted 8 lines
Quoted 4 lines

While CONFIG_FB is an interactive Kconfig option so a select can be dangerous to a correct dependency tree, it seems safe to do in this specific case because it seems to be a rather leaf entry with no dependencies.

I tried select FB. It's the right thing to do. It doesn't work. I posted to the mailing list two weeks ago about the insane dependency chain that kbuild comes up with and fails on when we do this, and got silence.

I tried what you had described in that email (from 2 weeks ago), got the same results that you did, but kbuild does seem very confused (to me).

Quoted 7 lines
Quoted 4 lines

reference email from 2+ weeks ago:
http://marc.info/?l=linux-kernel&m=123197341316461&w=2

Adding Sam to cc.

Check the patch i posted in this thread earlier today, it solves this problem.

I saw it. I'd rather kconfig be fixed instead, if possible.

Signature · 1 line
~Randy
#151 lore raw
Quoted 52 lines · Randy Dunlap
Quoted 51 lines · Ingo Molnar
Quoted 49 lines · Randy Dunlap
Quoted 34 lines · Eric Anholt
Quoted 33 lines · Ingo Molnar
Quoted 9 lines · Linus Torvalds
Quoted 4 lines · Norbert Preining

The problem is that if you have a configuration under 2.6.28 without CONFIG_FB and just call make oldconfig, or even make config and don't know that you loose the DRM. And I was using make oldconfig (there is a graphical config?? ;-))

Sure. It's inconvenient, no question about that. I asked the i915 people to look into not requiring CONFIG_FB, and I hope they will, but my point is that I don't think we can consider "small one-time inconvenience" to be a "regression".

if you mean that as a general principle, there's four very real downsides in my opinion.

Firstly, we could have done better (and still can do better), via various easy and non-intrusive measures:

- We could add a runtime warning:

for example a WARN_ONCE("please enable CONFIG_DRM_I915 and CONFIG_FB") that there's no DRM because CONFIG_FB is not selected and oldconfig loses the I915 setting silently - placed in a key DRM ioctl, would have gone a long way addressing the issue. Testers do notice kernel warnings that pop up when their X gets slow. (This approach might also have the added bonus of warning folks who enable the wrong driver for the hardware.)

- Or we could add a more thoughtful Kconfig migration:

Rename DRM_I915 to DRM_I915_FB [which it really is now], and keep DRM_I915 as a non-interactive migration helper: if set, it auto-selects both FB and DRM_I915_FB.

Quoted 8 lines
Quoted 4 lines

While CONFIG_FB is an interactive Kconfig option so a select can be dangerous to a correct dependency tree, it seems safe to do in this specific case because it seems to be a rather leaf entry with no dependencies.

I tried select FB. It's the right thing to do. It doesn't work. I posted to the mailing list two weeks ago about the insane dependency chain that kbuild comes up with and fails on when we do this, and got silence.

I tried what you had described in that email (from 2 weeks ago), got the same results that you did, but kbuild does seem very confused (to me).

reference email from 2+ weeks ago:
http://marc.info/?l=linux-kernel&m=123197341316461&w=2

Quoted 6 lines
Quoted 4 lines
Quoted 1 line

Adding Sam to cc.

Check the patch i posted in this thread earlier today, it solves this problem.

I saw it. I'd rather kconfig be fixed instead, if possible.

kconfig was not broken at all in this case. It detected a circular dependency and did its work well.

( kconfig is broken in some areas - for example its misfeature of not
propagating selects along dependency chains is annoying. It should at minimum warn when it sees such partial selects. But this is not one of those breakages. )

Ingo

#152 lore raw
Quoted 54 lines · Ingo Molnar
Quoted 52 lines · Randy Dunlap
Quoted 51 lines · Ingo Molnar
Quoted 49 lines · Randy Dunlap
Quoted 34 lines · Eric Anholt
Quoted 33 lines · Ingo Molnar
Quoted 9 lines · Linus Torvalds
Quoted 4 lines · Norbert Preining

The problem is that if you have a configuration under 2.6.28 without CONFIG_FB and just call make oldconfig, or even make config and don't know that you loose the DRM. And I was using make oldconfig (there is a graphical config?? ;-))

Sure. It's inconvenient, no question about that. I asked the i915 people to look into not requiring CONFIG_FB, and I hope they will, but my point is that I don't think we can consider "small one-time inconvenience" to be a "regression".

if you mean that as a general principle, there's four very real downsides in my opinion.

Firstly, we could have done better (and still can do better), via various easy and non-intrusive measures:

- We could add a runtime warning:

for example a WARN_ONCE("please enable CONFIG_DRM_I915 and CONFIG_FB") that there's no DRM because CONFIG_FB is not selected and oldconfig loses the I915 setting silently - placed in a key DRM ioctl, would have gone a long way addressing the issue. Testers do notice kernel warnings that pop up when their X gets slow. (This approach might also have the added bonus of warning folks who enable the wrong driver for the hardware.)

- Or we could add a more thoughtful Kconfig migration:

Rename DRM_I915 to DRM_I915_FB [which it really is now], and keep DRM_I915 as a non-interactive migration helper: if set, it auto-selects both FB and DRM_I915_FB.

Quoted 8 lines
Quoted 4 lines

While CONFIG_FB is an interactive Kconfig option so a select can be dangerous to a correct dependency tree, it seems safe to do in this specific case because it seems to be a rather leaf entry with no dependencies.

I tried select FB. It's the right thing to do. It doesn't work. I posted to the mailing list two weeks ago about the insane dependency chain that kbuild comes up with and fails on when we do this, and got silence.

I tried what you had described in that email (from 2 weeks ago), got the same results that you did, but kbuild does seem very confused (to me).

reference email from 2+ weeks ago:
http://marc.info/?l=linux-kernel&m=123197341316461&w=2

Quoted 7 lines
Quoted 4 lines
Quoted 3 lines
Quoted 1 line

Adding Sam to cc.

Check the patch i posted in this thread earlier today, it solves this problem.

I saw it. I'd rather kconfig be fixed instead, if possible.

kconfig was not broken at all in this case. It detected a circular dependency and did its work well.

Maybe. I haven't seen an explanation for the problems that Eric reported 2+ weeks ago.

Quoted 4 lines

( kconfig is broken in some areas - for example its misfeature of not
propagating selects along dependency chains is annoying. It should at minimum warn when it sees such partial selects. But this is not one of those breakages. )

Signature · 1 line
~Randy
#153 lore raw
Quoted 5 lines · Randy Dunlap
Quoted 2 lines

kconfig was not broken at all in this case. It detected a circular dependency and did its work well.

Maybe. I haven't seen an explanation for the problems that Eric reported 2+ weeks ago.

The thing he reported 2 weeks ago is a straightforward recursive dependency bug in the config entries:

drivers/gpu/drm/Kconfig:8:error: found recursive dependency:
DRM -> <choice> -> DRM_I915 -> FB -> FB_I810 -> AGP -> DRM

this is the precise circular dependency problem that the I810_FB and INTEL_FB changes in my patch solve.

Give my patch a try :-)

Ingo

#154 lore raw
Quoted 8 lines · Ingo Molnar
Quoted 5 lines · Randy Dunlap
Quoted 2 lines

kconfig was not broken at all in this case. It detected a circular dependency and did its work well.

Maybe. I haven't seen an explanation for the problems that Eric reported 2+ weeks ago.

Quoted 8 lines

The thing he reported 2 weeks ago is a straightforward recursive dependency bug in the config entries:

drivers/gpu/drm/Kconfig:8:error: found recursive dependency:

   DRM -> <choice> -> DRM_I915 -> FB -> FB_I810 -> AGP -> DRM

this is the precise circular dependency problem that the I810_FB and INTEL_FB changes in my patch solve.

I think the problem is that when looking at
select FB
or
depends on FB
I see "this thing depends on FB. If I said select, then instead of asking the user to enable it before seeing the option, just enable the option for them". Even after seeing your patch, I still don't see why the select means that AGP depends on DRM.

But big thanks for fixing it!

Signature · 2 lines
Eric Anholt
[email protected]                         [email protected]
#155 lore raw
In reply to Linus Torvalds
Quoted 9 lines · Linus Torvalds
Quoted 6 lines · Rafael J. Wysocki

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12599
Subject : dri /dev node disappeared with 2.6.29-rc1
Submitter : Norbert Preining <[email protected]>
Date : 2009-01-15 13:42 (21 days old)
References : http://marc.info/?l=linux-kernel&m=123202701026647&w=4

Quoted 7 lines

I think this one is simply that Intel DRI now requires CONFIG_FB.

So make sure that you say yes to the CONFIG_FB question (easiest fix: just edit the .config file and change the line "# CONFIG_FB is not set" to "CONFIG_FB=y" and do "make oldconfig" - I realize that people think that graphical config front-ends are simpler, but often just editing the file is the quickest way).

Well, I always have doubts whether to list such things or not and I usually end up deciding that listing is probably better than not listing.

I think I'll close this one, it's got enough attention IMO.

Quoted 8 lines
Quoted 6 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12591
Subject : NULL pointer dereference in blk_queue_io_stat
Submitter : Petr Vandrovec <[email protected]>
Date : 2009-01-31 23:58 (5 days old)
First-Bad-Commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=bc58ba9468d94d62c56ab9b47173583ec140b165

I think this one is fixed by commit fb8ec18c316d8.

Closed.

Quoted 10 lines
Quoted 6 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12511
Subject : WARNING: at drivers/dma/dmaengine.c:352
Submitter : Ingo Molnar <[email protected]>
Date : 2009-01-19 21:31 (17 days old)
References : http://marc.info/?l=linux-kernel&m=123240070614443&w=4

This warning just got removed as bogus.

Commit 83436a0560e9ef8af2f0796264dde4bed1415359

Closed.

Quoted 7 lines
Quoted 5 lines

Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12506
Subject : Undefined symbols when CONFIG_MFD_PCF50633 is enabled
Submitter : Ozan Çağlayan <[email protected]>
Date : 2009-01-17 10:56 (19 days old)
References : http://marc.info/?l=linux-kernel&m=123218991104592&w=4

Commit 9e6f8ed7c3a303d37eb119847dd3029701e37e28

Closed.

Thanks,
Rafael