Skip to contentSkip to conversation

patches added to release/test trees

DISCUSSION Luck, Tony 4 participants 18 messages Jan 27, 2006 – Jun 27, 2006

lore
Share & tools
lore

Luck, Tony

author
#0 lore raw

The following patches have been added to the release tree (destined to go into 2.6.16 when Linus gets back from linux.conf.au):

Dean Roe:
[IA64-SGI] add sn_feature_sets bit

Greg Edwards:
[IA64] sn2 maintainer update (Jes Sorensen)

Jack Steiner:
[IA64] Scaling fix for simultaneous unaligned accesses

Jes Sorensen:
[IA64-SGI] XPC remove unnecessary GFP_DMA flag

Keith Owens:
[IA64] Set the correct default OS status in the MCA handler

Prarit Bhargava:
[IA64-SGI] Add PROM feature set for device flush list

Takashi Iwai:
[IA64-SGI] sn_dma_alloc_coherent should use gfp flags

All of the above and these patches have been added to the test tree (destined for some -mm releases, and then 2.6.17):

Aaron Young:
[IA64-SGI] Handle SC env. powerdown events

Brent Casavant:
[IA64] hooks to wait for mmio writes to drain when migrating processes

Chen, Kenneth W:
[IA64] implement ia64 specific mutex primitives

Jack Steiner:
[IA64-SGI] Update TLB flushing code for SN platform

Jes Sorensen:
[IA64-SGI] sem2mutex ioc4.c
drivers/sn/ must be entered for CONFIG_SGI_IOC3

Keith Owens:
[IA64-SGI] Recursive flags do not work for selective builds [IA64] Delete MCA/INIT sigdelayed code

Mark Maule:
[IA64-SGI] driver bugfixes and hardware workarounds for CE1.0 asic

author
#1 lore raw

The following patches have been added to the release tree:

Bjorn Helgaas:
[IA64] avoid broken SAL_CACHE_FLUSH implementations

Chen, Kenneth W:
[IA64] remove staled comments in asm/system.h

Jack Steiner:
[IA64-SGI] Update TLB flushing code for SN platform (pulled from test)

Jes Sorensen:
drivers/sn/ must be entered for CONFIG_SGI_IOC3 (pulled from test) [IA64-SGI] sn2 housekeeping
[IA64-SGI] include/asm-ia64/sn/intr.h more sn2 housekeeping

Kyle McMartin:
[IA64] Remove stale comment from ia64/Kconfig

Mark Maule:
[IA64-SGI] fix smp_affinity redirection when using CONFIG_PCI_MSI [IA64-SGI] disable msi for all altix pci devices

Robin Holt:
[IA64-SGI] Fix XPC code which sleeps with spin_lock_irqsave().

author
#2 lore raw
In reply to Luck, Tony

The following patches have been added to the release/test trees:

Jes Sorensen:
[IA64] prevent sn2 specific code to be run in generic kernels

Tony Luck:
[IA64] Fix CONFIG_PRINTK_TIME

Zou Nan hai:
[IA64] Fix a possible buffer overflow in efi.c [IA64] Fix wrong use of memparse in efi.c

The following patches have been added to the test tree:

Chen, Kenneth W:
[IA64] map ia64_hint definition to intel compiler intrinsic [IA64] clean up asm/intel_intrin.h
[IA64] add __builtin_trap definition for icc build [IA64] use icc defined constant

Jack Steiner:
[IA64-SGI] - Eliminate SN pio_phys_xxx macros. Move to assembly

Keith Owens:
[IA64] MCA: print messages in MCA handler [IA64] MCA: update MCA comm field for user space tasks [IA64] MCA: remove obsolete ifdef

author
#3 lore raw
In reply to Luck, Tony

The following patches have been added to my release/test trees:

Bjorn Helgaas:
[IA64] update HP CSR space discovery via ACPI

KAMEZAWA Hiroyuki:

      [IA64] for_each_possible_cpu: ia64

Tony Luck:
[IA64] Wire up new syscall sync_file_range()
[IA64] 'msg' may be used uninitialized in xpc_initiate_allocate() [IA64] Wire up new syscalls {set,get}_robust_list

author
#4 lore raw
In reply to Luck, Tony

I just did a quick skim through my backlog of patches and pushed all the "obviously correct" (famous last words) ones into the release tree. I'll send them to Linus on Friday (unless I hear about any problems). Remind me about any other "obvious" stuff that my eyes skipped over.

The following patches have been added to the release tree:

Alex Williamson:
[IA64] add vmlinuz target
[IA64] sanity check reserved region usage

Bjorn Helgaas:
[IA64] rework memory attribute aliasing

David Mosberger-Tang:
[IA64] make efi_stub.S fit in 80 cols

Ian Wienand:
[IA64] SKI Simulator boot

Jack Steiner:
[IA64-SGI] SN topology fix for large systems

Keith Owens:
[IA64] Sanitize assembler code for ia64_sal_os_state

Kenji Kaneshige:
[IA64] Make PCI Express support selectable

Mike Habeck:
[IA64-SGI] fix SGI Altix tioce_bus_fixup() bug

Russ Anderson:
[IA64-SGI] Remove SN SAL error handling feature bit that is no longer needed

Additionally the following patches have been added to the test tree. These will probably go to Linus before we get to -rc1.

David Mosberger-Tang:

      [IA64] esi-support

Tony Luck:
[IA64] Add "model name" to /proc/cpuinfo

Horms and Ian. Thanks for all the kexec patches you've posted. I'll be folding them in on top of the current kexec stuff soon. There still seem to be a lot of open issues on this, and hardly any reports of any testing going on. At this pace kexec/kdump will be sitting in "test" (and -mm) for at least another release cycle.

#5 lore raw
In reply to Luck, Tony
Quoted 3 lines · "Luck, Tony" (on Wed, 21 Jun 2006 16:19:58 -0700) wrote:

I just did a quick skim through my backlog of patches and pushed all the "obviously correct" (famous last words) ones into the release tree.

I just did a git-fetch on git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6.git
and the log for the release branch shows no changes to arch/ia64 since
May 16. The Makefile is for 2.6.17-rc5. Is there a propogation delay
on your git updates before they hit git.kernel.org?

BTW, is there any reason that the aegl git tree has no tags?

#6 lore raw
In reply to Luck, Tony
Quoted 9 lines · Keith Owens (on Thu, 22 Jun 2006 10:33:24 +1000) wrote:
Quoted 3 lines · "Luck, Tony" (on Wed, 21 Jun 2006 16:19:58 -0700) wrote:

I just did a quick skim through my backlog of patches and pushed all the "obviously correct" (famous last words) ones into the release tree.

I just did a git-fetch on git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6.git
and the log for the release branch shows no changes to arch/ia64 since
May 16. The Makefile is for 2.6.17-rc5. Is there a propogation delay
on your git updates before they hit git.kernel.org?

Never mind, a later git-fetch got the data. Must be a propogation delay between master.kernel.org and git.kernel.org.

#7 lore raw
In reply to Luck, Tony
Quoted 6 lines · In article <[email protected]> you wrote:

Horms and Ian. Thanks for all the kexec patches you've posted. I'll be folding them in on top of the current kexec stuff soon. There still seem to be a lot of open issues on this, and hardly any reports of any testing going on. At this pace kexec/kdump will be sitting in "test" (and -mm) for at least another release cycle.

Hi Tony,

thanks for the feedback. I understand that kexec/kdump is quite green/flakey and having it sit in test as you suggest is quite approtiate in my oppinion too. Hopefully its quality can improve over time.

As for the patches that I have sent, let me know if you need any of them rediffed or reworked. I'm more than happy to do that kind of leg work, as some of the patches conflict in minor ways, and some of them are against the kdump patch before it went into your tree, and thus may need some minor merging.

Thanks

Signature · 1 line
Horms                                           http://www.vergenet.net/~horms/
author
#8 lore raw
In reply to Luck, Tony
Quoted 2 lines

May 16. The Makefile is for 2.6.17-rc5. Is there a propogation delay on your git updates before they hit git.kernel.org?

Yes, there is a delay. I update to a backend system at kernel.org which regularly rsyncs the data to the two front-end systems: {www1,www2}.kernel.org The delay varies depending on load on the systems, and the overall internet weather. It can be as short as a couple of minutes. When it goes above a couple of hours it indicates there is a problem. The updates to the two front-end systems are independent ... you can force access to one or the other by using the www1, www2 names to see whether one managed to get the update.

Quoted 1 line

BTW, is there any reason that the aegl git tree has no tags?

I haven't thought of any useful way to use tags. But perhaps it's becuase I mostly think of my GIT trees as conduits to Linus/Andrew. If you have suggestions on how I can make my trees more useful to you by tagging things, then let me know.

-Tony

#9 lore raw
In reply to Luck, Tony
Quoted 6 lines · "Luck, Tony" (on Thu, 22 Jun 2006 10:34:06 -0700) wrote:
Quoted 1 line

BTW, is there any reason that the aegl git tree has no tags?

I haven't thought of any useful way to use tags. But perhaps it's becuase I mostly think of my GIT trees as conduits to Linus/Andrew. If you have suggestions on how I can make my trees more useful to you by tagging things, then let me know.

Not adding your own tags, but Linus's tags are missing from the ia64 tree. For some reason, just copying .git/refs/tags from the torvalds git tree does not work, git-branch complains about invalid objects. I thought that tags were portable between trees with common ancestors.

git-clone .../git/torvalds/linux-2.6.git linux-2.6 git-clone .../git/aegl/linux-2.6.git linux-2.6-ia64
cp -a linux-2.6/.git/refs/tags/* linux-2.6-ia64/.git/refs/tags/ cd linux-2.6-ia64
git-branch
error: refs/tags/v2.6.11 does not point to a valid commit object!
error: refs/tags/v2.6.11-tree does not point to a valid commit object!
...
error: refs/tags/v2.6.17-rc5 does not point to a valid commit object!
error: refs/tags/v2.6.17-rc6 does not point to a valid commit object!
origin
* release
test

#10 lore raw
In reply to Luck, Tony
Quoted 4 lines · Keith Owens

Not adding your own tags, but Linus's tags are missing from the ia64 tree. For some reason, just copying .git/refs/tags from the torvalds git tree does not work, git-branch complains about invalid objects. I thought that tags were portable between trees with common ancestors.

You have only copied the references, but not the objects itself. The easiest way to do that is to point objects/info/alternates to your copy of Linus' object store.

Andreas.

Signature · 4 lines
Andreas Schwab, SuSE Labs, [email protected]
SuSE Linux Products GmbH, Maxfeldstraße 5, 90409 Nürnberg, Germany
PGP key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."
#11 lore raw
In reply to Luck, Tony
Quoted 10 lines · Andreas Schwab (on Fri, 23 Jun 2006 02:11:55 +0200) wrote:
Quoted 4 lines · Keith Owens

Not adding your own tags, but Linus's tags are missing from the ia64 tree. For some reason, just copying .git/refs/tags from the torvalds git tree does not work, git-branch complains about invalid objects. I thought that tags were portable between trees with common ancestors.

You have only copied the references, but not the objects itself. The easiest way to do that is to point objects/info/alternates to your copy of Linus' object store.

That makes sense. But why is the ia64 git tree missing the Linus objects and tags in the first place?

author
#12 lore raw
In reply to Luck, Tony
Quoted 2 lines

That makes sense. But why is the ia64 git tree missing the Linus objects and tags in the first place?

"tags" is a bit ambiguous here, there are two possible things that you might be talking about when you say "tag" w.r.t. a GIT tree.

1) A file in .git/refs/tags/ that contains the SHA1 of a git object.

2) A git object of type "tag".

The files in .git/refs/tags refer to GIT tag objects. Usually a tag object is a pointer to a commit object ... but they can point at anything (the v2.6.11 tag object points to a "tree").

Things get more complex from here. My kernel.org tree contains none of the .git/refs/tags files (because I "git push" the heads of the release/test branches up from my local tree, anf this won't copy any tags files up ... even though they are all in my local tree from my "git fetch linus" pulls from Linus).

The git tag objects actually are in my kernel.org tree! But only because I share packfiles with Linus to reduce load on kernel.org mirroring system.

You don't get to see them though. When you do a fetch/pull from my tree (using "git://" transfer method), git will construct the list of commit/tree/blob objects, build a packfile, and send it to you. Any tag objects in my (really Linus') packfile will just be ignored. If you used "rsync://", it would copy the whole packfile. Not sure what an http:// pull would do.

Is that as clear as mud now?

-Tony

#13 lore raw
In reply to Luck, Tony
Quoted 7 lines · "Luck, Tony" (on Fri, 23 Jun 2006 10:08:22 -0700) wrote:
Quoted 2 lines

That makes sense. But why is the ia64 git tree missing the Linus objects and tags in the first place?

"tags" is a bit ambiguous here, there are two possible things that you might be talking about when you say "tag" w.r.t. a GIT tree.

1) A file in .git/refs/tags/ that contains the SHA1 of a git object.

Them.

Quoted 11 lines

2) A git object of type "tag".

The files in .git/refs/tags refer to GIT tag objects. Usually a tag object is a pointer to a commit object ... but they can point at anything (the v2.6.11 tag object points to a "tree").

Things get more complex from here. My kernel.org tree contains none of the .git/refs/tags files (because I "git push" the heads of the release/test branches up from my local tree, anf this won't copy any tags files up ... even though they are all in my local tree from my "git fetch linus" pulls from Linus).

Could you push the refs/tags from your tree as well? Many of the git trees on kernel.org have copies of Linus's tags, which makes it a bit easier to cross reference.

author
#14 lore raw
In reply to Luck, Tony
Quoted 3 lines

Could you push the refs/tags from your tree as well? Many of the git trees on kernel.org have copies of Linus's tags, which makes it a bit easier to cross reference.

I linked his tags directory to my tree ... lets see how that works.

-Tony

#15 lore raw
In reply to Luck, Tony
Quoted 5 lines · In article <617E1C2C70743745A92448908E030B2A22FFE9@scsmsx411.amr.corp.intel.com…
Quoted 3 lines

Could you push the refs/tags from your tree as well? Many of the git trees on kernel.org have copies of Linus's tags, which makes it a bit easier to cross reference.

I linked his tags directory to my tree ... lets see how that works.

I noticed (to my surprise) a bunch of tags show up while using gitweb this morning. So I guess its working, at least to some extent.

Signature · 2 lines
Horms                                           
H: http://www.vergenet.net/~horms/          W: http://www.valinux.co.jp/en/
author
#16 lore raw
In reply to Luck, Tony
Quoted 4 lines
Quoted 1 line

I linked his tags directory to my tree ... lets see how that works.

I noticed (to my surprise) a bunch of tags show up while using gitweb this morning. So I guess its working, at least to some extent.

The only bit that I'm worried about is what will happen when Linus adds a new tag. Until I do a pull&push to catch up, I'll have a "bogus" tag (pointing to an object that I don't have). Whether this is an annoyance, or a major problem, remains to be seen.

The 2.6.17 tag is now 8 days old, so if Linus sticks to the 2-week merge window, we'll find out in ~6days when 2.6.18-rc1 comes out.

-Tony

#17 lore raw
In reply to Luck, Tony
Quoted 10 lines · "Luck, Tony" (on Mon, 26 Jun 2006 10:23:19 -0700) wrote:
Quoted 4 lines
Quoted 1 line

I linked his tags directory to my tree ... lets see how that works.

I noticed (to my surprise) a bunch of tags show up while using gitweb this morning. So I guess its working, at least to some extent.

The only bit that I'm worried about is what will happen when Linus adds a new tag. Until I do a pull&push to catch up, I'll have a "bogus" tag (pointing to an object that I don't have). Whether this is an annoyance, or a major problem, remains to be seen.

Annoyance only. git-branch will list it as a dangling tag but that is all.