Can you include the x86_64.org patches? I will attach them. I have applied them to the current gentoo-dev-sources and they apply cleaning in the proper order. Reproducible: Always Steps to Reproduce: 1. 2. 3.
Created attachment 33614 [details] x86_64-2.6.7-1.bz2
Comment on attachment 33614 [details] x86_64-2.6.7-1.bz2 First x86-64.org patchset
Created attachment 33615 [details] x86_64-2.6.7-2.patch Second x86-64.org patch for 2.6.7
Created attachment 33736 [details] x86_64-2.6.7-2.bz2 Newer patchset, from the x86-64.org mailing list: A new x86-64 patchkit for the Linux v2.6 has been released. ftp://ftp.x86-64.org/pub/linux/v2.6/x86_64-2.6.7-2.bz2 03db8fef878268d0e905c158f132db56 x86_64-2.6.7-2.bz2 This version has some more machine check handler fixes (thanks to Eric Morton and Paul Devriendt for suggestions). The machine check handler should work well now. Please don't forget to run mcelog from your crontab to log its output. It doesn't use backwards rep ; movsb for memmove anymore to work around a new (extremly unlikely) Opteron erratum. A new lazy merging scheme suggested by Ben Herrenschmidt has been implemented in the IOMMU. IOMMU level merging can give improved performance with some IO devices (in particular SCSI adapters). Previously the merging implementation would tell the block layer to assume merging for everything. Problem was that when the IOMMU was unable to merge later for some reason there was no way to undo this without erroring out the request. This made it not really suitable for general usage. The new lazy merging scheme doesn't involve the block layer anymore, but just merges at driver level. This is less efficient, but more reliable. This is enabled by default now when the IOMMU is in use. If you have a block IO limited load it may be interesting to try iommu=force with the new kernel. For networking it seems to make things slower. If you do this it's suggest to enlarge the AGP aperture or IOMMU option in the BIOS (if you don't have AGP) to 256-512MB or more. If you see any improvements from this please report back. I changed the bitmap search functions to use 64 bit accesses. This may improve performance on file systems using bitmaps (like ext3/3, reiserfs) Also this release has various other fixes, see the ChangeLog for details. -Andi
I'll be glad to include x86-64 specific bug fixes, but to just drop andi's whole patchset into this tree isn't always a good thing to have. So, I'm going to reject this, and if you have any specific bug fixes you want to pick out of the big patch, feel free to reopen this, and add just those patches.
Well, the patches aren't broken out, so it's kinda hard to see which relate to which, but I would definitely like to have the 64-bit accessed bitmaps that could potential increase filesystem performance. The machine check handler stuff would be nice, but it looks like a fair number of files were touched to patch that. Don't care about IOMMU as I don't have >4GB of RAM ;)
"Nice to have" is not going to work, sorry :) We need bugfix only patches, in order to stay sane and to take advantage of the kernel.org developers.
wait... when did you guys stop including the x86-64 patches? from 2.6.1: * Applying 125_x86_64_org_patches_2.6.1_rc3-brad.patch... [ ok ] the reason why gentoo-dev-sources is the suggested kernel in the amd64 docs is because of this patchset... it's not so much of an issue anymore, but back when development-sources didnt work at all, g-d-s did. if this isnt in g-d-s anymore, do you think it might be a good idea to have an x86-64-sources? it might be useful to not have 'dev' in the name of the only kernel considered stable on amd64 :) perhaps amd64-sources so that we can also include the pax patch by default, allowing for easier control of amd64's native NX support?
The Machine Check Handler is a bugfix no? Either it works or it doesn't. Previously didn't work, patch makes it work. *shrugs* I can just apply the x86-64.org patches by hand.
Ok, as the patchset does not touch any non x86-64 files, and this is the stable kernel for that arch, I've added it back to the package, and released a 2.6.7-r4 version for it. Sorry about the confusion.
you -rock- :) thanks much for the re-add.
