Date: Fri, 10 Aug 2001 22:49:31 +0200
From: Andrea Arcangeli <firstname.lastname@example.org>
Only in 2.4.8pre8aa1: 00_alpha-iommu_fixup-1
Really reserve the last mbytes of 32bit pci space for cypress magics.
Only in 2.4.8pre8aa1: 00_alpha-sg_fill-1
When splitting one virtual consecutive sg slot into many physically
consecutive slots update the dma_length of ex-leader and ex-mergers
Only in 2.4.8pre8aa1: 00_debug-vm-lru-mm-corruption-1
Add debugging code, during heavy load a BUG() triggered in DEBUG_ADD_PAGE
in page_launder while removing a buffer from the inactive_dirty,
despite we cleared the inactive_dirty bitflag from the page->flags
some other bitflag was still set, this could be random memory corruption
but it could also be a race condition in the vm. Unfortunately the
printk in BUG() clobbered %eax so I couldn't see the contents of page->flags
from the bugreport so now I'm printing page->flags explicitly before
BUG() so next time we'll see if page->flags was just random or in which
other list the page was supposed to be too.
Only in 2.4.8pre8aa1: 00_drm_vm-1
Never use page->virtual directly, always use page_address(page)
to get that information, so we keep saving ram when highmem is
not selected and on 64bit archs in general.
Only in 2.4.8pre8aa1: 00_module-oops-1
Andi extracted the module_oops code from -ac, ready to be applied
Only in 2.4.8pre6aa1: 40_blkdev-pagecache-10
Only in 2.4.8pre8aa1: 40_blkdev-pagecache-11
Fixed trivial rejects, fsync_no_super lines.
Only in 2.4.8pre6aa1: 50_uml-2.4.7-3.bz2
Only in 2.4.8pre8aa1: 50_uml-2.4.7-5.bz2
Picked new -5 revision from sourceforge.
Some of the products that appear on this site are from companies from which QuinStreet receives compensation. This compensation may impact how and where products appear on this site including, for example, the order in which they appear. QuinStreet does not include all companies or all types of products available in the marketplace.