The patch titled
Subject: revert mm-do-not-stall-in-synchronous-compaction-for-thp-allocations
has been removed from the -mm tree. Its filename was
revert-mm-do-not-stall-in-synchronous-compaction-for-thp-allocations.patch
This patch was dropped because an updated version will be merged
The current -mm tree may be found at http://userweb.kernel.org/~akpm/mmotm/
------------------------------------------------------
From: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
Subject: revert mm-do-not-stall-in-synchronous-compaction-for-thp-allocations
Take this away again to amke way for "Reduce compaction-related stalls and
improve asynchronous migration of dirty pages v6".
But mm-do-not-stall-in-synchronous-compaction-for-thp-allocations migth
still be appl;icable for earlier kernel versions.
Cc: Mel Gorman <mgorman@xxxxxxx>
Cc: Andy Isaacson <adi@xxxxxxxxxxxxx>
Cc: Johannes Weiner <hannes@xxxxxxxxxxx>
Cc: Johannes Weiner <hannes@xxxxxxxxxxx>
Cc: Andrea Arcangeli <aarcange@xxxxxxxxxx>
Cc: Alan Cox <alan@xxxxxxxxxxxxxxxxxxx>
Cc: <stable@xxxxxxxxxxxxxxx>
Cc: Minchan Kim <minchan.kim@xxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
include/linux/gfp.h | 11 -----------
mm/page_alloc.c | 9 +--------
2 files changed, 1 insertion(+), 19 deletions(-)
diff -puN include/linux/gfp.h~revert-mm-do-not-stall-in-synchronous-compaction-for-thp-allocations include/linux/gfp.h
--- a/include/linux/gfp.h~revert-mm-do-not-stall-in-synchronous-compaction-for-thp-allocations
+++ a/include/linux/gfp.h
@@ -84,18 +84,7 @@ struct vm_area_struct;
#define __GFP_RECLAIMABLE ((__force gfp_t)___GFP_RECLAIMABLE) /* Page is reclaimable */
#define __GFP_NOTRACK ((__force gfp_t)___GFP_NOTRACK) /* Don't track with kmemcheck */
-/*
- * __GFP_NO_KSWAPD indicates that the VM should favour failing the allocation
- * over excessive disruption of the system. Currently this means
- * 1. Do not wake kswapd (hence the flag name)
- * 2. Do not use stall in synchronous compaction for high-order allocations
- * as this may cause the caller to stall writing out pages
- *
- * This flag it primarily intended for use with transparent hugepage support.
- * If the flag is used outside the VM, linux-mm should be cc'd for review.
- */
#define __GFP_NO_KSWAPD ((__force gfp_t)___GFP_NO_KSWAPD)
-
#define __GFP_OTHER_NODE ((__force gfp_t)___GFP_OTHER_NODE) /* On behalf of other node */
#define __GFP_WRITE ((__force gfp_t)___GFP_WRITE) /* Allocator intends to dirty page */
diff -puN mm/page_alloc.c~revert-mm-do-not-stall-in-synchronous-compaction-for-thp-allocations mm/page_alloc.c
--- a/mm/page_alloc.c~revert-mm-do-not-stall-in-synchronous-compaction-for-thp-allocations
+++ a/mm/page_alloc.c
@@ -2285,14 +2285,7 @@ rebalance:
sync_migration);
if (page)
goto got_pg;
-
- /*
- * Do not use sync migration if __GFP_NO_KSWAPD is used to indicate
- * the system should not be heavily disrupted. In practice, this is
- * to avoid THP callers being stalled in writeback during migration
- * as it's preferable for the the allocations to fail than to stall
- */
- sync_migration = !(gfp_mask & __GFP_NO_KSWAPD);
+ sync_migration = true;
/* Try direct reclaim and then allocating */
page = __alloc_pages_direct_reclaim(gfp_mask, order,
_
Patches currently in -mm which might be from akpm@xxxxxxxxxxxxxxxxxxxx are
origin.patch
linux-next.patch
linux-next-git-rejects.patch
i-need-old-gcc.patch
arch-alpha-kernel-systblss-remove-debug-check.patch
arch-x86-platform-iris-irisc-register-a-platform-device-and-a-platform-driver.patch
x86-olpc-xo15-sci-enable-lid-close-wakeup-control-through-sysfs-fix.patch
mm-vmallocc-eliminate-extra-loop-in-pcpu_get_vm_areas-error-path-fix.patch
drivers-platform-x86-sony-laptopc-fix-scancodes-checkpatch-fixes.patch
drivers-platform-x86-sony-laptopc-fix-scancodes-v2-checkpatch-fixes.patch
slub-document-setting-min-order-with-debug_guardpage_minorder-0-checkpatch-fixes.patch
mm.patch
mm-add-extra-free-kbytes-tunable-update-checkpatch-fixes.patch
mm-reduce-the-amount-of-work-done-when-updating-min_free_kbytes-checkpatch-fixes.patch
mm-hugetlbc-fix-virtual-address-handling-in-hugetlb-fault-fix.patch
mm-more-intensive-memory-corruption-debug-fix.patch
mm-exclude-reserved-pages-from-dirtyable-memory-fix.patch
mm-simplify-find_vma_prev-fix.patch
frv-duplicate-output_buffer-of-e03-checkpatch-fixes.patch
hpet-factor-timer-allocate-from-open.patch
treewide-convert-uses-of-attrib_noreturn-to-__noreturn-checkpatch-fixes.patch
leds-add-driver-for-tca6507-led-controller-fix.patch
leds-add-driver-for-tca6507-led-controller-checkpatch-fixes.patch
checkpatch-improve-memset-and-min-max-with-cast-checking-fix.patch
checkpatch-catch-all-occurances-of-type-and-cast-spacing-errors-per-line-checkpatch-fixes.patch
drivers-rtc-rtc-mxcc-fix-setting-time-for-mx1-soc-fix.patch
drivers-rtc-rtc-mxcc-make-alarm-work-fix.patch
rtc-ab8500-add-calibration-attribute-to-ab8500-rtc-checkpatch-fixes.patch
rtc-ab8500-add-calibration-attribute-to-ab8500-rtc-v3-checkpatch-fixes.patch
mm-vmscan-distinguish-between-memcg-triggering-reclaim-and-memcg-being-scanned-checkpatch-fixes.patch
memcg-make-mem_cgroup_split_huge_fixup-more-efficient-fix.patch
memcg-clear-pc-mem_cgorup-if-necessary-fix.patch
procfs-introduce-the-proc-pid-map_files-directory-checkpatch-fixes.patch
workqueue-make-alloc_workqueue-take-printf-fmt-and-args-for-name-fix.patch
kexec-remove-kmsg_dump_kexec.patch
panic-dont-print-redundant-backtraces-on-oops-fix.patch
scatterlist-new-helper-functions.patch
scatterlist-new-helper-functions-update-fix.patch
memstick-add-support-for-legacy-memorysticks-fix.patch
dio-optimize-cache-misses-in-the-submission-path-v2-checkpatch-fixes.patch
selftests-new-very-basic-kernel-selftests-directory.patch
c-r-procfs-add-start_data-end_data-start_brk-members-to-proc-pid-stat-v4-fix.patch
c-r-prctl-add-pr_set_mm-codes-to-set-up-mm_struct-entries-fix.patch
fixed-use-of-rounddown_pow_of_two-in-ramoops.patch
notify_change-check-that-i_mutex-is-held.patch
journal_add_journal_head-debug.patch
mutex-subsystem-synchro-test-module-fix.patch
slab-leaks3-default-y.patch
put_bh-debug.patch
--
To unsubscribe from this list: send the line "unsubscribe mm-commits" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
