mm: get rid of __alloc_pages_high_priority
Subsystems (2)
PATCH 1/1 mm: get rid of __alloc_pages_high_priority
From: Michal Hocko <[email protected]>
__alloc_pages_high_priority doesn't do anything special other than it calls get_page_from_freelist and loops around GFP_NOFAIL allocation until it succeeds. It would be better if the first part was done in __alloc_pages_slowpath where we modify the zonelist because this would be easier to read and understand. And do the retry at the very same place because retrying without even attempting to do any reclaim is fragile because we are basically relying on somebody else to make the reclaim (be it the direct reclaim or OOM killer) for us. The caller might be holding resources (e.g. locks) which block other other reclaimers from making any progress for example.
Remove the helper and open code it into its only user. We have to be careful about __GFP_NOFAIL allocations from the PF_MEMALLOC context even though this is a very bad idea to begin with because no progress can be gurateed at all. We shouldn't break the __GFP_NOFAIL semantic here though. It could be argued that this is essentially GFP_NOWAIT context which we do not support but PF_MEMALLOC is much harder to check for existing users because they might happen deep down the code path performed much later after setting the flag so we cannot really rule out there is no kernel path triggering this combination.
---
Hi,
I think that this is more a cleanup than any functional change. We are rarely screwed so much that __alloc_pages_high_priority would fail. Yet I think that __alloc_pages_high_priority is obscuring the overal intention more than it is helpful. Another motivation is to reduce wait_iff_congested call to a single one in the allocator. I plan to do other changes in that area and get rid of it altogether.
mm/page_alloc.c | 48 +++++++++++++++++------------------------------- 1 file changed, 17 insertions(+), 31 deletions(-)
mm/page_alloc.c +17 -31
@@ -2902,28 +2902,6 @@ __alloc_pages_direct_reclaim(gfp_t gfp_mask, unsigned int order,return page; } -/* -* This is called in the allocator slow-path if the allocation request is of -* sufficient urgency to ignore watermarks and take other desperate measures -*/ -static inline struct page * -__alloc_pages_high_priority (gfp_t gfp_mask,unsigned int order, -const struct alloc_context *ac) -{ -struct page *page; - -do { - page =get_page_from_freelist (gfp_mask, order, -ALLOC_NO_WATERMARKS , ac); - -if (!page && gfp_mask & __GFP_NOFAIL) -wait_iff_congested (ac->preferred_zone,BLK_RW_ASYNC , -HZ /50 ); - }while (!page && (gfp_mask & __GFP_NOFAIL)); - -return page; -} -static void wake_all_kswapds (unsigned int order,const struct alloc_context *ac) {struct zoneref *z;
@@ -3068,28 +3046,36 @@ __alloc_pages_slowpath(gfp_t gfp_mask, unsigned int order,* allocations are system rather than user orientated */ ac->zonelist =node_zonelist (numa_node_id (), gfp_mask); - - page =__alloc_pages_high_priority (gfp_mask, order, ac); - -if (page) { + page =get_page_from_freelist (gfp_mask, order, +ALLOC_NO_WATERMARKS , ac); +if (page) goto got_pg; - } }/* Caller is not willing to reclaim, we can't balance anything */ if (!can_direct_reclaim) {/* -* All existing users of the deprecated __GFP_NOFAIL are -* blockable, so warn of any new users that actually allow this -* type of allocation to fail. +* All existing users of the __GFP_NOFAIL are blockable, so warn +* of any new users that actually allow this type of allocation +* to fail. */ WARN_ON_ONCE (gfp_mask & __GFP_NOFAIL); goto nopage; }/* Avoid recursion of direct reclaim */ -if (current->flags &PF_MEMALLOC ) +if (current->flags &PF_MEMALLOC ) { +/* +* __GFP_NOFAIL request from this context is rather bizarre +* because we cannot reclaim anything and only can loop waiting +* for somebody to do a work for us. +*/ +if (WARN_ON_ONCE (gfp_mask & __GFP_NOFAIL)) { +cond_resched (); + goto retry; + } goto nopage; + }/* Avoid allocations with no watermarks from looping endlessly */ if (test_thread_flag (TIF_MEMDIE ) && !(gfp_mask & __GFP_NOFAIL))
Signature · 7 lines
2.6.2 -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to [email protected]. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"[email protected]"> [email protected] </a>
Quoted 26 lines · On Thu, 12 Nov 2015, [email protected] wrote:
From: Michal Hocko <[email protected]>
__alloc_pages_high_prioritydoesn't do anything special other than it callsget_page_from_freelistand loops aroundGFP_NOFAILallocation until it succeeds. It would be better if the first part was done in__alloc_pages_slowpathwhere we modify the zonelist because this would be easier to read and understand. And do the retry at the very same place because retrying without even attempting to do any reclaim is fragile because we are basically relying on somebody else to make the reclaim (be it the direct reclaim or OOM killer) for us. The caller might be holding resources (e.g. locks) which block other other reclaimers from making any progress for example.Remove the helper and open code it into its only user. We have to be careful about
__GFP_NOFAILallocations from thePF_MEMALLOCcontext even though this is a very bad idea to begin with because no progress can be gurateed at all. We shouldn't break the__GFP_NOFAILsemantic here though. It could be argued that this is essentiallyGFP_NOWAITcontext which we do not support butPF_MEMALLOCis much harder to check for existing users because they might happen deep down the code path performed much later after setting the flag so we cannot really rule out there is no kernel path triggering this combination.Signed-off-by Michal Hocko <[email protected]>---
Quoted 7 lines
Hi,
I think that this is more a cleanup than any functional change. We are rarely screwed so much that__alloc_pages_high_prioritywould fail. Yet I think that__alloc_pages_high_priorityis obscuring the overal intention more than it is helpful. Another motivation is to reducewait_iff_congestedcall to a single one in the allocator. I plan to do other changes in that area and get rid of it altogether.
I think it's a combination of a cleanup (the inlining of __alloc_pages_high_priority) and a functional change (no longer looping infinitely around a get_page_from_freelist() call). I'd suggest doing the inlining in one patch and then the reworking of __GFP_NOFAIL when ALLOC_NO_WATERMARKS fails just so we could easily revert the latter if necessary.
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to [email protected]. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"[email protected]"> [email protected] </a>
Quoted 1 line · David Rientjes
On Thu, 12 Nov 2015, [email protected] wrote:
[...]
Quoted 8 lines
Quoted 7 lines
Hi,
I think that this is more a cleanup than any functional change. We are rarely screwed so much that__alloc_pages_high_prioritywould fail. Yet I think that__alloc_pages_high_priorityis obscuring the overal intention more than it is helpful. Another motivation is to reducewait_iff_congestedcall to a single one in the allocator. I plan to do other changes in that area and get rid of it altogether.
Quoted 6 lines
I think it's a combination of a cleanup (the inlining of
__alloc_pages_high_priority) and a functional change (no longer looping infinitely around aget_page_from_freelist()call). I'd suggest doing the inlining in one patch and then the reworking of__GFP_NOFAILwhenALLOC_NO_WATERMARKSfails just so we could easily revert the latter if necessary.
I can split it up if this is really preferable of course.
Signature · 8 lines
Michal Hocko SUSE Labs -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to [email protected]. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"[email protected]"> [email protected] </a>
Quoted 8 lines · Michal Hocko
Quoted 8 lines
Quoted 7 lines
Hi,
I think that this is more a cleanup than any functional change. We are rarely screwed so much that__alloc_pages_high_prioritywould fail. Yet I think that__alloc_pages_high_priorityis obscuring the overal intention more than it is helpful. Another motivation is to reducewait_iff_congestedcall to a single one in the allocator. I plan to do other changes in that area and get rid of it altogether.
Quoted 8 lines
Quoted 6 lines
I think it's a combination of a cleanup (the inlining of
__alloc_pages_high_priority) and a functional change (no longer looping infinitely around aget_page_from_freelist()call). I'd suggest doing the inlining in one patch and then the reworking of__GFP_NOFAILwhenALLOC_NO_WATERMARKSfails just so we could easily revert the latter if necessary.I can split it up if this is really preferable of course.
I think it's preferable.
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to [email protected]. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"[email protected]"> [email protected] </a>
Quoted 23 lines · On Thu, Nov 12, 2015 at 04:53:38PM +0100, [email protected] wrote:
From: Michal Hocko <[email protected]>
__alloc_pages_high_prioritydoesn't do anything special other than it callsget_page_from_freelistand loops aroundGFP_NOFAILallocation until it succeeds. It would be better if the first part was done in__alloc_pages_slowpathwhere we modify the zonelist because this would be easier to read and understand. And do the retry at the very same place because retrying without even attempting to do any reclaim is fragile because we are basically relying on somebody else to make the reclaim (be it the direct reclaim or OOM killer) for us. The caller might be holding resources (e.g. locks) which block other other reclaimers from making any progress for example.Remove the helper and open code it into its only user. We have to be careful about
__GFP_NOFAILallocations from thePF_MEMALLOCcontext even though this is a very bad idea to begin with because no progress can be gurateed at all. We shouldn't break the__GFP_NOFAILsemantic here though. It could be argued that this is essentiallyGFP_NOWAITcontext which we do not support butPF_MEMALLOCis much harder to check for existing users because they might happen deep down the code path performed much later after setting the flag so we cannot really rule out there is no kernel path triggering this combination.
Quoted 1 line
Signed-off-by Michal Hocko <[email protected]>
Signature · 8 lines
Mel Gorman SUSE Labs -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to [email protected]. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"[email protected]"> [email protected] </a>