aboutsummaryrefslogtreecommitdiffstats
path: root/tools/perf/scripts/python/event_analyzing_sample.py
diff options
context:
space:
mode:
authorDave Chinner <dchinner@redhat.com>2013-01-21 23:53:54 +1100
committerBen Myers <bpm@sgi.com>2013-01-28 12:50:50 -0600
commitf2a459565b02b60408f3f2e5ca992a031319712b (patch)
tree1521bd5d0e47e5bd23ecfcde1e3763a4491ed2ae /tools/perf/scripts/python/event_analyzing_sample.py
parentxfs: fix _xfs_buf_find oops on blocks beyond the filesystem end (diff)
downloadlinux-f2a459565b02b60408f3f2e5ca992a031319712b.tar.gz
linux-f2a459565b02b60408f3f2e5ca992a031319712b.zip
xfs: limit speculative prealloc near ENOSPC thresholds
There is a window on small filesytsems where specualtive preallocation can be larger than that ENOSPC throttling thresholds, resulting in specualtive preallocation trying to reserve more space than there is space available. This causes immediate ENOSPC to be triggered, prealloc to be turned off and flushing to occur. One the next write (i.e. next 4k page), we do exactly the same thing, and so effective drive into synchronous 4k writes by triggering ENOSPC flushing on every page while in the window between the prealloc size and the ENOSPC prealloc throttle threshold. Fix this by checking to see if the prealloc size would consume all free space, and throttle it appropriately to avoid premature ENOSPC... Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Brian Foster <bfoster@redhat.com> Signed-off-by: Ben Myers <bpm@sgi.com>
Diffstat (limited to 'tools/perf/scripts/python/event_analyzing_sample.py')
0 files changed, 0 insertions, 0 deletions
=acb499f0ddfbea1176a70d93eb53943c325a073b&follow=1'>sh: Move vsyscall_init() defs up one level.Paul Mundt2-6/+6 2008-01-28sh: Nopped out p3_cache_init() on SH-5 also.Paul Mundt1-0/+1 2008-01-28sh: Fix up VMALLOC_START for SH-5.Paul Mundt1-0/+4 2008-01-28sh: Move over the SH-5 head.S and tlb.h.Paul Mundt6-44/+24 2008-01-28sh: Have SH-5 provide an {en,dis}able_fpu() impl.Paul Mundt3-10/+21 2008-01-28sh: Disable initial cache flush on SH-5.Paul Mundt1-14/+18 2008-01-28sh: Don't reference UBC code in CPU init on sh64.Paul Mundt1-1/+6 2008-01-28sh: imask IRQ depends on sh32.Paul Mundt1-1/+2 2008-01-28sh: Move in the SH-5 traps.c impl.Paul Mundt4-3/+2