[buildbot tests] [PR5062, PR5404, PR5449, PR5285, LRU for ABD, zvol: reduce linear list search]

[kernelOfTruth]

The following pull-requests are tested on top of latest master which includes ABD changes,

pushing this through a battery of tests to make sure it's stable for my "primetime" (bleeding edge production) testing

#5062 Godfather's child I/Os that may can't done and lost chance to reexecute again
#5404 OpenZFS 7303 - dynamic metaslab selection
#5449 OpenZFS 6569 - large file delete can starve out write ops
#5285 Allow zfs to send replication streams when missing snapshots in the hierarchy
https://github.com/jxiong/zfs/commits/abd_LRU [rebased]
tuxoko@0a7c4b7 zvol: reduce linear list search

Thanks

[mention-bot]

@kernelOfTruth, thanks for your PR! By analyzing the history of the files in this pull request, we identified @grwilson, @behlendorf and @don-brady to be potential reviewers.

[kernelOfTruth]

amended latest commit with no changes == different hash,

hopefully the buildbots this time are fine

edit:

/tmp/zfs-build-buildbot-H3lFoW5P/BUILD/zfs-kmod-0.7.0/_kmod_build_4.4.30-32.54.amzn1.x86_64/../zfs-0.7.0/module/zfs/abd.c:122:26: fatal error: sys/spinlock.h: No such file or directory
 #include <sys/spinlock.h>
                          ^
compilation terminated.

alright, need to complete the patchset, seems I oversaw something ...

edit2:

introduced by

LRU for ABD

edit3:

debug builds are broken

../../module/zfs/abd.c: In function ‘abd_lru_put’:
../../module/zfs/abd.c:339:26: error: incompatible type for argument 1 of ‘refcount_is_zero’
  ASSERT(refcount_is_zero(abd->abd_children));
                          ^
../../lib/libspl/include/assert.h:48:13: note: in definition of macro ‘VERIFY’
  (void) ((!(cond)) &&      \
             ^
../../module/zfs/abd.c:339:2: note: in expansion of macro ‘ASSERT’
  ASSERT(refcount_is_zero(abd->abd_children));
  ^
In file included from ../../include/sys/abd.h:32:0,
                 from ../../module/zfs/abd.c:117:
../../include/sys/refcount.h:68:5: note: expected ‘refcount_t * {aka struct refcount *}’ but argument is of type ‘refcount_t {aka struct refcount}’
 int refcount_is_zero(refcount_t *rc);
     ^
In file included from ../../lib/libspl/include/sys/debug.h:30:0,
                 from ../../include/sys/abd.h:31,
                 from ../../module/zfs/abd.c:117:
../../module/zfs/abd.c: In function ‘abd_lru_cpu_purge’:
../../module/zfs/abd.c:397:11: error: implicit declaration of function ‘ABD_IS_SCATTER’ [-Werror=implicit-function-declaration]
    ASSERT(ABD_IS_SCATTER(obj));
           ^
../../lib/libspl/include/assert.h:48:13: note: in definition of macro ‘VERIFY’
  (void) ((!(cond)) &&      \
             ^
../../module/zfs/abd.c:397:4: note: in expansion of macro ‘ASSERT’
    ASSERT(ABD_IS_SCATTER(obj));
    ^
cc1: all warnings being treated as errors

[kernelOfTruth]

I'm not certain whether that spinlock.h error is a limitation of the buildbots or a bogus error message,

LRU for ABD needs more testing and a PR anyway, it's very interesting 😄

CC: @jxiong

[kernelOfTruth]

Closing for the moment