Commit Graph

442 Commits

Author SHA1 Message Date
Brian Behlendorf 1315c88437 Coverity 9649, 9650, 9651: Uninit
This check was originally added to detect double initializations
of mutex types (which it did find).  Unfortunately, Coverity is
right that there is a very small chance we could trigger the
assertion by accident because an uninitialized stack variable
happens to contain the mutex magic.  This is particularly unlikely
since we do poison the mutexs when destroyed but still possible.
Therefore I'm simply removing the assertion.
2009-02-18 09:48:07 -08:00
Brian Behlendorf bb01879ebe Coverity 9654, 9654: Use After Free
Because vmem_free() was implemented as a macro using the ','
operator to evaluate both arguments and we performed the free
before evaluating size we would deference the free'd pointer.
To resolve the problem we just invert the ordering and evaluate
size first just as if it was evaluated by the caller when being
passed to this function.  This ensure that if the caller is
doing something reckless like performing an assignment as
part of the size argument we still perform it and it simply
doesn't get removed by the macro.  Oh course nobody should
be doing this sort of thing, but just in case.
2009-02-17 16:51:19 -08:00
Brian Behlendorf 15dc8b072e Coverity 9652, 9653: No Effect
Removed 2 ASSERT()s which had no effect because by definition
size_t is always an unsigned type thus is always >= 0.
2009-02-17 16:30:58 -08:00
Brian Behlendorf 014b1d6f54 Coverity 9641: Buffer Size
When SPLAT_TEST_INIT() initialized SPLAT_KMEM_TEST11_NAME the short
short test name overran the static length buffer of SPLAT_NAME_SIZE.
This was fixed by increasing the buffer length from 16 to 20 bytes.
2009-02-17 16:24:26 -08:00
Brian Behlendorf 9b1b8e4c24 kmem slab magazine ageing deadlock
- The previous magazine ageing sceme relied on the on_each_cpu()
  function to call spl_magazine_age() on each cpu.  It turns out
  this could deadlock with do_flush_tlb_all() which also relies
  on the IPI based on_each_cpu().  To avoid this problem a per-
  magazine delayed work item is created and indepentantly
  scheduled to the correct cpu removing the need for on_each_cpu().
- Additionally two unused fields were removed from the type
  spl_kmem_cache_t, they were hold overs from previous cleanup.
    - struct work_struct work
    - struct timer_list timer
2009-02-17 15:52:18 -08:00
Brian Behlendorf 1a944a7d0b kmem slab fixes
- spl_slab_reclaim() 'continue' changed back to 'break' from commit
  37db7d8cf9.  The original was correct,
  I have added a comment to ensure this does not happen again.
- spl_slab_reclaim() further optimized by moving the destructor call
  in spl_slab_free() outside the skc->skc_lock.  This minimizes the
  length of time the spin lock is held, allows the destructors to
  be invoked concurrently for different objects, and as a bonus makes
  it safe (although unwise) to sleep in the destructors.
2009-02-13 10:28:55 -08:00
Brian Behlendorf fce5ef8306 Build system update
- Added default build flags for kernel modules:
  -Wstrict-prototypes -Werror
2009-02-12 15:04:36 -08:00
Brian Behlendorf f6c5d4ff88 Build system update
- Added default build flags:
  -Wall -Wstrict-prototypes -Werror -Wshadow
- Added missing Makefile's for include/ subdirectories.
2009-02-12 14:45:22 -08:00
Brian Behlendorf 37db7d8cf9 kmem slab fixes
- Default SPL_KMEM_CACHE_DELAY changed to 15 to match Solaris.
- Aged out slab checking occurs every SPL_KMEM_CACHE_DELAY / 3.
- skc->skc_reap tunable added whichs allows callers of
  spl_slab_reclaim() to cap the number of slabs reclaimed.
  On Solaris all eligible slabs are always reclaimed, and this
  is still the default behavior.  However, I suspect that is
  not always wise for reasons such as in the next comment.
- spl_slab_reclaim() added cond_resched() while walking the
  slab/object free lists.  Soft lockups were observed when
  freeing large numbers of vmalloc'd slabs/objets.
- spl_slab_reclaim() 'sks->sks_ref > 0' check changes from
  incorrect 'break' to 'continue' to ensure all slabs are
  checked.
- spl_cache_age() reworked to avoid a deadlock with
  do_flush_tlb_all() which occured because we slept waiting
  for completion in spl_cache_age().  To waiting for magazine
  reclamation to finish is not required so we no longer wait.
- spl_magazine_create() and spl_magazine_destroy() shifted
  back to using for_each_online_cpu() instead of the
  spl_on_each_cpu() approach which was of course a bad idea
  due to memory allocations which Ricardo pointed out.
2009-02-12 13:32:10 -08:00
Ricardo M. Correia f500ccff35 Minor bug fix due to MAXOFFSET_T constant being too large on 32-bit systems. 2009-02-07 00:53:39 +00:00
Brian Behlendorf e50ad76da5 Prep for 0.4.2 tag 2009-02-05 13:43:45 -08:00
Brian Behlendorf 4ab13d3b5c Additional Linux VM integration
Added support for Solaris swapfs_minfree, and swapfs_reserve tunables.
In additional availrmem is now available and return a reasonable value
which is reasonably analogous to the Solaris meaning.  On linux we
return the sun of free and inactive pages since these are all easily
reclaimable.

All tunables are available in /proc/sys/kernel/spl/vm/* and they may
need a little adjusting once we observe the real behavior.  Some of
the defaults are mapped to similar linux counterparts, others are
straight from the OpenSolaris defaults.
2009-02-05 12:26:34 -08:00
Brian Behlendorf 36b313dacf Linux VM integration / device special files
Support added to provide reasonable values for the global Solaris
VM variables: minfree, desfree, lotsfree, needfree.  These values
are set to the sum of their per-zone linux counterparts which
should be close enough for Solaris consumers.

When a non-GPL app links against the SPL we cannot use the udev
interfaces, which means non of the device special files are created.
Because of this I had added a poor mans udev which cause the SPL
to invoke an upcall and create the basic devices when a minor
is registered.  When a minor is unregistered we use the vnode
interface to unlink the special file.
2009-02-04 15:15:41 -08:00
Brian Behlendorf 31a033ecd4 2.6.27+ portability changes
- Added SPL_AC_3ARGS_ON_EACH_CPU configure check to determine
  if the older 4 argument version of on_each_cpu() should be
  used or the new 3 argument version.  The retry argument was
  dropped in the new API which was never used anyway.
- Updated work queue compatibility wrappers.  The old way this
  worked was to pass a data point when initialized the workqueue.
  The new API assumed the work item is embedding in a structure
  and we us container_of() to find that data pointer.
- Updated skc->skc_flags to be an unsigned long which is now
  type checked in the bit operations.  This silences the warnings.
- Updated autogen products and splat tests accordingly
2009-02-02 15:12:30 -08:00
Brian Behlendorf f220894e1f Make the number of system taskq threads based on the node of cores in the node, as is done for most linux system tasks 2009-02-02 08:53:53 -08:00
Brian Behlendorf 10a4be0f03 Update thread tests to have max_time 2009-01-30 21:24:42 -08:00
Brian Behlendorf 416bae036b Add new workqueue header 2009-01-30 21:11:42 -08:00
Brian Behlendorf ea3e6ca9e5 kmem_cache hardening and performance improvements
- Added slab work queue task which gradually ages and free's slabs
  from the cache which have not been used recently.
- Optimized slab packing algorithm to ensure each slab contains the
  maximum number of objects without create to large a slab.
- Fix deadlock, we can never call kv_free() under the skc_lock.  We
  now unlink the objects and slabs from the cache itself and attach
  them to a private work list.  The contents of the list are then
  subsequently freed outside the spin lock.
- Move magazine create/destroy operation on to local cpu.
- Further performace optimizations by minimize the usage of the large
  per-cache skc_lock.  This includes the addition of KMC_BIT_REAPING
  bit mask which is used to prevent concurrent reaping, and to defer
  new slab creation when reaping is occuring.
- Add KMC_BIT_DESTROYING bit mask which is set when the cache is being
  destroyed, this is used to catch any task accessing the cache while
  it is being destroyed.
- Add comments to all the functions and additional comments to try
  and make everything as clear as possible.
- Major cleanup and additions to the SPLAT kmem tests to more
  rigerously stress the cache implementation and look for any problems.
  This includes correctness and performance tests.
- Updated portable work queue interfaces
2009-01-30 20:54:49 -08:00
Brian Behlendorf 34e71c9e97 Remove debug check was was accidentally left in place an prevent the slab cache from working on systems with >4 cores 2009-01-26 20:10:23 -08:00
Brian Behlendorf 0f233eac33 Pull the blkdev header in to the sunldi for some useful structure definitions and helper functions 2009-01-26 16:47:49 -08:00
Brian Behlendorf 48e0606a52 Implement kmem cache alignment argument 2009-01-26 09:02:04 -08:00
Brian Behlendorf e4f3ea278e Remove stray ` from macro 2009-01-23 08:59:11 -08:00
Brian Behlendorf 3f4126739d Sleep uninteruptibly, waking up early may result in a crash 2009-01-22 09:58:48 -08:00
Brian Behlendorf 511176398c Update debug.h to standardize VERIFY3_IMPL error messages in debug and non-debug mode 2009-01-22 09:41:47 -08:00
Brian Behlendorf 064bbffb63 Prep for 0.4.1 tag 2009-01-21 11:46:02 -08:00
Brian Behlendorf b6b2acc66e Minor fix for compiler warning when KMEM_TRACKING is enabled 2009-01-20 13:39:35 -08:00
Brian Behlendorf ae3b87f908 KMEM_TRACKING turned up a missing free in list test 6, fix the leak 2009-01-20 12:47:53 -08:00
Brian Behlendorf 15270e003e Ensure -NDEBUG does not get added to spl_config.h and is only set in the build options. This allows other kernel modules to use spl_config to leverage the reset of the config checks without getting confused with the debug options 2009-01-20 11:59:47 -08:00
Brian Behlendorf 5566ec0959 Refresh libtool 2009-01-15 10:47:24 -08:00
Brian Behlendorf 617d5a673c Rename modules to module and update references 2009-01-15 10:44:54 -08:00
Brian Behlendorf f6a19c0d37 Make the splat load message caps just for consistency 2009-01-13 11:45:02 -08:00
Brian Behlendorf b172b6dfde TASKQ_DYNAMIC not yet support, do not create the global taskq with that flag or we crash with debug enabled. Also don't bother dumping debug when debugging is diabled, it's pointless 2009-01-13 11:43:05 -08:00
Brian Behlendorf b871b8cdef Rework ddi_strtox calls to a native implementation which actuall supports the EINVAL, ERANGE error handling, plus add a regression suite to ensure I got it atleast mostly right 2009-01-13 09:30:59 -08:00
Brian Behlendorf 1e4ed6c990 Add missing stub headers 2009-01-09 16:04:44 -08:00
Brian Behlendorf 121d48c97d Add basic ksid_lookupdomain and ksiddomain_rele support, just allocations 2009-01-09 15:30:53 -08:00
Brian Behlendorf f590d7d374 Make sure we export ddi_quiesce_not_needed 2009-01-09 14:30:30 -08:00
Brian Behlendorf 0e41414946 Add two new stub headers 2009-01-09 14:04:13 -08:00
Brian Behlendorf 97735c39e3 Add VOP_SEEK 2009-01-09 13:59:39 -08:00
Brian Behlendorf d83ba26e18 Add missing policy includes, add missing sun ddi bits 2009-01-09 10:49:47 -08:00
Brian Behlendorf 70997fb4b1 Add share.h stub 2009-01-09 10:06:18 -08:00
Brian Behlendorf 71c8ab9c68 Drat fix missing ; 2009-01-09 10:05:03 -08:00
Brian Behlendorf 23f5c4c281 Add missing callback_context_t and fid_t types 2009-01-09 10:03:37 -08:00
Brian Behlendorf 703e7a3cf4 Add stubs for three more includes 2009-01-09 09:47:27 -08:00
Brian Behlendorf 434d1d0f8f Add active test for splat list tests 2009-01-07 13:48:36 -08:00
Brian Behlendorf d702c04ff1 Add 5 splat tests for list handling 2009-01-07 12:54:03 -08:00
Brian Behlendorf 4c18c39ecb Add include/sys/compress.h header 2009-01-06 09:47:00 -08:00
Brian Behlendorf 160c63ab76 Add P2BOUNDARY macro 2009-01-06 09:23:13 -08:00
Brian Behlendorf 7adbea4141 Pull in some default page typedefs 2009-01-05 16:14:38 -08:00
Brian Behlendorf 0f37204417 Add DTRACE_PROBE(a) 2009-01-05 16:09:21 -08:00
Brian Behlendorf b53c565e65 Stub u8_textprep.h for inclusion purposes 2009-01-05 15:37:07 -08:00