zfs/config/spl-build.m4

1651 lines
42 KiB
Plaintext
Raw Normal View History

###############################################################################
# Copyright (C) 2007-2010 Lawrence Livermore National Security, LLC.
# Copyright (C) 2007 The Regents of the University of California.
# Written by Brian Behlendorf <behlendorf1@llnl.gov>.
###############################################################################
# SPL_AC_CONFIG_KERNEL: Default SPL kernel configuration.
###############################################################################
AC_DEFUN([SPL_AC_CONFIG_KERNEL], [
SPL_AC_KERNEL
if test "${LINUX_OBJ}" != "${LINUX}"; then
KERNELMAKE_PARAMS="$KERNELMAKE_PARAMS O=$LINUX_OBJ"
fi
AC_SUBST(KERNELMAKE_PARAMS)
2010-09-02 19:12:39 +00:00
KERNELCPPFLAGS="$KERNELCPPFLAGS -Wstrict-prototypes"
AC_SUBST(KERNELCPPFLAGS)
SPL_AC_DEBUG
SPL_AC_DEBUG_LOG
SPL_AC_DEBUG_KMEM
Autoconf --enable-debug-* cleanup Cleanup the --enable-debug-* configure options, this has been pending for quite some time and I am glad I finally got to it. To summerize: 1) All SPL_AC_DEBUG_* macros were updated to be a more autoconf friendly. This mainly involved shift to the GNU approved usage of AC_ARG_ENABLE and ensuring AS_IF is used rather than directly using an if [ test ] construct. 2) --enable-debug-kmem=yes by default. This simply enabled keeping a running tally of total memory allocated and freed and reporting a memory leak if there was one at module unload. Additionally, it ensure /proc/spl/kmem/slab will exist by default which is handy. The overhead is low for this and it should not impact performance. 3) --enable-debug-kmem-tracking=no by default. This option was added to provide a configure option to enable to detailed memory allocation tracking. This support was always there but you had to know where to turn it on. By default this support is disabled because it is known to badly hurt performence, however it is invaluable when chasing a memory leak. 4) --enable-debug-kstat removed. After further reflection I can't see why you would ever really want to turn this support off. It is now always on which had the nice side effect of simplifying the proc handling code in spl-proc.c. We can now always assume the top level directory will be there. 5) --enable-debug-callb removed. This never really did anything, it was put in provisionally because it might have been needed. It turns out it was not so I am just removing it to prevent confusion.
2009-10-30 20:58:51 +00:00
SPL_AC_DEBUG_KMEM_TRACKING
SPL_AC_TEST_MODULE
SPL_AC_ATOMIC_SPINLOCK
SPL_AC_SHRINKER_CALLBACK
SPL_AC_CTL_NAME
SPL_AC_PDE_DATA
Reimplement mutexs for Linux lock profiling/analysis For a generic explanation of why mutexs needed to be reimplemented to work with the kernel lock profiling see commits: e811949a57044d60d12953c5c3b808a79a7d36ef and d28db80fd0fd4fd63aec09037c44408e51a222d6 The specific changes made to the mutex implemetation are as follows. The Linux mutex structure is now directly embedded in the kmutex_t. This allows a kmutex_t to be directly case to a mutex struct and passed directly to the Linux primative. Just like with the rwlocks it is critical that these functions be implemented as '#defines to ensure the location information is preserved. The preprocessor can then do a direct replacement of the Solaris primative with the linux primative. Just as with the rwlocks we need to track the lock owner. Here things get a little more interesting because depending on your kernel version, and how you've built your kernel Linux may already do this for you. If your running a 2.6.29 or newer kernel on a SMP system the lock owner will be tracked. This was added to Linux to support adaptive mutexs, more on that shortly. Alternately, your kernel might track the lock owner if you've set CONFIG_DEBUG_MUTEXES in the kernel build. If neither of the above things is true for your kernel the kmutex_t type will include and track the lock owner to ensure correct behavior. This is all handled by a new autoconf check called SPL_AC_MUTEX_OWNER. Concerning adaptive mutexs these are a very recent development and they did not make it in to either the latest FC11 of SLES11 kernels. Ideally, I'd love to see this kernel change appear in one of these distros because it does help performance. From Linux kernel commit: 0d66bf6d3514b35eb6897629059443132992dbd7 "Testing with Ingo's test-mutex application... gave a 345% boost for VFS scalability on my testbox" However, if you don't want to backport this change yourself you can still simply export the task_curr() symbol. The kmutex_t implementation will use this symbol when it's available to provide it's own adaptive mutexs. Finally, DEBUG_MUTEX support was removed including the proc handlers. This was done because now that we are cleanly integrated with the kernel profiling all this information and much much more is available in debug kernel builds. This code was now redundant. Update mutexs validated on: - SLES10 (ppc64) - SLES11 (x86_64) - CHAOS4.2 (x86_64) - RHEL5.3 (x86_64) - RHEL6 (x86_64) - FC11 (x86_64)
2009-09-25 21:47:01 +00:00
SPL_AC_MUTEX_OWNER
SPL_AC_MUTEX_OWNER_TASK_STRUCT
SPL_AC_SET_FS_PWD_WITH_CONST
SPL_AC_2ARGS_VFS_UNLINK
SPL_AC_4ARGS_VFS_RENAME
SPL_AC_VFS_FSYNC
SPL_AC_2ARGS_VFS_FSYNC
SPL_AC_INODE_TRUNCATE_RANGE
SPL_AC_FS_STRUCT_SPINLOCK
SPL_AC_KUIDGID_T
Reimplement rwlocks for Linux lock profiling/analysis. It turns out that the previous rwlock implementation worked well but did not integrate properly with the upstream kernel lock profiling/ analysis tools. This is a major problem since it would be awfully nice to be able to use the automatic lock checker and profiler. The problem is that the upstream lock tools use the pre-processor to create a lock class for each uniquely named locked. Since the rwsem was embedded in a wrapper structure the name was always the same. The effect was that we only ended up with one lock class for the entire SPL which caused the lock dependency checker to flag nearly everything as a possible deadlock. The solution was to directly map a krwlock to a Linux rwsem using a typedef there by eliminating the wrapper structure. This was not done initially because the rwsem implementation is specific to the arch. To fully implement the Solaris krwlock API using only the provided rwsem API is not possible. It can only be done by directly accessing some of the internal data member of the rwsem structure. For example, the Linux API provides a different function for dropping a reader vs writer lock. Whereas the Solaris API uses the same function and the caller does not pass in what type of lock it is. This means to properly drop the lock we need to determine if the lock is currently a reader or writer lock. Then we need to call the proper Linux API function. Unfortunately, there is no provided API for this so we must extracted this information directly from arch specific lock implementation. This is all do able, and what I did, but it does complicate things considerably. The good news is that in addition to the profiling benefits of this change. We may see performance improvements due to slightly reduced overhead when creating rwlocks and manipulating them. The only function I was forced to sacrafice was rw_owner() because this information is simply not stored anywhere in the rwsem. Luckily this appears not to be a commonly used function on Solaris, and it is my understanding it is mainly used for debugging anyway. In addition to the core rwlock changes, extensive updates were made to the rwlock regression tests. Each class of test was extended to provide more API coverage and to be more rigerous in checking for misbehavior. This is a pretty significant change and with that in mind I have been careful to validate it on several platforms before committing. The full SPLAT regression test suite was run numberous times on all of the following platforms. This includes various kernels ranging from 2.6.16 to 2.6.29. - SLES10 (ppc64) - SLES11 (x86_64) - CHAOS4.2 (x86_64) - RHEL5.3 (x86_64) - RHEL6 (x86_64) - FC11 (x86_64)
2009-09-18 23:09:47 +00:00
SPL_AC_PUT_TASK_STRUCT
SPL_AC_5ARGS_PROC_HANDLER
SPL_AC_KVASPRINTF
Correctly handle rwsem_is_locked() behavior A race condition in rwsem_is_locked() was fixed in Linux 2.6.33 and the fix was backported to RHEL5 as of kernel 2.6.18-190.el5. Details can be found here: https://bugzilla.redhat.com/show_bug.cgi?id=526092 The race condition was fixed in the kernel by acquiring the semaphore's wait_lock inside rwsem_is_locked(). The SPL worked around the race condition by acquiring the wait_lock before calling that function, but with the fix in place it must not do that. This commit implements an autoconf test to detect whether the fixed version of rwsem_is_locked() is present. The previous version of rwsem_is_locked() was an inline static function while the new version is exported as a symbol which we can check for in module.symvers. Depending on the result we correctly implement the needed compatibility macros for proper spinlock handling. Finally, we do the right thing with spin locks in RW_*_HELD() by using the new compatibility macros. We only only acquire the semaphore's wait_lock if it is calling a rwsem_is_locked() that does not itself try to acquire the lock. Some new overhead and a small harmless race is introduced by this change. This is because RW_READ_HELD() and RW_WRITE_HELD() now acquire and release the wait_lock twice: once for the call to rwsem_is_locked() and once for the call to rw_owner(). This can't be avoided if calling a rwsem_is_locked() that takes the wait_lock, as it will in more recent kernels. The other case which only occurs in legacy kernels could be optimized by taking the lock only once, as was done prior to this commit. However, I decided that the performance gain probably wasn't significant enough to justify the messy special cases required. The function spl_rw_get_owner() was only used to enable the afore-mentioned optimization. Since it is no longer used, I removed it. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
2010-08-10 18:01:46 +00:00
SPL_AC_EXPORTED_RWSEM_IS_LOCKED
SPL_AC_KERNEL_FALLOCATE
SPL_AC_KERN_PATH
SPL_AC_CONFIG_ZLIB_INFLATE
SPL_AC_CONFIG_ZLIB_DEFLATE
SPL_AC_2ARGS_ZLIB_DEFLATE_WORKSPACESIZE
SPL_AC_SHRINK_CONTROL_STRUCT
SPL_AC_RWSEM_SPINLOCK_IS_RAW
SPL_AC_SCHED_RT_HEADER
SPL_AC_2ARGS_VFS_GETATTR
SPL_AC_USLEEP_RANGE
Add KMC_SLAB cache type For small objects the Linux slab allocator has several advantages over its counterpart in the SPL. These include: 1) It is more memory-efficient and packs objects more tightly. 2) It is continually tuned to maximize performance. Therefore it makes sense to layer the SPLs slab allocator on top of the Linux slab allocator. This allows us to leverage the advantages above while preserving the Illumos semantics we depend on. However, there are some things we need to be careful of: 1) The Linux slab allocator was never designed to work well with large objects. Because the SPL slab must still handle this use case a cut off limit was added to transition from Linux slab backed objects to kmem or vmem backed slabs. spl_kmem_cache_slab_limit - Objects less than or equal to this size in bytes will be backed by the Linux slab. By default this value is zero which disables the Linux slab functionality. Reasonable values for this cut off limit are in the range of 4096-16386 bytes. spl_kmem_cache_kmem_limit - Objects less than or equal to this size in bytes will be backed by a kmem slab. Objects over this size will be vmem backed instead. This value defaults to 1/8 a page, or 512 bytes on an x86_64 architecture. 2) Be aware that using the Linux slab may inadvertently introduce new deadlocks. Care has been taken previously to ensure that all allocations which occur in the write path use GFP_NOIO. However, there may be internal allocations performed in the Linux slab which do not honor these flags. If this is the case a deadlock may occur. The path forward is definitely to start relying on the Linux slab. But for that to happen we need to start building confidence that there aren't any unexpected surprises lurking for us. And ideally need to move completely away from using the SPLs slab for large memory allocations. This patch is a first step. NOTES: 1) The KMC_NOMAGAZINE flag was leveraged to support the Linux slab backed caches but it is not supported for kmem/vmem backed caches. 2) Regardless of the spl_kmem_cache_*_limit settings a cache may be explicitly set to a given type by passed the KMC_KMEM, KMC_VMEM, or KMC_SLAB flags during cache creation. 3) The constructors, destructors, and reclaim callbacks are all functional and will be called regardless of the cache type. 4) KMC_SLAB caches will not appear in /proc/spl/kmem/slab due to the issues involved in presenting correct object accounting. Instead they will appear in /proc/slabinfo under the same names. 5) Several kmem SPLAT tests needed to be fixed because they relied incorrectly on internal kmem slab accounting. With the updated test cases all the SPLAT tests pass as expected. 6) An autoconf test was added to ensure that the __GFP_COMP flag was correctly added to the default flags used when allocating a slab. This is required to ensure all pages in higher order slabs are properly refcounted, see ae16ed9. 7) When using the SLUB allocator there is no need to attempt to set the __GFP_COMP flag. This has been the default behavior for the SLUB since Linux 2.6.25. 8) When using the SLUB it may be desirable to set the slub_nomerge kernel parameter to prevent caches from being merged. Original-patch-by: DHE <git@dehacked.net> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Prakash Surya <surya1@llnl.gov> Signed-off-by: Tim Chase <tim@chase2k.com> Signed-off-by: DHE <git@dehacked.net> Signed-off-by: Chunwei Chen <tuxoko@gmail.com> Closes #356
2013-12-08 22:01:45 +00:00
SPL_AC_KMEM_CACHE_ALLOCFLAGS
SPL_AC_WAIT_ON_BIT
])
AC_DEFUN([SPL_AC_MODULE_SYMVERS], [
modpost=$LINUX/scripts/Makefile.modpost
AC_MSG_CHECKING([kernel file name for module symbols])
if test "x$enable_linux_builtin" != xyes -a -f "$modpost"; then
if grep -q Modules.symvers $modpost; then
LINUX_SYMBOLS=Modules.symvers
else
LINUX_SYMBOLS=Module.symvers
fi
if ! test -f "$LINUX_OBJ/$LINUX_SYMBOLS"; then
AC_MSG_ERROR([
*** Please make sure the kernel devel package for your distribution
*** is installed. If you are building with a custom kernel, make sure the
*** kernel is configured, built, and the '--with-linux=PATH' configure
*** option refers to the location of the kernel source.])
fi
else
LINUX_SYMBOLS=NONE
fi
AC_MSG_RESULT($LINUX_SYMBOLS)
AC_SUBST(LINUX_SYMBOLS)
])
AC_DEFUN([SPL_AC_KERNEL], [
AC_ARG_WITH([linux],
AS_HELP_STRING([--with-linux=PATH],
[Path to kernel source]),
[kernelsrc="$withval"])
AC_ARG_WITH([linux-obj],
AS_HELP_STRING([--with-linux-obj=PATH],
[Path to kernel build objects]),
[kernelbuild="$withval"])
AC_MSG_CHECKING([kernel source directory])
if test -z "$kernelsrc"; then
if test -e "/lib/modules/$(uname -r)/source"; then
headersdir="/lib/modules/$(uname -r)/source"
sourcelink=$(readlink -f "$headersdir")
elif test -e "/lib/modules/$(uname -r)/build"; then
headersdir="/lib/modules/$(uname -r)/build"
sourcelink=$(readlink -f "$headersdir")
else
sourcelink=$(ls -1d /usr/src/kernels/* \
/usr/src/linux-* \
2>/dev/null | grep -v obj | tail -1)
fi
if test -n "$sourcelink" && test -e ${sourcelink}; then
kernelsrc=`readlink -f ${sourcelink}`
else
kernelsrc="[Not found]"
fi
else
if test "$kernelsrc" = "NONE"; then
kernsrcver=NONE
fi
fi
AC_MSG_RESULT([$kernelsrc])
if test ! -d "$kernelsrc"; then
AC_MSG_ERROR([
*** Please make sure the kernel devel package for your distribution
*** is installed and then try again. If that fails, you can specify the
*** location of the kernel source with the '--with-linux=PATH' option.])
fi
AC_MSG_CHECKING([kernel build directory])
if test -z "$kernelbuild"; then
if test -e "/lib/modules/$(uname -r)/build"; then
kernelbuild=`readlink -f /lib/modules/$(uname -r)/build`
elif test -d ${kernelsrc}-obj/${target_cpu}/${target_cpu}; then
kernelbuild=${kernelsrc}-obj/${target_cpu}/${target_cpu}
elif test -d ${kernelsrc}-obj/${target_cpu}/default; then
kernelbuild=${kernelsrc}-obj/${target_cpu}/default
elif test -d `dirname ${kernelsrc}`/build-${target_cpu}; then
kernelbuild=`dirname ${kernelsrc}`/build-${target_cpu}
else
kernelbuild=${kernelsrc}
fi
fi
AC_MSG_RESULT([$kernelbuild])
AC_MSG_CHECKING([kernel source version])
utsrelease1=$kernelbuild/include/linux/version.h
utsrelease2=$kernelbuild/include/linux/utsrelease.h
utsrelease3=$kernelbuild/include/generated/utsrelease.h
if test -r $utsrelease1 && fgrep -q UTS_RELEASE $utsrelease1; then
utsrelease=linux/version.h
elif test -r $utsrelease2 && fgrep -q UTS_RELEASE $utsrelease2; then
utsrelease=linux/utsrelease.h
elif test -r $utsrelease3 && fgrep -q UTS_RELEASE $utsrelease3; then
utsrelease=generated/utsrelease.h
fi
if test "$utsrelease"; then
kernsrcver=`(echo "#include <$utsrelease>";
echo "kernsrcver=UTS_RELEASE") |
cpp -I $kernelbuild/include |
grep "^kernsrcver=" | cut -d \" -f 2`
if test -z "$kernsrcver"; then
AC_MSG_RESULT([Not found])
AC_MSG_ERROR([*** Cannot determine kernel version.])
fi
else
AC_MSG_RESULT([Not found])
if test "x$enable_linux_builtin" != xyes; then
AC_MSG_ERROR([*** Cannot find UTS_RELEASE definition.])
else
AC_MSG_ERROR([
*** Cannot find UTS_RELEASE definition.
*** Please run 'make prepare' inside the kernel source tree.])
fi
fi
AC_MSG_RESULT([$kernsrcver])
LINUX=${kernelsrc}
LINUX_OBJ=${kernelbuild}
LINUX_VERSION=${kernsrcver}
AC_SUBST(LINUX)
AC_SUBST(LINUX_OBJ)
AC_SUBST(LINUX_VERSION)
SPL_AC_MODULE_SYMVERS
])
dnl #
dnl # Default SPL user configuration
dnl #
AC_DEFUN([SPL_AC_CONFIG_USER], [])
dnl #
dnl # Check for rpm+rpmbuild to build RPM packages. If these tools
dnl # are missing, it is non-fatal, but you will not be able to build
dnl # RPM packages and will be warned if you try too.
dnl #
dnl # By default, the generic spec file will be used because it requires
dnl # minimal dependencies. Distribution specific spec files can be
dnl # placed under the 'rpm/<distribution>' directory and enabled using
dnl # the --with-spec=<distribution> configure option.
dnl #
AC_DEFUN([SPL_AC_RPM], [
RPM=rpm
RPMBUILD=rpmbuild
AC_MSG_CHECKING([whether $RPM is available])
AS_IF([tmp=$($RPM --version 2>/dev/null)], [
RPM_VERSION=$(echo $tmp | $AWK '/RPM/ { print $[3] }')
HAVE_RPM=yes
AC_MSG_RESULT([$HAVE_RPM ($RPM_VERSION)])
],[
HAVE_RPM=no
AC_MSG_RESULT([$HAVE_RPM])
])
AC_MSG_CHECKING([whether $RPMBUILD is available])
AS_IF([tmp=$($RPMBUILD --version 2>/dev/null)], [
RPMBUILD_VERSION=$(echo $tmp | $AWK '/RPM/ { print $[3] }')
HAVE_RPMBUILD=yes
AC_MSG_RESULT([$HAVE_RPMBUILD ($RPMBUILD_VERSION)])
],[
HAVE_RPMBUILD=no
AC_MSG_RESULT([$HAVE_RPMBUILD])
])
RPM_DEFINE_COMMON='--define "$(DEBUG_SPL) 1" --define "$(DEBUG_LOG) 1" --define "$(DEBUG_KMEM) 1" --define "$(DEBUG_KMEM_TRACKING) 1"'
RPM_DEFINE_UTIL=
RPM_DEFINE_KMOD='--define "kernels $(LINUX_VERSION)"'
RPM_DEFINE_DKMS=
SRPM_DEFINE_COMMON='--define "build_src_rpm 1"'
SRPM_DEFINE_UTIL=
SRPM_DEFINE_KMOD=
SRPM_DEFINE_DKMS=
RPM_SPEC_DIR="rpm/generic"
AC_ARG_WITH([spec],
AS_HELP_STRING([--with-spec=SPEC],
[Spec files 'generic|fedora']),
[RPM_SPEC_DIR="rpm/$withval"])
AC_MSG_CHECKING([whether spec files are available])
AC_MSG_RESULT([yes ($RPM_SPEC_DIR/*.spec.in)])
AC_SUBST(HAVE_RPM)
AC_SUBST(RPM)
AC_SUBST(RPM_VERSION)
AC_SUBST(HAVE_RPMBUILD)
AC_SUBST(RPMBUILD)
AC_SUBST(RPMBUILD_VERSION)
AC_SUBST(RPM_SPEC_DIR)
AC_SUBST(RPM_DEFINE_UTIL)
AC_SUBST(RPM_DEFINE_KMOD)
AC_SUBST(RPM_DEFINE_DKMS)
AC_SUBST(RPM_DEFINE_COMMON)
AC_SUBST(SRPM_DEFINE_UTIL)
AC_SUBST(SRPM_DEFINE_KMOD)
AC_SUBST(SRPM_DEFINE_DKMS)
AC_SUBST(SRPM_DEFINE_COMMON)
])
dnl #
dnl # Check for dpkg+dpkg-buildpackage to build DEB packages. If these
dnl # tools are missing it is non-fatal but you will not be able to build
dnl # DEB packages and will be warned if you try too.
dnl #
AC_DEFUN([SPL_AC_DPKG], [
DPKG=dpkg
DPKGBUILD=dpkg-buildpackage
AC_MSG_CHECKING([whether $DPKG is available])
AS_IF([tmp=$($DPKG --version 2>/dev/null)], [
DPKG_VERSION=$(echo $tmp | $AWK '/Debian/ { print $[7] }')
HAVE_DPKG=yes
AC_MSG_RESULT([$HAVE_DPKG ($DPKG_VERSION)])
],[
HAVE_DPKG=no
AC_MSG_RESULT([$HAVE_DPKG])
])
AC_MSG_CHECKING([whether $DPKGBUILD is available])
AS_IF([tmp=$($DPKGBUILD --version 2>/dev/null)], [
DPKGBUILD_VERSION=$(echo $tmp | \
$AWK '/Debian/ { print $[4] }' | cut -f-4 -d'.')
HAVE_DPKGBUILD=yes
AC_MSG_RESULT([$HAVE_DPKGBUILD ($DPKGBUILD_VERSION)])
],[
HAVE_DPKGBUILD=no
AC_MSG_RESULT([$HAVE_DPKGBUILD])
])
AC_SUBST(HAVE_DPKG)
AC_SUBST(DPKG)
AC_SUBST(DPKG_VERSION)
AC_SUBST(HAVE_DPKGBUILD)
AC_SUBST(DPKGBUILD)
AC_SUBST(DPKGBUILD_VERSION)
])
dnl #
dnl # Until native packaging for various different packing systems
dnl # can be added the least we can do is attempt to use alien to
dnl # convert the RPM packages to the needed package type. This is
dnl # a hack but so far it has worked reasonable well.
dnl #
AC_DEFUN([SPL_AC_ALIEN], [
ALIEN=alien
AC_MSG_CHECKING([whether $ALIEN is available])
AS_IF([tmp=$($ALIEN --version 2>/dev/null)], [
ALIEN_VERSION=$(echo $tmp | $AWK '{ print $[3] }')
HAVE_ALIEN=yes
AC_MSG_RESULT([$HAVE_ALIEN ($ALIEN_VERSION)])
],[
HAVE_ALIEN=no
AC_MSG_RESULT([$HAVE_ALIEN])
])
AC_SUBST(HAVE_ALIEN)
AC_SUBST(ALIEN)
AC_SUBST(ALIEN_VERSION)
])
dnl #
dnl # Using the VENDOR tag from config.guess set the default
dnl # package type for 'make pkg': (rpm | deb | tgz)
dnl #
AC_DEFUN([SPL_AC_DEFAULT_PACKAGE], [
AC_MSG_CHECKING([linux distribution])
if test -f /etc/toss-release ; then
VENDOR=toss ;
elif test -f /etc/fedora-release ; then
VENDOR=fedora ;
elif test -f /etc/redhat-release ; then
VENDOR=redhat ;
elif test -f /etc/gentoo-release ; then
VENDOR=gentoo ;
elif test -f /etc/arch-release ; then
VENDOR=arch ;
elif test -f /etc/SuSE-release ; then
VENDOR=sles ;
elif test -f /etc/slackware-version ; then
VENDOR=slackware ;
elif test -f /etc/lunar.release ; then
VENDOR=lunar ;
elif test -f /etc/lsb-release ; then
VENDOR=ubuntu ;
elif test -f /etc/debian_version ; then
VENDOR=debian ;
else
VENDOR= ;
fi
AC_MSG_RESULT([$VENDOR])
AC_SUBST(VENDOR)
AC_MSG_CHECKING([default package type])
case "$VENDOR" in
toss) DEFAULT_PACKAGE=rpm ;;
redhat) DEFAULT_PACKAGE=rpm ;;
fedora) DEFAULT_PACKAGE=rpm ;;
gentoo) DEFAULT_PACKAGE=tgz ;;
arch) DEFAULT_PACKAGE=tgz ;;
sles) DEFAULT_PACKAGE=rpm ;;
slackware) DEFAULT_PACKAGE=tgz ;;
lunar) DEFAULT_PACKAGE=tgz ;;
ubuntu) DEFAULT_PACKAGE=deb ;;
debian) DEFAULT_PACKAGE=deb ;;
*) DEFAULT_PACKAGE=rpm ;;
esac
AC_MSG_RESULT([$DEFAULT_PACKAGE])
AC_SUBST(DEFAULT_PACKAGE)
])
dnl #
dnl # Default SPL user configuration
dnl #
AC_DEFUN([SPL_AC_PACKAGE], [
SPL_AC_DEFAULT_PACKAGE
SPL_AC_RPM
SPL_AC_DPKG
SPL_AC_ALIEN
])
AC_DEFUN([SPL_AC_LICENSE], [
AC_MSG_CHECKING([spl author])
AC_MSG_RESULT([$SPL_META_AUTHOR])
AC_MSG_CHECKING([spl license])
AC_MSG_RESULT([$SPL_META_LICENSE])
])
AC_DEFUN([SPL_AC_CONFIG], [
SPL_CONFIG=all
AC_ARG_WITH([config],
AS_HELP_STRING([--with-config=CONFIG],
[Config file 'kernel|user|all|srpm']),
[SPL_CONFIG="$withval"])
AC_ARG_ENABLE([linux-builtin],
[AC_HELP_STRING([--enable-linux-builtin],
[Configure for builtin in-tree kernel modules @<:@default=no@:>@])],
[],
[enable_linux_builtin=no])
AC_MSG_CHECKING([spl config])
AC_MSG_RESULT([$SPL_CONFIG]);
AC_SUBST(SPL_CONFIG)
case "$SPL_CONFIG" in
kernel) SPL_AC_CONFIG_KERNEL ;;
user) SPL_AC_CONFIG_USER ;;
all) SPL_AC_CONFIG_KERNEL
SPL_AC_CONFIG_USER ;;
srpm) ;;
*)
AC_MSG_RESULT([Error!])
AC_MSG_ERROR([Bad value "$SPL_CONFIG" for --with-config,
user kernel|user|all|srpm]) ;;
esac
AM_CONDITIONAL([CONFIG_USER],
[test "$SPL_CONFIG" = user -o "$SPL_CONFIG" = all])
AM_CONDITIONAL([CONFIG_KERNEL],
[test "$SPL_CONFIG" = kernel -o "$SPL_CONFIG" = all] &&
[test "x$enable_linux_builtin" != xyes ])
])
Autoconf --enable-debug-* cleanup Cleanup the --enable-debug-* configure options, this has been pending for quite some time and I am glad I finally got to it. To summerize: 1) All SPL_AC_DEBUG_* macros were updated to be a more autoconf friendly. This mainly involved shift to the GNU approved usage of AC_ARG_ENABLE and ensuring AS_IF is used rather than directly using an if [ test ] construct. 2) --enable-debug-kmem=yes by default. This simply enabled keeping a running tally of total memory allocated and freed and reporting a memory leak if there was one at module unload. Additionally, it ensure /proc/spl/kmem/slab will exist by default which is handy. The overhead is low for this and it should not impact performance. 3) --enable-debug-kmem-tracking=no by default. This option was added to provide a configure option to enable to detailed memory allocation tracking. This support was always there but you had to know where to turn it on. By default this support is disabled because it is known to badly hurt performence, however it is invaluable when chasing a memory leak. 4) --enable-debug-kstat removed. After further reflection I can't see why you would ever really want to turn this support off. It is now always on which had the nice side effect of simplifying the proc handling code in spl-proc.c. We can now always assume the top level directory will be there. 5) --enable-debug-callb removed. This never really did anything, it was put in provisionally because it might have been needed. It turns out it was not so I am just removing it to prevent confusion.
2009-10-30 20:58:51 +00:00
dnl #
dnl # Enable if the SPL should be compiled with internal debugging enabled.
dnl # By default this support is disabled.
dnl #
AC_DEFUN([SPL_AC_DEBUG], [
AC_MSG_CHECKING([whether debugging is enabled])
Autoconf --enable-debug-* cleanup Cleanup the --enable-debug-* configure options, this has been pending for quite some time and I am glad I finally got to it. To summerize: 1) All SPL_AC_DEBUG_* macros were updated to be a more autoconf friendly. This mainly involved shift to the GNU approved usage of AC_ARG_ENABLE and ensuring AS_IF is used rather than directly using an if [ test ] construct. 2) --enable-debug-kmem=yes by default. This simply enabled keeping a running tally of total memory allocated and freed and reporting a memory leak if there was one at module unload. Additionally, it ensure /proc/spl/kmem/slab will exist by default which is handy. The overhead is low for this and it should not impact performance. 3) --enable-debug-kmem-tracking=no by default. This option was added to provide a configure option to enable to detailed memory allocation tracking. This support was always there but you had to know where to turn it on. By default this support is disabled because it is known to badly hurt performence, however it is invaluable when chasing a memory leak. 4) --enable-debug-kstat removed. After further reflection I can't see why you would ever really want to turn this support off. It is now always on which had the nice side effect of simplifying the proc handling code in spl-proc.c. We can now always assume the top level directory will be there. 5) --enable-debug-callb removed. This never really did anything, it was put in provisionally because it might have been needed. It turns out it was not so I am just removing it to prevent confusion.
2009-10-30 20:58:51 +00:00
AC_ARG_ENABLE([debug],
[AS_HELP_STRING([--enable-debug],
[Enable generic debug support @<:@default=no@:>@])],
[],
[enable_debug=no])
AS_IF([test "x$enable_debug" = xyes],
[
KERNELCPPFLAGS="${KERNELCPPFLAGS} -DDEBUG -Werror"
DEBUG_CFLAGS="-DDEBUG -Werror"
DEBUG_SPL="_with_debug"
], [
KERNELCPPFLAGS="${KERNELCPPFLAGS} -DNDEBUG"
DEBUG_CFLAGS="-DNDEBUG"
DEBUG_SPL="_without_debug"
])
Autoconf --enable-debug-* cleanup Cleanup the --enable-debug-* configure options, this has been pending for quite some time and I am glad I finally got to it. To summerize: 1) All SPL_AC_DEBUG_* macros were updated to be a more autoconf friendly. This mainly involved shift to the GNU approved usage of AC_ARG_ENABLE and ensuring AS_IF is used rather than directly using an if [ test ] construct. 2) --enable-debug-kmem=yes by default. This simply enabled keeping a running tally of total memory allocated and freed and reporting a memory leak if there was one at module unload. Additionally, it ensure /proc/spl/kmem/slab will exist by default which is handy. The overhead is low for this and it should not impact performance. 3) --enable-debug-kmem-tracking=no by default. This option was added to provide a configure option to enable to detailed memory allocation tracking. This support was always there but you had to know where to turn it on. By default this support is disabled because it is known to badly hurt performence, however it is invaluable when chasing a memory leak. 4) --enable-debug-kstat removed. After further reflection I can't see why you would ever really want to turn this support off. It is now always on which had the nice side effect of simplifying the proc handling code in spl-proc.c. We can now always assume the top level directory will be there. 5) --enable-debug-callb removed. This never really did anything, it was put in provisionally because it might have been needed. It turns out it was not so I am just removing it to prevent confusion.
2009-10-30 20:58:51 +00:00
AC_SUBST(DEBUG_CFLAGS)
AC_SUBST(DEBUG_SPL)
Autoconf --enable-debug-* cleanup Cleanup the --enable-debug-* configure options, this has been pending for quite some time and I am glad I finally got to it. To summerize: 1) All SPL_AC_DEBUG_* macros were updated to be a more autoconf friendly. This mainly involved shift to the GNU approved usage of AC_ARG_ENABLE and ensuring AS_IF is used rather than directly using an if [ test ] construct. 2) --enable-debug-kmem=yes by default. This simply enabled keeping a running tally of total memory allocated and freed and reporting a memory leak if there was one at module unload. Additionally, it ensure /proc/spl/kmem/slab will exist by default which is handy. The overhead is low for this and it should not impact performance. 3) --enable-debug-kmem-tracking=no by default. This option was added to provide a configure option to enable to detailed memory allocation tracking. This support was always there but you had to know where to turn it on. By default this support is disabled because it is known to badly hurt performence, however it is invaluable when chasing a memory leak. 4) --enable-debug-kstat removed. After further reflection I can't see why you would ever really want to turn this support off. It is now always on which had the nice side effect of simplifying the proc handling code in spl-proc.c. We can now always assume the top level directory will be there. 5) --enable-debug-callb removed. This never really did anything, it was put in provisionally because it might have been needed. It turns out it was not so I am just removing it to prevent confusion.
2009-10-30 20:58:51 +00:00
AC_MSG_RESULT([$enable_debug])
])
dnl #
dnl # Enabled by default it provides a basic debug log infrastructure.
dnl # Each subsystem registers itself with a name and logs messages
dnl # using predefined types. If the debug mask it set to allow the
dnl # message type it will be written to the internal log. The log
dnl # can be dumped to a file by echoing 1 to the 'dump' proc entry,
dnl # after dumping the log it must be decoded using the spl utility.
dnl #
dnl # echo 1 >/proc/sys/kernel/spl/debug/dump
dnl # spl /tmp/spl-log.xxx.yyy /tmp/spl-log.xxx.yyy.txt
dnl #
AC_DEFUN([SPL_AC_DEBUG_LOG], [
AC_ARG_ENABLE([debug-log],
[AS_HELP_STRING([--enable-debug-log],
[Enable basic debug logging @<:@default=yes@:>@])],
[],
[enable_debug_log=yes])
AS_IF([test "x$enable_debug_log" = xyes],
[
KERNELCPPFLAGS="${KERNELCPPFLAGS} -DDEBUG_LOG"
DEBUG_LOG="_with_debug_log"
AC_DEFINE([DEBUG_LOG], [1],
[Define to 1 to enable basic debug logging])
], [
DEBUG_LOG="_without_debug_log"
])
AC_SUBST(DEBUG_LOG)
AC_MSG_CHECKING([whether basic debug logging is enabled])
AC_MSG_RESULT([$enable_debug_log])
])
Autoconf --enable-debug-* cleanup Cleanup the --enable-debug-* configure options, this has been pending for quite some time and I am glad I finally got to it. To summerize: 1) All SPL_AC_DEBUG_* macros were updated to be a more autoconf friendly. This mainly involved shift to the GNU approved usage of AC_ARG_ENABLE and ensuring AS_IF is used rather than directly using an if [ test ] construct. 2) --enable-debug-kmem=yes by default. This simply enabled keeping a running tally of total memory allocated and freed and reporting a memory leak if there was one at module unload. Additionally, it ensure /proc/spl/kmem/slab will exist by default which is handy. The overhead is low for this and it should not impact performance. 3) --enable-debug-kmem-tracking=no by default. This option was added to provide a configure option to enable to detailed memory allocation tracking. This support was always there but you had to know where to turn it on. By default this support is disabled because it is known to badly hurt performence, however it is invaluable when chasing a memory leak. 4) --enable-debug-kstat removed. After further reflection I can't see why you would ever really want to turn this support off. It is now always on which had the nice side effect of simplifying the proc handling code in spl-proc.c. We can now always assume the top level directory will be there. 5) --enable-debug-callb removed. This never really did anything, it was put in provisionally because it might have been needed. It turns out it was not so I am just removing it to prevent confusion.
2009-10-30 20:58:51 +00:00
dnl #
dnl # Enabled by default it provides a minimal level of memory tracking.
dnl # A total count of bytes allocated is kept for each alloc and free.
dnl # Then at module unload time a report to the console will be printed
dnl # if memory was leaked. Additionally, /proc/spl/kmem/slab will exist
dnl # and provide an easy way to inspect the kmem based slab.
dnl #
AC_DEFUN([SPL_AC_DEBUG_KMEM], [
Autoconf --enable-debug-* cleanup Cleanup the --enable-debug-* configure options, this has been pending for quite some time and I am glad I finally got to it. To summerize: 1) All SPL_AC_DEBUG_* macros were updated to be a more autoconf friendly. This mainly involved shift to the GNU approved usage of AC_ARG_ENABLE and ensuring AS_IF is used rather than directly using an if [ test ] construct. 2) --enable-debug-kmem=yes by default. This simply enabled keeping a running tally of total memory allocated and freed and reporting a memory leak if there was one at module unload. Additionally, it ensure /proc/spl/kmem/slab will exist by default which is handy. The overhead is low for this and it should not impact performance. 3) --enable-debug-kmem-tracking=no by default. This option was added to provide a configure option to enable to detailed memory allocation tracking. This support was always there but you had to know where to turn it on. By default this support is disabled because it is known to badly hurt performence, however it is invaluable when chasing a memory leak. 4) --enable-debug-kstat removed. After further reflection I can't see why you would ever really want to turn this support off. It is now always on which had the nice side effect of simplifying the proc handling code in spl-proc.c. We can now always assume the top level directory will be there. 5) --enable-debug-callb removed. This never really did anything, it was put in provisionally because it might have been needed. It turns out it was not so I am just removing it to prevent confusion.
2009-10-30 20:58:51 +00:00
AC_ARG_ENABLE([debug-kmem],
[AS_HELP_STRING([--enable-debug-kmem],
[Enable basic kmem accounting @<:@default=yes@:>@])],
[],
[enable_debug_kmem=yes])
Autoconf --enable-debug-* cleanup Cleanup the --enable-debug-* configure options, this has been pending for quite some time and I am glad I finally got to it. To summerize: 1) All SPL_AC_DEBUG_* macros were updated to be a more autoconf friendly. This mainly involved shift to the GNU approved usage of AC_ARG_ENABLE and ensuring AS_IF is used rather than directly using an if [ test ] construct. 2) --enable-debug-kmem=yes by default. This simply enabled keeping a running tally of total memory allocated and freed and reporting a memory leak if there was one at module unload. Additionally, it ensure /proc/spl/kmem/slab will exist by default which is handy. The overhead is low for this and it should not impact performance. 3) --enable-debug-kmem-tracking=no by default. This option was added to provide a configure option to enable to detailed memory allocation tracking. This support was always there but you had to know where to turn it on. By default this support is disabled because it is known to badly hurt performence, however it is invaluable when chasing a memory leak. 4) --enable-debug-kstat removed. After further reflection I can't see why you would ever really want to turn this support off. It is now always on which had the nice side effect of simplifying the proc handling code in spl-proc.c. We can now always assume the top level directory will be there. 5) --enable-debug-callb removed. This never really did anything, it was put in provisionally because it might have been needed. It turns out it was not so I am just removing it to prevent confusion.
2009-10-30 20:58:51 +00:00
AS_IF([test "x$enable_debug_kmem" = xyes],
[
KERNELCPPFLAGS="${KERNELCPPFLAGS} -DDEBUG_KMEM"
DEBUG_KMEM="_with_debug_kmem"
AC_DEFINE([DEBUG_KMEM], [1],
Autoconf --enable-debug-* cleanup Cleanup the --enable-debug-* configure options, this has been pending for quite some time and I am glad I finally got to it. To summerize: 1) All SPL_AC_DEBUG_* macros were updated to be a more autoconf friendly. This mainly involved shift to the GNU approved usage of AC_ARG_ENABLE and ensuring AS_IF is used rather than directly using an if [ test ] construct. 2) --enable-debug-kmem=yes by default. This simply enabled keeping a running tally of total memory allocated and freed and reporting a memory leak if there was one at module unload. Additionally, it ensure /proc/spl/kmem/slab will exist by default which is handy. The overhead is low for this and it should not impact performance. 3) --enable-debug-kmem-tracking=no by default. This option was added to provide a configure option to enable to detailed memory allocation tracking. This support was always there but you had to know where to turn it on. By default this support is disabled because it is known to badly hurt performence, however it is invaluable when chasing a memory leak. 4) --enable-debug-kstat removed. After further reflection I can't see why you would ever really want to turn this support off. It is now always on which had the nice side effect of simplifying the proc handling code in spl-proc.c. We can now always assume the top level directory will be there. 5) --enable-debug-callb removed. This never really did anything, it was put in provisionally because it might have been needed. It turns out it was not so I am just removing it to prevent confusion.
2009-10-30 20:58:51 +00:00
[Define to 1 to enable basic kmem accounting])
], [
DEBUG_KMEM="_without_debug_kmem"
])
Autoconf --enable-debug-* cleanup Cleanup the --enable-debug-* configure options, this has been pending for quite some time and I am glad I finally got to it. To summerize: 1) All SPL_AC_DEBUG_* macros were updated to be a more autoconf friendly. This mainly involved shift to the GNU approved usage of AC_ARG_ENABLE and ensuring AS_IF is used rather than directly using an if [ test ] construct. 2) --enable-debug-kmem=yes by default. This simply enabled keeping a running tally of total memory allocated and freed and reporting a memory leak if there was one at module unload. Additionally, it ensure /proc/spl/kmem/slab will exist by default which is handy. The overhead is low for this and it should not impact performance. 3) --enable-debug-kmem-tracking=no by default. This option was added to provide a configure option to enable to detailed memory allocation tracking. This support was always there but you had to know where to turn it on. By default this support is disabled because it is known to badly hurt performence, however it is invaluable when chasing a memory leak. 4) --enable-debug-kstat removed. After further reflection I can't see why you would ever really want to turn this support off. It is now always on which had the nice side effect of simplifying the proc handling code in spl-proc.c. We can now always assume the top level directory will be there. 5) --enable-debug-callb removed. This never really did anything, it was put in provisionally because it might have been needed. It turns out it was not so I am just removing it to prevent confusion.
2009-10-30 20:58:51 +00:00
AC_SUBST(DEBUG_KMEM)
Autoconf --enable-debug-* cleanup Cleanup the --enable-debug-* configure options, this has been pending for quite some time and I am glad I finally got to it. To summerize: 1) All SPL_AC_DEBUG_* macros were updated to be a more autoconf friendly. This mainly involved shift to the GNU approved usage of AC_ARG_ENABLE and ensuring AS_IF is used rather than directly using an if [ test ] construct. 2) --enable-debug-kmem=yes by default. This simply enabled keeping a running tally of total memory allocated and freed and reporting a memory leak if there was one at module unload. Additionally, it ensure /proc/spl/kmem/slab will exist by default which is handy. The overhead is low for this and it should not impact performance. 3) --enable-debug-kmem-tracking=no by default. This option was added to provide a configure option to enable to detailed memory allocation tracking. This support was always there but you had to know where to turn it on. By default this support is disabled because it is known to badly hurt performence, however it is invaluable when chasing a memory leak. 4) --enable-debug-kstat removed. After further reflection I can't see why you would ever really want to turn this support off. It is now always on which had the nice side effect of simplifying the proc handling code in spl-proc.c. We can now always assume the top level directory will be there. 5) --enable-debug-callb removed. This never really did anything, it was put in provisionally because it might have been needed. It turns out it was not so I am just removing it to prevent confusion.
2009-10-30 20:58:51 +00:00
AC_MSG_CHECKING([whether basic kmem accounting is enabled])
AC_MSG_RESULT([$enable_debug_kmem])
])
Autoconf --enable-debug-* cleanup Cleanup the --enable-debug-* configure options, this has been pending for quite some time and I am glad I finally got to it. To summerize: 1) All SPL_AC_DEBUG_* macros were updated to be a more autoconf friendly. This mainly involved shift to the GNU approved usage of AC_ARG_ENABLE and ensuring AS_IF is used rather than directly using an if [ test ] construct. 2) --enable-debug-kmem=yes by default. This simply enabled keeping a running tally of total memory allocated and freed and reporting a memory leak if there was one at module unload. Additionally, it ensure /proc/spl/kmem/slab will exist by default which is handy. The overhead is low for this and it should not impact performance. 3) --enable-debug-kmem-tracking=no by default. This option was added to provide a configure option to enable to detailed memory allocation tracking. This support was always there but you had to know where to turn it on. By default this support is disabled because it is known to badly hurt performence, however it is invaluable when chasing a memory leak. 4) --enable-debug-kstat removed. After further reflection I can't see why you would ever really want to turn this support off. It is now always on which had the nice side effect of simplifying the proc handling code in spl-proc.c. We can now always assume the top level directory will be there. 5) --enable-debug-callb removed. This never really did anything, it was put in provisionally because it might have been needed. It turns out it was not so I am just removing it to prevent confusion.
2009-10-30 20:58:51 +00:00
dnl #
dnl # Disabled by default it provides detailed memory tracking. This
dnl # feature also requires --enable-debug-kmem to be set. When enabled
dnl # not only will total bytes be tracked but also the location of every
dnl # alloc and free. When the SPL module is unloaded a list of all leaked
dnl # addresses and where they were allocated will be dumped to the console.
dnl # Enabling this feature has a significant impact on performance but it
dnl # makes finding memory leaks pretty straight forward.
dnl #
AC_DEFUN([SPL_AC_DEBUG_KMEM_TRACKING], [
AC_ARG_ENABLE([debug-kmem-tracking],
[AS_HELP_STRING([--enable-debug-kmem-tracking],
[Enable detailed kmem tracking @<:@default=no@:>@])],
[],
[enable_debug_kmem_tracking=no])
AS_IF([test "x$enable_debug_kmem_tracking" = xyes],
[
KERNELCPPFLAGS="${KERNELCPPFLAGS} -DDEBUG_KMEM_TRACKING"
DEBUG_KMEM_TRACKING="_with_debug_kmem_tracking"
AC_DEFINE([DEBUG_KMEM_TRACKING], [1],
Autoconf --enable-debug-* cleanup Cleanup the --enable-debug-* configure options, this has been pending for quite some time and I am glad I finally got to it. To summerize: 1) All SPL_AC_DEBUG_* macros were updated to be a more autoconf friendly. This mainly involved shift to the GNU approved usage of AC_ARG_ENABLE and ensuring AS_IF is used rather than directly using an if [ test ] construct. 2) --enable-debug-kmem=yes by default. This simply enabled keeping a running tally of total memory allocated and freed and reporting a memory leak if there was one at module unload. Additionally, it ensure /proc/spl/kmem/slab will exist by default which is handy. The overhead is low for this and it should not impact performance. 3) --enable-debug-kmem-tracking=no by default. This option was added to provide a configure option to enable to detailed memory allocation tracking. This support was always there but you had to know where to turn it on. By default this support is disabled because it is known to badly hurt performence, however it is invaluable when chasing a memory leak. 4) --enable-debug-kstat removed. After further reflection I can't see why you would ever really want to turn this support off. It is now always on which had the nice side effect of simplifying the proc handling code in spl-proc.c. We can now always assume the top level directory will be there. 5) --enable-debug-callb removed. This never really did anything, it was put in provisionally because it might have been needed. It turns out it was not so I am just removing it to prevent confusion.
2009-10-30 20:58:51 +00:00
[Define to 1 to enable detailed kmem tracking])
], [
DEBUG_KMEM_TRACKING="_without_debug_kmem_tracking"
])
Autoconf --enable-debug-* cleanup Cleanup the --enable-debug-* configure options, this has been pending for quite some time and I am glad I finally got to it. To summerize: 1) All SPL_AC_DEBUG_* macros were updated to be a more autoconf friendly. This mainly involved shift to the GNU approved usage of AC_ARG_ENABLE and ensuring AS_IF is used rather than directly using an if [ test ] construct. 2) --enable-debug-kmem=yes by default. This simply enabled keeping a running tally of total memory allocated and freed and reporting a memory leak if there was one at module unload. Additionally, it ensure /proc/spl/kmem/slab will exist by default which is handy. The overhead is low for this and it should not impact performance. 3) --enable-debug-kmem-tracking=no by default. This option was added to provide a configure option to enable to detailed memory allocation tracking. This support was always there but you had to know where to turn it on. By default this support is disabled because it is known to badly hurt performence, however it is invaluable when chasing a memory leak. 4) --enable-debug-kstat removed. After further reflection I can't see why you would ever really want to turn this support off. It is now always on which had the nice side effect of simplifying the proc handling code in spl-proc.c. We can now always assume the top level directory will be there. 5) --enable-debug-callb removed. This never really did anything, it was put in provisionally because it might have been needed. It turns out it was not so I am just removing it to prevent confusion.
2009-10-30 20:58:51 +00:00
AC_SUBST(DEBUG_KMEM_TRACKING)
Autoconf --enable-debug-* cleanup Cleanup the --enable-debug-* configure options, this has been pending for quite some time and I am glad I finally got to it. To summerize: 1) All SPL_AC_DEBUG_* macros were updated to be a more autoconf friendly. This mainly involved shift to the GNU approved usage of AC_ARG_ENABLE and ensuring AS_IF is used rather than directly using an if [ test ] construct. 2) --enable-debug-kmem=yes by default. This simply enabled keeping a running tally of total memory allocated and freed and reporting a memory leak if there was one at module unload. Additionally, it ensure /proc/spl/kmem/slab will exist by default which is handy. The overhead is low for this and it should not impact performance. 3) --enable-debug-kmem-tracking=no by default. This option was added to provide a configure option to enable to detailed memory allocation tracking. This support was always there but you had to know where to turn it on. By default this support is disabled because it is known to badly hurt performence, however it is invaluable when chasing a memory leak. 4) --enable-debug-kstat removed. After further reflection I can't see why you would ever really want to turn this support off. It is now always on which had the nice side effect of simplifying the proc handling code in spl-proc.c. We can now always assume the top level directory will be there. 5) --enable-debug-callb removed. This never really did anything, it was put in provisionally because it might have been needed. It turns out it was not so I am just removing it to prevent confusion.
2009-10-30 20:58:51 +00:00
AC_MSG_CHECKING([whether detailed kmem tracking is enabled])
AC_MSG_RESULT([$enable_debug_kmem_tracking])
])
dnl #
dnl # SPL_LINUX_CONFTEST
dnl #
AC_DEFUN([SPL_LINUX_CONFTEST], [
cat confdefs.h - <<_ACEOF >conftest.c
$1
_ACEOF
])
dnl #
dnl # SPL_LANG_PROGRAM(C)([PROLOGUE], [BODY])
dnl #
m4_define([SPL_LANG_PROGRAM], [
$1
int
main (void)
{
dnl Do *not* indent the following line: there may be CPP directives.
dnl Don't move the `;' right after for the same reason.
$2
;
return 0;
}
])
dnl #
dnl # SPL_LINUX_COMPILE_IFELSE / like AC_COMPILE_IFELSE
dnl #
AC_DEFUN([SPL_LINUX_COMPILE_IFELSE], [
m4_ifvaln([$1], [SPL_LINUX_CONFTEST([$1])])
rm -Rf build && mkdir -p build && touch build/conftest.mod.c
echo "obj-m := conftest.o" >build/Makefile
modpost_flag=''
test "x$enable_linux_builtin" = xyes && modpost_flag='modpost=true' # fake modpost stage
AS_IF(
[AC_TRY_COMMAND(cp conftest.c build && make [$2] -C $LINUX_OBJ EXTRA_CFLAGS="-Werror-implicit-function-declaration $EXTRA_KCFLAGS" $ARCH_UM M=$PWD/build $modpost_flag) >/dev/null && AC_TRY_COMMAND([$3])],
[$4],
[_AC_MSG_LOG_CONFTEST m4_ifvaln([$5],[$5])]
)
rm -Rf build
])
dnl #
dnl # SPL_LINUX_TRY_COMPILE like AC_TRY_COMPILE
dnl #
AC_DEFUN([SPL_LINUX_TRY_COMPILE],
[SPL_LINUX_COMPILE_IFELSE(
[AC_LANG_SOURCE([SPL_LANG_PROGRAM([[$1]], [[$2]])])],
[modules],
[test -s build/conftest.o],
[$3], [$4])
])
dnl #
dnl # SPL_CHECK_SYMBOL_EXPORT
dnl # check symbol exported or not
dnl #
AC_DEFUN([SPL_CHECK_SYMBOL_EXPORT], [
grep -q -E '[[[:space:]]]$1[[[:space:]]]' \
$LINUX_OBJ/Module*.symvers 2>/dev/null
rc=$?
if test $rc -ne 0; then
export=0
for file in $2; do
grep -q -E "EXPORT_SYMBOL.*($1)" \
"$LINUX_OBJ/$file" 2>/dev/null
rc=$?
if test $rc -eq 0; then
export=1
break;
fi
done
if test $export -eq 0; then :
$4
else :
$3
fi
else :
$3
fi
])
dnl #
dnl # SPL_LINUX_TRY_COMPILE_SYMBOL
dnl # like SPL_LINUX_TRY_COMPILE, except SPL_CHECK_SYMBOL_EXPORT
dnl # is called if not compiling for builtin
dnl #
AC_DEFUN([SPL_LINUX_TRY_COMPILE_SYMBOL], [
SPL_LINUX_TRY_COMPILE([$1], [$2], [rc=0], [rc=1])
if test $rc -ne 0; then :
$6
else
if test "x$enable_linux_builtin" != xyes; then
SPL_CHECK_SYMBOL_EXPORT([$3], [$4], [rc=0], [rc=1])
fi
if test $rc -ne 0; then :
$6
else :
$5
fi
fi
])
dnl #
dnl # SPL_CHECK_SYMBOL_HEADER
dnl # check if a symbol prototype is defined in listed headers.
dnl #
AC_DEFUN([SPL_CHECK_SYMBOL_HEADER], [
AC_MSG_CHECKING([whether symbol $1 exists in header])
header=0
for file in $3; do
grep -q "$2" "$LINUX/$file" 2>/dev/null
rc=$?
if test $rc -eq 0; then
header=1
break;
fi
done
if test $header -eq 0; then
AC_MSG_RESULT([no])
$5
else
AC_MSG_RESULT([yes])
$4
fi
])
dnl #
dnl # SPL_CHECK_HEADER
dnl # check whether header exists and define HAVE_$2_HEADER
dnl #
AC_DEFUN([SPL_CHECK_HEADER],
[AC_MSG_CHECKING([whether header $1 exists])
SPL_LINUX_TRY_COMPILE([
#include <$1>
],[
return 0;
],[
AC_DEFINE(HAVE_$2_HEADER, 1, [$1 exists])
AC_MSG_RESULT(yes)
$3
],[
AC_MSG_RESULT(no)
$4
])
])
dnl #
dnl # Basic toolchain sanity check. Verify that kernel modules can
dnl # be built and which symbols can be used.
dnl #
AC_DEFUN([SPL_AC_TEST_MODULE],
[AC_MSG_CHECKING([whether modules can be built])
SPL_LINUX_TRY_COMPILE([],[],[
AC_MSG_RESULT([yes])
],[
AC_MSG_RESULT([no])
if test "x$enable_linux_builtin" != xyes; then
AC_MSG_ERROR([*** Unable to build an empty module.])
else
AC_MSG_ERROR([
*** Unable to build an empty module.
*** Please run 'make scripts' inside the kernel source tree.])
fi
])
AC_RUN_IFELSE([
AC_LANG_PROGRAM([
#include "$LINUX/include/linux/license.h"
], [
return !license_is_gpl_compatible("$SPL_META_LICENSE");
])
], [
AC_DEFINE([SPL_IS_GPL_COMPATIBLE], [1],
[Define to 1 if GPL-only symbols can be used])
], [
])
])
dnl #
dnl # Use the atomic implemenation based on global spinlocks. This
dnl # should only be needed by 32-bit kernels which do not provide
dnl # the atomic64_* API. It may be optionally enabled as a fallback
dnl # if problems are observed with the direct mapping to the native
dnl # Linux atomic operations. You may not disable atomic spinlocks
dnl # if you kernel does not an atomic64_* API.
dnl #
AC_DEFUN([SPL_AC_ATOMIC_SPINLOCK], [
AC_ARG_ENABLE([atomic-spinlocks],
[AS_HELP_STRING([--enable-atomic-spinlocks],
[Atomic types use spinlocks @<:@default=check@:>@])],
[],
[enable_atomic_spinlocks=check])
SPL_LINUX_TRY_COMPILE([
#include <linux/fs.h>
],[
atomic64_t *ptr __attribute__ ((unused));
],[
have_atomic64_t=yes
AC_DEFINE(HAVE_ATOMIC64_T, 1,
[kernel defines atomic64_t])
],[
have_atomic64_t=no
])
AS_IF([test "x$enable_atomic_spinlocks" = xcheck], [
AS_IF([test "x$have_atomic64_t" = xyes], [
enable_atomic_spinlocks=no
],[
enable_atomic_spinlocks=yes
])
])
AS_IF([test "x$enable_atomic_spinlocks" = xyes], [
AC_DEFINE([ATOMIC_SPINLOCK], [1],
[Atomic types use spinlocks])
],[
AS_IF([test "x$have_atomic64_t" = xno], [
AC_MSG_FAILURE(
[--disable-atomic-spinlocks given but required atomic64 support is unavailable])
])
])
AC_MSG_CHECKING([whether atomic types use spinlocks])
AC_MSG_RESULT([$enable_atomic_spinlocks])
AC_MSG_CHECKING([whether kernel defines atomic64_t])
AC_MSG_RESULT([$have_atomic64_t])
])
AC_DEFUN([SPL_AC_SHRINKER_CALLBACK],[
tmp_flags="$EXTRA_KCFLAGS"
EXTRA_KCFLAGS="-Werror"
dnl #
dnl # 2.6.23 to 2.6.34 API change
dnl # ->shrink(int nr_to_scan, gfp_t gfp_mask)
dnl #
AC_MSG_CHECKING([whether old 2-argument shrinker exists])
SPL_LINUX_TRY_COMPILE([
#include <linux/mm.h>
int shrinker_cb(int nr_to_scan, gfp_t gfp_mask);
],[
struct shrinker cache_shrinker = {
.shrink = shrinker_cb,
.seeks = DEFAULT_SEEKS,
};
register_shrinker(&cache_shrinker);
],[
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_2ARGS_OLD_SHRINKER_CALLBACK, 1,
[old shrinker callback wants 2 args])
],[
AC_MSG_RESULT(no)
dnl #
dnl # 2.6.35 - 2.6.39 API change
dnl # ->shrink(struct shrinker *,
dnl # int nr_to_scan, gfp_t gfp_mask)
dnl #
AC_MSG_CHECKING([whether old 3-argument shrinker exists])
SPL_LINUX_TRY_COMPILE([
#include <linux/mm.h>
int shrinker_cb(struct shrinker *, int nr_to_scan,
gfp_t gfp_mask);
],[
struct shrinker cache_shrinker = {
.shrink = shrinker_cb,
.seeks = DEFAULT_SEEKS,
};
register_shrinker(&cache_shrinker);
],[
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_3ARGS_SHRINKER_CALLBACK, 1,
[old shrinker callback wants 3 args])
],[
AC_MSG_RESULT(no)
dnl #
dnl # 3.0 - 3.11 API change
dnl # ->shrink(struct shrinker *,
dnl # struct shrink_control *sc)
dnl #
AC_MSG_CHECKING(
[whether new 2-argument shrinker exists])
SPL_LINUX_TRY_COMPILE([
#include <linux/mm.h>
int shrinker_cb(struct shrinker *,
struct shrink_control *sc);
],[
struct shrinker cache_shrinker = {
.shrink = shrinker_cb,
.seeks = DEFAULT_SEEKS,
};
register_shrinker(&cache_shrinker);
],[
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_2ARGS_NEW_SHRINKER_CALLBACK, 1,
[new shrinker callback wants 2 args])
],[
AC_MSG_RESULT(no)
dnl #
dnl # 3.12 API change,
dnl # ->shrink() is logically split in to
dnl # ->count_objects() and ->scan_objects()
dnl #
AC_MSG_CHECKING(
[whether ->count_objects callback exists])
SPL_LINUX_TRY_COMPILE([
#include <linux/mm.h>
unsigned long shrinker_cb(
struct shrinker *,
struct shrink_control *sc);
],[
struct shrinker cache_shrinker = {
.count_objects = shrinker_cb,
.scan_objects = shrinker_cb,
.seeks = DEFAULT_SEEKS,
};
register_shrinker(&cache_shrinker);
],[
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_SPLIT_SHRINKER_CALLBACK,
1, [->count_objects exists])
],[
AC_MSG_ERROR(error)
])
])
])
])
EXTRA_KCFLAGS="$tmp_flags"
])
dnl #
dnl # 2.6.33 API change,
dnl # Removed .ctl_name from struct ctl_table.
dnl #
AC_DEFUN([SPL_AC_CTL_NAME], [
AC_MSG_CHECKING([whether struct ctl_table has ctl_name])
SPL_LINUX_TRY_COMPILE([
#include <linux/sysctl.h>
],[
struct ctl_table ctl __attribute__ ((unused));
ctl.ctl_name = 0;
],[
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_CTL_NAME, 1, [struct ctl_table has ctl_name])
],[
AC_MSG_RESULT(no)
])
])
Reimplement mutexs for Linux lock profiling/analysis For a generic explanation of why mutexs needed to be reimplemented to work with the kernel lock profiling see commits: e811949a57044d60d12953c5c3b808a79a7d36ef and d28db80fd0fd4fd63aec09037c44408e51a222d6 The specific changes made to the mutex implemetation are as follows. The Linux mutex structure is now directly embedded in the kmutex_t. This allows a kmutex_t to be directly case to a mutex struct and passed directly to the Linux primative. Just like with the rwlocks it is critical that these functions be implemented as '#defines to ensure the location information is preserved. The preprocessor can then do a direct replacement of the Solaris primative with the linux primative. Just as with the rwlocks we need to track the lock owner. Here things get a little more interesting because depending on your kernel version, and how you've built your kernel Linux may already do this for you. If your running a 2.6.29 or newer kernel on a SMP system the lock owner will be tracked. This was added to Linux to support adaptive mutexs, more on that shortly. Alternately, your kernel might track the lock owner if you've set CONFIG_DEBUG_MUTEXES in the kernel build. If neither of the above things is true for your kernel the kmutex_t type will include and track the lock owner to ensure correct behavior. This is all handled by a new autoconf check called SPL_AC_MUTEX_OWNER. Concerning adaptive mutexs these are a very recent development and they did not make it in to either the latest FC11 of SLES11 kernels. Ideally, I'd love to see this kernel change appear in one of these distros because it does help performance. From Linux kernel commit: 0d66bf6d3514b35eb6897629059443132992dbd7 "Testing with Ingo's test-mutex application... gave a 345% boost for VFS scalability on my testbox" However, if you don't want to backport this change yourself you can still simply export the task_curr() symbol. The kmutex_t implementation will use this symbol when it's available to provide it's own adaptive mutexs. Finally, DEBUG_MUTEX support was removed including the proc handlers. This was done because now that we are cleanly integrated with the kernel profiling all this information and much much more is available in debug kernel builds. This code was now redundant. Update mutexs validated on: - SLES10 (ppc64) - SLES11 (x86_64) - CHAOS4.2 (x86_64) - RHEL5.3 (x86_64) - RHEL6 (x86_64) - FC11 (x86_64)
2009-09-25 21:47:01 +00:00
dnl #
dnl # 2.6.29 API change,
dnl # Adaptive mutexs were introduced which track the mutex owner. The
dnl # mutex wrappers leverage this functionality to avoid tracking the
dnl # owner multipe times.
Reimplement mutexs for Linux lock profiling/analysis For a generic explanation of why mutexs needed to be reimplemented to work with the kernel lock profiling see commits: e811949a57044d60d12953c5c3b808a79a7d36ef and d28db80fd0fd4fd63aec09037c44408e51a222d6 The specific changes made to the mutex implemetation are as follows. The Linux mutex structure is now directly embedded in the kmutex_t. This allows a kmutex_t to be directly case to a mutex struct and passed directly to the Linux primative. Just like with the rwlocks it is critical that these functions be implemented as '#defines to ensure the location information is preserved. The preprocessor can then do a direct replacement of the Solaris primative with the linux primative. Just as with the rwlocks we need to track the lock owner. Here things get a little more interesting because depending on your kernel version, and how you've built your kernel Linux may already do this for you. If your running a 2.6.29 or newer kernel on a SMP system the lock owner will be tracked. This was added to Linux to support adaptive mutexs, more on that shortly. Alternately, your kernel might track the lock owner if you've set CONFIG_DEBUG_MUTEXES in the kernel build. If neither of the above things is true for your kernel the kmutex_t type will include and track the lock owner to ensure correct behavior. This is all handled by a new autoconf check called SPL_AC_MUTEX_OWNER. Concerning adaptive mutexs these are a very recent development and they did not make it in to either the latest FC11 of SLES11 kernels. Ideally, I'd love to see this kernel change appear in one of these distros because it does help performance. From Linux kernel commit: 0d66bf6d3514b35eb6897629059443132992dbd7 "Testing with Ingo's test-mutex application... gave a 345% boost for VFS scalability on my testbox" However, if you don't want to backport this change yourself you can still simply export the task_curr() symbol. The kmutex_t implementation will use this symbol when it's available to provide it's own adaptive mutexs. Finally, DEBUG_MUTEX support was removed including the proc handlers. This was done because now that we are cleanly integrated with the kernel profiling all this information and much much more is available in debug kernel builds. This code was now redundant. Update mutexs validated on: - SLES10 (ppc64) - SLES11 (x86_64) - CHAOS4.2 (x86_64) - RHEL5.3 (x86_64) - RHEL6 (x86_64) - FC11 (x86_64)
2009-09-25 21:47:01 +00:00
dnl #
AC_DEFUN([SPL_AC_MUTEX_OWNER], [
AC_MSG_CHECKING([whether struct mutex has owner])
SPL_LINUX_TRY_COMPILE([
#include <linux/mutex.h>
],[
struct mutex mtx __attribute__ ((unused));
Reimplement mutexs for Linux lock profiling/analysis For a generic explanation of why mutexs needed to be reimplemented to work with the kernel lock profiling see commits: e811949a57044d60d12953c5c3b808a79a7d36ef and d28db80fd0fd4fd63aec09037c44408e51a222d6 The specific changes made to the mutex implemetation are as follows. The Linux mutex structure is now directly embedded in the kmutex_t. This allows a kmutex_t to be directly case to a mutex struct and passed directly to the Linux primative. Just like with the rwlocks it is critical that these functions be implemented as '#defines to ensure the location information is preserved. The preprocessor can then do a direct replacement of the Solaris primative with the linux primative. Just as with the rwlocks we need to track the lock owner. Here things get a little more interesting because depending on your kernel version, and how you've built your kernel Linux may already do this for you. If your running a 2.6.29 or newer kernel on a SMP system the lock owner will be tracked. This was added to Linux to support adaptive mutexs, more on that shortly. Alternately, your kernel might track the lock owner if you've set CONFIG_DEBUG_MUTEXES in the kernel build. If neither of the above things is true for your kernel the kmutex_t type will include and track the lock owner to ensure correct behavior. This is all handled by a new autoconf check called SPL_AC_MUTEX_OWNER. Concerning adaptive mutexs these are a very recent development and they did not make it in to either the latest FC11 of SLES11 kernels. Ideally, I'd love to see this kernel change appear in one of these distros because it does help performance. From Linux kernel commit: 0d66bf6d3514b35eb6897629059443132992dbd7 "Testing with Ingo's test-mutex application... gave a 345% boost for VFS scalability on my testbox" However, if you don't want to backport this change yourself you can still simply export the task_curr() symbol. The kmutex_t implementation will use this symbol when it's available to provide it's own adaptive mutexs. Finally, DEBUG_MUTEX support was removed including the proc handlers. This was done because now that we are cleanly integrated with the kernel profiling all this information and much much more is available in debug kernel builds. This code was now redundant. Update mutexs validated on: - SLES10 (ppc64) - SLES11 (x86_64) - CHAOS4.2 (x86_64) - RHEL5.3 (x86_64) - RHEL6 (x86_64) - FC11 (x86_64)
2009-09-25 21:47:01 +00:00
mtx.owner = NULL;
],[
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_MUTEX_OWNER, 1, [struct mutex has owner])
],[
AC_MSG_RESULT(no)
])
])
dnl #
dnl # 2.6.39 API change,
dnl # Owner type change. A Linux mutex prior to 2.6.39 would store
dnl # the owner as a thread_info pointer when CONFIG_DEBUG_MUTEXES
dnl # was defined. As of 2.6.39 this was changed to a task_struct
dnl # pointer which frankly makes a lot more sense.
dnl #
AC_DEFUN([SPL_AC_MUTEX_OWNER_TASK_STRUCT], [
AC_MSG_CHECKING([whether struct mutex owner is a task_struct])
tmp_flags="$EXTRA_KCFLAGS"
EXTRA_KCFLAGS="-Werror"
SPL_LINUX_TRY_COMPILE([
#include <linux/mutex.h>
#include <linux/sched.h>
],[
struct mutex mtx __attribute__ ((unused));
mtx.owner = current;
],[
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_MUTEX_OWNER_TASK_STRUCT, 1,
[struct mutex owner is a task_struct])
],[
AC_MSG_RESULT(no)
])
EXTRA_KCFLAGS="$tmp_flags"
])
dnl #
dnl # 3.10 API change,
dnl # PDE is replaced by PDE_DATA
dnl #
AC_DEFUN([SPL_AC_PDE_DATA], [
AC_MSG_CHECKING([whether PDE_DATA() is available])
SPL_LINUX_TRY_COMPILE_SYMBOL([
#include <linux/proc_fs.h>
], [
PDE_DATA(NULL);
], [PDE_DATA], [], [
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_PDE_DATA, 1, [yes])
],[
AC_MSG_RESULT(no)
])
])
dnl #
dnl # 3.9 API change
dnl # set_fs_pwd takes const struct path *
dnl #
AC_DEFUN([SPL_AC_SET_FS_PWD_WITH_CONST],
tmp_flags="$EXTRA_KCFLAGS"
EXTRA_KCFLAGS="-Werror"
[AC_MSG_CHECKING([whether set_fs_pwd() requires const struct path *])
SPL_LINUX_TRY_COMPILE([
#include <linux/spinlock.h>
#include <linux/fs_struct.h>
#include <linux/path.h>
void (*const set_fs_pwd_func)
(struct fs_struct *, const struct path *)
= set_fs_pwd;
],[
return 0;
],[
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_SET_FS_PWD_WITH_CONST, 1,
[set_fs_pwd() needs const path *])
],[
SPL_LINUX_TRY_COMPILE([
#include <linux/spinlock.h>
#include <linux/fs_struct.h>
#include <linux/path.h>
void (*const set_fs_pwd_func)
(struct fs_struct *, struct path *)
= set_fs_pwd;
],[
return 0;
],[
AC_MSG_RESULT(no)
],[
AC_MSG_ERROR(unknown)
])
])
EXTRA_KCFLAGS="$tmp_flags"
])
dnl #
dnl # 3.13 API change
dnl # vfs_unlink() updated to take a third delegated_inode argument.
dnl #
AC_DEFUN([SPL_AC_2ARGS_VFS_UNLINK],
[AC_MSG_CHECKING([whether vfs_unlink() wants 2 args])
SPL_LINUX_TRY_COMPILE([
#include <linux/fs.h>
],[
vfs_unlink((struct inode *) NULL, (struct dentry *) NULL);
],[
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_2ARGS_VFS_UNLINK, 1,
[vfs_unlink() wants 2 args])
],[
AC_MSG_RESULT(no)
dnl #
dnl # Linux 3.13 API change
dnl # Added delegated inode
dnl #
AC_MSG_CHECKING([whether vfs_unlink() wants 3 args])
SPL_LINUX_TRY_COMPILE([
#include <linux/fs.h>
],[
vfs_unlink((struct inode *) NULL,
(struct dentry *) NULL,
(struct inode **) NULL);
],[
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_3ARGS_VFS_UNLINK, 1,
[vfs_unlink() wants 3 args])
],[
AC_MSG_ERROR(no)
])
])
])
dnl #
dnl # 3.13 and 3.15 API changes
dnl # Added delegated inode and flags argument.
dnl #
AC_DEFUN([SPL_AC_4ARGS_VFS_RENAME],
[AC_MSG_CHECKING([whether vfs_rename() wants 4 args])
SPL_LINUX_TRY_COMPILE([
#include <linux/fs.h>
],[
vfs_rename((struct inode *) NULL, (struct dentry *) NULL,
(struct inode *) NULL, (struct dentry *) NULL);
],[
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_4ARGS_VFS_RENAME, 1,
[vfs_rename() wants 4 args])
],[
AC_MSG_RESULT(no)
dnl #
dnl # Linux 3.13 API change
dnl # Added delegated inode
dnl #
AC_MSG_CHECKING([whether vfs_rename() wants 5 args])
SPL_LINUX_TRY_COMPILE([
#include <linux/fs.h>
],[
vfs_rename((struct inode *) NULL,
(struct dentry *) NULL,
(struct inode *) NULL,
(struct dentry *) NULL,
(struct inode **) NULL);
],[
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_5ARGS_VFS_RENAME, 1,
[vfs_rename() wants 5 args])
],[
AC_MSG_RESULT(no)
dnl #
dnl # Linux 3.15 API change
dnl # Added flags
dnl #
AC_MSG_CHECKING([whether vfs_rename() wants 6 args])
SPL_LINUX_TRY_COMPILE([
#include <linux/fs.h>
],[
vfs_rename((struct inode *) NULL,
(struct dentry *) NULL,
(struct inode *) NULL,
(struct dentry *) NULL,
(struct inode **) NULL,
(unsigned int) 0);
],[
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_6ARGS_VFS_RENAME, 1,
[vfs_rename() wants 6 args])
],[
AC_MSG_ERROR(no)
])
])
])
])
dnl #
dnl # 2.6.36 API change,
dnl # The 'struct fs_struct->lock' was changed from a rwlock_t to
dnl # a spinlock_t to improve the fastpath performance.
dnl #
AC_DEFUN([SPL_AC_FS_STRUCT_SPINLOCK], [
AC_MSG_CHECKING([whether struct fs_struct uses spinlock_t])
tmp_flags="$EXTRA_KCFLAGS"
EXTRA_KCFLAGS="-Werror"
SPL_LINUX_TRY_COMPILE([
#include <linux/sched.h>
#include <linux/fs_struct.h>
],[
struct fs_struct fs;
spin_lock_init(&fs.lock);
],[
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_FS_STRUCT_SPINLOCK, 1,
[struct fs_struct uses spinlock_t])
],[
AC_MSG_RESULT(no)
])
EXTRA_KCFLAGS="$tmp_flags"
])
dnl #
dnl # User namespaces, use kuid_t in place of uid_t
dnl # where available. Not strictly a user namespaces thing
dnl # but it should prevent surprises
dnl #
AC_DEFUN([SPL_AC_KUIDGID_T], [
AC_MSG_CHECKING([whether kuid_t/kgid_t is available])
SPL_LINUX_TRY_COMPILE([
#include <linux/uidgid.h>
], [
kuid_t userid = KUIDT_INIT(0);
kgid_t groupid = KGIDT_INIT(0);
],[
SPL_LINUX_TRY_COMPILE([
#include <linux/uidgid.h>
], [
kuid_t userid = 0;
kgid_t groupid = 0;
],[
AC_MSG_RESULT(yes; optional)
],[
AC_MSG_RESULT(yes; mandatory)
AC_DEFINE(HAVE_KUIDGID_T, 1, [kuid_t/kgid_t in use])
])
],[
AC_MSG_RESULT(no)
])
])
Reimplement rwlocks for Linux lock profiling/analysis. It turns out that the previous rwlock implementation worked well but did not integrate properly with the upstream kernel lock profiling/ analysis tools. This is a major problem since it would be awfully nice to be able to use the automatic lock checker and profiler. The problem is that the upstream lock tools use the pre-processor to create a lock class for each uniquely named locked. Since the rwsem was embedded in a wrapper structure the name was always the same. The effect was that we only ended up with one lock class for the entire SPL which caused the lock dependency checker to flag nearly everything as a possible deadlock. The solution was to directly map a krwlock to a Linux rwsem using a typedef there by eliminating the wrapper structure. This was not done initially because the rwsem implementation is specific to the arch. To fully implement the Solaris krwlock API using only the provided rwsem API is not possible. It can only be done by directly accessing some of the internal data member of the rwsem structure. For example, the Linux API provides a different function for dropping a reader vs writer lock. Whereas the Solaris API uses the same function and the caller does not pass in what type of lock it is. This means to properly drop the lock we need to determine if the lock is currently a reader or writer lock. Then we need to call the proper Linux API function. Unfortunately, there is no provided API for this so we must extracted this information directly from arch specific lock implementation. This is all do able, and what I did, but it does complicate things considerably. The good news is that in addition to the profiling benefits of this change. We may see performance improvements due to slightly reduced overhead when creating rwlocks and manipulating them. The only function I was forced to sacrafice was rw_owner() because this information is simply not stored anywhere in the rwsem. Luckily this appears not to be a commonly used function on Solaris, and it is my understanding it is mainly used for debugging anyway. In addition to the core rwlock changes, extensive updates were made to the rwlock regression tests. Each class of test was extended to provide more API coverage and to be more rigerous in checking for misbehavior. This is a pretty significant change and with that in mind I have been careful to validate it on several platforms before committing. The full SPLAT regression test suite was run numberous times on all of the following platforms. This includes various kernels ranging from 2.6.16 to 2.6.29. - SLES10 (ppc64) - SLES11 (x86_64) - CHAOS4.2 (x86_64) - RHEL5.3 (x86_64) - RHEL6 (x86_64) - FC11 (x86_64)
2009-09-18 23:09:47 +00:00
dnl #
dnl # 2.6.x API change,
dnl # __put_task_struct() was exported in RHEL5 but unavailable elsewhere.
dnl #
AC_DEFUN([SPL_AC_PUT_TASK_STRUCT],
[AC_MSG_CHECKING([whether __put_task_struct() is available])
SPL_LINUX_TRY_COMPILE_SYMBOL([
#include <linux/sched.h>
], [
__put_task_struct(NULL);
], [__put_task_struct], [], [
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_PUT_TASK_STRUCT, 1,
[__put_task_struct() is available])
], [
AC_MSG_RESULT(no)
])
Reimplement rwlocks for Linux lock profiling/analysis. It turns out that the previous rwlock implementation worked well but did not integrate properly with the upstream kernel lock profiling/ analysis tools. This is a major problem since it would be awfully nice to be able to use the automatic lock checker and profiler. The problem is that the upstream lock tools use the pre-processor to create a lock class for each uniquely named locked. Since the rwsem was embedded in a wrapper structure the name was always the same. The effect was that we only ended up with one lock class for the entire SPL which caused the lock dependency checker to flag nearly everything as a possible deadlock. The solution was to directly map a krwlock to a Linux rwsem using a typedef there by eliminating the wrapper structure. This was not done initially because the rwsem implementation is specific to the arch. To fully implement the Solaris krwlock API using only the provided rwsem API is not possible. It can only be done by directly accessing some of the internal data member of the rwsem structure. For example, the Linux API provides a different function for dropping a reader vs writer lock. Whereas the Solaris API uses the same function and the caller does not pass in what type of lock it is. This means to properly drop the lock we need to determine if the lock is currently a reader or writer lock. Then we need to call the proper Linux API function. Unfortunately, there is no provided API for this so we must extracted this information directly from arch specific lock implementation. This is all do able, and what I did, but it does complicate things considerably. The good news is that in addition to the profiling benefits of this change. We may see performance improvements due to slightly reduced overhead when creating rwlocks and manipulating them. The only function I was forced to sacrafice was rw_owner() because this information is simply not stored anywhere in the rwsem. Luckily this appears not to be a commonly used function on Solaris, and it is my understanding it is mainly used for debugging anyway. In addition to the core rwlock changes, extensive updates were made to the rwlock regression tests. Each class of test was extended to provide more API coverage and to be more rigerous in checking for misbehavior. This is a pretty significant change and with that in mind I have been careful to validate it on several platforms before committing. The full SPLAT regression test suite was run numberous times on all of the following platforms. This includes various kernels ranging from 2.6.16 to 2.6.29. - SLES10 (ppc64) - SLES11 (x86_64) - CHAOS4.2 (x86_64) - RHEL5.3 (x86_64) - RHEL6 (x86_64) - FC11 (x86_64)
2009-09-18 23:09:47 +00:00
])
dnl #
dnl # 2.6.32 API change,
dnl # Unused 'struct file *' removed from prototype.
dnl #
AC_DEFUN([SPL_AC_5ARGS_PROC_HANDLER], [
AC_MSG_CHECKING([whether proc_handler() wants 5 args])
SPL_LINUX_TRY_COMPILE([
#include <linux/sysctl.h>
],[
proc_dostring(NULL, 0, NULL, NULL, NULL);
],[
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_5ARGS_PROC_HANDLER, 1,
[proc_handler() wants 5 args])
],[
AC_MSG_RESULT(no)
])
])
dnl #
dnl # 2.6.x API change,
dnl # kvasprintf() function added.
dnl #
AC_DEFUN([SPL_AC_KVASPRINTF],
[AC_MSG_CHECKING([whether kvasprintf() is available])
SPL_LINUX_TRY_COMPILE_SYMBOL([
#include <linux/kernel.h>
], [
kvasprintf(0, NULL, *((va_list*)NULL));
], [kvasprintf], [], [
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_KVASPRINTF, 1, [kvasprintf() is available])
], [
AC_MSG_RESULT(no)
])
])
dnl #
dnl # 2.6.29 API change,
dnl # vfs_fsync() funcation added, prior to this use file_fsync().
dnl #
AC_DEFUN([SPL_AC_VFS_FSYNC],
[AC_MSG_CHECKING([whether vfs_fsync() is available])
SPL_LINUX_TRY_COMPILE_SYMBOL([
#include <linux/fs.h>
], [
(void) vfs_fsync;
], [vfs_fsync], [fs/sync.c], [
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_VFS_FSYNC, 1, [vfs_fsync() is available])
], [
AC_MSG_RESULT(no)
])
])
dnl #
dnl # 2.6.35 API change,
dnl # Unused 'struct dentry *' removed from vfs_fsync() prototype.
dnl #
AC_DEFUN([SPL_AC_2ARGS_VFS_FSYNC], [
AC_MSG_CHECKING([whether vfs_fsync() wants 2 args])
SPL_LINUX_TRY_COMPILE([
#include <linux/fs.h>
],[
vfs_fsync(NULL, 0);
],[
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_2ARGS_VFS_FSYNC, 1, [vfs_fsync() wants 2 args])
],[
AC_MSG_RESULT(no)
])
])
Correctly handle rwsem_is_locked() behavior A race condition in rwsem_is_locked() was fixed in Linux 2.6.33 and the fix was backported to RHEL5 as of kernel 2.6.18-190.el5. Details can be found here: https://bugzilla.redhat.com/show_bug.cgi?id=526092 The race condition was fixed in the kernel by acquiring the semaphore's wait_lock inside rwsem_is_locked(). The SPL worked around the race condition by acquiring the wait_lock before calling that function, but with the fix in place it must not do that. This commit implements an autoconf test to detect whether the fixed version of rwsem_is_locked() is present. The previous version of rwsem_is_locked() was an inline static function while the new version is exported as a symbol which we can check for in module.symvers. Depending on the result we correctly implement the needed compatibility macros for proper spinlock handling. Finally, we do the right thing with spin locks in RW_*_HELD() by using the new compatibility macros. We only only acquire the semaphore's wait_lock if it is calling a rwsem_is_locked() that does not itself try to acquire the lock. Some new overhead and a small harmless race is introduced by this change. This is because RW_READ_HELD() and RW_WRITE_HELD() now acquire and release the wait_lock twice: once for the call to rwsem_is_locked() and once for the call to rw_owner(). This can't be avoided if calling a rwsem_is_locked() that takes the wait_lock, as it will in more recent kernels. The other case which only occurs in legacy kernels could be optimized by taking the lock only once, as was done prior to this commit. However, I decided that the performance gain probably wasn't significant enough to justify the messy special cases required. The function spl_rw_get_owner() was only used to enable the afore-mentioned optimization. Since it is no longer used, I removed it. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
2010-08-10 18:01:46 +00:00
dnl #
dnl # 3.5 API change,
dnl # inode_operations.truncate_range removed
dnl #
AC_DEFUN([SPL_AC_INODE_TRUNCATE_RANGE], [
AC_MSG_CHECKING([whether truncate_range() inode operation is available])
SPL_LINUX_TRY_COMPILE([
#include <linux/fs.h>
],[
struct inode_operations ops;
ops.truncate_range = NULL;
],[
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_INODE_TRUNCATE_RANGE, 1,
[truncate_range() inode operation is available])
],[
AC_MSG_RESULT(no)
])
])
dnl #
dnl # Linux 2.6.38 - 3.x API
dnl #
AC_DEFUN([SPL_AC_KERNEL_FILE_FALLOCATE], [
AC_MSG_CHECKING([whether fops->fallocate() exists])
SPL_LINUX_TRY_COMPILE([
#include <linux/fs.h>
],[
long (*fallocate) (struct file *, int, loff_t, loff_t) = NULL;
struct file_operations fops __attribute__ ((unused)) = {
.fallocate = fallocate,
};
],[
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_FILE_FALLOCATE, 1, [fops->fallocate() exists])
],[
AC_MSG_RESULT(no)
])
])
dnl #
dnl # Linux 2.6.x - 2.6.37 API
dnl #
AC_DEFUN([SPL_AC_KERNEL_INODE_FALLOCATE], [
AC_MSG_CHECKING([whether iops->fallocate() exists])
SPL_LINUX_TRY_COMPILE([
#include <linux/fs.h>
],[
long (*fallocate) (struct inode *, int, loff_t, loff_t) = NULL;
struct inode_operations fops __attribute__ ((unused)) = {
.fallocate = fallocate,
};
],[
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_INODE_FALLOCATE, 1, [fops->fallocate() exists])
],[
AC_MSG_RESULT(no)
])
])
dnl #
dnl # PaX Linux 2.6.38 - 3.x API
dnl #
AC_DEFUN([SPL_AC_PAX_KERNEL_FILE_FALLOCATE], [
AC_MSG_CHECKING([whether fops->fallocate() exists])
SPL_LINUX_TRY_COMPILE([
#include <linux/fs.h>
],[
long (*fallocate) (struct file *, int, loff_t, loff_t) = NULL;
struct file_operations_no_const fops __attribute__ ((unused)) = {
.fallocate = fallocate,
};
],[
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_FILE_FALLOCATE, 1, [fops->fallocate() exists])
],[
AC_MSG_RESULT(no)
])
])
dnl #
dnl # The fallocate callback was moved from the inode_operations
dnl # structure to the file_operations structure.
dnl #
AC_DEFUN([SPL_AC_KERNEL_FALLOCATE], [
SPL_AC_KERNEL_FILE_FALLOCATE
SPL_AC_KERNEL_INODE_FALLOCATE
SPL_AC_PAX_KERNEL_FILE_FALLOCATE
])
Correctly handle rwsem_is_locked() behavior A race condition in rwsem_is_locked() was fixed in Linux 2.6.33 and the fix was backported to RHEL5 as of kernel 2.6.18-190.el5. Details can be found here: https://bugzilla.redhat.com/show_bug.cgi?id=526092 The race condition was fixed in the kernel by acquiring the semaphore's wait_lock inside rwsem_is_locked(). The SPL worked around the race condition by acquiring the wait_lock before calling that function, but with the fix in place it must not do that. This commit implements an autoconf test to detect whether the fixed version of rwsem_is_locked() is present. The previous version of rwsem_is_locked() was an inline static function while the new version is exported as a symbol which we can check for in module.symvers. Depending on the result we correctly implement the needed compatibility macros for proper spinlock handling. Finally, we do the right thing with spin locks in RW_*_HELD() by using the new compatibility macros. We only only acquire the semaphore's wait_lock if it is calling a rwsem_is_locked() that does not itself try to acquire the lock. Some new overhead and a small harmless race is introduced by this change. This is because RW_READ_HELD() and RW_WRITE_HELD() now acquire and release the wait_lock twice: once for the call to rwsem_is_locked() and once for the call to rw_owner(). This can't be avoided if calling a rwsem_is_locked() that takes the wait_lock, as it will in more recent kernels. The other case which only occurs in legacy kernels could be optimized by taking the lock only once, as was done prior to this commit. However, I decided that the performance gain probably wasn't significant enough to justify the messy special cases required. The function spl_rw_get_owner() was only used to enable the afore-mentioned optimization. Since it is no longer used, I removed it. Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
2010-08-10 18:01:46 +00:00
dnl #
dnl # 2.6.33 API change. Also backported in RHEL5 as of 2.6.18-190.el5.
dnl # Earlier versions of rwsem_is_locked() were inline and had a race
dnl # condition. The fixed version is exported as a symbol. The race
dnl # condition is fixed by acquiring sem->wait_lock, so we must not
dnl # call that version while holding sem->wait_lock.
dnl #
AC_DEFUN([SPL_AC_EXPORTED_RWSEM_IS_LOCKED],
[AC_MSG_CHECKING([whether rwsem_is_locked() acquires sem->wait_lock])
SPL_LINUX_TRY_COMPILE_SYMBOL([
#include <linux/rwsem.h>
int rwsem_is_locked(struct rw_semaphore *sem) { return 0; }
], [], [rwsem_is_locked], [lib/rwsem-spinlock.c], [
AC_MSG_RESULT(yes)
AC_DEFINE(RWSEM_IS_LOCKED_TAKES_WAIT_LOCK, 1,
[rwsem_is_locked() acquires sem->wait_lock])
], [
AC_MSG_RESULT(no)
])
])
dnl #
dnl # 2.6.28 API change
dnl # The kern_path() function has been introduced. We adopt it as the new way
dnl # of looking up paths. When it is not available, we emulate it using the
dnl # older interfaces.
dnl #
AC_DEFUN([SPL_AC_KERN_PATH],
[AC_MSG_CHECKING([whether kern_path() is available])
SPL_LINUX_TRY_COMPILE_SYMBOL([
#include <linux/namei.h>
], [
int r = kern_path(NULL, 0, NULL);
], [kern_path], [fs/namei.c], [
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_KERN_PATH, 1,
[kern_path() is available])
], [
AC_MSG_RESULT(no)
AC_MSG_CHECKING([whether path_lookup() is available])
SPL_LINUX_TRY_COMPILE_SYMBOL([
#include <linux/namei.h>
], [
int r = path_lookup(NULL, 0, NULL);
], [path_lookup], [fs/namei.c], [
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_KERN_PATH, 1,
[kern_path() is available])
], [
AC_MSG_RESULT(no)
AC_MSG_ERROR([
*** Neither kern_path() nor path_lookup() is available.
*** Please file an issue:
*** https://github.com/zfsonlinux/spl/issues/new])
])
])
])
dnl #
dnl # zlib inflate compat,
dnl # Verify the kernel has CONFIG_ZLIB_INFLATE support enabled.
dnl #
AC_DEFUN([SPL_AC_CONFIG_ZLIB_INFLATE], [
AC_MSG_CHECKING([whether CONFIG_ZLIB_INFLATE is defined])
SPL_LINUX_TRY_COMPILE([
#if !defined(CONFIG_ZLIB_INFLATE) && \
!defined(CONFIG_ZLIB_INFLATE_MODULE)
#error CONFIG_ZLIB_INFLATE not defined
#endif
],[ ],[
AC_MSG_RESULT([yes])
],[
AC_MSG_RESULT([no])
AC_MSG_ERROR([
*** This kernel does not include the required zlib inflate support.
*** Rebuild the kernel with CONFIG_ZLIB_INFLATE=y|m set.])
])
])
dnl #
dnl # zlib deflate compat,
dnl # Verify the kernel has CONFIG_ZLIB_DEFLATE support enabled.
dnl #
AC_DEFUN([SPL_AC_CONFIG_ZLIB_DEFLATE], [
AC_MSG_CHECKING([whether CONFIG_ZLIB_DEFLATE is defined])
SPL_LINUX_TRY_COMPILE([
#if !defined(CONFIG_ZLIB_DEFLATE) && \
!defined(CONFIG_ZLIB_DEFLATE_MODULE)
#error CONFIG_ZLIB_DEFLATE not defined
#endif
],[ ],[
AC_MSG_RESULT([yes])
],[
AC_MSG_RESULT([no])
AC_MSG_ERROR([
*** This kernel does not include the required zlib deflate support.
*** Rebuild the kernel with CONFIG_ZLIB_DEFLATE=y|m set.])
])
])
dnl #
dnl # 2.6.39 API compat,
dnl # The function zlib_deflate_workspacesize() now take 2 arguments.
dnl # This was done to avoid always having to allocate the maximum size
dnl # workspace (268K). The caller can now specific the windowBits and
dnl # memLevel compression parameters to get a smaller workspace.
dnl #
AC_DEFUN([SPL_AC_2ARGS_ZLIB_DEFLATE_WORKSPACESIZE],
[AC_MSG_CHECKING([whether zlib_deflate_workspacesize() wants 2 args])
SPL_LINUX_TRY_COMPILE([
#include <linux/zlib.h>
],[
return zlib_deflate_workspacesize(MAX_WBITS, MAX_MEM_LEVEL);
],[
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_2ARGS_ZLIB_DEFLATE_WORKSPACESIZE, 1,
[zlib_deflate_workspacesize() wants 2 args])
],[
AC_MSG_RESULT(no)
])
])
dnl #
dnl # 2.6.39 API change,
dnl # Shrinker adjust to use common shrink_control structure.
dnl #
AC_DEFUN([SPL_AC_SHRINK_CONTROL_STRUCT], [
AC_MSG_CHECKING([whether struct shrink_control exists])
SPL_LINUX_TRY_COMPILE([
#include <linux/mm.h>
],[
struct shrink_control sc __attribute__ ((unused));
sc.nr_to_scan = 0;
sc.gfp_mask = GFP_KERNEL;
],[
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_SHRINK_CONTROL_STRUCT, 1,
[struct shrink_control exists])
],[
AC_MSG_RESULT(no)
])
])
dnl #
dnl # 3.1 API Change
dnl #
dnl # The rw_semaphore.wait_lock member was changed from spinlock_t to
dnl # raw_spinlock_t at commit ddb6c9b58a19edcfac93ac670b066c836ff729f1.
dnl #
AC_DEFUN([SPL_AC_RWSEM_SPINLOCK_IS_RAW], [
AC_MSG_CHECKING([whether struct rw_semaphore member wait_lock is raw])
tmp_flags="$EXTRA_KCFLAGS"
EXTRA_KCFLAGS="-Werror"
SPL_LINUX_TRY_COMPILE([
#include <linux/rwsem.h>
],[
struct rw_semaphore dummy_semaphore __attribute__ ((unused));
raw_spinlock_t dummy_lock __attribute__ ((unused));
dummy_semaphore.wait_lock = dummy_lock;
],[
AC_MSG_RESULT(yes)
AC_DEFINE(RWSEM_SPINLOCK_IS_RAW, 1,
[struct rw_semaphore member wait_lock is raw_spinlock_t])
],[
AC_MSG_RESULT(no)
])
EXTRA_KCFLAGS="$tmp_flags"
])
dnl #
dnl # 3.9 API change,
dnl # Moved things from linux/sched.h to linux/sched/rt.h
dnl #
AC_DEFUN([SPL_AC_SCHED_RT_HEADER],
[AC_MSG_CHECKING([whether header linux/sched/rt.h exists])
SPL_LINUX_TRY_COMPILE([
#include <linux/sched.h>
#include <linux/sched/rt.h>
],[
return 0;
],[
AC_DEFINE(HAVE_SCHED_RT_HEADER, 1, [linux/sched/rt.h exists])
AC_MSG_RESULT(yes)
],[
AC_MSG_RESULT(no)
])
])
dnl #
dnl # 3.9 API change,
dnl # vfs_getattr() uses 2 args
dnl # It takes struct path * instead of struct vfsmount * and struct dentry *
dnl #
AC_DEFUN([SPL_AC_2ARGS_VFS_GETATTR], [
AC_MSG_CHECKING([whether vfs_getattr() wants])
SPL_LINUX_TRY_COMPILE([
#include <linux/fs.h>
],[
vfs_getattr((struct path *) NULL,
(struct kstat *)NULL);
],[
AC_MSG_RESULT(2 args)
AC_DEFINE(HAVE_2ARGS_VFS_GETATTR, 1,
[vfs_getattr wants 2 args])
],[
SPL_LINUX_TRY_COMPILE([
#include <linux/fs.h>
],[
vfs_getattr((struct vfsmount *)NULL,
(struct dentry *)NULL,
(struct kstat *)NULL);
],[
AC_MSG_RESULT(3 args)
],[
AC_MSG_ERROR(unknown)
])
])
])
dnl #
dnl # 2.6.36 API compatibility.
dnl # Added usleep_range timer.
dnl # usleep_range is a finer precision implementation of msleep
dnl # designed to be a drop-in replacement for udelay where a precise
dnl # sleep / busy-wait is unnecessary.
dnl #
AC_DEFUN([SPL_AC_USLEEP_RANGE], [
AC_MSG_CHECKING([whether usleep_range() is available])
SPL_LINUX_TRY_COMPILE([
#include <linux/delay.h>
],[
usleep_range(0, 0);
],[
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_USLEEP_RANGE, 1,
[usleep_range is available])
],[
AC_MSG_RESULT(no)
])
])
Add KMC_SLAB cache type For small objects the Linux slab allocator has several advantages over its counterpart in the SPL. These include: 1) It is more memory-efficient and packs objects more tightly. 2) It is continually tuned to maximize performance. Therefore it makes sense to layer the SPLs slab allocator on top of the Linux slab allocator. This allows us to leverage the advantages above while preserving the Illumos semantics we depend on. However, there are some things we need to be careful of: 1) The Linux slab allocator was never designed to work well with large objects. Because the SPL slab must still handle this use case a cut off limit was added to transition from Linux slab backed objects to kmem or vmem backed slabs. spl_kmem_cache_slab_limit - Objects less than or equal to this size in bytes will be backed by the Linux slab. By default this value is zero which disables the Linux slab functionality. Reasonable values for this cut off limit are in the range of 4096-16386 bytes. spl_kmem_cache_kmem_limit - Objects less than or equal to this size in bytes will be backed by a kmem slab. Objects over this size will be vmem backed instead. This value defaults to 1/8 a page, or 512 bytes on an x86_64 architecture. 2) Be aware that using the Linux slab may inadvertently introduce new deadlocks. Care has been taken previously to ensure that all allocations which occur in the write path use GFP_NOIO. However, there may be internal allocations performed in the Linux slab which do not honor these flags. If this is the case a deadlock may occur. The path forward is definitely to start relying on the Linux slab. But for that to happen we need to start building confidence that there aren't any unexpected surprises lurking for us. And ideally need to move completely away from using the SPLs slab for large memory allocations. This patch is a first step. NOTES: 1) The KMC_NOMAGAZINE flag was leveraged to support the Linux slab backed caches but it is not supported for kmem/vmem backed caches. 2) Regardless of the spl_kmem_cache_*_limit settings a cache may be explicitly set to a given type by passed the KMC_KMEM, KMC_VMEM, or KMC_SLAB flags during cache creation. 3) The constructors, destructors, and reclaim callbacks are all functional and will be called regardless of the cache type. 4) KMC_SLAB caches will not appear in /proc/spl/kmem/slab due to the issues involved in presenting correct object accounting. Instead they will appear in /proc/slabinfo under the same names. 5) Several kmem SPLAT tests needed to be fixed because they relied incorrectly on internal kmem slab accounting. With the updated test cases all the SPLAT tests pass as expected. 6) An autoconf test was added to ensure that the __GFP_COMP flag was correctly added to the default flags used when allocating a slab. This is required to ensure all pages in higher order slabs are properly refcounted, see ae16ed9. 7) When using the SLUB allocator there is no need to attempt to set the __GFP_COMP flag. This has been the default behavior for the SLUB since Linux 2.6.25. 8) When using the SLUB it may be desirable to set the slub_nomerge kernel parameter to prevent caches from being merged. Original-patch-by: DHE <git@dehacked.net> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Prakash Surya <surya1@llnl.gov> Signed-off-by: Tim Chase <tim@chase2k.com> Signed-off-by: DHE <git@dehacked.net> Signed-off-by: Chunwei Chen <tuxoko@gmail.com> Closes #356
2013-12-08 22:01:45 +00:00
dnl #
dnl # 2.6.35 API change,
dnl # The cachep->gfpflags member was renamed cachep->allocflags. These are
dnl # private allocation flags which are applied when allocating a new slab
dnl # in kmem_getpages(). Unfortunately there is no public API for setting
dnl # non-default flags.
dnl #
AC_DEFUN([SPL_AC_KMEM_CACHE_ALLOCFLAGS], [
AC_MSG_CHECKING([whether struct kmem_cache has allocflags])
SPL_LINUX_TRY_COMPILE([
#include <linux/slab.h>
],[
struct kmem_cache cachep __attribute__ ((unused));
cachep.allocflags = GFP_KERNEL;
],[
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_KMEM_CACHE_ALLOCFLAGS, 1,
[struct kmem_cache has allocflags])
],[
AC_MSG_RESULT(no)
AC_MSG_CHECKING([whether struct kmem_cache has gfpflags])
SPL_LINUX_TRY_COMPILE([
#include <linux/slab.h>
],[
struct kmem_cache cachep __attribute__ ((unused));
cachep.gfpflags = GFP_KERNEL;
],[
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_KMEM_CACHE_GFPFLAGS, 1,
[struct kmem_cache has gfpflags])
],[
AC_MSG_RESULT(no)
])
])
])
dnl #
dnl # 3.17 API change,
dnl # wait_on_bit() no longer requires an action argument. The former
dnl # "wait_on_bit" interface required an 'action' function to be provided
dnl # which does the actual waiting. There were over 20 such functions in the
dnl # kernel, many of them identical, though most cases can be satisfied by one
dnl # of just two functions: one which uses io_schedule() and one which just
dnl # uses schedule(). This API change was made to consolidate all of those
dnl # redundant wait functions.
dnl #
AC_DEFUN([SPL_AC_WAIT_ON_BIT], [
AC_MSG_CHECKING([whether wait_on_bit() takes an action])
SPL_LINUX_TRY_COMPILE([
#include <linux/wait.h>
],[
int (*action)(void *) = NULL;
wait_on_bit(NULL, 0, action, 0);
],[
AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_WAIT_ON_BIT_ACTION, 1, [yes])
],[
AC_MSG_RESULT(no)
])
])