search for: target_cpu

I am trying to set up a flexible infrastructure for the assembly code. Basically what I want is configure.in determination of basic machine type (intel/compatible, alpha, ppc), then within that (say intel) the code will detect variants like MMX, SSE, and use the right routines. I know how to do the second, but what is a good way to do the first? Linux/Cygwin/Solaris seem to support the MACHTYPE

...ou. Use AC_CANONICAL_TARGET in configure.in, and it will place canonical system type for the target in the shell variable 'target' in the form CPU_TYPE-MANUFACTURER-KERNEL-OPERATING_SYSTEM. Be sure to use the target, not the host, to allow cross-compiling. It will also be broken out into target_cpu, target_vendor and target_os. So you can use $target_cpu for your tests. A simple: case $target_cpu in i*86) blahblah ;; powerpc) blahblah ;; sparc) blahblah ;; esac should do the trick. If you'd rather do the computation in the makefile, you can just AC_SUBST these in (automake migh...

...we check the yield_count and > + * return directly if the holder_cpu is running. > + * IOW. do NOT code like below. > + * yield_count = be32_to_cpu(lppaca_of(holder_cpu).yield_count); > + * if ((yield_count & 1) == 0) > + * return; > + * > + * a PROD hcall marks the target_cpu proded, which cause the next cede > + * or confer called on the target_cpu invalid. > + */ > + plpar_hcall_norets(H_PROD, > + get_hard_smp_processor_id(holder_cpu)); > +} > +EXPORT_SYMBOL_GPL(__spin_wake_cpu); > + > #ifndef CONFIG_QUEUED_SPINLOCKS > void __spin_yield...

...we check the yield_count and > + * return directly if the holder_cpu is running. > + * IOW. do NOT code like below. > + * yield_count = be32_to_cpu(lppaca_of(holder_cpu).yield_count); > + * if ((yield_count & 1) == 0) > + * return; > + * > + * a PROD hcall marks the target_cpu proded, which cause the next cede > + * or confer called on the target_cpu invalid. > + */ > + plpar_hcall_norets(H_PROD, > + get_hard_smp_processor_id(holder_cpu)); > +} > +EXPORT_SYMBOL_GPL(__spin_wake_cpu); > + > #ifndef CONFIG_QUEUED_SPINLOCKS > void __spin_yield...

...* + * So we might lose a wake if we check the yield_count and + * return directly if the holder_cpu is running. + * IOW. do NOT code like below. + * yield_count = be32_to_cpu(lppaca_of(holder_cpu).yield_count); + * if ((yield_count & 1) == 0) + * return; + * + * a PROD hcall marks the target_cpu proded, which cause the next cede + * or confer called on the target_cpu invalid. + */ + plpar_hcall_norets(H_PROD, + get_hard_smp_processor_id(holder_cpu)); +} +EXPORT_SYMBOL_GPL(__spin_wake_cpu); + #ifndef CONFIG_QUEUED_SPINLOCKS void __spin_yield(arch_spinlock_t *lock) { -- 2.4.11

This patch set should make it easier to maintain PV-GRUB2 installations. The general idea is based on discussions I had with Xen developers (mainly Ian Jackson) at the Ubuntu Developer Summit in May 2011; though I never did manage to get the core port done and Vladimir beat me to that, I think the configuration approach we discussed there is still valid and useful. The idea here is that people

...s/patch-Makefile_in Wed May 31 08:29:37 2006 +++ ./patches/patch-Makefile_in Wed Dec 31 19:00:00 1969 @@ -1,12 +0,0 @@ -$OpenBSD: patch-Makefile_in,v 1.1.1.1 2005/07/09 00:35:02 jolan Exp $ ---- Makefile.in.orig Wed Dec 15 14:06:17 2004 -+++ Makefile.in Fri Jul 8 19:03:16 2005 -@@ -206,7 +206,7 @@ target_cpu = @target_cpu@ - target_os = @target_os@ - target_vendor = @target_vendor@ - AUTOMAKE_OPTIONS = foreign 1.6 dist-zip dist-bzip2 --SUBDIRS = lib include doc examples debian -+SUBDIRS = lib include - EXTRA_DIST = COPYING autogen.sh win32 libtheora.spec libtheora.spec.in \ - theora-uninstalled.pc.in...

...to disable processor-specific optimizations. Some people need to be able to build (for example) generic i386 packages, no matter whether the build machine happens to support SSE etc. > I know how to do the second, but what is a good way to do the first? Follow Matt's suggestion. Note that target_cpu = i386 doesn't (portably) imply an actual i386 but rather an x86-family processor. > Is uname -m or uname -p more reliable? You don't want to have to deal with those. Their output can vary widely on different platforms, e.g.: Solaris: -m sun4m -p sparc OpenBSD: -m sparc -p...

..._transform_name='s,x,x,' rootsbindir='${SBINDIR}' sbindir='${BINDIR}' selftest_prefix='./' sharedstatedir='${prefix}/com' smbtorture4_path='' swatdir='${prefix}/swat' sysconfdir='${prefix}/etc' target='' target_alias='' target_cpu='' target_os='' target_vendor='' ## ----------- ## ## confdefs.h. ## ## ----------- ## #define PACKAGE_BUGREPORT "" #define PACKAGE_NAME "" #define PACKAGE_STRING "" #define PACKAGE_TARNAME "" #define PACKAGE_VERSION ""...

...fix='NONE' program_transform_name='s,x,x,' psdir='${docdir}' sbindir='${exec_prefix}/sbin' sharedstatedir='${prefix}/com' subdirs=' projects/sample' sysconfdir='${prefix}/etc' target='powerpc-apple-darwin8.11.0' target_alias='' target_cpu='powerpc' target_os='darwin8.11.0' target_vendor='apple' ## ----------- ## ## confdefs.h. ## ## ----------- ## #define PACKAGE_NAME "llvm" #define PACKAGE_TARNAME "-llvm-" #define PACKAGE_VERSION "2.5" #define PACKAGE_STRING "llvm 2.5&quo...

Hi All, this is the fairlock patchset. You can apply them and build successfully. patches are based on linux-next qspinlock can avoid waiter starved issue. It has about the same speed in single-thread and it can be much faster in high contention situations especially when the spinlock is embedded within the data structure to be protected. v7 -> v8: add one patch to drop a function call

Hi All, this is the fairlock patchset. You can apply them and build successfully. patches are based on linux-next qspinlock can avoid waiter starved issue. It has about the same speed in single-thread and it can be much faster in high contention situations especially when the spinlock is embedded within the data structure to be protected. v7 -> v8: add one patch to drop a function call

This is a rebase/resend of earlier patches. I skipped the pure cosmetic patches for now. Mostly this is consolidation earlier changes, removing dead code etc. The important part is the change for allowing a vmbus channel to get callback directly in interrupt mode; this is necessary for NAPI support. Stephen Hemminger (14): vmbus: use kernel bitops for traversing interrupt mask vmbus: drop

This is a rebase/resend of earlier patches. I skipped the pure cosmetic patches for now. Mostly this is consolidation earlier changes, removing dead code etc. The important part is the change for allowing a vmbus channel to get callback directly in interrupt mode; this is necessary for NAPI support. Stephen Hemminger (14): vmbus: use kernel bitops for traversing interrupt mask vmbus: drop

...,' rootsbindir='${SBINDIR}' sbindir='${exec_prefix}/sbin' selftest_prefix='' sharedstatedir='${prefix}/com' smbtorture4_path='' swatdir='/usr/share/swat' sysconfdir='${prefix}/etc' target='i686-pc-linux-gnu' target_alias='' target_cpu='i686' target_os='linux-gnu' target_vendor='pc' ## ----------- ## ## confdefs.h. ## ## ----------- ## #define CONFIG_H_IS_FROM_SAMBA 1 #define PACKAGE_BUGREPORT "" #define PACKAGE_NAME "" #define PACKAGE_STRING "" #define PACKAGE_TARNAME &quot...

.../locks' prefix='NONE' privatedir='${prefix}/private' program_transform_name='s,x,x,' sbindir='${exec_prefix}/sbin' sharedstatedir='${prefix}/com' swatdir='${prefix}/swat' sysconfdir='${prefix}/etc' target='' target_alias='' target_cpu='' target_os='' target_vendor='' ## ----------- ## ## confdefs.h. ## ## ----------- ## #define PACKAGE_BUGREPORT "" #define PACKAGE_NAME "" #define PACKAGE_STRING "" #define PACKAGE_TARNAME "" #define PACKAGE_VERSION "" confi...

...'${SBINDIR}' > sbindir='${BINDIR}' > selftest_prefix='./' > sharedstatedir='${prefix}/com' > smbtorture4_path='' > swatdir='${prefix}/swat' > sysconfdir='${prefix}/etc' > target='' > target_alias='' > target_cpu='' > target_os='' > target_vendor='' > > ## ----------- ## > ## confdefs.h. ## > ## ----------- ## > > #define PACKAGE_BUGREPORT "" > #define PACKAGE_NAME "" > #define PACKAGE_STRING "" > #define PACKAGE_TARN...

...e='s,x,x,' > psdir='${docdir}' > sbindir='${exec_prefix}/sbin' > sharedstatedir='${prefix}/com' > subdirs=' projects/sample' > sysconfdir='${prefix}/etc' > target='powerpc-apple-darwin8.11.0' > target_alias='' > target_cpu='powerpc' > target_os='darwin8.11.0' > target_vendor='apple' > > ## ----------- ## > ## confdefs.h. ## > ## ----------- ## > > #define PACKAGE_NAME "llvm" > #define PACKAGE_TARNAME "-llvm-" > #define PACKAGE_VERSION "2.5...

...9; pid_t_fmt='' prefix='NONE' program_transform_name='s,x,x,' psdir='${docdir}' sbindir='${exec_prefix}/sbin' sharedstatedir='${prefix}/com' sysconfdir='${prefix}/etc' target='x86_64-unknown-linux-gnu' target_alias='' target_cpu='x86_64' target_os='linux-gnu' target_vendor='unknown' uint32_t_fmt='' uint32_t_hex_fmt='' uint64_t_fmt='' uint64_t_hex_fmt='' ## ----------- ## ## confdefs.h. ## ## ----------- ## #define PACKAGE_NAME "" #define PACKA...

Hi All, this is the fairlock patchset. You can apply them and build successfully. patches are based on linux-next qspinlock can avoid waiter starved issue. It has about the same speed in single-thread and it can be much faster in high contention situations especially when the spinlock is embedded within the data structure to be protected. v8 -> v9: mv qspinlocm config entry to