Upgrade hwloc to v1.2.1.
This commit was SVN r25088.
Этот коммит содержится в:
Jeff Squyres
родитель
f46ef05c6e
Коммит
495ceef60d
@ -17,6 +17,28 @@ bug fixes (and other actions) for each version of hwloc since version
|
||||
in v0.9.1).
|
||||
|
||||
|
||||
Version 1.2.1
|
||||
-------------
|
||||
* Improve support of AMD Bulldozer "Compute-Unit" modules by detecting
|
||||
logical processors with different core IDs on Linux.
|
||||
* Fix hwloc-ps crash when listing processes from another Linux cpuset.
|
||||
Thanks to Carl Smith for reporting the problem.
|
||||
* Fix build on AIX and Solaris. Thanks to Carl Smith and Andreas Kupries
|
||||
for reporting the problems.
|
||||
* Fix cache size detection on Darwin. Thanks to Erkcan Özcan for reporting
|
||||
the problem.
|
||||
* Make configure fail if --enable-xml or --enable-cairo is given and
|
||||
proper support cannot be found. Thanks to Andreas Kupries for reporting
|
||||
the XML problem.
|
||||
* Fix spurious L1 cache detection on AIX. Thanks to Hendryk Bockelmann
|
||||
for reporting the problem.
|
||||
* Fix hwloc_get_last_cpu_location(THREAD) on Linux. Thanks to Gabriele
|
||||
Fatigati for reporting the problem.
|
||||
* Fix object distance detection on Solaris.
|
||||
* Add pthread_self weak symbol to ease static linking.
|
||||
* Minor documentation fixes.
|
||||
|
||||
|
||||
Version 1.2.0
|
||||
-------------
|
||||
* Major features
|
||||
|
||||
@ -89,7 +89,7 @@ CLI Examples
|
||||
On a 4-socket 2-core machine with hyperthreading, the lstopo tool may show the
|
||||
following graphical output:
|
||||
|
||||
dudley.png
|
||||
dudley.png
|
||||
|
||||
Here's the equivalent output in textual form:
|
||||
|
||||
@ -193,7 +193,7 @@ line), and only socket #0 is shown for brevity:
|
||||
On a 4-socket 2-core Opteron NUMA machine, the lstopo tool may show the
|
||||
following graphical output:
|
||||
|
||||
hagrid.png
|
||||
hagrid.png
|
||||
|
||||
Here's the equivalent output in textual form:
|
||||
|
||||
@ -263,7 +263,7 @@ added and only PU #0 is shown for brevity:
|
||||
On a 2-socket quad-core Xeon (pre-Nehalem, with 2 dual-core dies into each
|
||||
socket):
|
||||
|
||||
emmett.png
|
||||
emmett.png
|
||||
|
||||
Here's the same output in textual form:
|
||||
|
||||
@ -381,12 +381,12 @@ interfaces for them are usually available on the supported Operating Systems.
|
||||
Here's the graphical output from lstopo on this platform when Simultaneous
|
||||
Multi-Threading (SMT) is enabled:
|
||||
|
||||
ppc64-with-smt.png
|
||||
ppc64-with-smt.png
|
||||
|
||||
And here's the graphical output from lstopo on this platform when SMT is
|
||||
disabled:
|
||||
|
||||
ppc64-without-smt.png
|
||||
ppc64-without-smt.png
|
||||
|
||||
Notice that hwloc only sees half the PUs when SMT is disabled. PU #15, for
|
||||
example, seems to change location from NUMA node #0 to #1. In reality, no PUs
|
||||
@ -403,7 +403,7 @@ Note that upgrading the Linux kernel on the same PPC64 system mentioned above
|
||||
to 2.6.34, hwloc is able to discover all the topology information. The
|
||||
following picture shows the entire topology layout when SMT is enabled:
|
||||
|
||||
ppc64-full-with-smt.png
|
||||
ppc64-full-with-smt.png
|
||||
|
||||
Developers using the hwloc API or XML output for portable applications should
|
||||
therefore be extremely careful to not make any assumptions about the structure
|
||||
@ -684,5 +684,5 @@ Make sure to have had a look at those too!
|
||||
|
||||
-------------------------------------------------------------------------------
|
||||
|
||||
Generated on Thu Apr 14 2011 22:34:49 for Hardware Locality (hwloc) by doxygen
|
||||
1.7.3
|
||||
Generated on Tue Aug 16 2011 19:37:04 for Hardware Locality (hwloc) by doxygen
|
||||
1.7.4
|
||||
|
||||
@ -7,7 +7,7 @@
|
||||
|
||||
major=1
|
||||
minor=2
|
||||
release=0
|
||||
release=1
|
||||
|
||||
# greek is used for alpha or beta release tags. If it is non-empty,
|
||||
# it will be appended to the version number. It does not have to be
|
||||
@ -41,11 +41,11 @@ want_repo_rev=0
|
||||
# from an SVN checkout, if repo_rev=-1, then its value is replaced
|
||||
# with the output of "svnversion".
|
||||
|
||||
repo_rev=r3454
|
||||
repo_rev=r3619
|
||||
|
||||
# The date when this release was created
|
||||
|
||||
date="Apr 14, 2011"
|
||||
date="Aug 16, 2011"
|
||||
|
||||
# The shared library version of hwloc's public library. This version
|
||||
# is maintained in accordance with the "Library Interface Versions"
|
||||
@ -58,4 +58,4 @@ date="Apr 14, 2011"
|
||||
# 2. Version numbers are described in the Libtool current:revision:age
|
||||
# format.
|
||||
|
||||
libhwloc_so_version=3:0:3
|
||||
libhwloc_so_version=3:1:3
|
||||
|
||||
@ -524,15 +524,21 @@ EOF])
|
||||
)
|
||||
|
||||
# XML support
|
||||
hwloc_xml_happy=
|
||||
if test "x$enable_xml" != "xno"; then
|
||||
HWLOC_PKG_CHECK_MODULES([XML], [libxml-2.0], [xmlNewDoc], [:], [enable_xml="no"])
|
||||
HWLOC_PKG_CHECK_MODULES([XML], [libxml-2.0], [xmlNewDoc],
|
||||
[hwloc_xml_happy=yes],
|
||||
[hwloc_xml_happy=no])
|
||||
fi
|
||||
if test "x$enable_xml" != "xno"; then
|
||||
if test "x$hwloc_xml_happy" = "xyes"; then
|
||||
HWLOC_REQUIRES="libxml-2.0 $HWLOC_REQUIRES"
|
||||
AC_DEFINE([HWLOC_HAVE_XML], [1], [Define to 1 if you have the `xml' library.])
|
||||
AC_SUBST([HWLOC_HAVE_XML], [1])
|
||||
else
|
||||
AC_SUBST([HWLOC_HAVE_XML], [0])
|
||||
AS_IF([test "$enable_xml" = "yes"],
|
||||
[AC_MSG_WARN([--enable-xml requested, but XML support was not found])
|
||||
AC_MSG_ERROR([Cannot continue])])
|
||||
fi
|
||||
HWLOC_CFLAGS="$HWLOC_CFLAGS $HWLOC_XML_CFLAGS"
|
||||
|
||||
@ -623,7 +629,7 @@ AC_DEFUN([HWLOC_DO_AM_CONDITIONALS],[
|
||||
AM_CONDITIONAL([HWLOC_HAVE_CUDART],
|
||||
[test "x$hwloc_have_cudart" = "xyes"])
|
||||
AM_CONDITIONAL([HWLOC_HAVE_CAIRO], [test "x$enable_cairo" != "xno"])
|
||||
AM_CONDITIONAL([HWLOC_HAVE_XML], [test "x$enable_xml" != "xno"])
|
||||
AM_CONDITIONAL([HWLOC_HAVE_XML], [test "$hwloc_xml_happy" = "yes"])
|
||||
AM_CONDITIONAL([HWLOC_HAVE_SET_MEMPOLICY], [test "x$enable_set_mempolicy" != "xno"])
|
||||
AM_CONDITIONAL([HWLOC_HAVE_MBIND], [test "x$enable_mbind" != "xno"])
|
||||
AM_CONDITIONAL([HWLOC_HAVE_BUNZIPP], [test "x$BUNZIPP" != "xfalse"])
|
||||
|
||||
@ -1,7 +1,7 @@
|
||||
dnl -*- Autoconf -*-
|
||||
dnl
|
||||
dnl Copyright (c) 2009 INRIA. All rights reserved.
|
||||
dnl Copyright (c) 2009 Université Bordeaux 1
|
||||
dnl Copyright (c) 2009, 2011 Université Bordeaux 1
|
||||
dnl Copyright (c) 2004-2005 The Trustees of Indiana University and Indiana
|
||||
dnl University Research and Technology
|
||||
dnl Corporation. All rights reserved.
|
||||
@ -211,9 +211,12 @@ EOF
|
||||
hwloc_build_utils=yes
|
||||
|
||||
# Cairo support
|
||||
hwloc_cairo_happy=
|
||||
if test "x$enable_cairo" != "xno"; then
|
||||
HWLOC_PKG_CHECK_MODULES([CAIRO], [cairo], [cairo_fill], [:], [enable_cairo=no])
|
||||
if test "x$enable_cairo" != "xno"; then
|
||||
HWLOC_PKG_CHECK_MODULES([CAIRO], [cairo], [cairo_fill],
|
||||
[hwloc_cairo_happy=yes],
|
||||
[hwloc_cairo_happy=no])
|
||||
if test "x$hwloc_cairo_happy" = "xyes"; then
|
||||
AC_PATH_XTRA
|
||||
CFLAGS_save=$CFLAGS
|
||||
LIBS_save=$LIBS
|
||||
@ -232,6 +235,7 @@ EOF
|
||||
)
|
||||
if test "x$enable_X11" != "xyes"; then
|
||||
AC_MSG_WARN([X11 headers not found, Cairo/X11 back-end disabled])
|
||||
hwloc_cairo_happy=no
|
||||
fi
|
||||
|
||||
CFLAGS=$CFLAGS_save
|
||||
@ -239,8 +243,12 @@ EOF
|
||||
fi
|
||||
fi
|
||||
|
||||
if test "x$enable_cairo" != "xno"; then
|
||||
if test "x$hwloc_cairo_happy" = "xyes"; then
|
||||
AC_DEFINE([HWLOC_HAVE_CAIRO], [1], [Define to 1 if you have the `cairo' library.])
|
||||
else
|
||||
AS_IF([test "$enable_cairo" = "yes"],
|
||||
[AC_MSG_WARN([--enable-cairo requested, but Cairo/X11 support was not found])
|
||||
AC_MSG_ERROR([Cannot continue])])
|
||||
fi
|
||||
|
||||
AC_CHECK_TYPES([wchar_t], [
|
||||
@ -254,16 +262,33 @@ EOF
|
||||
AC_CHECK_FUNCS([nl_langinfo])
|
||||
])
|
||||
hwloc_old_LIBS="$LIBS"
|
||||
LIBS=
|
||||
AC_CHECK_HEADERS([curses.h], [
|
||||
AC_CHECK_HEADERS([term.h], [
|
||||
AC_SEARCH_LIBS([tparm], [termcap ncursesw ncurses curses], [
|
||||
AC_SUBST([HWLOC_TERMCAP_LIBS], ["$LIBS"])
|
||||
AC_DEFINE([HWLOC_HAVE_LIBTERMCAP], [1],
|
||||
[Define to 1 if you have a library providing the termcap interface])
|
||||
])
|
||||
], [], [[#include <curses.h>]])
|
||||
])
|
||||
chosen_curses=""
|
||||
for curses in ncurses curses
|
||||
do
|
||||
for lib in "" -ltermcap -l${curses}w -l$curses
|
||||
do
|
||||
AC_MSG_CHECKING(termcap support using $curses and $lib)
|
||||
LIBS="$hwloc_old_LIBS $lib"
|
||||
AC_LINK_IFELSE([AC_LANG_PROGRAM([[
|
||||
#include <$curses.h>
|
||||
#include <term.h>
|
||||
]], [[tparm(NULL, 0, 0, 0, 0, 0, 0, 0, 0, 0)]])], [
|
||||
AC_MSG_RESULT(yes)
|
||||
AC_SUBST([HWLOC_TERMCAP_LIBS], ["$LIBS"])
|
||||
AC_DEFINE([HWLOC_HAVE_LIBTERMCAP], [1],
|
||||
[Define to 1 if you have a library providing the termcap interface])
|
||||
chosen_curses=$curses
|
||||
], [
|
||||
AC_MSG_RESULT(no)
|
||||
])
|
||||
test "x$chosen_curses" != "x" && break
|
||||
done
|
||||
test "x$chosen_curses" != "x" && break
|
||||
done
|
||||
if test "$chosen_curses" = ncurses
|
||||
then
|
||||
AC_DEFINE([HWLOC_USE_NCURSES], [1], [Define to 1 if ncurses works, preferred over curses])
|
||||
fi
|
||||
LIBS="$hwloc_old_LIBS"
|
||||
unset hwloc_old_LIBS
|
||||
|
||||
@ -357,9 +382,10 @@ EOF
|
||||
hwloc_config_prefix[tests/linux/gather/test-gather-topology.sh]
|
||||
hwloc_config_prefix[tests/linux/test-topology.sh]
|
||||
hwloc_config_prefix[tests/xml/test-topology.sh]
|
||||
hwloc_config_prefix[utils/test-hwloc-calc.sh]
|
||||
hwloc_config_prefix[utils/test-hwloc-distrib.sh])
|
||||
|
||||
AC_CONFIG_COMMANDS([chmoding-scripts], [chmod +x ]hwloc_config_prefix[tests/linux/test-topology.sh ]hwloc_config_prefix[tests/xml/test-topology.sh ]hwloc_config_prefix[tests/linux/hwloc-gather-topology ]hwloc_config_prefix[tests/linux/gather/test-gather-topology.sh ]hwloc_config_prefix[utils/test-hwloc-distrib.sh])
|
||||
AC_CONFIG_COMMANDS([chmoding-scripts], [chmod +x ]hwloc_config_prefix[tests/linux/test-topology.sh ]hwloc_config_prefix[tests/xml/test-topology.sh ]hwloc_config_prefix[tests/linux/hwloc-gather-topology ]hwloc_config_prefix[tests/linux/gather/test-gather-topology.sh ]hwloc_config_prefix[utils/test-hwloc-calc.sh ]hwloc_config_prefix[utils/test-hwloc-distrib.sh])
|
||||
|
||||
# These links are only needed in standalone mode. It would
|
||||
# be nice to m4 foreach this somehow, but whenever I tried
|
||||
|
||||
@ -894,10 +894,10 @@ HWLOC_DECLSPEC int hwloc_topology_is_thissystem(hwloc_topology_t __hwloc_restri
|
||||
* @{
|
||||
*/
|
||||
|
||||
/** \brief Returns the topology object at index \p index from depth \p depth */
|
||||
/** \brief Returns the topology object at logical index \p idx from depth \p depth */
|
||||
HWLOC_DECLSPEC hwloc_obj_t hwloc_get_obj_by_depth (hwloc_topology_t topology, unsigned depth, unsigned idx) __hwloc_attribute_pure;
|
||||
|
||||
/** \brief Returns the topology object at index \p index with type \p type
|
||||
/** \brief Returns the topology object at logical index \p idx with type \p type
|
||||
*
|
||||
* If no object for that type exists, \c NULL is returned.
|
||||
* If there are several levels with objects of that type, \c NULL is returned
|
||||
@ -1044,6 +1044,9 @@ hwloc_obj_get_info_by_name(hwloc_obj_t obj, const char *name)
|
||||
*
|
||||
* \note On some operating systems, CPU binding may have effects on memory binding, see
|
||||
* ::HWLOC_CPUBIND_NOMEMBIND
|
||||
*
|
||||
* Running lstopo --top can be a very convenient tool to check how binding
|
||||
* actually happened.
|
||||
* @{
|
||||
*/
|
||||
|
||||
@ -1117,6 +1120,9 @@ typedef enum {
|
||||
HWLOC_DECLSPEC int hwloc_set_cpubind(hwloc_topology_t topology, hwloc_const_cpuset_t set, int flags);
|
||||
|
||||
/** \brief Get current process or thread binding
|
||||
*
|
||||
* Writes into \p set the cpuset which the process or thread (according to \e
|
||||
* flags) was last bound to.
|
||||
*/
|
||||
HWLOC_DECLSPEC int hwloc_get_cpubind(hwloc_topology_t topology, hwloc_cpuset_t set, int flags);
|
||||
|
||||
@ -1139,14 +1145,14 @@ HWLOC_DECLSPEC int hwloc_set_proc_cpubind(hwloc_topology_t topology, hwloc_pid_t
|
||||
HWLOC_DECLSPEC int hwloc_get_proc_cpubind(hwloc_topology_t topology, hwloc_pid_t pid, hwloc_cpuset_t set, int flags);
|
||||
|
||||
#ifdef hwloc_thread_t
|
||||
/** \brief Bind a thread \p tid on cpus given in bitmap \p set
|
||||
/** \brief Bind a thread \p thread on cpus given in bitmap \p set
|
||||
*
|
||||
* \note hwloc_thread_t is pthread_t on unix platforms, and HANDLE on native
|
||||
* Windows platforms
|
||||
*
|
||||
* \note HWLOC_CPUBIND_PROCESS can not be used in \p flags.
|
||||
*/
|
||||
HWLOC_DECLSPEC int hwloc_set_thread_cpubind(hwloc_topology_t topology, hwloc_thread_t tid, hwloc_const_cpuset_t set, int flags);
|
||||
HWLOC_DECLSPEC int hwloc_set_thread_cpubind(hwloc_topology_t topology, hwloc_thread_t thread, hwloc_const_cpuset_t set, int flags);
|
||||
#endif
|
||||
|
||||
#ifdef hwloc_thread_t
|
||||
|
||||
@ -1,7 +1,7 @@
|
||||
/*
|
||||
* Copyright © 2009 CNRS
|
||||
* Copyright © 2009-2011 INRIA. All rights reserved.
|
||||
* Copyright © 2009-2010 Université Bordeaux 1
|
||||
* Copyright © 2009-2011 Université Bordeaux 1
|
||||
* Copyright © 2009-2010 Cisco Systems, Inc. All rights reserved.
|
||||
* See COPYING in top-level directory.
|
||||
*/
|
||||
@ -292,7 +292,7 @@ hwloc_get_next_obj_inside_cpuset_by_type (hwloc_topology_t topology, hwloc_const
|
||||
return hwloc_get_next_obj_inside_cpuset_by_depth(topology, set, depth, prev);
|
||||
}
|
||||
|
||||
/** \brief Return the \p index -th object at depth \p depth included in CPU set \p set.
|
||||
/** \brief Return the (logically) \p idx -th object at depth \p depth included in CPU set \p set.
|
||||
*/
|
||||
static __hwloc_inline hwloc_obj_t __hwloc_attribute_pure
|
||||
hwloc_get_obj_inside_cpuset_by_depth (hwloc_topology_t topology, hwloc_const_cpuset_t set,
|
||||
@ -529,8 +529,8 @@ HWLOC_DECLSPEC unsigned hwloc_get_closest_objs (hwloc_topology_t topology, hwloc
|
||||
/** \brief Find an object below another object, both specified by types and indexes.
|
||||
*
|
||||
* Start from the top system object and find object of type \p type1
|
||||
* and index \p idx1. Then look below this object and find another
|
||||
* object of type \p type2 and index \p idx2. Indexes are specified
|
||||
* and logical index \p idx1. Then look below this object and find another
|
||||
* object of type \p type2 and logical index \p idx2. Indexes are specified
|
||||
* within the parent, not withing the entire system.
|
||||
*
|
||||
* For instance, if type1 is SOCKET, idx1 is 2, type2 is CORE and idx2
|
||||
@ -557,7 +557,7 @@ hwloc_get_obj_below_by_type (hwloc_topology_t topology,
|
||||
* Arrays \p typev and \p idxv must contain \p nr types and indexes.
|
||||
*
|
||||
* Start from the top system object and walk the arrays \p typev and \p idxv.
|
||||
* For each type and index couple in the arrays, look under the previously found
|
||||
* For each type and logical index couple in the arrays, look under the previously found
|
||||
* object to find the index-th object of the given type.
|
||||
* Indexes are specified within the parent, not withing the entire system.
|
||||
*
|
||||
|
||||
@ -1,7 +1,7 @@
|
||||
/*
|
||||
* Copyright © 2009 CNRS
|
||||
* Copyright © 2009-2010 INRIA. All rights reserved.
|
||||
* Copyright © 2009-2010 Université Bordeaux 1
|
||||
* Copyright © 2009-2011 Université Bordeaux 1
|
||||
* See COPYING in top-level directory.
|
||||
*/
|
||||
|
||||
@ -48,7 +48,7 @@ HWLOC_DECLSPEC int hwloc_linux_set_tid_cpubind(hwloc_topology_t topology, pid_t
|
||||
|
||||
/** \brief Get the current binding of thread \p tid
|
||||
*
|
||||
* The behavior is exactly the same as the Linux sched_setaffinity system call,
|
||||
* The behavior is exactly the same as the Linux sched_getaffinity system call,
|
||||
* but uses a hwloc cpuset.
|
||||
*/
|
||||
HWLOC_DECLSPEC int hwloc_linux_get_tid_cpubind(hwloc_topology_t topology, pid_t tid, hwloc_cpuset_t set);
|
||||
|
||||
@ -42,9 +42,6 @@
|
||||
/* Define to 1 if you have the <cuda_runtime_api.h> header file. */
|
||||
#undef HAVE_CUDA_RUNTIME_API_H
|
||||
|
||||
/* Define to 1 if you have the <curses.h> header file. */
|
||||
#undef HAVE_CURSES_H
|
||||
|
||||
/* Define to 1 if you have the declaration of `CTL_HW', and to 0 if you don't.
|
||||
*/
|
||||
#undef HAVE_DECL_CTL_HW
|
||||
@ -261,9 +258,6 @@
|
||||
/* Define to 1 if you have the <sys/utsname.h> header file. */
|
||||
#undef HAVE_SYS_UTSNAME_H
|
||||
|
||||
/* Define to 1 if you have the <term.h> header file. */
|
||||
#undef HAVE_TERM_H
|
||||
|
||||
/* Define to 1 if you have the `uname' function. */
|
||||
#undef HAVE_UNAME
|
||||
|
||||
@ -477,6 +471,9 @@
|
||||
/* Define to 1 on unsupported systems */
|
||||
#undef HWLOC_UNSUPPORTED_SYS
|
||||
|
||||
/* Define to 1 if ncurses works, preferred over curses */
|
||||
#undef HWLOC_USE_NCURSES
|
||||
|
||||
/* Define to 1 on WINDOWS */
|
||||
#undef HWLOC_WIN_SYS
|
||||
|
||||
|
||||
@ -165,9 +165,11 @@ static void hwloc_get_type_distances_from_string(struct hwloc_topology *topology
|
||||
} else {
|
||||
/* parse a comma separated list of distances */
|
||||
for(i=0; i<nbobjs*nbobjs; i++) {
|
||||
distances[i] = strtof(tmp, &next);
|
||||
tmp = next+1;
|
||||
if (!*next && i!=nbobjs*nbobjs-1) {
|
||||
distances[i] = atof(tmp);
|
||||
next = strchr(tmp, ',');
|
||||
if (next) {
|
||||
tmp = next+1;
|
||||
} else if (i!=nbobjs*nbobjs-1) {
|
||||
fprintf(stderr, "Ignoring %s distances from environment variable, not enough values (%u out of %u)\n",
|
||||
hwloc_obj_type_string(type), i+1, nbobjs*nbobjs);
|
||||
free(indexes);
|
||||
|
||||
@ -475,6 +475,7 @@ look_rset(int sdl, hwloc_obj_type_t type, struct hwloc_topology *topology, int l
|
||||
break;
|
||||
case HWLOC_OBJ_GROUP:
|
||||
obj->attr->group.depth = level;
|
||||
break;
|
||||
case HWLOC_OBJ_CORE:
|
||||
{
|
||||
hwloc_obj_t obj2 = hwloc_alloc_setup_object(HWLOC_OBJ_CACHE, i);
|
||||
@ -505,6 +506,8 @@ hwloc_look_aix(struct hwloc_topology *topology)
|
||||
int i;
|
||||
/* TODO: R_LGPGDEF/R_LGPGFREE for large pages */
|
||||
|
||||
hwloc_debug("Note: SMPSDL is at %d\n", rs_getinfo(NULL, R_SMPSDL, 0));
|
||||
|
||||
for (i=0; i<=rs_getinfo(NULL, R_MAXSDL, 0); i++)
|
||||
{
|
||||
int known = 0;
|
||||
|
||||
@ -133,8 +133,8 @@ hwloc_look_darwin(struct hwloc_topology *topology)
|
||||
cacheconfig[i] = cacheconfig32[i];
|
||||
}
|
||||
|
||||
memset(cachesize, 0, sizeof(cachesize));
|
||||
size = sizeof(cachesize);
|
||||
memset(cachesize, 0, sizeof(uint64_t) * n);
|
||||
size = sizeof(uint64_t) * n;
|
||||
if (sysctlbyname("hw.cachesize", cachesize, &size, NULL, 0)) {
|
||||
if (n > 0)
|
||||
cachesize[0] = memsize;
|
||||
|
||||
@ -27,6 +27,7 @@
|
||||
#include <sched.h>
|
||||
#include <pthread.h>
|
||||
#include <sys/mman.h>
|
||||
#include <sys/syscall.h>
|
||||
#if defined HWLOC_HAVE_SET_MEMPOLICY || defined HWLOC_HAVE_MBIND
|
||||
#define migratepages migrate_pages /* workaround broken migratepages prototype in numaif.h before libnuma 2.0.2 */
|
||||
#include <numaif.h>
|
||||
@ -641,6 +642,7 @@ hwloc_linux_get_thisthread_cpubind(hwloc_topology_t topology, hwloc_bitmap_t hwl
|
||||
|
||||
#if HAVE_DECL_PTHREAD_SETAFFINITY_NP
|
||||
#pragma weak pthread_setaffinity_np
|
||||
#pragma weak pthread_self
|
||||
|
||||
static int
|
||||
hwloc_linux_set_thread_cpubind(hwloc_topology_t topology, pthread_t tid, hwloc_const_bitmap_t hwloc_set, int flags __hwloc_attribute_unused)
|
||||
@ -652,11 +654,15 @@ hwloc_linux_set_thread_cpubind(hwloc_topology_t topology, pthread_t tid, hwloc_c
|
||||
return -1;
|
||||
}
|
||||
|
||||
if (!pthread_self) {
|
||||
/* ?! Application uses set_thread_cpubind, but doesn't link against libpthread ?! */
|
||||
errno = ENOSYS;
|
||||
return -1;
|
||||
}
|
||||
if (tid == pthread_self())
|
||||
return hwloc_linux_set_tid_cpubind(topology, 0, hwloc_set);
|
||||
|
||||
if (!pthread_setaffinity_np) {
|
||||
/* ?! Application uses set_thread_cpubind, but doesn't link against libpthread ?! */
|
||||
errno = ENOSYS;
|
||||
return -1;
|
||||
}
|
||||
@ -735,6 +741,7 @@ hwloc_linux_set_thread_cpubind(hwloc_topology_t topology, pthread_t tid, hwloc_c
|
||||
|
||||
#if HAVE_DECL_PTHREAD_GETAFFINITY_NP
|
||||
#pragma weak pthread_getaffinity_np
|
||||
#pragma weak pthread_self
|
||||
|
||||
static int
|
||||
hwloc_linux_get_thread_cpubind(hwloc_topology_t topology, pthread_t tid, hwloc_bitmap_t hwloc_set, int flags __hwloc_attribute_unused)
|
||||
@ -746,11 +753,15 @@ hwloc_linux_get_thread_cpubind(hwloc_topology_t topology, pthread_t tid, hwloc_b
|
||||
return -1;
|
||||
}
|
||||
|
||||
if (!pthread_self) {
|
||||
/* ?! Application uses set_thread_cpubind, but doesn't link against libpthread ?! */
|
||||
errno = ENOSYS;
|
||||
return -1;
|
||||
}
|
||||
if (tid == pthread_self())
|
||||
return hwloc_linux_get_tid_cpubind(topology, 0, hwloc_set);
|
||||
|
||||
if (!pthread_getaffinity_np) {
|
||||
/* ?! Application uses get_thread_cpubind, but doesn't link against libpthread ?! */
|
||||
errno = ENOSYS;
|
||||
return -1;
|
||||
}
|
||||
@ -845,17 +856,27 @@ hwloc_linux_get_tid_last_cpu_location(hwloc_topology_t topology __hwloc_attribut
|
||||
FILE *file;
|
||||
int i;
|
||||
|
||||
if (!tid)
|
||||
strcpy(name, "/proc/self/stat");
|
||||
else
|
||||
snprintf(name, sizeof(name), "/proc/%lu/stat", (unsigned long) tid);
|
||||
if (!tid) {
|
||||
#ifdef SYS_gettid
|
||||
tid = syscall(SYS_gettid);
|
||||
#else
|
||||
errno = ENOSYS;
|
||||
return -1;
|
||||
#endif
|
||||
}
|
||||
|
||||
snprintf(name, sizeof(name), "/proc/%lu/stat", (unsigned long) tid);
|
||||
file = fopen(name, "r");
|
||||
if (!file) {
|
||||
errno = ENOSYS;
|
||||
return -1;
|
||||
}
|
||||
fgets(buf, sizeof(buf), file);
|
||||
tmp = fgets(buf, sizeof(buf), file);
|
||||
fclose(file);
|
||||
if (!tmp) {
|
||||
errno = ENOSYS;
|
||||
return -1;
|
||||
}
|
||||
|
||||
tmp = strrchr(buf, ')');
|
||||
if (!tmp) {
|
||||
@ -940,7 +961,11 @@ hwloc_linux_get_thisproc_last_cpu_location(hwloc_topology_t topology, hwloc_bitm
|
||||
static int
|
||||
hwloc_linux_get_thisthread_last_cpu_location(hwloc_topology_t topology, hwloc_bitmap_t hwloc_set, int flags __hwloc_attribute_unused)
|
||||
{
|
||||
return hwloc_linux_get_tid_last_cpu_location(topology, topology->pid, hwloc_set);
|
||||
if (topology->pid) {
|
||||
errno = ENOSYS;
|
||||
return -1;
|
||||
}
|
||||
return hwloc_linux_get_tid_last_cpu_location(topology, 0, hwloc_set);
|
||||
}
|
||||
|
||||
|
||||
@ -2397,6 +2422,7 @@ look_sysfscpu(struct hwloc_topology *topology, const char *path)
|
||||
struct hwloc_obj *sock, *core, *thread;
|
||||
hwloc_bitmap_t socketset, coreset, threadset, savedcoreset;
|
||||
unsigned mysocketid, mycoreid;
|
||||
int threadwithcoreid = 0;
|
||||
|
||||
/* look at the socket */
|
||||
mysocketid = 0; /* shut-up the compiler */
|
||||
@ -2424,9 +2450,31 @@ look_sysfscpu(struct hwloc_topology *topology, const char *path)
|
||||
sprintf(str, "%s/cpu%d/topology/thread_siblings", path, i);
|
||||
coreset = hwloc_parse_cpumap(str, topology->backend_params.sysfs.root_fd);
|
||||
savedcoreset = coreset; /* store it for later work-arounds */
|
||||
if (coreset && hwloc_bitmap_first(coreset) == i) {
|
||||
|
||||
if (coreset && hwloc_bitmap_weight(coreset) > 1) {
|
||||
/* check if this is hyperthreading or different coreids */
|
||||
unsigned siblingid, siblingcoreid;
|
||||
hwloc_bitmap_t set = hwloc_bitmap_dup(coreset);
|
||||
hwloc_bitmap_clr(set, i);
|
||||
siblingid = hwloc_bitmap_first(set);
|
||||
siblingcoreid = mycoreid;
|
||||
sprintf(str, "%s/cpu%d/topology/core_id", path, siblingid);
|
||||
hwloc_parse_sysfs_unsigned(str, &siblingcoreid, topology->backend_params.sysfs.root_fd);
|
||||
threadwithcoreid = (siblingcoreid != mycoreid);
|
||||
hwloc_bitmap_free(set);
|
||||
}
|
||||
|
||||
|
||||
if (coreset && (hwloc_bitmap_first(coreset) == i || threadwithcoreid)) {
|
||||
/* regular core */
|
||||
core = hwloc_alloc_setup_object(HWLOC_OBJ_CORE, mycoreid);
|
||||
core->cpuset = coreset;
|
||||
if (threadwithcoreid) {
|
||||
/* amd multicore compute-unit, create one core per thread */
|
||||
core->cpuset = hwloc_bitmap_alloc();
|
||||
hwloc_bitmap_set(core->cpuset, i);
|
||||
} else {
|
||||
core->cpuset = coreset;
|
||||
}
|
||||
hwloc_debug_1arg_bitmap("os core %u has cpuset %s\n",
|
||||
mycoreid, coreset);
|
||||
hwloc_insert_object_by_cpuset(topology, core);
|
||||
|
||||
@ -419,7 +419,7 @@ hwloc_look_lgrp(struct hwloc_topology *topology)
|
||||
#ifdef HAVE_LGRP_LATENCY_COOKIE
|
||||
{
|
||||
float *distances = calloc(curlgrp*curlgrp, sizeof(float));
|
||||
unsigned *indexes;
|
||||
unsigned *indexes = calloc(curlgrp,sizeof(unsigned));
|
||||
unsigned i, j;
|
||||
for (i = 0; i < curlgrp; i++) {
|
||||
indexes[i] = glob_lgrps[i]->os_index;
|
||||
@ -650,15 +650,15 @@ hwloc_set_solaris_hooks(struct hwloc_topology *topology)
|
||||
topology->get_proc_cpubind = hwloc_solaris_get_proc_cpubind;
|
||||
topology->get_thisproc_cpubind = hwloc_solaris_get_thisproc_cpubind;
|
||||
topology->get_thisthread_cpubind = hwloc_solaris_get_thisthread_cpubind;
|
||||
#endif /* HAVE_LIBLGRP */
|
||||
#ifdef MADV_ACCESS_LWP
|
||||
topology->set_area_membind = hwloc_solaris_set_area_membind;
|
||||
topology->set_proc_membind = hwloc_solaris_set_proc_membind;
|
||||
topology->set_thisproc_membind = hwloc_solaris_set_thisproc_membind;
|
||||
topology->set_thisthread_membind = hwloc_solaris_set_thisthread_membind;
|
||||
topology->get_proc_membind = hwloc_solaris_get_proc_membind;
|
||||
topology->get_thisproc_membind = hwloc_solaris_get_thisproc_membind;
|
||||
topology->get_thisthread_membind = hwloc_solaris_get_thisthread_membind;
|
||||
#endif /* HAVE_LIBLGRP */
|
||||
#ifdef MADV_ACCESS_LWP
|
||||
topology->set_area_membind = hwloc_solaris_set_area_membind;
|
||||
topology->support.membind->firsttouch_membind = 1;
|
||||
topology->support.membind->bind_membind = 1;
|
||||
topology->support.membind->interleave_membind = 1;
|
||||
|
||||
Загрузка…
x
Ссылка в новой задаче
Block a user