Re: [openib-general] [patch] libsdp typo in config_parser
On Fri, Aug 18, 2006 at 10:05:35PM +0300, Eitan Zahavi wrote: >Hi Bernhard > >SDP traffic will not show on the IPoIB counters. It does no go through >IPoIB. That's what i thought, thanks for confirming. >You can use >lsmod | grep ib_sdp >to see how many connections are made over SDP. Running lam via 2 nodes, on 2 CPUs each, i see: # lsmod | grep ib_sdp ib_sdp 28184 4 rdma_cm27912 1 ib_sdp ib_core53632 12 ib_ucm,ib_uverbs,ib_sdp,rdma_cm,ib_cm,ib_local_sa,ib_umad,ib_ipoib,ib_multicast,ib_sa,ib_mthca,ib_mad I did start lamboot with libsdp.so preloaded: $ LD_PRELOAD=/usr/local/lib64/libsdp.so lamboot l $ lamnodes C -c -n node13ib.infiniband node13ib.infiniband node15ib.infiniband node15ib.infiniband $ LD_PRELOAD=/usr/local/lib64/libsdp.so mpirun -np 4 /there/vasp/20060503/vasp.4.6/vasp.mpi Still, ifconfig ib0 (which hosts node??ib.infiniband on 10.100.0.0/24) shows that the communication is being sent over ipoib as ifconfigs counters constantly go up when communicating (only one user is active on the system). $ /sbin/ifconfig ib0 ib0 Link encap:UNSPEC HWaddr 00-00-04-04-FE-80-00-00-00-00-00-00-00-00-00-00 inet addr:10.100.0.13 Bcast:10.100.0.255 Mask:255.255.255.0 UP BROADCAST RUNNING MULTICAST MTU:2044 Metric:1 RX packets:182037964 errors:0 dropped:0 overruns:0 frame:0 TX packets:183607689 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:128 RX bytes:189334244937 (180563.2 Mb) TX bytes:194777918565 (185754.6 Mb) My libsdp.conf looks like this: $ cat /usr/local/etc/libsdp.conf #log min-level 1 destination file libsdp.log use bothconnect * 10.100.0.0/24:* use bothserver * 10.100.0.0/24:* So i fear i'm missing something crucial. Ideas? >Exact number of packets and data can flowing through the IB port can be >obtained by : >/sys/class/infiniband/mthca0/ports/1/counters/port_rcv_packets >/sys/class/infiniband/mthca0/ports/1/counters/port_xmit_packets $ for i in /sys/class/infiniband/mthca0/ports/1/counters/*packets;do echo -n $i:' ' ; cat $i;done /sys/class/infiniband/mthca0/ports/1/counters/port_rcv_packets: 185010549 /sys/class/infiniband/mthca0/ports/1/counters/port_xmit_packets: 186584856 PS: The different pingpong test (which have outdated names in the openib wiki, btw) do work just fine if run from the very same user, so i think that the basic verbs communication would work proper. ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] [patch] libsdp typo in config_parser
On Fri, Aug 18, 2006 at 05:54:04PM +0300, Eitan Zahavi wrote: >Committed to the trunk (not 1.1 branch) >Thanks. > Thank you. PS: I think there is another occurance in srp_daemon that i forgot to include in the diff, fwiw. PPS: IIRC the traffic sent via SDP did not show up in the packet-counters of the corresponding ipoib device last time i looked. Is this still the case? Asking because i'm seeing it accounted to my ib0 device now although libsdp's log indicates that the app is using SDP.. TIA for any hint ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] [patch] libsdp typo in config_parser
Hi, The attached trivial patch fixes a typo in the debugging output of libsdp's config parser. Please apply. Signed-off-by: Bernhard Fischer <[EMAIL PROTECTED]> Index: libsdp/src/config_parser.c === --- libsdp/src/config_parser.c (revision 9003) +++ libsdp/src/config_parser.c (working copy) @@ -198,7 +198,7 @@ extern int __sdp_min_level; /* dump the current state in readable format */ static void __sdp_dump_config_state() { char buf[1024]; - sprintf(buf, "CONIFG: use %s %s %s", + sprintf(buf, "CONFIG: use %s %s %s", __sdp_get_family_str(__sdp_rule.target_family), __sdp_get_role_str( current_role ), __sdp_rule.prog_name_expr); Index: libsdp/src/config_parser.y === --- libsdp/src/config_parser.y (revision 9003) +++ libsdp/src/config_parser.y (working copy) @@ -143,7 +143,7 @@ extern int __sdp_min_level; /* dump the current state in readable format */ static void __sdp_dump_config_state() { char buf[1024]; - sprintf(buf, "CONIFG: use %s %s %s", + sprintf(buf, "CONFIG: use %s %s %s", __sdp_get_family_str(__sdp_rule.target_family), __sdp_get_role_str( current_role ), __sdp_rule.prog_name_expr); ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] ibv_reg_mr fails: bad address [was: Re: Problem with mca_mpool_openib_register - Cannot allocate memory]
On Mon, Jun 19, 2006 at 11:11:12AM -0400, Bill Wichser wrote: >Running the openib stack from Redhat on a 2.6.9-34.ELsmp kernel, dual >Xeon. Running with openmpi v1.0.2 compiled w/gcc. > >While we still have the problem with btl_openib_endpoint.c returning 0 >byte(s) for max inline data, and realize that another IB stack addresses >this, another problem when running across more than a single host pops >up generating huge amounts of error messages. > >The errors go something like this: > >mca_mpool_openib_register: ibv_reg_mr(0x2ac2622000,1052672) failed with >error: Cannot allocate memory >[0,1,1][btl_openib.c:496:mca_btl_openib_prepare_dst] >mpool_register(0x2ac2622040,1048576) failed: base 0x2ac040 lb 0 >offset 4194304 while 8MB (as Bill stated below) is likely to fail, i'm getting: [x86-64n001:07622] mca_mpool_openib_register: ibv_reg_mr(0x717f2000,2113536) failed with error: Bad address Signal:11 info.si_errno:0(Success) si_code:1(SEGV_MAPERR) Failing at addr:0x717f1ff8 two boxes involved, each with 2GB memory, my ulimits are ok: $ ulimit -l;rsh 10.100.0.44 "ulimit -l" unlimited unlimited Any hint on this one? TIA, Bernhard $ cat /sys/class/infiniband/mthca0/fw_ver 4.7.400 # lspci -vvxxx -s 01:00.0 01:00.0 InfiniBand: Mellanox Technologies MT25208 InfiniHost III Ex HCA (Tavor compatibility mode) (rev a0) Subsystem: Mellanox Technologies MT25208 InfiniHost III Ex HCA (Tavor compatibility mode) Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- SERR- >We fixed the /etc/security/limits.conf problem but I don't know what to >do about this one. The job seems to complete without error on 2 nodes >(4 processors) but to scale any larger just generates megabyte files of >these types of error messages. > >Any insights for this problem? All searches lead me to the limits.conf >which we have set to 8192. These are 8G machines if that makes any Bill, 8192 is just 8MB. See http://www.open-mpi.org/faq/?category=infiniband#ib-locked-pages and also make sure to have session requiredpam_limits.so in your rsh, rlogin and rexec files in pam.d >difference. > >Thanks, >Bill ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] [patch] update uverbs_event_get_sb() to new get_sb_pseudo()
On Thu, Jul 13, 2006 at 05:32:07PM +0200, Bernhard Fischer wrote: >On Thu, Jul 13, 2006 at 05:15:18PM +0300, Michael S. Tsirkin wrote: >>Quoting r. Bernhard Fischer <[EMAIL PROTECTED]>: >>> Subject: [patch] update uverbs_event_get_sb() to new get_sb_pseudo() >>> >>> Hi, >>> >>> get_sb_pseudo was changed but uverbs was not updated. >> >>svn only supports latest kernel.org release. >AFAICT, i have 2.6.18-rc1-git6 (fetched about 1 hour ago..) duh. Sorry for the noise. ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] [patch] update uverbs_event_get_sb() to new get_sb_pseudo()
On Thu, Jul 13, 2006 at 05:15:18PM +0300, Michael S. Tsirkin wrote: >Quoting r. Bernhard Fischer <[EMAIL PROTECTED]>: >> Subject: [patch] update uverbs_event_get_sb() to new get_sb_pseudo() >> >> Hi, >> >> get_sb_pseudo was changed but uverbs was not updated. > >svn only supports latest kernel.org release. AFAICT, i have 2.6.18-rc1-git6 (fetched about 1 hour ago..) grep get_sb_pseudo include/linux/fs.h -A2 extern int get_sb_pseudo(struct file_system_type *, char *, struct super_operations *ops, unsigned long, struct vfsmount *mnt); openib gen2 has grep get_sb_pseudo ../cvs/openib.gen2/upstream/gen2/trunk/src/linux-kernel/infiniband/core/uverbs_main.c -A1 return get_sb_pseudo(fs_type, "infinibandevent:", NULL, INFINIBANDEVENTFS_MAGIC); > >> >> PS: should nowadays the in-kernel impl be used instead of the one in >> svn? >> >> Signed-off-by: Bernhard Fischer <[EMAIL PROTECTED]> > >Please test 2.6.18-rc1 if possible. There's also a way to port just >the infiniband bits from 2.6.18 to older kernels if you want to. I copied src/linux-kernel/infiniband into the drivers/ dir like i always did (after rm'ing the in-kernel drivers/infiniband dir). Neither uverbs_main.c did compile (hence the trivial patchlet in this thread) nor ipoib (complaining about xmit_lock not in netdev or the like). Is using the linux-kernel/infiniband driver from svn still the proper thing to do? Does ipoib from svn build for the current kernel? hm.. Could it be that the current git repo is not supported but only the last "prepatch"? ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] [patch] update uverbs_event_get_sb() to new get_sb_pseudo()
Hi, get_sb_pseudo was changed but uverbs was not updated. PS: should nowadays the in-kernel impl be used instead of the one in svn? Signed-off-by: Bernhard Fischer <[EMAIL PROTECTED]> Index: uverbs_main.c === --- uverbs_main.c (revision 8507) +++ uverbs_main.c (working copy) @@ -816,10 +816,10 @@ } static struct super_block *uverbs_event_get_sb(struct file_system_type *fs_type, int flags, - const char *dev_name, void *data) + const char *dev_name, void *data, struct vfsmount *mnt) { return get_sb_pseudo(fs_type, "infinibandevent:", NULL, -INFINIBANDEVENTFS_MAGIC); +INFINIBANDEVENTFS_MAGIC, mnt); } static struct file_system_type uverbs_event_fs = { ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] Re: mods to tvflash/configure.in broke build
On Wed, Mar 01, 2006 at 06:33:16PM +0200, Jack Morgenstein wrote: >On Wednesday 01 March 2006 18:08, Bernhard Fischer wrote: >> >> # dpkg -S /usr/include/pci/pci.h >> pciutils-dev: /usr/include/pci/pci.h >> >> respectively on a SuSE box >> # rpm -qf /usr/include/pci/pci.h >> pciutils-devel-2.1.99_test9-3 >> Perhaps you need to install the pciutils-devel package? >Before the patch, I noted that the search for pci/pci.h failed in our builds, >but this did not break the configure -- configure just continued to run. So tvflash was never built since you don't have the proper headers installed. Configure now detects this proper and errors out, so everything is ok. > >We don't install the pciutils-devel package as a rule. Well, if you don't have pciutils-devel on your build-host, then you cannot build tvflash. If you request to configure and build tvflash and don't have the prerequisites installed, then this is an error. ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] Re: mods to tvflash/configure.in broke build
On Wed, Mar 01, 2006 at 05:18:43PM +0200, Jack Morgenstein wrote: >Hi, >Last night, you checked in a mod to tvflash/configure.in which makes > a required file: > >svn log -v trunk/src/userspace/tvflashyields: > >r5518 | roland | 2006-02-28 07:03:32 +0200 (Tue, 28 Feb 2006) | 6 lines >Changed paths: > M /gen2/trunk/src/userspace/tvflash/configure.in > >- check for pci/pci.h and bail out if it is not found. >- check for needed functions provided by libpci > >Signed-off-by: Bernhard Fischer <[EMAIL PROTECTED]> >Signed-off-by: Roland Dreier <[EMAIL PROTECTED]> > > > >This has broken our build, since is not in the userspace include >search list. # dpkg -S /usr/include/pci/pci.h pciutils-dev: /usr/include/pci/pci.h respectively on a SuSE box # rpm -qf /usr/include/pci/pci.h pciutils-devel-2.1.99_test9-3 Which pci/pci.h did you use on your box when it compiled successfully before the recent checkin? Does tvflash build if you revert that patch? >The only place does appear is in the kernel, in pci source code >directories: under drivers/pci, and arch/i386/pci, and that does not sound >correct for outside inclusion: > >The default search path for gcc is: (verified by running gcc -v) > > /usr/local/include > /usr/lib/gcc/i586-suse-linux/4.0.2/include > /usr/lib/gcc/i586-suse-linux/4.0.2/../../../../i586-suse-linux/include > /usr/include > >Maybe you wished to check , which does exist in userspace includes? Perhaps you need to install the pciutils-devel package? ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] Re: [PATCH] osm: support for trivial PKey manager
On Wed, Mar 01, 2006 at 08:54:41AM -0500, Hal Rosenstock wrote: >Hi Ofer, > >On Wed, 2006-03-01 at 08:42, Ofer Gigi wrote: >> Hi Hal, >> >> Please add an empty file >> management/osm/include/opensm/osm_svn_revision.h >> >> This file is needed since during the build process we write into it the >> revision. When the file doesn't exists the writing fails. > >What svn version are you at ? > >I thought we already went through a cycle on this. Either way I thought >the makefile generated creates this file: > >if test '!' -d './.svn'; then \ >echo -n Exported revision >> > ./../include/opensm/osm_svn_revision_new.h ; \ >else \ >svnversion -n ./.. >> ./../include/opensm/osm_svn_revision_new.h ; \ >fi ; \ > >Is there some permission problem which prevents this ? My copy of opensm/Makefile.am reads e.g.: $(srcdir)/../include/opensm/osm_svn_revision.h: always Not sure if srcdir points to the objdir or not and i havent tried, admittedly, but if the srcdir is read-only then osm_svn_revision.h should most likely be generated in the (writable) objdir, no? Just curious.. ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] [patch] tvflash configure checks for libpci
Hi, tvflash does check if pci/pci.h is installed but does not error out if it is not found. This leads to build-errors later on as the include is not found. Attached patch would - check for pci/pci.h and bail out if it is not found. - check for needed functions provided by libpci Signed-off-by: Bernhard Fischer <[EMAIL PROTECTED]> Please consider applying something to this effect. Thank you Index: configure.in === --- configure.in(revision 5203) +++ configure.in(working copy) @@ -11,16 +11,25 @@ # Checks for programs. AC_PROG_CC -# Checks for libraries. -AC_CHECK_LIB(pci, pci_alloc) - # Checks for header files. AC_HEADER_STDC -AC_CHECK_HEADERS(fcntl.h limits.h stdlib.h string.h unistd.h pci/pci.h) +AC_CHECK_HEADERS(fcntl.h limits.h stdlib.h string.h unistd.h) +AC_CHECK_HEADER([pci/pci.h], [], +AC_MSG_ERROR([ not found.])) + # Checks for typedefs, structures, and compiler characteristics. # Checks for library functions. AC_FUNC_MALLOC AC_CHECK_FUNCS(memset strchr strtoul) +AC_CHECK_LIB(pci, pci_init, [], +AC_MSG_ERROR([libpci not found. +])) +AC_CHECK_FUNCS(pci_init pci_scan_bus pci_alloc \ + pci_cleanup pci_fill_info pci_write_long pci_read_long \ + pci_read_word, [], +AC_MSG_ERROR([libpci functions not found. +])) + AC_OUTPUT(Makefile tvflash.spec) ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] imgen/mic.cpp compilation error
On Mon, Jan 30, 2006 at 01:05:12PM +0200, Michael S. Tsirkin wrote: > >https://openib.org/tiki/tiki-index.php?page=Installation+Cheat+Sheet I'm getting a compilation error in mic.cpp since the bool for hinting whether to add an info section is missing: g++ -Wall -W -Werror -g -O2 -MP -MD '-DBLD_VER_STR="devel"' '-DIBADM_VER_STR=""' -fno-exceptions -c -o mic.o mic.cpp mic.cpp: In function 'int main(int, char**)': mic.cpp:452: error: no matching function for call to 'TImage::TImage(ParamList*, char*&, char*&, const char*)' TImage.h:84: note: candidates are: TImage::TImage(ParamList*, const char*, const char*, const char*, bool) TImage.h:78: note: TImage::TImage(const TImage&) make: *** [mic.o] Error 1 Looks like there should be a command-line switch to toggle this. Perhaps mic.cpp wasn't updated when you "Update from MFT 1.0.1 (IBG2.0.1)" TImage.h ? The reason i think about updating my FW (currently at 4.6.2) is that i get errors from mvapich2 from openib svn (rev 5204) like this: $ mpdrun -np 2 ./cpi cannot create cq Fail to init hca rank 1 in job 11 x86-64n001_57976 caused collective abort of all ranks exit status of rank 1: killed by signal 11 RDMA_DEFAULT_MAX_CQ_SIZE seems to default to 4, so: $ export RDMA_DEFAULT_MAX_CQ_SIZE=64 $ mpdrun -np 2 ./cpi [Init] Fail to create qp for rank 0 Fail to init hca rank 1 in job 12 x86-64n001_57976 caused collective abort of all ranks exit status of rank 1: killed by signal 11 The ibv_*_pingpong tests do work properly even with the old firmware. Does mvapich2 from svn work for anybody else? ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] [patch] convert ipath_driver from i_sem to i_mutex
Hi, I don't know when exactly the i_sem to i_mutex conversion was done. In 2.6.16 this patch is needed to build the ipath driver. Signed-off-by: Bernhard Fischer <[EMAIL PROTECTED]> Please apply. Index: infiniband/hw/ipath/ipath_driver.c === --- infiniband/hw/ipath/ipath_driver.c (revision 5203) +++ infiniband/hw/ipath/ipath_driver.c (working copy) @@ -2754,7 +2754,12 @@ static loff_t ipath_llseek(struct file * loff_t ret; /* range checking is done where offset is used, not here. */ +/* XXX remove this compatibility hack */ +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,16) down(&fp->f_dentry->d_inode->i_sem); +#else + mutex_lock(&fp->f_dentry->d_inode->i_mutex); +#endif if (!whence) ret = fp->f_pos = off; else if (whence == 1) { @@ -2762,7 +2767,11 @@ static loff_t ipath_llseek(struct file * ret = fp->f_pos; } else ret = -EINVAL; +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,16) up(&fp->f_dentry->d_inode->i_sem); +#else + mutex_unlock(&fp->f_dentry->d_inode->i_mutex); +#endif _IPATH_DBG("New offset %llx from seek %llx whence=%d\n", fp->f_pos, off, whence); ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] Errors in compilation with 2.6.14.4!!
n `cm_rtu_handler': >> drivers/infiniband/core/cm.c:1732: warning: passing arg 2 of >> `ib_cancel_mad' makes integer from pointer without a cast >> drivers/infiniband/core/cm.c: In function `ib_send_cm_dreq': >> drivers/infiniband/core/cm.c:1790: error: structure has no member >> named `timeout_ms' >> drivers/infiniband/core/cm.c:1793: warning: passing arg 1 of >> `ib_post_send_mad' from incompatible pointer type >> drivers/infiniband/core/cm.c:1793: error: too few arguments to >> function `ib_post_send_mad' >> drivers/infiniband/core/cm.c: In function `ib_send_cm_drep': >> drivers/infiniband/core/cm.c:1856: warning: passing arg 1 of >> `ib_post_send_mad' from incompatible pointer type >> drivers/infiniband/core/cm.c:1856: error: too few arguments to >> function `ib_post_send_mad' >> drivers/infiniband/core/cm.c: In function `cm_dreq_handler': >> drivers/infiniband/core/cm.c:1891: warning: passing arg 2 of >> `ib_cancel_mad' makes integer from pointer without a cast >> drivers/infiniband/core/cm.c:1905: warning: passing arg 1 of >> `ib_post_send_mad' from incompatible pointer type >> drivers/infiniband/core/cm.c:1905: error: too few arguments to >> function `ib_post_send_mad' >> drivers/infiniband/core/cm.c: In function `cm_drep_handler': >> drivers/infiniband/core/cm.c:1952: warning: passing arg 2 of >> `ib_cancel_mad' makes integer from pointer without a cast >> drivers/infiniband/core/cm.c: In function `ib_send_cm_rej': >> drivers/infiniband/core/cm.c:2020: warning: passing arg 1 of >> `ib_post_send_mad' from incompatible pointer type >> drivers/infiniband/core/cm.c:2020: error: too few arguments to >> function `ib_post_send_mad' >> drivers/infiniband/core/cm.c: In function `cm_rej_handler': >> drivers/infiniband/core/cm.c:2096: warning: passing arg 2 of >> `ib_cancel_mad' makes integer from pointer without a cast >> drivers/infiniband/core/cm.c:2106: warning: passing arg 2 of >> `ib_cancel_mad' makes integer from pointer without a cast >> drivers/infiniband/core/cm.c: In function `ib_send_cm_mra': >> drivers/infiniband/core/cm.c:2164: warning: passing arg 1 of >> `ib_post_send_mad' from incompatible pointer type >> drivers/infiniband/core/cm.c:2164: error: too few arguments to >> function `ib_post_send_mad' >> drivers/infiniband/core/cm.c:2177: warning: passing arg 1 of >> `ib_post_send_mad' from incompatible pointer type >> drivers/infiniband/core/cm.c:2177: error: too few arguments to >> function `ib_post_send_mad' >> drivers/infiniband/core/cm.c:2190: warning: passing arg 1 of >> `ib_post_send_mad' from incompatible pointer type >> drivers/infiniband/core/cm.c:2190: error: too few arguments to >> function `ib_post_send_mad' >> drivers/infiniband/core/cm.c: In function `cm_mra_handler': >> drivers/infiniband/core/cm.c:2252: warning: passing arg 2 of >> `ib_modify_mad' makes integer from pointer without a cast >> drivers/infiniband/core/cm.c:2259: warning: passing arg 2 of >> `ib_modify_mad' makes integer from pointer without a cast >> drivers/infiniband/core/cm.c:2267: warning: passing arg 2 of >> `ib_modify_mad' makes integer from pointer without a cast >> drivers/infiniband/core/cm.c: In function `ib_send_cm_lap': >> drivers/infiniband/core/cm.c:2350: error: structure has no member >> named `timeout_ms' >> drivers/infiniband/core/cm.c:2353: warning: passing arg 1 of >> `ib_post_send_mad' from incompatible pointer type >> drivers/infiniband/core/cm.c:2353: error: too few arguments to >> function `ib_post_send_mad' >> drivers/infiniband/core/cm.c: In function `cm_lap_handler': >> drivers/infiniband/core/cm.c:2430: warning: passing arg 1 of >> `ib_post_send_mad' from incompatible pointer type >> drivers/infiniband/core/cm.c:2430: error: too few arguments to >> function `ib_post_send_mad' >> drivers/infiniband/core/cm.c: In function `ib_send_cm_apr': >> drivers/infiniband/core/cm.c:2508: warning: passing arg 1 of >> `ib_post_send_mad' from incompatible pointer type >> drivers/infiniband/core/cm.c:2508: error: too few arguments to >> function `ib_post_send_mad' >> drivers/infiniband/core/cm.c: In function `cm_apr_handler': >> drivers/infiniband/core/cm.c:2547: warning: passing arg 2 of >> `ib_cancel_mad' makes integer from pointer without a cast >> drivers/infiniband/core/cm.c: In function `ib_send_cm_sidr_req': >> drivers/infiniband/c
Re: [openib-general] Re: [PATCH] libibmad: configure option to skip library test
On Mon, Aug 08, 2005 at 11:29:07AM -0400, Hal Rosenstock wrote: >On Mon, 2005-08-08 at 10:26, Michael S. Tsirkin wrote: >> Hal, I'm trying to split the build process to configure/make/install >> steps. >Seems reasonable. Will you be doing the same to umad and common too ? > >> Add option to skip >> >> Signed-off-by: Michael S. Tsirkin <[EMAIL PROTECTED]> > >Thanks. Applied. Hal, please correct this: presense -> presence TIA, Bernhard ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] Re: [PATCH][dapl] cleanup dapl_cookie
On Thu, Jun 30, 2005 at 10:04:36AM -0400, James Lentini wrote: > > >On Wed, 29 Jun 2005, Bernhard Fischer wrote: > >>On Tue, Jun 28, 2005 at 03:39:59PM -0400, James Lentini wrote: >>> >>>Hi Bernhard, >>> >>>The changes look fine. Why the additional copyright? I need to be able >>>to explain it to my legal department. >> >>My legaleeze states that whatever i do during work-time is contributed >>to work and whatever is related to work done during leasure time has to >>be attributed to /me _at_ _least_. As that snippet (which was a >>test-balloon >>for that category) clearly was done in my spare time, i'm forced to >>attribute it accordingly :-/ >> >>Does that answer your question satisfactorily? > >Thanks Bernhard. That makes sense to me. My legal inquired about the >"all rights reserved" qualifier. All the copyrights I found in the >OpenIB tree (including NetApp's) use that language. I'll run this by >them. > As rev. 2934 i do not see this patch applied. To recap, it removed some unneeded local variables (which my compiler wasn't smart enough to eleminate on it's own -- gcc-4.0 and gcc-HEAD) and simplified some conditionals and branches. Back then, i only submitted the changes to dapl_cookie.c to see if such kind of code simplifications would be accepted or not. James, can you please elaborate why the patch was rejected? Thank you, Bernhard ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] Gen2 build/update process
On Mon, Jul 11, 2005 at 12:10:48PM -0700, Tom Duffy wrote: >Sayantan Sur wrote: > >>Hi, >> >>Is there any plan to have any formal install procedure instead of the >>current process of going into subdirectories and building? >> >>If someone has build scripts to share, that'll be real nice. >> >> > >What are you trying to build? Kernel build should not need to do this. .. and src/userspace/management has a 'automake' target. In case it's helpful for you, i'm attaching a simple script which can be easily extended to cover the rest. I usually call it like this: nohup ./mk.openib2.usr.sh 01a It's not pretty but you get the idea. hth, Bernhard mk.openib2.usr.sh Description: Bourne shell script ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] Re: [PATCH] ibdm nits and a few questions
On Thu, Jul 07, 2005 at 05:05:01PM +0300, Eitan Zahavi wrote: >Hi Ber, > >I plan a huge change in the simulator core. So I do not want to waste your >time in cleaning up this code. I have some branch that I use now for >debugging. > >I appreciate your check and contribution but as the code I'm about to commit >has many diffs I think it make sense to leave the IBMgtSim for 1 week or so. Ok, thanks. ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] Re: [PATCH] ibdm nits and a few questions
On Thu, Jul 07, 2005 at 10:15:15AM +0300, Eitan Zahavi wrote: >Hi Ber, > >Thanks for the patch and the word list. I will apply these right away. massage -> message randesvous -> rendezvous menager -> manager contant -> content accoumulate -> accumulate somthing -> something Constractor -> Constructor naem -> name > >> >> What about removing all the generated scripts (configure et al) and >> auto-tools helpers (aclocal.m4 etc and most of config/*) from the >> repository? >[EZ] I thought I have done that. I will double check. In both ibdm,IBMgtSim there is a configure config/share config/libtoolize. ibdm/config/depcomp libtool.m4 install-sh are (or should be) all generated. > >> Note that the other userspace IB libraries do not have these in the >> repository but each do provide a toplevel ./autogen.sh -- as opposed >> to ibdm's ./config/bootstrap. >[EZ] We can rename the bootstrap autogen.sh but from what I learned about >auto tools many use ./config/bootstrap ... >Is it really annoying? It's not consistent, but i don't really care either way. >> There are a number of other unused variables. >[EZ] We should fix all of those! IBMgtSim/src/sim.cpp does not compile for me. I tried gcc-3.x as well as 4.x to no avail. Ideas? sim.cpp:201: error: expected `}' at end of input >> Is ibdm supposed to be compilable with pure ISO c++ compilers, btw? >> There is at least one place which uses a variable sized array which AFAIK >> is not valid ISO c++. >[EZ] I'm not sure what the answer to that is. I guess it's a question of policy. If gcc can deal with it, it's fine with me, fwiw. o IBMSClientInMsgs::proccessClientMsg() strip clientSock which from a short glance seems to be unused. Attached patch does: in ibmgtsim - turn ibms_disconnect() and ibms_disconnect()) into a void functions - add static in few places - opensm off gen2 doesn't seem to have osm_build_id.h so check for osm_base.h instead - correct offset in configure.in's --enable-debug help text Signed-off-by: Bernhard Fischer <[EMAIL PROTECTED]> Please review the patch and advise on the sim.cpp compile failure i get with pristine sources. TIA, Bernhard diff -X excl -rduNp gen2.2810.oorig/utils/src/linux-user/IBMgtSim/config/osm.m4 gen2.2810/utils/src/linux-user/IBMgtSim/config/osm.m4 --- gen2.2810.oorig/utils/src/linux-user/IBMgtSim/config/osm.m4 2005-03-24 17:56:41.0 +0100 +++ gen2.2810/utils/src/linux-user/IBMgtSim/config/osm.m4 2005-07-06 17:40:30.0 +0200 @@ -35,14 +35,14 @@ if test "x$with_osm" = xnone; then fi dnl validate the defined path -AC_CHECK_FILE($with_osm/include/opensm/osm_build_id.h,, - AC_MSG_ERROR([ could not find $with_osm/include/opensm/osm_build_id.h])) +AC_CHECK_FILE($with_osm/include/infiniband/opensm/osm_base.h,, + AC_MSG_ERROR([ could not find $with_osm/include/infiniband/opensm/osm_base.h])) AC_SUBST(with_osm) dnl now figure out somehow if the build was for debug or not dnl the header file should have the string debug if OpenSM is debug mode: -if test `grep debug $with_osm/include/opensm/osm_build_id.h | wc -l` = 1; then +if test "$debug" != "false"; then dnl why did they need so many ??? OSM_DEBUG_FLAGS='-DDEBUG -D_DEBUG -D_DEBUG_ -DDBG' AC_MSG_NOTICE(OSM compiled in DEBUG mode) diff -X excl -rduNp gen2.2810.oorig/utils/src/linux-user/IBMgtSim/configure.in gen2.2810/utils/src/linux-user/IBMgtSim/configure.in --- gen2.2810.oorig/utils/src/linux-user/IBMgtSim/configure.in 2005-03-24 17:56:44.0 +0100 +++ gen2.2810/utils/src/linux-user/IBMgtSim/configure.in2005-07-06 17:39:30.0 +0200 @@ -68,7 +68,7 @@ AC_CHECK_FUNCS([strchr strstr strtol str dnl Define an input config option to control debug compile AC_ARG_ENABLE(debug, -[ --enable-debugTurn on debugging], +[ --enable-debug Turn on debugging], [case "${enableval}" in yes) debug=true ;; no) debug=false ;; diff -X excl -rduNp gen2.2810.oorig/utils/src/linux-user/IBMgtSim/src/client.cpp gen2.2810/utils/src/linux-user/IBMgtSim/src/client.cpp --- gen2.2810.oorig/utils/src/linux-user/IBMgtSim/src/client.cpp 2005-03-24 17:55:58.0 +0100 +++ gen2.2810/utils/src/linux-user/IBMgtSim/src/client.cpp 2005-07-07 14:18:38.0 +0200 @@ -302,7 +302,7 @@ ibms_send( } /* disconnect from the simulator */ -int +void ibms_disconnect( ibms_conn_handle_t conHdl) { diff -X excl -rduNp gen2.2810.oorig/utils/src/linux-user/IBMgtSim/src/ibms_client_api.h gen2.2810/utils/src/linux-user/IBMgtSim/src/ibms_client_api.h --- gen2.2810.oorig/utils/src/linux-user/IBMgtSim/src/ibms_client_api.h 2005-03-24 17:55:59.0 +0100 +++ gen2.2810/utils/src/linux-user/IBMgtSim/src/ibms_client_api.h 2005-07-07 14:18:46.0 +0200 @
[openib-general] [PATCH][osm] install iba include
On Thu, Jul 07, 2005 at 10:15:15AM +0300, Eitan Zahavi wrote: >> IBMgtSim doesn't compile for me as iba/ib_types.h isn't installed, so >[EZ] You are missing OpenSM installation on the machine. It is required ... I did have opensm installed. >You can use --with-osm to point to where it is installed ... - install iba/ib_types.h which is needed for IBMgtSim. Signed-off-by: Bernhard Fischer <[EMAIL PROTECTED]> Index: trunk/src/userspace/management/Makefile === --- trunk/src/userspace/management/Makefile (revision 2810) +++ trunk/src/userspace/management/Makefile (working copy) @@ -1,7 +1,7 @@ #LIBS:=$(wildcard lib*) LIBS:=libibcommon libibumad libibmad OSM:=osm -OSMLIBS:=complib libvendor +OSMLIBS:= include complib libvendor UTIL:=$(wildcard util/*) DIAG:=$(wildcard diags/*) ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] [PATCH] ibdm nits and a few questions
Hi Eitan and Hal, On Tue, Jul 05, 2005 at 02:23:57PM +0300, Eitan Zahavi wrote: >Hi Ber, > >Thanks. >I have used an ex script to fix the following typos on all the code tree's I >own. >In the future - it is much better from my perspective to simply get the list >of words I need to check for. >I hope you will be able to simply send me the list of words. It will make my >life easier. Sure. missm -> mism optential -> potential finaly -> finally adn -> and simplicty -> simplicity consistant -> consistent the[[:space:]]*the[[:space:]] -> the (i.e. double 'the' but not 'the theory') incomming -> incoming seperat -> separat unseccessful -> unsuccessful fucntions -> functions acording -> according (acordingly -> accordingly) callabcks -> callbacks distructor -> destructor Distractor -> Destructor reciev -> receiv (recieving -> receiving) (Recieved -> Received) What about removing all the generated scripts (configure et al) and auto-tools helpers (aclocal.m4 etc and most of config/*) from the repository? Note that the other userspace IB libraries do not have these in the repository but each do provide a toplevel ./autogen.sh -- as opposed to ibdm's ./config/bootstrap. There are a number of other unused variables. Is ibdm supposed to be compilable with pure ISO c++ compilers, btw? There is at least one place which uses a variable sized array which AFAIK is not valid ISO c++. Attached patch does: IBSystem::guid: As default nowadays is public, move guid into the private section explicitely. configure.in: fix indentation of --enable-debug help text. IBFabric::parseFdbFile(): initialize p_node IBFabric::parseMCFdbFile(): ditto IBFabric::IBLinkWidth: remove comma at end of enumerator list IBFabric::IBLinkSpeed: ditto IBFabric:: remove 'IBFabric::' qualifier from member function parseMCFdbFile() Fabric.h: remove extra ';' from member functions datamodel/ibdmsh_wrap.cpp::ibdmGetObjPtrByTclName(): avoid using uninitialized name datamodel/ibdm_wrap.cpp::ibdmGetObjPtrByTclName(): ditto datamodel/SubnMgt.cpp::SubnMgtCalcMinHopTables(): make sure worstHopLid is initialized datamodel/SubnMgt.cpp::SubnReportNonUpDownCa2CaPaths(): make sure p_prevNode is initialized datamodel/TopoMatch.cpp::TopoMatchNodeByAdjacentMatched(): make sure succeedMatch is initialized datamodel/TopoMatch.cpp::TopoMatchSpecNodesByAdjacentNode(): add missing return numMatched; datamodel/TraceRoute.cpp::TraceDRPathRoute(): remove check which is always false due to limited datatype src/osm_check.cpp::main(): remove unused variable 'status' IBMgtSim doesn't compile for me as iba/ib_types.h isn't installed, so the above is only compile-tested. Hal, will you take care of installing the missing headers? Thank you, Bernhard diff -X excl -rduNp gen2.2810.oorig/utils/src/linux-user/ibdm/configure.in gen2.2810/utils/src/linux-user/ibdm/configure.in --- gen2.2810.oorig/utils/src/linux-user/ibdm/configure.in 2005-03-24 17:55:47.0 +0100 +++ gen2.2810/utils/src/linux-user/ibdm/configure.in2005-07-06 13:42:52.0 +0200 @@ -73,7 +73,7 @@ dnl AC_SUBST(LTLIBOBJS) dnl Define an input config option to control debug compile AC_ARG_ENABLE(debug, -[ --enable-debugTurn on debugging], +[ --enable-debug Turn on debugging], [case "${enableval}" in yes) debug=true ;; no) debug=false ;; diff -X excl -rduNp gen2.2810.oorig/utils/src/linux-user/ibdm/datamodel/Fabric.cpp gen2.2810/utils/src/linux-user/ibdm/datamodel/Fabric.cpp --- gen2.2810.oorig/utils/src/linux-user/ibdm/datamodel/Fabric.cpp 2005-07-06 11:52:49.0 +0200 +++ gen2.2810/utils/src/linux-user/ibdm/datamodel/Fabric.cpp2005-07-06 14:31:01.0 +0200 @@ -1593,7 +1593,7 @@ IBFabric::parseFdbFile(string fn) { cout << "-I- Parsing OpenSM FDBs file:" << fn.c_str() << endl; - IBNode *p_node; + IBNode *p_node = NULL; int anyErr = 0; while (f.good()) { @@ -1665,7 +1665,7 @@ IBFabric::parseMCFdbFile(string fn) { cout << "-I- Parsing OpenSM Multicast FDBs file:" << fn.c_str() << endl; - IBNode *p_node; + IBNode *p_node = NULL; int anyErr = 0; while (f.good()) { diff -X excl -rduNp gen2.2810.oorig/utils/src/linux-user/ibdm/datamodel/Fabric.h gen2.2810/utils/src/linux-user/ibdm/datamodel/Fabric.h --- gen2.2810.oorig/utils/src/linux-user/ibdm/datamodel/Fabric.h 2005-07-06 11:52:49.0 +0200 +++ gen2.2810/utils/src/linux-user/ibdm/datamodel/Fabric.h 2005-07-06 14:16:07.0 +0200 @@ -132,7 +132,7 @@ typedef enum {IB_UNKNOWN_LINK_WIDTH = 0, IB_LINK_WIDTH_1X = 1, IB_LINK_WIDTH_4X = 2, IB_LINK_WIDTH_8X = 4, - IB_LINK_WIDTH_12X =8, + IB_LINK_WIDTH_12X =8 } IBLinkWidth; static inline IBLinkWidth char2width( const char *w) @@ -143,7 +143,7 @@ static inline IBLinkWidth char2width( co if (!strcmp(w,"8x")) return IB_LINK_WIDTH_8X; if (!strcmp(w,"12
Re: [openib-general] OpenIB Releases
On Tue, Jul 05, 2005 at 03:28:16PM -0400, Hal Rosenstock wrote: >On Tue, 2005-07-05 at 15:22, Bernhard Fischer wrote: >> On Tue, Jul 05, 2005 at 02:08:41PM -0400, Hal Rosenstock wrote: >> >Hi, >> [] >> >User Space >> [] >> >MPI ? >> I started to port LAM/MPI to gen2's verbs but due to lack of HW and >> time it is still not completely finished > >How far along is it ? Does it compile ? Does it link ? It compiles and links, fwiw. > >> RTR->RTS state handling has a bug). > >Can you elaborate on this bug ? It's a pure bug in my port, not openib. IIRC it overflows when setting wr_id, but i didn't track it down. >> I, personally, would be very happy if there finally was a working >> MPI implementation perusing gen2 (which can be obtained without filling >> in some odd HTML form). ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] OpenIB Releases
On Tue, Jul 05, 2005 at 02:08:41PM -0400, Hal Rosenstock wrote: >Hi, [] >User Space [] > MPI ? I started to port LAM/MPI to gen2's verbs but due to lack of HW and time it is still not completely finished (RTR->RTS state handling has a bug). I did choose LAM because it's available publically and right now, as opposed to other eventually forthcoming solutions. I, personally, would be very happy if there finally was a working MPI implementation perusing gen2 (which can be obtained without filling in some odd HTML form). >Key to all this is testers picking these releases up and testing and >reporting back to the list. > >Comments ? > >-- Hal Bernhard ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [Openib-windows] Re: [openib-general] Re: IB Diagnositic Tool s
On Mon, Jul 04, 2005 at 11:54:18PM +0300, Eitan Zahavi wrote: >Hi Ben > >The patch try to start scanning the ports from i=1 rather then i=0. Hm? My patch started to scan from 0 and not 1; The patch i sent to Hal in February kindly started at 1 (i.e. current opensm), but i wonder what port 0 in this case does denotes. Please clarify. > >status = osm_vendor_get_all_port_attr(p_osm->p_vendor, attr_array, >&num_ports ); > >But this function to return the port attributes of all available ports does >not use the "port number" as index into the returned array of port >attributes. It simply starts with i=0 and adds on to it. The port numbers >are available in the attributes struct. > >Eitan Zahavi >Design Technology Director >Mellanox Technologies LTD >Tel:+972-4-9097208 >Fax:+972-4-9593245 >P.O. Box 586 Yokneam 20692 ISRAEL Bernhard ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] [PATCH][IBMgtSim] commentary typos
Hi, - commentary typos. Signed-off-by: Bernhard Fischer <[EMAIL PROTECTED]> No obj-code changes. PS: Eitan; oh and yes, so far i didn't even try to build one of ibdm nor IBMgtSim, fwiw.. I'll most likely get back to you on those. thank you, Bernhard diff -X excl -rduNp gen2.2780.oorig/utils/src/linux-user/IBMgtSim/src/dispatcher.cpp gen2.2780/utils/src/linux-user/IBMgtSim/src/dispatcher.cpp --- gen2.2780.oorig/utils/src/linux-user/IBMgtSim/src/dispatcher.cpp 2005-04-26 20:57:17.0 +0200 +++ gen2.2780/utils/src/linux-user/IBMgtSim/src/dispatcher.cpp 2005-07-04 21:12:07.0 +0200 @@ -69,7 +69,7 @@ IBMSDispatcher::IBMSDispatcher( MSG_EXIT_FUNC; } -/* distructor */ +/* destructor */ IBMSDispatcher::~IBMSDispatcher() { MSG_ENTER_FUNC; diff -X excl -rduNp gen2.2780.oorig/utils/src/linux-user/IBMgtSim/src/sma.cpp gen2.2780/utils/src/linux-user/IBMgtSim/src/sma.cpp --- gen2.2780.oorig/utils/src/linux-user/IBMgtSim/src/sma.cpp 2005-07-04 20:19:54.0 +0200 +++ gen2.2780/utils/src/linux-user/IBMgtSim/src/sma.cpp 2005-07-04 21:30:50.0 +0200 @@ -58,7 +58,7 @@ ibms_dump_mad( { if (dir == RCV) { -MSGREG(inf100, 'M', "\nRecieved Mad Dump\n$", "ibms_dump_mad"); +MSGREG(inf100, 'M', "\nReceived Mad Dump\n$", "ibms_dump_mad"); MSGSND(inf100, ibms_get_mad_header_str(madMsg.header)); } else @@ -1423,7 +1423,7 @@ int IBMSSma::processMad(uint8_t inPort, if (madMsg.header.method == IB_MAD_METHOD_TRAP_REPRESS) { -MSGREG(inf0, 'I', "--- Recieved Trap Repress ---", "processMad"); +MSGREG(inf0, 'I', "--- Received Trap Repress ---", "processMad"); MSGSND(inf0); MSG_EXIT_FUNC; diff -X excl -rduNp gen2.2780.oorig/utils/src/linux-user/IBMgtSim/src/sma.h gen2.2780/utils/src/linux-user/IBMgtSim/src/sma.h --- gen2.2780.oorig/utils/src/linux-user/IBMgtSim/src/sma.h 2005-04-26 20:57:17.0 +0200 +++ gen2.2780/utils/src/linux-user/IBMgtSim/src/sma.h 2005-07-04 21:30:32.0 +0200 @@ -63,7 +63,7 @@ * * DESCRIPTION * upper level function of printing all mad contant in -* recieving or sending of Mad +* receiving or sending of Mad * * SYNOPSIS */ ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] [PATCH][iser] commentary typos
To whom it may concern, - commentary typos. Signed-off-by: Bernhard Fischer <[EMAIL PROTECTED]> No obj-code changes. thank you, Bernhard diff -X excl -rduNp gen2.2780.oorig/ulps/iser/datamover/iser_conn.c gen2.2780/ulps/iser/datamover/iser_conn.c --- gen2.2780.oorig/ulps/iser/datamover/iser_conn.c 2005-04-26 20:57:58.0 +0200 +++ gen2.2780/ulps/iser/datamover/iser_conn.c 2005-07-04 21:30:11.0 +0200 @@ -276,7 +276,7 @@ iser_connection_accept(iser_conn_request iser_cr_reject(p_iser_conn); ret_val = ISER_FAILURE; IERROR - ("post recieve pool == NULL, connection rejected\n"); + ("post receive pool == NULL, connection rejected\n"); goto iser_connection_accept_exit; } /* During the login phase data for sent PDUs is drawn from the same buffer pool as recv */ @@ -1186,7 +1186,7 @@ iser_post_receive_control(struct iser_co ITRACE(ISER_TRACE_CONN, "Added data buffer 0x%p to DTO as entry: %d\n", p_regd_buf, i); - ITRACE(ISER_TRACE_CONN, "Call post recieve with the DTO\n"); + ITRACE(ISER_TRACE_CONN, "Call post receive with the DTO\n"); if (iser_post_recv(p_recv_dto) != 0) { IERROR("Failed to post receive DTO\n"); ret_val = -1; ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] [PATCH][mad] commentary typos
To whom it may concern, - commentary typos. Signed-off-by: Bernhard Fischer <[EMAIL PROTECTED]> No obj-code changes. thank you, Bernhard diff -X excl -rduNp gen2.2780.oorig/trunk/src/userspace/management/libibmad/src/fields.c gen2.2780/trunk/src/userspace/management/libibmad/src/fields.c --- gen2.2780.oorig/trunk/src/userspace/management/libibmad/src/fields.c 2005-06-05 14:16:19.0 +0200 +++ gen2.2780/trunk/src/userspace/management/libibmad/src/fields.c 2005-07-04 21:19:48.0 +0200 @@ -44,9 +44,9 @@ #include /* - * BITSOFFS and BE_OFFS are required due the fact that the bit offsets are incosistently + * BITSOFFS and BE_OFFS are required due the fact that the bit offsets are inconsistently * encoded in the IB spec - IB headers are encoded such that the bit offsets - * are in big endian convension (BE_OFFS), while the SMI/GSI queries data fields bit + * are in big endian convention (BE_OFFS), while the SMI/GSI queries data fields bit * offsets are specified using real bit offset (?!) * The following macros normalize everything to big endian offsets. */ @@ -202,7 +202,7 @@ ib_field_t ib_mad_f [] = { [IB_LINEAR_FORW_TBL_F] {0, 512, "LinearForwTbl", mad_dump_array}, /* -* NodeDescription fileds: +* NodeDescription fields: */ [IB_NODE_DESC_F]{0, 64*8, "NodeDesc", mad_dump_string}, ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] [PATCH][dat] commentary typos
Hi, - commentary typos. Signed-off-by: Bernhard Fischer <[EMAIL PROTECTED]> No obj-code changes. thank you, Bernhard diff -X excl -rduNp gen2.2780.oorig/users/jlentini/linux-kernel/dat-provider/dapl_sp.c gen2.2780/users/jlentini/linux-kernel/dat-provider/dapl_sp.c --- gen2.2780.oorig/users/jlentini/linux-kernel/dat-provider/dapl_sp.c 2005-06-27 23:07:04.0 +0200 +++ gen2.2780/users/jlentini/linux-kernel/dat-provider/dapl_sp.c 2005-07-04 21:29:16.0 +0200 @@ -330,9 +330,9 @@ u32 dapl_rsp_query(struct dat_sp *dat_rs } /* - * Create a persistent Public Service Point that can recieve multiple + * Create a persistent Public Service Point that can receive multiple * requests for connections and generate multiple connection request - * instances that wil be delivered to the specified Event Dispatcher + * instances that will be delivered to the specified Event Dispatcher * in a notification event. Differs from dapl_psp_create() in that * the conn_qual is selected by the implementation and returned to * the user. @@ -435,9 +435,9 @@ bail: } /* - * Create a persistent Public Service Point that can recieve multiple + * Create a persistent Public Service Point that can receive multiple * requests for connections and generate multiple connection request - * instances that wil be delivered to the specified Event Dispatcher + * instances that will be delivered to the specified Event Dispatcher * in a notification event. */ u32 dapl_psp_create(struct dat_ia *dat_ia, DAT_CONN_QUAL conn_qual, diff -X excl -rduNp gen2.2780.oorig/users/jlentini/linux-kernel/dat-provider/dapl_srq.c gen2.2780/users/jlentini/linux-kernel/dat-provider/dapl_srq.c --- gen2.2780.oorig/users/jlentini/linux-kernel/dat-provider/dapl_srq.c 2005-06-27 23:07:04.0 +0200 +++ gen2.2780/users/jlentini/linux-kernel/dat-provider/dapl_srq.c 2005-07-04 21:22:07.0 +0200 @@ -207,7 +207,7 @@ u32 dapl_srq_free(struct dat_srq *srq) if (atomic_read(&srq_ptr->srq_ref_count) != 0) { /* * The DAPL 1.2 spec says to return DAT_SRQ_IN_USE, which does -* not exist. Have filed the following as an eratta. +* not exist. Have filed the following as an errata. */ status = DAT_ERROR(DAT_INVALID_STATE, DAT_INVALID_STATE_SRQ_IN_USE); @@ -338,7 +338,7 @@ u32 dapl_srq_query(struct dat_srq *srq, /* * dapl_srq_resize * - * Modify the size fo the event queue of a Shared Recieve Queue + * Modify the size of the event queue of a Shared Receive Queue * * Input: * srq diff -X excl -rduNp gen2.2780.oorig/users/jlentini/linux-kernel/test/dapltest/test/dapl_performance_client.c gen2.2780/users/jlentini/linux-kernel/test/dapltest/test/dapl_performance_client.c --- gen2.2780.oorig/users/jlentini/linux-kernel/test/dapltest/test/dapl_performance_client.c 2005-06-27 23:07:03.0 +0200 +++ gen2.2780/users/jlentini/linux-kernel/test/dapltest/test/dapl_performance_client.c 2005-07-04 21:29:42.0 +0200 @@ -409,7 +409,7 @@ DT_Performance_Test_Client_Exchange ( test_ptr->ep_context.ep, DT_PERF_SYNC_BUFF_SIZE, dto_cookie, - "Recieve Sync_Msg") ) + "Received Sync_Msg") ) { return FALSE; } diff -X excl -rduNp gen2.2780.oorig/users/jlentini/linux-kernel/test/dapltest/test/dapl_performance_server.c gen2.2780/users/jlentini/linux-kernel/test/dapltest/test/dapl_performance_server.c --- gen2.2780.oorig/users/jlentini/linux-kernel/test/dapltest/test/dapl_performance_server.c 2005-06-27 23:07:03.0 +0200 +++ gen2.2780/users/jlentini/linux-kernel/test/dapltest/test/dapl_performance_server.c 2005-07-04 21:29:27.0 +0200 @@ -384,7 +384,7 @@ DT_Performance_Test_Server_Exchange ( test_ptr->ep_context.ep, DT_PERF_SYNC_BUFF_SIZE, dto_cookie, - "Recieve Sync_Msg") ) + "Received Sync_Msg") ) { return FALSE; } diff -X excl -rduNp gen2.2780.oorig/users/jlentini/userspace/dapl/common/dapl_evd_resize.c gen2.2780/users/jlentini/userspace/dapl/common/dapl_evd_resize.c --- gen2.2780.oorig/users/jlentini/userspace/dapl/common/dapl_evd_resize.c 2005-04-26 20:57:43.0 +0200 +++ gen2.2780/users/jlentini/userspace/dapl/common/dapl_evd_resize.c 2005-07-04 21:04:18.0 +0200 @@ -47,7 +47,7 @@ * * DAPL Requirements Version xxx, 6.3.2.5 * - * Modify the size fo the event queue of an Event Dispatcher + * Modify the size of the event queue of an Event Dispatcher * * Input: * evd_handle diff -X excl -rduNp gen2.2780.oorig/users/jlentini/userspace/dapl/common/dapl_psp_create_any.c gen2.2780
[openib-general] [PATCH][osm] commentary typos
Hi, - commentary typos. Signed-off-by: Bernhard Fischer <[EMAIL PROTECTED]> No obj-code changes. thank you, Bernhard diff -X excl -rduNp gen2.2780.oorig/trunk/src/userspace/management/osm/include/opensm/osm_port.h gen2.2780/trunk/src/userspace/management/osm/include/opensm/osm_port.h --- gen2.2780.oorig/trunk/src/userspace/management/osm/include/opensm/osm_port.h 2005-05-10 18:25:28.0 +0200 +++ gen2.2780/trunk/src/userspace/management/osm/include/opensm/osm_port.h 2005-07-04 21:09:50.0 +0200 @@ -391,7 +391,7 @@ osm_link_is_healthy( * * DESCRIPTION * Sets the port health flag. TRUE means the port is healthy and -* should eb used for packet routing. FALSE means it should be avoided. +* should be used for packet routing. FALSE means it should be avoided. * * SYNOPSIS */ diff -X excl -rduNp gen2.2780.oorig/trunk/src/userspace/management/osm/include/opensm/osm_switch.h gen2.2780/trunk/src/userspace/management/osm/include/opensm/osm_switch.h --- gen2.2780.oorig/trunk/src/userspace/management/osm/include/opensm/osm_switch.h 2005-04-26 20:58:23.0 +0200 +++ gen2.2780/trunk/src/userspace/management/osm/include/opensm/osm_switch.h 2005-07-04 21:09:10.0 +0200 @@ -1128,12 +1128,12 @@ osm_switch_recommend_path( * [in] The maximum allowed number of target lids routed through * a specific port of the switch. If the port already assigned * (in the lfdb) this number of target lids - it will not be used -* even if it has teh smallest hops count to teh target lid. +* even if it has the smallest hops count to the target lid. * * ui_ucast_fdb_assign_func_defined * [in] If TRUE - this means that there is a ui ucast_fdb_assign table * function defined (in pfn_ui_ucast_fdb_assign in subnet opts). This -* Means that all current entries in the fdbs will be used. +* means that all current entries in the fdbs will be used. * If FALSE - such function isn't defined. Do the minimum hop checks * before accepting the current fdbs. * diff -X excl -rduNp gen2.2780.oorig/trunk/src/userspace/management/osm/include/vendor/osm_vendor_ts.h gen2.2780/trunk/src/userspace/management/osm/include/vendor/osm_vendor_ts.h --- gen2.2780.oorig/trunk/src/userspace/management/osm/include/vendor/osm_vendor_ts.h 2005-04-26 20:58:29.0 +0200 +++ gen2.2780/trunk/src/userspace/management/osm/include/vendor/osm_vendor_ts.h 2005-07-04 21:07:07.0 +0200 @@ -115,7 +115,7 @@ typedef struct _osm_ts_bind_info * the file handle to be used for sending the MADs * * hca_hndl - * Handle to the HCA provided by teh underlying VAPI + * Handle to the HCA provided by the underlying VAPI * * p_vend * Pointer to the vendor object. diff -X excl -rduNp gen2.2780.oorig/trunk/src/userspace/management/osm/libvendor/osm_vendor_mlx_anafa.c gen2.2780/trunk/src/userspace/management/osm/libvendor/osm_vendor_mlx_anafa.c --- gen2.2780.oorig/trunk/src/userspace/management/osm/libvendor/osm_vendor_mlx_anafa.c 2005-04-26 20:58:32.0 +0200 +++ gen2.2780/trunk/src/userspace/management/osm/libvendor/osm_vendor_mlx_anafa.c 2005-07-04 21:05:23.0 +0200 @@ -515,7 +515,7 @@ osm_vendor_send (IN osm_bind_handle_t h_ /* Only the requestor expects the error callback */ p_bo->send_err_cb (p_bo->cb_context, p_madw); } else { - /* put back teh mad - it is useless ... */ + /* put back the mad - it is useless ... */ osm_mad_pool_put (p_bo->p_osm_pool, p_madw); } } diff -X excl -rduNp gen2.2780.oorig/trunk/src/userspace/management/osm/libvendor/osm_vendor_mlx.c gen2.2780/trunk/src/userspace/management/osm/libvendor/osm_vendor_mlx.c --- gen2.2780.oorig/trunk/src/userspace/management/osm/libvendor/osm_vendor_mlx.c 2005-04-26 20:58:32.0 +0200 +++ gen2.2780/trunk/src/userspace/management/osm/libvendor/osm_vendor_mlx.c 2005-07-04 21:12:12.0 +0200 @@ -533,7 +533,7 @@ osm_vendor_send( } else { - /* put back teh mad - it is useless ... */ + /* put back the mad - it is useless ... */ osm_mad_pool_put(p_bo->p_osm_pool, p_madw ); } } @@ -542,7 +542,7 @@ osm_vendor_send( osmv_txn_unlock(p_bo); - /* we need the ref to happen after the unlock to prevent distruct of the unlock spinlock */ + /* we need the ref to happen after the unlock to prevent destruct of the unlock spinlock */ cl_atomic_dec(&p_bo->ref_count); if (TRUE == p_bo->is_closing) diff -X excl -rduNp gen2.2780.oorig/trunk/src/userspace/management/osm/libvendor/osm_vendor_mlx_ibmgt.c gen2.2780/trunk/src/userspace/management/osm/libvendor/osm_vendor_mlx_ibmgt.c --- gen2.2780.oorig/trunk/src/userspace/management/osm/libvendor/osm_vendor_mlx_ibmgt.c 20
Re: [Openib-windows] Re: [openib-general] Re: IB Diagnositic Tools
On Tue, Jun 28, 2005 at 09:22:00AM -0400, Hal Rosenstock wrote: >On Mon, 2005-06-27 at 17:30, Hal Rosenstock wrote: >> On Mon, 2005-06-27 at 17:13, Bernhard Fischer wrote: >> > Hal, >> > >> > This is kinda offtopic, >> >> Yes, this is different topic(s). >> >> > but (iirc) i once stumbled over the issue of >> > "port" vs. "mgmt port" [back then i had access to two 2-port cards] >> > where you may have said something along the lines of >> > \"There is clearly a bug for multi HCAs in osm_vendor_get_all_port_attr >> > which is in the vendor layer. This needs to be fixed and is our problem. >> > So I am close to being able to commit what I now have for this and fix >> > this later (as there are other multi HCA issues).\" >> > Just curious.. did somebody already have a chance to touch those or not? >> >> I don't think it has been fixed :-( I will need to refresh myself on >> this again. > >Just tried the diags and OpenSM on a multi HCA machine and they appear >to run fine. I will need to dig out from my notes as that comment came >from back on 2/16. Please note that that specific occurance was *not* triggered in a strict \"multi HCA\" setup but with two-port HCA(s e.g. opensm'd) back-to_back. Back then, i saw (can't reproduce it now since i do not have access to an appropriate setup; can't reproduce since even if i would have access to such a setup i most likely wouldn't have root access) a listing akin to : phy-port 1 phy-port 1 phy-port 2 e-o-f Please mail me the output of opensm -g [with opensm as primary SM -- but i guess it does literally everywhere, be it hosts or switches] with the attached patch. PS: Hal, if the moved question is true, please point me to a document and section where this is defined. I may be a pita, but i'd like to know that answer. TIA and . ..sorry for the noise -- Bernhard diff -X excl -rduNp gen2.2780.osm-local-port.oorig/trunk/src/userspace/management/osm/opensm/main.c gen2.2780.osm-local-port/trunk/src/userspace/management/osm/opensm/main.c --- gen2.2780.osm-local-port.oorig/trunk/src/userspace/management/osm/opensm/main.c 2005-04-26 20:58:16.0 +0200 +++ gen2.2780.osm-local-port/trunk/src/userspace/management/osm/opensm/main.c 2005-07-04 22:27:47.0 +0200 @@ -252,7 +252,8 @@ print_all_guids( } printf("\nListing GUIDs:\n"); - for (i = 1; i < num_ports; i++) { /* excluding logical mgmt port */ +/* "local ports" is(?) phys, shouldn't this exclude port 0 then ? */ + for (i = 0; i < num_ports; i++) { /* excluding logical mgmt port */ printf("Port %i: 0x%"PRIx64"\n", i, cl_hton64(attr_array[i].port_guid)); } } @@ -272,7 +273,7 @@ get_port_guid( Call the transport layer for a list of local port GUID values. */ -/* "local ports" is(?) phys, shouldn't this exclude port 0 then ? */ + status = osm_vendor_get_all_port_attr(p_osm->p_vendor, attr_array, &num_ports ); if ( status != IB_SUCCESS ) { ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] RE: [openib-commits] r2776 - gen2/utils/src/linux-user/ibdm/datam odel
On Mon, Jul 04, 2005 at 08:51:10PM +0200, Bernhard Fischer wrote: >On Mon, Jul 04, 2005 at 10:20:16AM +0300, Eitan Zahavi wrote: >>Thanks. Committed as 2779. > >I'm sorry but i overlooked another occurance of acor; attached. >Would you be so kind and do these in ibdm and IBMgtSim? TIA. >s/teh/the/g >s/eb/be/g >s/fo/of/g >this may or may not work for you: >egrep -ri "([[:space:]](fo|teh|eb)[[:space:]])" . >Index: utils/src/linux-user/IBMgtSim/utils/RunSimTest >=== >--- utils/src/linux-user/IBMgtSim/utils/RunSimTest (revision 2780) >+++ utils/src/linux-user/IBMgtSim/utils/RunSimTest (working copy) >@@ -112,18 +112,18 @@ proc Help {} { >-# The log fo all events is accoumulated in the global list: >+# The log of all events is accoumulated in the global list: s/accoumulate/accumulate/ ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] RE: [openib-commits] r2776 - gen2/utils/src/linux-user/ibdm/datam odel
On Mon, Jul 04, 2005 at 10:20:16AM +0300, Eitan Zahavi wrote: >Thanks. Committed as 2779. I'm sorry but i overlooked another occurance of acor; attached. Would you be so kind and do these in ibdm and IBMgtSim? TIA. s/teh/the/g s/eb/be/g s/fo/of/g this may or may not work for you: egrep -ri "([[:space:]](fo|teh|eb)[[:space:]])" . - fix a few typos. Signed-off-by: Bernhard Fischer <[EMAIL PROTECTED]> Index: utils/src/linux-user/IBMgtSim/src/ibdm.i === --- utils/src/linux-user/IBMgtSim/src/ibdm.i(revision 2780) +++ utils/src/linux-user/IBMgtSim/src/ibdm.i(working copy) @@ -732,7 +732,7 @@ int FabricUtilsVerboseLevel; IBDM exposes some of its internal objects. The objects identifiers returned by the various function calls are formatted - acording to the following rules: + according to the following rules: Fabric: fabric: System: system:: SysPort: sysport::: Index: utils/src/linux-user/IBMgtSim/utils/RunSimTest === --- utils/src/linux-user/IBMgtSim/utils/RunSimTest (revision 2780) +++ utils/src/linux-user/IBMgtSim/utils/RunSimTest (working copy) @@ -112,18 +112,18 @@ proc Help {} { # OpenSM log file analyzer # # Continuously monitor the OpenSM log file and generate a log of all the -# events reported in teh log file. +# events reported in the log file. # # On any event - it scans through the list of callbacks to be invoked -# and calls them acordingly. The list of callabcks is in osmLogCallbacks(logFile) +# and calls them accordingly. The list of callbacks is in osmLogCallbacks(logFile) # -# The log fo all events is accoumulated in the global list: +# The log of all events is accoumulated in the global list: # osmEventLog(logFile) # # The format of the event log list entry is: # # -# = SubnetUp, Standby, ERR, ErrorsInInitialization, ReportNoice, +# = SubnetUp, Standby, ERR, ErrorsInInitialization, ReportNotice, # = the local time in tracker. # = the complete log file line. # ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] [PATCH][osm] commentary typos
Hi Hal, - commentary typos Signed-off-by: Bernhard Fischer <[EMAIL PROTECTED]> Index: trunk/src/userspace/management/osm/opensm/osm_state_mgr.c === --- trunk/src/userspace/management/osm/opensm/osm_state_mgr.c (revision 2778) +++ trunk/src/userspace/management/osm/opensm/osm_state_mgr.c (working copy) @@ -518,7 +518,7 @@ __osm_state_mgr_reset_node_count( { osm_log( p_mgr->p_log, OSM_LOG_DEBUG, "__osm_state_mgr_reset_node_count: " - "Reseting discovery count for node 0x%" PRIx64 ".\n", + "Resetting discovery count for node 0x%" PRIx64 ".\n", cl_ntoh64( osm_node_get_node_guid( p_node ) )); } @@ -539,7 +539,7 @@ __osm_state_mgr_reset_port_count( { osm_log( p_mgr->p_log, OSM_LOG_DEBUG, "__osm_state_mgr_reset_port_count: " - "Reseting discovery count for port 0x%" PRIx64 ".\n", + "Resetting discovery count for port 0x%" PRIx64 ".\n", cl_ntoh64( osm_port_get_guid( p_port ) )); } @@ -560,7 +560,7 @@ __osm_state_mgr_reset_switch_count( { osm_log( p_mgr->p_log, OSM_LOG_DEBUG, "__osm_state_mgr_reset_switch_count: " - "Reseting discovery count for switch 0x%" PRIx64 ".\n", + "Resetting discovery count for switch 0x%" PRIx64 ".\n", cl_ntoh64( osm_node_get_node_guid( p_sw->p_node ) )); } Index: trunk/src/userspace/management/osm/opensm/osm_drop_mgr.c === --- trunk/src/userspace/management/osm/opensm/osm_drop_mgr.c(revision 2778) +++ trunk/src/userspace/management/osm/opensm/osm_drop_mgr.c(working copy) @@ -271,7 +271,7 @@ __osm_drop_mgr_remove_port( osm_port_discovery_count_reset( p_remote_port ); osm_log( p_mgr->p_log, OSM_LOG_DEBUG, "__osm_drop_mgr_remove_port: " - "reseting discovery count of node: " + "resetting discovery count of node: " "0x%016" PRIx64 " port num:%u.\n", cl_ntoh64( osm_node_get_node_guid( p_remote_node ) ), remote_port_num ); Index: trunk/src/userspace/management/osm/include/complib/cl_timer.h === --- trunk/src/userspace/management/osm/include/complib/cl_timer.h (revision 2778) +++ trunk/src/userspace/management/osm/include/complib/cl_timer.h (working copy) @@ -331,7 +331,7 @@ cl_timer_trim( * remaining time when the timer is set. * * If the new interval time is less than the remaining time, cl_timer_trim -* implicitly stops the timer before reseting it. +* implicitly stops the timer before resetting it. * * If the timer is reset, it is guaranteed to expire no sooner than the * new interval, but may take longer to expire. Index: trunk/src/userspace/management/osm/complib/cl_event.c === --- trunk/src/userspace/management/osm/complib/cl_event.c (revision 2778) +++ trunk/src/userspace/management/osm/complib/cl_event.c (working copy) @@ -102,7 +102,7 @@ cl_event_signal( cl_spinlock_acquire( &p_event->spinlock ); p_event->signaled = TRUE; - /* Wake up one or all depending on whether the event is auto-reseting. */ + /* Wake up one or all depending on whether the event is auto-resetting. */ if( p_event->manual_reset ) pthread_cond_broadcast( &p_event->condvar ); else ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] Re: [openib-commits] r2776 - gen2/utils/src/linux-user/ibdm/datamodel
On Sat, Jul 02, 2005 at 12:25:33PM -0700, [EMAIL PROTECTED] wrote: >Modified: gen2/utils/src/linux-user/ibdm/datamodel/Fabric.cpp Eitan, - fix a few typos. Signed-off-by: Bernhard Fischer <[EMAIL PROTECTED]> Index: utils/src/linux-user/ibdm/datamodel/ibdm.i === --- utils/src/linux-user/ibdm/datamodel/ibdm.i (revision 2778) +++ utils/src/linux-user/ibdm/datamodel/ibdm.i (working copy) @@ -780,7 +780,7 @@ int FabricUtilsVerboseLevel; IBDM exposes some of its internal objects. The objects identifiers returned by the various function calls are formatted - acording to the following rules: + according to the following rules: Fabric: fabric: System: system:: SysPort: sysport::: Index: utils/src/linux-user/ibdm/datamodel/Fabric.cpp === --- utils/src/linux-user/ibdm/datamodel/Fabric.cpp (revision 2778) +++ utils/src/linux-user/ibdm/datamodel/Fabric.cpp (working copy) @@ -34,7 +34,7 @@ /* IB Fabric Data Model -This file hodls implementation of the data model classes and methods +This file holds implementation of the data model classes and methods */ @@ -119,8 +119,8 @@ IBPort::connect (IBPort *p_otherPort, if (p_remotePort != p_otherPort) { cout << "-W- Disconnecting: " << p_remotePort->getName() << " previously connected to:" - << p_remotePort->getName() - << " whil econnecting:" << p_otherPort->getName() << endl; + << p_remotePort->getName() + << " while connecting:" << p_otherPort->getName() << endl; // the other side should be cleaned only if points here if (p_remotePort->p_remotePort == this) { p_remotePort->p_remotePort = NULL; @@ -580,7 +580,7 @@ IBSystem::getSysPort(string name) { // constructor: IBSystem::IBSystem(string n, class IBFabric *p_fab, string t) { if (p_fab->getSystem(n)) { -cerr << "Can' deal with double allocation of same system!" << endl; +cerr << "Can't deal with double allocation of same system!" << endl; abort(); } name = n; @@ -679,7 +679,7 @@ IBSystem::removeBoard (string boardName) // Warn if no match: if (matchedNodes.empty()) { -cout << "-W-(RemoveBoard) Fail to find any node in:" +cout << "-W- removeBoard : Fail to find any node in:" << sysNodePrefix << " while removing:" << boardName << endl; return 1; } @@ -1636,7 +1636,7 @@ IBFabric::parseFdbFile(string fn) { //cout << "-W- Ignoring line:" << sLine << endl; } - cout << "-I- Defined " << fdbLines << " fdb entires for:" + cout << "-I- Defined " << fdbLines << " fdb entries for:" << switches << " switches" << endl; f.close(); return anyErr; @@ -1715,7 +1715,7 @@ IBFabric::parseMCFdbFile(string fn) { //cout << "-W- Ignoring line:" << sLine << endl; } - cout << "-I- Defined " << fdbLines << " Multicast Fdb entires for:" + cout << "-I- Defined " << fdbLines << " Multicast Fdb entries for:" << switches << " switches" << endl; f.close(); return anyErr; Index: utils/src/linux-user/IBMgtSim/src/sma.cpp === --- utils/src/linux-user/IBMgtSim/src/sma.cpp (revision 2778) +++ utils/src/linux-user/IBMgtSim/src/sma.cpp (working copy) @@ -904,7 +904,7 @@ int IBMSSma::setIBPortBaseLid( unsigned intportLidIndex; MSG_ENTER_FUNC; - MSGREG(inf0, 'I', "Seting base_lid for node:$ port:$ to $", "setIBPortBaseLid"); + MSGREG(inf0, 'I', "Setting base_lid for node:$ port:$ to $", "setIBPortBaseLid"); pNode = pSimNode->getIBNode(); pPort = pNode->getPort(portNum); if (! pPort) @@ -925,7 +925,7 @@ int IBMSSma::setIBPortBaseLid( /* make sure the vector of port by lid has enough entries */ if (pNode->p_fabric->PortByLid.size() <= base_lid) { -/* we add 20 entires each time */ +/* we add 20 entries each time */ pNode->p_fabric->PortByLid.resize(base_lid+20); for ( portLidIndex = pNode->p_fabric->PortByLid.size(); portLidIndex < base_lid + 20; ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] Re: [PATCH] [uDAPL] Fix compile of dapl_bpool.c on x86_64
On Wed, Jun 29, 2005 at 05:18:53PM -0400, James Lentini wrote: > >Committed in revision 2751. > >On Tue, 28 Jun 2005, Hal Rosenstock wrote: > >halr> I was able to build udapltest for both x86 and x86-64 with the following >halr> change: >halr> >halr> uDAPL: Fix compile of dapl_bpool.c on x86_64 >halr> >halr> Signed-off-by: Hal Rosenstock <[EMAIL PROTECTED]> >halr> >halr> --- userspace/test/dapltest/test/dapl_bpool.c.orig 2005-01-18 >halr> 13:20:31.0 -0500 >halr> +++ userspace/test/dapltest/test/dapl_bpool.c 2005-06-24 >halr> 23:47:43.0 -0400 >halr> @@ -356,8 +356,8 @@ [snip full patch] James, Don't mean to sound rude, but imagine there *still* are modem-users around! That may sound odd to you as you're well connected, still those do exists.. May i kindly ask that you do *not* quote the patch you ACK'ed but just the subj/thread? Folks interrested still can readup in the mail-archives. Look at e.g. Roland, who also does not quote the patch itself but politely follows-up with something along the lines of 'comitted'. Thank you. Yours sincerely, ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] Re: [PATCH][dapl] cleanup dapl_cookie
On Tue, Jun 28, 2005 at 03:39:59PM -0400, James Lentini wrote: > >Hi Bernhard, > >The changes look fine. Why the additional copyright? I need to be able >to explain it to my legal department. My legaleeze states that whatever i do during work-time is contributed to work and whatever is related to work done during leasure time has to be attributed to /me _at_ _least_. As that snippet (which was a test-balloon for that category) clearly was done in my spare time, i'm forced to attribute it accordingly :-/ Does that answer your question satisfactorily? > >james Sorry, -- Bernhard ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] Re: IB Diagnositic Tools
On Mon, Jun 27, 2005 at 02:04:16PM -0400, Hal Rosenstock wrote: >Hi Eitan, > >On Sat, 2005-06-25 at 15:25, Eitan Zahavi wrote: >> Following the discussion about the debug tools, I would like to >> propose using OpenSM Vendor layer as a common layer for developing the >> debug tools. Hal, This is kinda offtopic, but (iirc) i once stumbled over the issue of "port" vs. "mgmt port" [back then i had access to two 2-port cards] where you may have said something along the lines of \"There is clearly a bug for multi HCAs in osm_vendor_get_all_port_attr which is in the vendor layer. This needs to be fixed and is our problem. So I am close to being able to commit what I now have for this and fix this later (as there are other multi HCA issues).\" Just curious.. did somebody already have a chance to touch those or not? On a related note (just the same thing, i tend to think) /* "local ports" is(?) phys, shouldn't this exclude port 0 then ? */ I'm not too familiar with these kindof questions, which might be define in a spec, so any hint on this would be well received, at least from my part. anyone? Eitan? -- thank you, Bernhard ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] [PATCH][dapl] cleanup dapl_cookie
Hi James, untested. - cleanup dapl_cookie.c: remove unneeded local variables and simplify branches to be consistent with dapl_rmr_cookie_alloc(). Signed-off-by: Bernhard Fischer <[EMAIL PROTECTED]> thank you, Bernhard Index: users/jlentini/linux-kernel/dat-provider/dapl_cookie.c === --- users/jlentini/linux-kernel/dat-provider/dapl_cookie.c (revision 2715) +++ users/jlentini/linux-kernel/dat-provider/dapl_cookie.c (working copy) @@ -1,6 +1,7 @@ /* * Copyright (c) 2002-2005, Network Appliance, Inc. All rights reserved. * Copyright (c) 2005 Sun Microsystems, Inc. All rights reserved. + * Copyright (c) 2005 Bernhard Fischer, All rights reserved. * * This Software is licensed under one of the following licenses: * @@ -136,8 +137,8 @@ u32 dapl_cb_create(struct dapl_cookie_bu } return DAT_SUCCESS; - } else - return DAT_INSUFFICIENT_RESOURCES; + } + return DAT_INSUFFICIENT_RESOURCES; } /* @@ -157,8 +158,7 @@ u32 dapl_cb_create(struct dapl_cookie_bu */ void dapl_cb_free(struct dapl_cookie_buffer *buffer) { - if (NULL != buffer->pool) - kfree(buffer->pool); + kfree(buffer->pool); } /* @@ -181,24 +181,19 @@ void dapl_cb_free(struct dapl_cookie_buf u32 dapl_cb_get(struct dapl_cookie_buffer *buffer, struct dapl_cookie **cookie_ptr) { - u32 dat_status; int new_head; BUG_ON(cookie_ptr == NULL); new_head = (atomic_read(&buffer->head) + 1) % buffer->pool_size; - if (new_head == atomic_read(&buffer->tail)) { - dat_status = DAT_INSUFFICIENT_RESOURCES; - goto bail; - } else { - atomic_set(&buffer->head, new_head); + if (new_head == atomic_read(&buffer->tail)) + return DAT_INSUFFICIENT_RESOURCES; - *cookie_ptr = &buffer->pool[atomic_read(&buffer->head)]; - dat_status = DAT_SUCCESS; - } -bail: - return dat_status; + atomic_set(&buffer->head, new_head); + + *cookie_ptr = &buffer->pool[atomic_read(&buffer->head)]; + return DAT_SUCCESS; } /* @@ -224,24 +219,19 @@ u32 dapl_rmr_cookie_alloc(struct dapl_co struct dapl_cookie **cookie_ptr) { struct dapl_cookie *cookie; - u32 dat_status; if (DAT_SUCCESS != dapl_cb_get(buffer, &cookie)) { *cookie_ptr = NULL; - dat_status = - DAT_ERROR(DAT_INSUFFICIENT_RESOURCES, DAT_RESOURCE_MEMORY); - goto bail; + return DAT_ERROR(DAT_INSUFFICIENT_RESOURCES, +DAT_RESOURCE_MEMORY); } - dat_status = DAT_SUCCESS; cookie->type = DAPL_COOKIE_TYPE_RMR; cookie->val.rmr.rmr = rmr; cookie->val.rmr.cookie = user_cookie; *cookie_ptr = cookie; - -bail: - return dat_status; + return DAT_SUCCESS; } /* ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] Re: [openib-commits] r2629 - in gen2/branches/shaharf-ibat/src/linux-kernel/infiniband: core include
Hal, On Wed, Jun 15, 2005 at 12:27:18PM -0700, [EMAIL PROTECTED] wrote: >Support callback for route by ip >Modified: gen2/branches/shaharf-ibat/src/linux-kernel/infiniband/core/uat.c >=== >--- gen2/branches/shaharf-ibat/src/linux-kernel/infiniband/core/uat.c >2005-06-15 18:22:01 UTC (rev 2628) >+++ gen2/branches/shaharf-ibat/src/linux-kernel/infiniband/core/uat.c >2005-06-15 19:27:17 UTC (rev 2629) >@@ -414,13 +495,30 @@ > printk(KERN_ERR "ib_uat_event: uevent %p ctx %p status > %d not completed\n", uevent, uevent->ctx, uevent->ctx->status); > uevent->ctx->rec_num = -EIO; > } >- /* Copy path records returned from SA back to userspace */ >- if (copy_to_user(uevent->ctx->user_path_arr, >- uevent->ctx->path_arr, >- uevent->ctx->user_length)) >- uevent->ctx->status = IB_USER_AT_STATUS_ERROR; >- kfree(uevent->ctx->path_arr); >- uevent->ctx->path_arr = NULL; >+ switch (uevent->type) { >+ case IB_UAT_PATH_EVENT: >+ /* Copy path records returned from SA to userspace */ >+ if (copy_to_user(uevent->ctx->user_path_arr, >+ uevent->ctx->path_arr, >+ uevent->ctx->user_length)) Should { result be set to FAULT here >+ uevent->ctx->status = IB_USER_AT_STATUS_ERROR; } >+ kfree(uevent->ctx->path_arr); >+ uevent->ctx->path_arr = NULL; >+ break; >+ case IB_UAT_ROUTE_EVENT: >+ if (copy_to_user(uevent->ctx->user_ib_route, >+ uevent->ctx->ib_route, >+ sizeof(*uevent->ctx->user_ib_route))) .. and here { >+ uevent->ctx->status = IB_USER_AT_STATUS_ERROR; } >+ kfree(uevent->ctx->ib_route); >+ uevent->ctx->ib_route = NULL; >+ break; >+ case IB_UAT_ATS_EVENT: >+ default: >+ printk(KERN_ERR "ib_uat_event: type %d not handled\n", >+ uevent->type); >+ break; >+ } > uevent->resp.callback = (u64)(unsigned > long)uevent->ctx->user_callback; > uevent->resp.context = (u64)(unsigned > long)uevent->ctx->user_context; > uevent->resp.req_id = uevent->ctx->req_id; > .. and take result into account before putting the response back to the user or is setting the status like you did above enough? Thanks for clarification. ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] Re: [PATCH] rdma_lat-09 and results
On Wed, Jun 22, 2005 at 09:28:56AM -0700, Grant Grundler wrote: >I also don't know how to express > if target == static then > MORECFLAGS = -rdynamic... > endif $ cat Makefile BAR:=-Os ifeq ($(MAKECMDGOALS),static) BAR += -rdynamic endif #default: all static: all all: @echo "making ${MAKECMDGOALS}: $(BAR)" hth, Bernhard ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] cycles_to_units is incorrect in rdma_lat, rdma_bw.
On Thu, Jun 09, 2005 at 04:40:10PM -0700, Roland Dreier wrote: >Shirley> I used cyclets_to_units() get Bandwidth peak (#0 to >Shirley> #972): 1852.72 MByte/sec > >Shirley> Which is far away from the 1X (2.5Gbit/s) throughtput. > >Actually it's not that far away -- the 2.5 Gbit/sec is after 8b/10b >coding, so the raw data rate is really 2 Gbit/sec. Getting > 92% link >utilization is pretty good considering packet header overhead and so >on. BTW I'm assuming "MByte/sec" is a typo for "Mbit/sec". I may be wrong and OT, but from what i got, nowadays, http://kerneltrap.org/node/340 ymmv, -- Bernhard ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] [mthca] debug strings
On Fri, Jun 10, 2005 at 09:34:10PM +0200, Bernhard Fischer wrote: >Hi, >- remove duplicate ': ' and prefix messages from sdp with ib_sdp > like ib_mthca does. hm. when loading mthca i see: ib_mthca: Mellanox InfiniBand HCA driver v0.06-pre (November 8, 2004) ib_mthca: Initializing Mellanox Technologies . ib_mthca :04:00.0: FW version 000400060002, max commands 64 ... which is a bit inconsistent. No functional obj-code changes. - peruse DRV_NAME - trim ':' from PFX to be consistent with dev_{info,warn,err} output. - whitespace in struct mthca_profile: replace spaces with tabs to be consistent with other member of that struct wrt spacing style. - mthca_dev_lim() remove tab (was that ment to be a reminder?) - add apparently missing newlines to stings emitted by dev_err. - add note whom to contact if old FW is found. In mthca_map_eq_regs(), i take it that it is of no interrest to know who of "..dev->fw.arbel.eq_"arm_base|eq_set_ci_base") + "(0|4) did trigger? Memfree vs. not would be obvious from previous msgs. Same in mthca_tune_pci for cap messages (which don't fit into width=80 terminals) find_mgm(): 'if (0)\n\t\tmthca_dbg(dev, "Hash for..' should not rely on dead code elimination but HEAVY_DEBUG. Haven't looked for other occurances of that, yet. Will you? thank you, -- Bernhard diff -X excl -rduNp gen2.2551.oorig/trunk/src/linux-kernel/infiniband/hw/mthca/mthca_dev.h gen2.2551/trunk/src/linux-kernel/infiniband/hw/mthca/mthca_dev.h --- gen2.2551.oorig/trunk/src/linux-kernel/infiniband/hw/mthca/mthca_dev.h 2005-05-27 11:06:10.0 +0200 +++ gen2.2551/trunk/src/linux-kernel/infiniband/hw/mthca/mthca_dev.h 2005-06-11 01:27:31.0 +0200 @@ -45,7 +45,7 @@ #include "mthca_doorbell.h" #define DRV_NAME "ib_mthca" -#define PFXDRV_NAME ": " +#define PFXDRV_NAME " " #define DRV_VERSION"0.06-pre" #define DRV_RELDATE"November 8, 2004" diff -X excl -rduNp gen2.2551.oorig/trunk/src/linux-kernel/infiniband/hw/mthca/mthca_main.c gen2.2551/trunk/src/linux-kernel/infiniband/hw/mthca/mthca_main.c --- gen2.2551.oorig/trunk/src/linux-kernel/infiniband/hw/mthca/mthca_main.c 2005-05-27 11:06:10.0 +0200 +++ gen2.2551/trunk/src/linux-kernel/infiniband/hw/mthca/mthca_main.c 2005-06-10 23:14:43.0 +0200 @@ -69,7 +69,7 @@ MODULE_PARM_DESC(msi, "attempt to use MS #endif /* CONFIG_PCI_MSI */ static const char mthca_version[] __devinitdata = - "ib_mthca: Mellanox InfiniBand HCA driver v" + DRV_NAME " Mellanox InfiniBand HCA driver v" DRV_VERSION " (" DRV_RELDATE ")\n"; static struct mthca_profile default_profile = { @@ -154,11 +154,11 @@ static int __devinit mthca_dev_lim(struc return -ENODEV; } - mdev->limits.num_ports = dev_lim->num_ports; + mdev->limits.num_ports = dev_lim->num_ports; mdev->limits.vl_cap = dev_lim->max_vl; mdev->limits.mtu_cap= dev_lim->max_mtu; - mdev->limits.gid_table_len = dev_lim->max_gids; - mdev->limits.pkey_table_len = dev_lim->max_pkeys; + mdev->limits.gid_table_len = dev_lim->max_gids; + mdev->limits.pkey_table_len = dev_lim->max_pkeys; mdev->limits.local_ca_ack_delay = dev_lim->local_ca_ack_delay; mdev->limits.max_sg = dev_lim->max_sg; mdev->limits.reserved_qps = dev_lim->reserved_qps; @@ -188,7 +188,7 @@ static int __devinit mthca_dev_lim(struc if (dev_lim->flags & DEV_LIM_FLAG_BAD_QKEY_CNTR) mdev->device_cap_flags |= IB_DEVICE_BAD_QKEY_CNTR; - + if (dev_lim->flags & DEV_LIM_FLAG_RAW_MULTI) mdev->device_cap_flags |= IB_DEVICE_RAW_MULTI; @@ -927,13 +927,13 @@ static int __devinit mthca_init_one(stru */ if (!(pci_resource_flags(pdev, 0) & IORESOURCE_MEM) || pci_resource_len(pdev, 0) != 1 << 20) { - dev_err(&pdev->dev, "Missing DCS, aborting."); + dev_err(&pdev->dev, "Missing DCS, aborting.\n"); err = -ENODEV; goto err_disable_pdev; } if (!(pci_resource_flags(pdev, 2) & IORESOURCE_MEM) || pci_resource_len(pdev, 2) != 1 << 23) { - dev_err(&pdev->dev, "Missing UAR, aborting."); + dev_err(&pdev->dev, "Missing UAR, aborting.\n"); err = -ENODEV; goto err_disable_pdev; } @@ -1032,6 +1032,7 @@ static int __devinit mthca_init_one(stru (int) (mthca_hca
Re: [openib-general] cycles_to_units is incorrect in rdma_lat, rdma_bw.
On Fri, Jun 10, 2005 at 02:59:52PM -0700, Grant Grundler wrote: >On Thu, Jun 09, 2005 at 05:33:52PM -0700, Grant Grundler wrote: >(Sorry - this is a complete diff including previous >changes that haven't been committed yet.) np, as long as you use at least 4*page_size and reflect this in usage() :) >- int size = 1; >+ int size = 4 * 1024; cheers, -- Bernhard ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] [osm] segfault in libibumad
On Fri, Jun 10, 2005 at 01:59:49PM -0700, Tom Duffy wrote: >On Fri, 2005-06-10 at 21:46 +0200, Bernhard Fischer wrote: >> Yes, but should it work? I don't know if its possible and implemented >> (in the long run) to get to and query the respective API versions in >> order not to rely on sysfs. > >Do you have a system without sysfs? Is there a reason why you would >not? For normal small systems, i do not see a requirements for sysfs. Artifical use case (i don't own HCAs nor euid==0 access to them): /me builds a well-connected firewall/router/whatever. I don't want to waste RAM/ROM for sysfs or sysfs support. I do, however, agree that normal nodes will have sysfs support enabled in the mid/short run. > >The behavior /could/ be that if it didn't find the sysfs entry, opensm >would continue on its merry way and just assume the API was correct. As Hal expressed, sysfs is a requirement for OpenIB; so be it. -- Bernhard ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] [osm] segfault in libibumad
On Fri, Jun 10, 2005 at 03:22:32PM -0400, Hal Rosenstock wrote: >On Fri, 2005-06-10 at 14:50, Bernhard Fischer wrote: >> I'm not sure if userspace is supposed to work without sysfs. What do >> you think? > >It shouldn't segv... Yes, but should it work? I don't know if its possible and implemented (in the long run) to get to and query the respective API versions in order not to rely on sysfs. >> In osm_vendor_init(), i'd set int r = -1, n_cas = -1; and would say >> else\nif ((n_cas = umad_get_cas_names. >> Also, in umad_get_cas_names() i guess only freeing namelist if >> scandir did not return <0 may be better.. > >Yes to both. >Can you try this patch ? Thanks. >- free(namelist); >+ if (n >= 0) >+ free(namelist); yes, should do. >- } >+ } else >+ r = n_cas = -1; yes, but why not just initialize r and n_cas to -1? thank you, -- Bernhard ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] [sdp] debug strings
Hi, Upon loading ib_sdp with debugging enabled, i see: INIT: : SDP module load. with attached patchlet, i see the (to me) more appealing string ib_sdp INIT: SDP module load. - remove duplicate ': ' and prefix messages from sdp with ib_sdp like ib_mthca does. - fix odd "INIT: : INIT: SDP module unload." debug message. thank you, diff -X excl -rduNp gen2.2551.oorig/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_inet.c gen2.2551/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_inet.c --- gen2.2551.oorig/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_inet.c 2005-05-17 07:36:11.0 + +++ gen2.2551/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_inet.c 2005-06-10 19:07:49.0 + @@ -79,7 +79,7 @@ MODULE_PARM_DESC(recv_post_max, "Set the receive completion queue size."); module_param(recv_buff_max, int, 0); -MODULE_PARM_DESC(recv_buff_max, +MODULE_PARM_DESC(recv_buff_max, "Set the maximum number of receives buffered."); module_param(send_post_max, int, 0); @@ -91,7 +91,7 @@ MODULE_PARM_DESC(send_buff_max, "Set the maximum number of sends buffered."); module_param(send_usig_max, int, 0); -MODULE_PARM_DESC(send_usig_max, +MODULE_PARM_DESC(send_usig_max, "Set the maximum consecutive unsignalled send events."); module_param(sdp_debug_level, int, 0); @@ -1571,7 +1571,7 @@ error_proc: */ static void __exit sdp_exit(void) { - sdp_dbg_init("INIT: SDP module unload."); + sdp_dbg_init("SDP module unload."); /* * unregister */ diff -X excl -rduNp gen2.2551.oorig/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_proto.h gen2.2551/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_proto.h --- gen2.2551.oorig/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_proto.h 2005-05-17 07:36:11.0 + +++ gen2.2551/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_proto.h 2005-06-10 16:26:10.0 + @@ -471,7 +471,7 @@ extern int sdp_debug_level; #define sdp_dbg_out(level, type, format, arg...) \ do { \ if (!(level > sdp_debug_level)) { \ -printk("<%d>%s: " format "\n", \ +printk("<%d>ib_sdp %s: " format "\n", \ level, type, ## arg); \ } \ } while (0) @@ -516,7 +516,7 @@ extern int sdp_debug_level; #define sdp_dbg_init(format, arg...) do { } while (0) #else #define sdp_dbg_init(format, arg...) \ -sdp_dbg_out(__SDP_DEBUG_INIT, "INIT: ", format, ## arg) +sdp_dbg_out(__SDP_DEBUG_INIT, "INIT", format, ## arg) #endif #if __SDP_DEBUG_LEVEL < __SDP_DEBUG_WARN ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] [osm] segfault in libibumad
Hi, I have no ib drivers loaded, starting opensm fails with a segfault. I'm not sure if userspace is supposed to work without sysfs. What do you think? In osm_vendor_init(), i'd set int r = -1, n_cas = -1; and would say else\nif ((n_cas = umad_get_cas_names. Also, in umad_get_cas_names() i guess only freeing namelist if scandir did not return <0 may be better.. (gdb) run Starting program: /opt/infiniband/ib/bin/opensm - OpenSM Rev:openib-1.0.0 Command Line Arguments: Log File: /var/log/osm.log - warn: [5900] umad_init: can't read ABI version from /sys/class/infiniband_mad/abi_version (No such file or directory): is ib_umad module loaded? Program received signal SIGSEGV, Segmentation fault. 0x400b8735 in free () from /lib/tls/libc.so.6 (gdb) bt #0 0x400b8735 in free () from /lib/tls/libc.so.6 #1 0x4002235a in umad_get_cas_names (cas=0x80acc1c, max=32) at umad.c:513 #2 0x4001afe1 in osm_vendor_init (p_vend=0x80acb88, p_log=0x80a979c, timeout=100) at osm_vendor_ibumad.c:418 #3 0x4001b0ba in osm_vendor_new (p_log=0x80a979c, timeout=100) at osm_vendor_ibumad.c:452 #4 0x0805d2e4 in osm_opensm_init (p_osm=0x80a8520, p_opt=0xbfb5dc40) at osm_opensm.c:234 #5 0x0804ca19 in main (argc=1, argv=0xbfb5de14) at main.c:632 -- Bernhard ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] Re: [openib-commits] r2480 - gen2/trunk/src/userspace/perftest
On Wed, Jun 08, 2005 at 04:46:37PM -0700, Grant Grundler wrote: >On Wed, Jun 08, 2005 at 04:40:57PM -0700, Grant Grundler wrote: >> Index: rdma_bw.c >> === >> --- rdma_bw.c(revision 2570) >> +++ rdma_bw.c(working copy) >> @@ -61,10 +61,6 @@ >... > >Sorry - just realized I forgot the: >Signed-off-by: Grant Grundler <[EMAIL PROTECTED]> >> -int size = 1; >> +int size = 4 * 1024; Please reflect this change in the help text too. I'd make that at least 4*page_size, but 4k is certainly better than 1 :) ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] [PATCHv4] kdapl: remove use of HANDLE's (vs. r2572)
On Wed, Jun 08, 2005 at 01:52:47PM -0700, Tom Duffy wrote: >On Wed, 2005-06-08 at 22:41 +0200, Bernhard Fischer wrote: >> On Wed, Jun 08, 2005 at 01:19:41PM -0700, Tom Duffy wrote: >> >On Wed, 2005-06-08 at 21:48 +0200, Bernhard Fischer wrote: >> >> >> 'if (NULL != this)' and 'if (0 != that)' occurances, too? I personally >> >> would appreciate it, since these variants are harder to read for me. >> > >> >What do you want to see instead? >> > >> >if (this) { >> >/* do stuff */ >> >} >> >> yes, this. > >I think James would take these types of patches. > >In any event, going through the code with fine toothed comb will happen >after all the major upheavals, I think. *I* want to fix all the major >architectural issues that are holding up acceptance before nit-picking. certainly. > >If I have the time or inclination, I will clean up these things. Index: TODO === --- TODO(revision 2572) +++ TODO(working copy) @@ -23,3 +23,7 @@ IB DAT PROVIDER *) cache remote peer's address so that it does not need to be retrieved in dapl_ep_query. *) Replace dapl linked lists with Linux linked lists +*) referencing + http://openib.org/pipermail/openib-general/2005-January/003149.html + and only then replace the checks turned up by + 'grep -r -E "NULL.*!=|!=.*NULL" *' with if(foo)... for better readability ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] [PATCHv4] kdapl: remove use of HANDLE's (vs. r2572)
On Wed, Jun 08, 2005 at 01:19:41PM -0700, Tom Duffy wrote: >On Wed, 2005-06-08 at 21:48 +0200, Bernhard Fischer wrote: >> 'if (NULL != this)' and 'if (0 != that)' occurances, too? I personally >> would appreciate it, since these variants are harder to read for me. > >What do you want to see instead? > >if (this) { > /* do stuff */ >} yes, this. ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] openib-commits administrative
Hi, May i as ask if diffs sent to the -commits list may be changed to be generated with the --show-c-fun option to diff? Something along the lines of svn diff --diff-cmd diff -x -urNp .. works for me. Knowing the function the change was based against simply makes it easier to follow. PS: yes, OT. But as #openib is completely orphaned (which is a pity, one would expect that folks at least once a week would check the logs they gather from reading there, but well) i know no other addr i could send that plea too. Sorry to the not concerned majority who i had to mux. PPS: having a cia-bot in #openib would be great, too, but i have no illusion that anyone would be willing to setup this. Prove me wrong. thank you, ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] [PATCHv4] kdapl: remove use of HANDLE's (vs. r2572)
On Wed, Jun 08, 2005 at 11:10:03AM -0700, Sean Hefty wrote: >Tom Duffy wrote: >>On Wed, 2005-06-08 at 12:35 -0400, James Lentini wrote: >> >>>Committed in revision 2572. >> >>Updated patch against 2572. >> >>Signed-off-by: Tom Duffy <[EMAIL PROTECTED]> > >James, is there a specific hold-up for this patch? > Tom, just curious.. Do you plan to iterate over those disturbing 'if (NULL != this)' and 'if (0 != that)' occurances, too? I personally would appreciate it, since these variants are harder to read for me. But then i also tend to prefer '* 4711e-0815' over '/ 0x10' ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] [PATCH] mthca, ibverbs userspace readability
On Thu, May 26, 2005 at 09:29:17AM -0700, Roland Dreier wrote: >Bernhard> - improve readability of mthca and ibverbs headers >Bernhard> installed to userspace. > >This patch seems to consist mostly of replacing tabs with 8 spaces. >Why is that an improvement for readability? It does indeed. But with the patch, it's easier to read and code for. Also, there are a couple of fixes for super-long lines, which are extremely disturbing if you happen to code userspace while referencing those headers :-/ ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] [PATCH] mthca, ibverbs userspace readability
On Thu, May 26, 2005 at 09:29:17AM -0700, Roland Dreier wrote: >Bernhard> - improve readability of mthca and ibverbs headers >Bernhard> installed to userspace. > >This patch seems to consist mostly of replacing tabs with 8 spaces. >Why is that an improvement for readability? care to read one of those headers with ts=2 and code for userspace? without that patch, it simply is unreadable. Still, as noted before, just take it or leave it (fear you'll decide to go for the former :/ but i will continue to propagate the latter, fwiw.). > >Thanks, > Roland > ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] OpenIB and OpenRDMA: Convergence on common RDMAAPIs and ULPs for Linux
On Thu, May 26, 2005 at 09:25:27AM -0700, Caitlin Bestler wrote: >The current OpenIB structure is, not surprisingly, tremendously >IB-centric. Any attempt to implement an iWARP RNIC through >them will have to go through several layers of obfuscation. > >States do not match. Error codes do not match. Completion >Statuses do not match. There is a lot of information that Statuses? stati >does not translate to iWARP, and there is information that >is missing. such as.. ? > >RNIC-PI is at least an attempt at providing full control over >both iWARP and IB while making as much common as >possible. Please point me to any infrastructure which allows any private peer to steek at RNIC. At least at peer-level with a working driver which includes a working ethtool -C infrastructure.. > >I believe it is possible to define a verb set that is largely >compatible with current OpenIB IB verbs and have the >same transport neutrality that RNIC-PI has. But this will >not be just poking around and tweaking one or two >elements in the OpenIB code. [What is RNIC anyway? dict rnic returns just NIL, so what? Don't mean to sound rude, but please provide at least a _little_ background for newcomers, will you?] > >Is there a real willingness in OpenIB to make a transport >neutral verb layer? Has anyone even *read* the RNIC-PI Not being involved (heared of \"community\" before?), still yes. Care to point (once again, as i'm buzzword agnostic) to the relevant IEEE/whatever RFCs? TIA. >for substance rather than "pure Linux" debating points? >___ >openib-general mailing list >openib-general@openib.org >http://openib.org/mailman/listinfo/openib-general > >To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general > ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] [PATCH] mthca, ibverbs userspace readability
Hi, No object code changes. linux-kernel/infiniband/ulp/sdp/sdp_conn.c |1 userspace/libibverbs/include/infiniband/driver.h | 34 +- userspace/libibverbs/include/infiniband/kern-abi.h |2 userspace/libibverbs/include/infiniband/verbs.h| 253 ++--- userspace/libmthca/src/cq.c|9 userspace/libmthca/src/mthca.h | 87 +++ 6 files changed, 198 insertions(+), 188 deletions(-) - improve readability of mthca and ibverbs headers installed to userspace. Signed-off-by Bernhard Fischer <[EMAIL PROTECTED]> Those were hard/impossible to read for ts!=8 before. Please at least consider; I'm aware that those are purely cosmetic, granted. TIA for applying nonetheless. diff -X excl -rduNp gen2.2491.oorig/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_conn.c gen2.2491/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_conn.c --- gen2.2491.oorig/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_conn.c 2005-05-16 22:09:24.0 +0200 +++ gen2.2491/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_conn.c 2005-05-26 16:29:31.197039104 +0200 @@ -842,7 +842,6 @@ void sdp_conn_relock(struct sdp_opt *con sdp_dbg_warn(conn, "Error <%d> rearming send CQ", result); - rearm = 0; } else break; /* exit CQ handler routine */ diff -X excl -rduNp gen2.2491.oorig/trunk/src/userspace/libibverbs/include/infiniband/driver.h gen2.2491/trunk/src/userspace/libibverbs/include/infiniband/driver.h --- gen2.2491.oorig/trunk/src/userspace/libibverbs/include/infiniband/driver.h 2005-05-16 22:09:23.0 +0200 +++ gen2.2491/trunk/src/userspace/libibverbs/include/infiniband/driver.h 2005-05-26 16:29:31.200038648 +0200 @@ -60,35 +60,37 @@ * ops member filled in. If the driver does not support the device, * it should return NULL from openib_driver_init(). */ - + typedef struct ibv_device *(*ibv_driver_init_func)(struct sysfs_class_device *); extern int ibv_cmd_get_context(int num_comp, struct ibv_context *context, - struct ibv_get_context *cmd, size_t cmd_size); + struct ibv_get_context *cmd, size_t cmd_size); extern int ibv_cmd_query_port(struct ibv_context *context, uint8_t port_num, - struct ibv_port_attr *port_attr, - struct ibv_query_port *cmd, size_t cmd_size); + struct ibv_port_attr *port_attr, + struct ibv_query_port *cmd, size_t cmd_size); extern int ibv_cmd_alloc_pd(struct ibv_context *context, struct ibv_pd *pd, - struct ibv_alloc_pd *cmd, size_t cmd_size); +struct ibv_alloc_pd *cmd, size_t cmd_size); extern int ibv_cmd_dealloc_pd(struct ibv_pd *pd); extern int ibv_cmd_reg_mr(struct ibv_pd *pd, void *addr, size_t length, - uint64_t hca_va, enum ibv_access_flags access, - struct ibv_mr *mr, struct ibv_reg_mr *cmd, - size_t cmd_size); + uint64_t hca_va, enum ibv_access_flags access, + struct ibv_mr *mr, struct ibv_reg_mr *cmd, + size_t cmd_size); extern int ibv_cmd_dereg_mr(struct ibv_mr *mr); extern int ibv_cmd_create_cq(struct ibv_context *context, int cqe, -struct ibv_cq *cq, -struct ibv_create_cq *cmd, size_t cmd_size); + struct ibv_cq *cq, + struct ibv_create_cq *cmd, size_t cmd_size); extern int ibv_cmd_destroy_cq(struct ibv_cq *cq); extern int ibv_cmd_create_qp(struct ibv_pd *pd, -struct ibv_qp *qp, struct ibv_qp_init_attr *attr, -struct ibv_create_qp *cmd, size_t cmd_size); + struct ibv_qp *qp, struct ibv_qp_init_attr *attr, + struct ibv_create_qp *cmd, size_t cmd_size); extern int ibv_cmd_modify_qp(struct ibv_qp *qp, struct ibv_qp_attr *attr, -enum ibv_qp_attr_mask attr_mask, -struct ibv_modify_qp *cmd, size_t cmd_size); + enum ibv_qp_attr_mask attr_mask, + struct ibv_modify_qp *cmd, size_t cmd_size); extern int ibv_cmd_destroy_qp(struct ibv_qp *qp); -extern int ibv_cmd_attach_mcast(struct ibv_qp *qp, union ibv_gid *gid, uint16_t lid); -extern int ibv_cmd_detach_mcast(struct ibv_qp *qp, union ibv_gid *gid, uint16_t lid); +extern int ibv_cmd_attach_mcast(struct ibv_qp *qp, union ibv_gid *gid, +uint16_t lid); +extern int ibv_cmd
[openib-general] [PATCH] mthca: cosmetic fix for handle_error_cqe()
Hi, - remove unneeded if-condition in mthca's cq error path Signed-off-by: Bernhard Fischer <[EMAIL PROTECTED]> diff -X excl -rduNp gen2.2423.oorig/trunk/src/linux-kernel/infiniband/hw/mthca/mthca_cq.c gen2.2423/trunk/src/linux-kernel/infiniband/hw/mthca/mthca_cq.c --- gen2.2423.oorig/trunk/src/linux-kernel/infiniband/hw/mthca/mthca_cq.c 2005-04-08 10:22:08.0 +0200 +++ gen2.2423/trunk/src/linux-kernel/infiniband/hw/mthca/mthca_cq.c 2005-05-23 14:26:08.0 +0200 @@ -280,7 +280,7 @@ static int handle_error_cqe(struct mthca int dbd; u32 new_wqe; - if (1 && cqe->syndrome != SYNDROME_WR_FLUSH_ERR) { + if (cqe->syndrome != SYNDROME_WR_FLUSH_ERR) { int j; mthca_dbg(dev, "%x/%d: error CQE -> QPN %06x, WQE @ %08x\n", ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] Re: [openib-commits] r2426 - gen2/users/jlentini/linux-kernel/dat-provider
On Mon, May 23, 2005 at 10:32:12AM -0400, James Lentini wrote: > >Bernhard, > >Thank you for pointing these items out. Comments below: > >On Sat, 21 May 2005, Bernhard Fischer wrote: > >>On Fri, May 20, 2005 at 03:05:08PM -0700, [EMAIL PROTECTED] wrote: >>>Author: jlentini >>>Date: 2005-05-20 15:05:07 -0700 (Fri, 20 May 2005) >>>New Revision: 2426 >>>-if (DAPL_GET_CQE_OPTYPE(cqe_ptr) == OP_RECEIVE) { >>>+ "\t\t work_req_id %lli\n", cqe_ptr->wr_id); >>[[:space:]] --^ ? > >Which space are you referring to? The one between \t\t and >work_req_id? Yes, i ment the one between \t\t and work_req_id. I now see that those are part of the formatting, so i retract this comment. Sorry. >>>@@ -667,21 +657,21 @@ >>> >>>#ifdef DAPL_DBG >>> /* Current gen2 mthca is not setting the opcode in >>> seccesful cqe */ s/secces/succes/g >>>-/* The opcode will be OP_SEND or OP_RECEIVE acording >>>the is_send bit */ >>>+/* The opcode will be IB_WC_SEND or IB_WC_RECV >>>acording the is_send bit */ >>"acording" -enoparse: s/acord/accord/g >> >>also in: >>gen2/branches/shaharf-ibat/src/userspace/management/osm/opensm/osm_state_mgr.c >>gen2/branches/roland-uverbs/src/userspace/management/osm/opensm/osm_sa_mcmember_record.c >>gen2/utils/src/linux-user/ibdm/datamodel/ibdm.i >>gen2/utils/src/linux-user/IBMgtSim/src/ibdm.i >>gen2/utils/src/linux-user/IBMgtSim/utils/RunSimTest >>there: callabcks: -enoparse; >>continue with acord: >>gen2/users/jlentini/linux-kernel/dat-provider/dapl_evd_util.c >>gen2/trunk/src/userspace/management/osm/opensm/osm_state_mgr.c > >I've fixed my error: > >gen2/users/jlentini/linux-kernel/dat-provider/dapl_evd_util.c Ok, thank you. > >The others belong to Roland, Shahar, and Eitan. >>>-{ >>>-struct dat_ep_attr ep_attr; >>>-struct dat_named_attr ep_state; >>>+(void) dapl_modify_qp_state_to_error(ep_ptr->qp_handle); >> >>sun will hopefuly fix all of >>egrep -ri "(^[[:space:]]*\(void\))" >>openib.gen2/upstream/gen2/trunk/|egrep -v "svn-(text|base)" >>so i won't comment on that single occurance above ;) > >I didn't realize the convention was to not have spaces before a cast. >This isn't in Documentation/CodingStyle. I know I've been adding these >here and there. I'll fix these as I see them. I ment that Tom Duffy and you will take care of converting a couple of those into void functions, so the casts will no longer be needed. It's not worth the effort to remove the space after the cast. In contrast, personally, i find that the space makes to code easier to read. cheers, Bernhard ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] Re: [openib-commits] r2426 - gen2/users/jlentini/linux-kernel/dat-provider
On Fri, May 20, 2005 at 03:05:08PM -0700, [EMAIL PROTECTED] wrote: >Author: jlentini >Date: 2005-05-20 15:05:07 -0700 (Fri, 20 May 2005) >New Revision: 2426 Just tersely parsing, no mind to delve into any API implications whatsoever. [Late, no motivation to think, nor even read, not even speaking of writing anything publically. I'm over quota with my telephone-IP provider since 11th, which nag's our scanty pocket-money crucially, etc] >Modified: gen2/users/jlentini/linux-kernel/dat-provider/dapl.h >=== >--- gen2/users/jlentini/linux-kernel/dat-provider/dapl.h 2005-05-20 >17:22:17 UTC (rev 2425) >+++ gen2/users/jlentini/linux-kernel/dat-provider/dapl.h 2005-05-20 >22:05:07 UTC (rev 2426) >@@ -144,6 +140,24 @@ > * * > */ > >+typedef void (*ib_async_handler_t) (struct ib_event *, void *); >+ >+/* >+ * ib_hca_transport_t structure. Define fields specific to this >+ * provider implementation necessary to maintain the HCA >+ * >+ * NOTE: null_ib_hca_handle is strictly internal to IB. DAPL allows a >+ * NULL DTO EVD handle, but IB does not. Create a CQ under the >+ * hood and make sure an error is generated if the user every "every tries to post"? -enoparse; please be more specific -- simpler. 'ever' would still be obfuscated. IOW please rephrase and put abbreviations into the wiki. >+ * tries to post, by setting the WQ length to 0 in ep_create >+ * and/or ep_modify. >+ */ >@@ -394,7 +403,7 @@ > > /* maintenence fields */ maintenance. also in: users/jlentini/userspace/dapl/include/dapl.h > boolean_t listening;/* PSP is registered & active */ >- ib_cm_srvc_handle_t cm_srvc_handle; /* Used by CM */ >+ struct ib_cm_id *cm_srvc_handle;/* Used by CM */ > DAPL_LLIST_HEAD cr_list_head; /* CR pending queue */ > int cr_list_count; /* count of CRs on queue */ > }; >--- gen2/users/jlentini/linux-kernel/dat-provider/dapl_evd_util.c >2005-05-20 17:22:17 UTC (rev 2425) >+++ gen2/users/jlentini/linux-kernel/dat-provider/dapl_evd_util.c >2005-05-20 22:05:07 UTC (rev 2426) >@@ -331,22 +322,22 @@ > dapl_dbg_log(DAPL_DBG_TYPE_CALLBACK, >"\t dapl_evd_dto_callback : CQE \n"); > dapl_dbg_log(DAPL_DBG_TYPE_CALLBACK, >- "\t\t work_req_id %lli\n", DAPL_GET_CQE_WRID(cqe_ptr)); >- if (DAPL_GET_CQE_STATUS(cqe_ptr) == 0) { >- if (DAPL_GET_CQE_OPTYPE(cqe_ptr) == OP_RECEIVE) { >+ "\t\t work_req_id %lli\n", cqe_ptr->wr_id); [[:space:]] --^ ? >+ if (cqe_ptr->status == 0) { >+ if (cqe_ptr->opcode == IB_WC_RECV) { > dapl_dbg_log(DAPL_DBG_TYPE_CALLBACK, >- "\t\t op_type: OP_RECEIVE\n"); >+ "\t\t op_type: IB_WC_RECV\n"); ditto > dapl_dbg_log(DAPL_DBG_TYPE_CALLBACK, >"\t\t bytes_num %d\n", ditto >- DAPL_GET_CQE_BYTESNUM(cqe_ptr)); >+ cqe_ptr->byte_len); > } else { > dapl_dbg_log(DAPL_DBG_TYPE_CALLBACK, >"\t\t op_type: %s\n", ditto >- optable[DAPL_GET_CQE_OPTYPE(cqe_ptr)]); >+ optable[cqe_ptr->opcode]); > } > } > dapl_dbg_log(DAPL_DBG_TYPE_CALLBACK, >- "\t\t status %d\n", DAPL_GET_CQE_STATUS(cqe_ptr)); >+ "\t\t status %d\n", cqe_ptr->status); ditto > dapl_dbg_log(DAPL_DBG_TYPE_CALLBACK, >"\t >>><<<\n"); fooobar! merge errors? also in: gen2/branches/shaharf-ibat/src/userspace/management/libibmad/src/rpc.c gen2/ulps/iser/make.conf gen2/users/jlentini/linux-kernel/dat-provider/dapl_evd_util.c (multiple times) gen2/users/jlentini/linux-kernel/dat-provider/dapl_rsp_free.c (multiple gen2/users/jlentini/linux-kernel/dat-provider/dapl_rsp_create.c times) gen2/users/jlentini/linux-kernel/dat-provider/dapl_openib_cm.c (multiple times) gen2/users/jlentini/userspace/dapl/common/dapl_evd_util.c (ditto) gen2/users/jlentini/userspace/dapl/common/dapl_rsp_free.c (ditto) gen2/users/jlentini/userspace/dapl/common/dapl_rsp_create.c (ditto) gen2/trunk/src/userspace/management/libibmad/src/rpc.c > #endif >@@ -667,21 +657,21 @@ > > #ifdef DAPL_DBG > /* Current gen2 mthca is not setting the opcode in > seccesful cqe */ >- /* The opcode will be OP_SEND or OP_RECEIVE acording >the is_send bit */ >+ /* The opcode will be IB_WC_SEND or IB_WC_RECV acording >the is_send bit */ "acordi
[openib-general] [PATCH] auto*() && ./configure -!ac_default_prefix
Hi, [bt\nGPL throughout, on my side,.. thusly] * trunk/src/userspace/lib{mthca,ibcm}/configure.in: if prefix != ac_default_prefix use prefix for checks. fixup for e.g. SuSE when giving a non standart --prefix; Should be fixed upstream, non-debian will need a while to pick those up, though; therefore, please consider.. :-/ PS: the ../ib/. defaults are not covered by this but w{ould,ill} need seperate, explicit treatement. Not happy with this, but crucial reality, it seems. PPS: anyone ever tested to install chain into --prefix=/scratch/infiniband after a rm -rf /usr/local/* ? Drats. diff -X excl -rduNp gen2.2423.oorig/trunk/src/userspace/libibcm/configure.in gen2.2423/trunk/src/userspace/libibcm/configure.in --- gen2.2423.oorig/trunk/src/userspace/libibcm/configure.in2005-05-04 12:20:03.0 + +++ gen2.2423/trunk/src/userspace/libibcm/configure.in 2005-05-20 10:18:12.0 + @@ -12,6 +12,14 @@ AM_PROG_LIBTOOL dnl Checks for programs AC_PROG_CC +dnl if prefix != ac_default_prefix use prefix for checks +dnl fixup for e.g. SuSE when giving a non standard --prefix +if test "x$prefix" != "xNONE" -a "$prefix" != "$ac_default_prefix" +then + CPPFLAGS="$CPPFLAGS -I$prefix/include" + LDFLAGS="$LDFLAGS -L$prefix/lib" +fi + dnl Checks for libraries dnl Checks for header files. diff -X excl -rduNp gen2.2423.oorig/trunk/src/userspace/libmthca/configure.in gen2.2423/trunk/src/userspace/libmthca/configure.in --- gen2.2423.oorig/trunk/src/userspace/libmthca/configure.in 2005-05-04 12:20:05.0 + +++ gen2.2423/trunk/src/userspace/libmthca/configure.in 2005-05-20 10:12:57.0 + @@ -11,6 +11,14 @@ AM_PROG_LIBTOOL dnl Checks for programs AC_PROG_CC +dnl if prefix != ac_default_prefix use prefix for checks +dnl fixup for e.g. SuSE when giving a non standard --prefix +if test "x$prefix" != "xNONE" -a "$prefix" != "$ac_default_prefix" +then + CPPFLAGS="$CPPFLAGS -I$prefix/include" + LDFLAGS="$LDFLAGS -L$prefix/lib" +fi + dnl Checks for libraries dnl Checks for header files. ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] in need of a simple ulp
On Fri, May 20, 2005 at 01:40:42PM -0700, Sean Hefty wrote: >Jeff Carr wrote: >>There must also be some limit to how many cqe's you can allocate with >>ib_post_recv(). (?) > >Each receive posted will generate a cqe. That limit is controlled by the >message_count variable. > >>There must be some way to delete/free these? They don't get re-used I >>take it? Surely it wasn't intended that ib_post_recv() be initially run >>for each transfer expected in the lifetime of the connection. :) > >There needs to be one cqe reserved for each possible completion that can be >outstanding at one time. If you know that the number of completions will >be high initially, and then drop to a smaller value, there is a verb that >can be used to resize the CQ. (I don't believe it is implemented in the >openib stack at this time.) > >>There must also be some information about what is known about these >>cqe's. How do we know if one of them was used for a transfer from the >>server to the client or from the client letting the server know the >>transfer was recieved? > >In this case you won't know what's on a cqe until you poll the completion. >An alternative way of structuring the code is to create two CQs, and tie >one to the send queue, the other to the receive queue. hmz, yes, see "Quick unrelated question:" in http://openib.org/pipermail/openib-general/2005-May/006796.html I do have ...s_cq and r_cq. For two nodes, using cqe=64 does obviously work, but when i've got, say, a setup where 786 nodes each have r and s? This currently w{on't,ill in no way way} work (for me)!? Help!! Also, Sean, any thought on the qp limit question referenced there? TIA ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] perftest on i386
Hi, On i386 UP, i see odd results during runs. The machine is basically idle. The worst latency is always somewhere around exchange 12-80. Is this expected and if not, does somebody have an idea what may cause this? Thank you, $ ./rdma_lat -i 2 -U -s 1 tst09 -t 1 | egrep -v ", 3.(1|2|3)" local address: LID 0x02, QPN 0x1a0406, PSN 0x5379e1 RKey 0x1040436 VAddr 0x0008051001 remote address: LID 0x01, QPN 0x1d0406, PSN 0xd7aed6, RKey 0x188043c VAddr 0x0008051001 #, usec 1, 12.8368 2, 4.94677 12, 7.0644 13, 87.8933 26, 64.8241 109, 7.10019 166, 5.88006 193, 3.47097 250, 11.6861 305, 11.9567 388, 6.57047 445, 5.39661 529, 6.8158 536, 3.46172 586, 5.7806 671, 4.3199 728, 5.78413 813, 4.36494 870, 38.4934 902, 3.40266 944, 6.93949 Size: 1, TX-depth: 1, Exchanges: 1000 Latency typical: 3.26619 usec Latency best : 3.17244 usec Latency worst : 87.8933 usec ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
diff-perftest-08 [was: Re: [openib-general] Re: Re: diff-perftest-07 replace pp_get_local_lid()]
On Wed, May 18, 2005 at 08:56:31PM +0200, Bernhard Fischer wrote: >On Wed, May 18, 2005 at 09:44:36PM +0300, Michael S. Tsirkin wrote: >>Quoting r. Bernhard Fischer <[EMAIL PROTECTED]>: >>> Subject: Re: Re: diff-perftest-07 replace pp_get_local_lid() >>Size 0 doesnt make sense for this test since I'm polling on data: >>no data = nothing to poll upon. > >If size 0 and -1 (not 1 as you seem to have read) are illegal (they >obviously are for malloc) you should check if(size <1){usage();exit(1);} >Same for the other arguments. I'll send a patchlet if noone did fix >that by next week. > >>-s 1 should work - I'll check tomorrow. Looking at the code, 1 is the default and not 4096 as advertised (see attached patch to reflect default of 1 in usage). o Change usage() to match the actual default of size. o sparse warning amout non-ANSI function decl of get_cpu_mhz(), whitespace in printf while at it. o Add printing Size, TX-depth and Iters to print_report(). Without a short description of the workload, it is bit hard to interpret the results if one cannot see the command-line used to run the tests. o Check arguments -s and -t o Trim some whitespace o Remove odd 'write(sockfd,"foo"...' (What was that good for?) Signed-off-by: Bernhard Fischer <[EMAIL PROTECTED]> > >-s 1 did work for me last time i looked. >In fact, back then only 1 worked, i got errors for bigger values i tried >(42,128,256,512,1k,2k,4k,16k,32k,64k,128k,42M etc). The problem seems to be that i (wrongly) believed in the default size of 4k, so assumed i could run the server without args and pass -s 2 to the client. This is a wrong assertion as size really defaults to 1. I can specify -s 940 as the maximum, using 941 gives me "cannot create QP" in rdma_lat.c. Is this expected? What are the contraints and limits for size here? Quick unrelated question: Is there any limit on the number of cqe's i can pass to mthca_create_cq? I initially tried 40960 or 4096 but that did not work. I'm now using 64 but would like to use more than those in the future.. diff -X excl -rduNp gen2.2423.oorig/trunk/src/userspace/perftest/get_clock.c gen2.2423/trunk/src/userspace/perftest/get_clock.c --- gen2.2423.oorig/trunk/src/userspace/perftest/get_clock.c2005-05-17 14:38:30.0 + +++ gen2.2423/trunk/src/userspace/perftest/get_clock.c 2005-05-20 11:47:06.0 + @@ -35,7 +35,7 @@ #include #include -double get_cpu_mhz() +double get_cpu_mhz(void) { FILE* f; char buf[256]; @@ -55,7 +55,7 @@ double get_cpu_mhz() continue; } if (mhz != m) { - fprintf(stderr,"Conflicting CPU frequency values " + fprintf(stderr,"Conflicting CPU frequency values" " detected: %lf != %lf\n", mhz, m); return 0.0; diff -X excl -rduNp gen2.2423.oorig/trunk/src/userspace/perftest/rdma_lat.c gen2.2423/trunk/src/userspace/perftest/rdma_lat.c --- gen2.2423.oorig/trunk/src/userspace/perftest/rdma_lat.c 2005-05-20 07:06:11.0 + +++ gen2.2423/trunk/src/userspace/perftest/rdma_lat.c 2005-05-20 12:43:51.0 + @@ -434,7 +434,7 @@ static void usage(const char *argv0) printf(" -p, --port= listen on/connect to port (default 18515)\n"); printf(" -d, --ib-dev= use IB device (default first device found)\n"); printf(" -i, --ib-port= use port of IB device (default 1)\n"); - printf(" -s, --size= size of message to exchange (default 4096)\n"); + printf(" -s, --size= size of message to exchange (default 1)\n"); printf(" -t, --tx-depth= size of tx queue (default 50)\n"); printf(" -n, --iters=number of exchanges (at least 2, default 1000)\n"); printf(" -C, --report-cyclesreport times in cpu cycle units (default microseconds)\n"); @@ -468,7 +468,8 @@ static int cycles_compare(const void * a } static void print_report(struct report_options * options, -unsigned int iters, cycles_t *tstamp) +unsigned int iters, cycles_t *tstamp, +struct pingpong_context *ctx) { double cycles_to_units; cycles_t median; @@ -507,6 +508,9 @@ static void print_report(struct report_o printf("%d, %g\n", i + 1, delta[i] / cycles_to_units / 2); } + printf("Size: %d, TX-depth: %d, Exchanges: %d\n", + ctx->size, ctx->tx_depth, iters); + median = get_median(iters, delta); printf("Latency typical: %g %s\n", median /
[openib-general] Re: Re: diff-perftest-07 replace pp_get_local_lid()
On Wed, May 18, 2005 at 09:44:36PM +0300, Michael S. Tsirkin wrote: >Quoting r. Bernhard Fischer <[EMAIL PROTECTED]>: >> Subject: Re: Re: diff-perftest-07 replace pp_get_local_lid() >> >> On Wed, May 18, 2005 at 11:06:20AM -0700, Grant Grundler wrote: >> >hrm...that would work too though it's not as intuitive. >> >> Would only work partly.. > >Sorry, you need > > ssh server $PWD/rdma_lat & ( sleep 1 && $PWD/rdma_lat server) > >> try: >> ssh server $PWD/rdma_lat & ( sleep 1 && $PWD/rdma_lat -s 0 or -1) >> and your server process will just hang. >> I know.. don't do that then (checking if size<1 may just paper over >> missing checks in verbs. i don't know). >> > >Size 0 doesnt make sense for this test since I'm polling on data: >no data = nothing to poll upon. If size 0 and -1 (not 1 as you seem to have read) are illegal (they obviously are for malloc) you should check if(size <1){usage();exit(1);} Same for the other arguments. I'll send a patchlet if noone did fix that by next week. >-s 1 should work - I'll check tomorrow. -s 1 did work for me last time i looked. In fact, back then only 1 worked, i got errors for bigger values i tried (42,128,256,512,1k,2k,4k,16k,32k,64k,128k,42M etc). ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] Re: [openib-commits] r2384 - gen2/trunk/src/userspace/management/libibmad/src
On Wed, May 18, 2005 at 11:14:45AM -0700, [EMAIL PROTECTED] wrote: >Author: halr >Date: 2005-05-18 11:14:44 -0700 (Wed, 18 May 2005) >New Revision: 2384 > >Modified: > gen2/trunk/src/userspace/management/libibmad/src/sa.c Hal, I think this typo does also occur in gen2/trunk/src/userspace/management/osm/include/iba/ib_types.h Thank you, >Log: >Fix typo in #define >-#define IB_PR_COMPMASK_PFTLIFETIME (1ull<<21) >+#define IB_PR_COMPMASK_PKTLIFETIME (1ull<<21) ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] Re: diff-perftest-07 replace pp_get_local_lid()
On Wed, May 18, 2005 at 11:06:20AM -0700, Grant Grundler wrote: >On Wed, May 18, 2005 at 04:51:30PM +0300, Michael S. Tsirkin wrote: >> > I'm thinking about how to keep the server running and iterating. >> > The goal is to be able to run a sequence of tests just >> > from the client side. >> >> I just do ssh server $PWD/rdma_lat & ( sleep 1 && $PWD/rdma_lat) > >hrm...that would work too though it's not as intuitive. Would only work partly.. try: ssh server $PWD/rdma_lat & ( sleep 1 && $PWD/rdma_lat -s 0 or -1) and your server process will just hang. I know.. don't do that then (checking if size<1 may just paper over missing checks in verbs. i don't know). ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] [PATCH][RFC][0/4] InfiniBand userspace verbs implementation
On Mon, Apr 04, 2005 at 03:09:00PM -0700, Roland Dreier wrote: >Here is an initial implementation of InfiniBand userspace verbs. I >plan to commit this code to the OpenIB repository shortly, and submit >it for inclusion during the 2.6.13 cycle, so I am posting it early for >comments. [snip] >Please take a look at this code if you have a chance. I would >appreciate high-level criticism of the design and implementation as >well as nitpicky complaints about coding style and typos. Can we eventually have checks for NULL before dereferencing members? I'd be happy if they would be wrapped in #ifdef DEBUG or the like, so common use would not suffer the additional penalty those checks imply. e.g.: ibv_modify_qp(NULL,...) etc, etc i.e. ibv_() and also the API exposed by mthca. Something like ibverbs:verbs.c:145: Error: ibv_modify_qp called with qp==NULL Not pretty, agree, but would be convenient for testing new users of those libraries. Just a suggestion. > >In particular, the memory pinning code in in uverbs_mem.c could stand >a looking over. In addition, a sanity check of the write()-based >scheme for passing commands into the kernel in uverbs_main.c and >uverbs_cmd.c is probably worthwhile. ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] Gen2 test suites
On Tue, May 17, 2005 at 10:43:20AM -0700, Sean Hefty wrote: >Shirley Ma wrote: >>>I think an automated test suite would be extremely useful, but of >>>course it would be a lot of work to implement and maintain. >> >>We would have more resouce this year, we could work on this with >>community's help if we think it's important. > >I think having this would be great. [Community help or a testsuite? ;) Late here, so please excuse me] Let me just point anyone interrested to qmtest. YMMV. ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] ibv_reg_mr limits?
On Tue, May 17, 2005 at 08:07:32AM -0700, Roland Dreier wrote: >Arlin> Is there a memory registration limit with ibv_reg_mr? My >Arlin> pingpong test returns error when trying to register 100 MB >Arlin> of memory on my server with 1GB of memory. > >mthca (in the kernel) currently implements memory registration in a >fairly stupid way that requires it to allocate a contiguous chunk of Any timeframe on when this will be changed? I'm right now seeing a bunch of "rank XYZ reg mr failed!" in my test-applet which are my fault, but i wonder if -- after having fixed my test-applet -- i should go and boot with mem=384M and make mthca handle the rest of the mem on its own, as a temporary workaround >memory to hold the entire page table of the memory region being >registered. This can be fixed but right now I would expect the limit >on a single registration to be ~64 MB (or less if memory is fragmented). > > - R. >___ >openib-general mailing list >openib-general@openib.org >http://openib.org/mailman/listinfo/openib-general > >To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general > ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] Re: user-mode verbs on Itanium
On Thu, May 12, 2005 at 08:55:50AM -0700, Grant Grundler wrote: >On Wed, May 11, 2005 at 10:21:38AM +0300, Michael S. Tsirkin wrote: >> You mean just sample the clock before the first post? > >Yes. > >> That would be easy, but lets look at what does it measure: >> >> On the client, you just get the time it takes to perform the >> first post_send operation. If this is interesting to you, I agree, >> but lets take the timestamps around each post send operation to >> make it statistically relevant. >> Let me know - and need to think of a way to do this without >> affecting latency numbers. > >Yes, I'd like to try that. >But a few more of my pending patches need to either be rejected >or (preferably) accepted. Grant, please add the perftest.05 patch to the batch since without it the test is broken for i386. TIA, Bernhard ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [PATCH] Re: [openib-general] 0 op factor
On Thu, May 12, 2005 at 03:48:57PM -0700, Libor Michalek wrote: >On Tue, May 10, 2005 at 05:47:00PM -0700, Libor Michalek wrote: >> On Sat, May 07, 2005 at 07:47:18PM +0200, Bernhard Fischer wrote: >> >> > - remove expect from _sdp_cm_path_complete(). >> >> When I said that the SDP_EXPECT should eventually be removed, I meant >> that the functions which have their return values checked by SDP_EXPECT >> should either be turned into void return functions, or something >> intelligent should be done with the return value, such as error recovery >> or propogation. The functions which should be turned into void functions >> are the ones which will never return anything but success. > > To expand on the last point. There are a lot of functions, for >example those in sdp_buff.c, which check for incorrect function >usage, such as checking that a buffer is not already in a queue before >inserting it into a queue. These checks could be removed entirely or >the checks could be made and a call to BUG() in case the condition >is met. I prefer the later since the former would corrupt the entire >queue. I'm not sure which is preferable to everyone else, either way >the result is that the function becomes a void. BUG_ON would be fine with me, fwiw. Guess others might have a stronger opinion regarding either, though. ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [PATCH] Re: [openib-general] 0 op factor
On Tue, May 10, 2005 at 05:47:00PM -0700, Libor Michalek wrote: >On Sat, May 07, 2005 at 07:47:18PM +0200, Bernhard Fischer wrote: >> - remove '0 operator factor' statements. >> - a bit of whitespace removal. >> - remove return at end of void functions. > > If you're removing the 'return' why not remove the blank line before >it as well. Should have done that too, yes. Would you ditch those blank lines while ci the whitespace stuff? If you want, i could send you a patchlet which does remove them sometimes not before end of next week. Please just drop the removal of the expect, it shouldn't have been it in the patch i sent. Thank you, Bernhard ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] [PATCH] dapltest typo in documentation
Hi James, - typo in documentation Signed-off-by: Bernhard Fischer <[EMAIL PROTECTED]> diff -rup test.oorig/dapltest/README test/dapltest/README --- test.oorig/dapltest/README 2005-05-04 12:19:58.0 + +++ test/dapltest/README2005-05-11 17:05:42.0 + @@ -175,7 +175,7 @@ USAGE - Performance test client of data transfers will be made over each endpoint. Default: 1000 -[ -p pipeline ] Specify the pipline length, valid arguments are in +[ -p pipeline ] Specify the pipeline length, valid arguments are in the range [0,MAX_SEND_DTOS]. If a value greater than MAX_SEND_DTOS is requested the value will be adjusted down to MAX_SEND_DTOS. diff -rup test.oorig/dapltest/cmd/dapl_performance_cmd.c test/dapltest/cmd/dapl_performance_cmd.c --- test.oorig/dapltest/cmd/dapl_performance_cmd.c 2005-05-09 09:34:06.0 + +++ test/dapltest/cmd/dapl_performance_cmd.c2005-05-11 17:05:17.0 + @@ -37,7 +37,7 @@ DT_Performance_Cmd_Usage (void) DT_Mdep_printf ("USAGE: [-D ]\n"); DT_Mdep_printf ("USAGE: [-d] : debug (zero)\n"); DT_Mdep_printf ("USAGE: [-i ] : (1, 000)\n"); -DT_Mdep_printf ("USAGE: [-p ]\n"); +DT_Mdep_printf ("USAGE: [-p ]\n"); DT_Mdep_printf ("USAGE: [-R ]\n"); DT_Mdep_printf ("USAGE: (BE == QOS_BEST_EFFORT - Default)\n"); DT_Mdep_printf ("USAGE: (HT == QOS_HIGH_THROUGHPUT)\n"); @@ -213,12 +213,12 @@ DT_Performance_Cmd_Parse ( cmd->num_iterations = atol (opts->optarg); break; } - case 'p': /* pipline size */ + case 'p': /* pipeline size */ { len = strspn (opts->optarg, "0123456789"); if (len == 0 || len != strlen (opts->optarg)) { - DT_Mdep_printf ("Syntax Error -p option\n"); + DT_Mdep_printf ("Syntax Error -p option\n"); DT_Performance_Cmd_Usage (); return (false); } ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] perftest-03: put C code in .c file
On Wed, May 11, 2005 at 06:25:07PM +0200, Bernhard Fischer wrote: >On Wed, May 11, 2005 at 08:11:05AM -0700, Grant Grundler wrote: >>On Wed, May 11, 2005 at 03:00:04PM +0200, Bernhard Fischer wrote: >>> >- gcc $(CFLAGS) -o rdma_lat rdma_lat.c -libverbs >>> >>> please replace that gcc with $(CC) >> >>yes, good idea. Patch below (perftest-04) removes the line. >>Applies on top of perftest-03. > >With the attached (let's call it perftest-05 as it's on top of -04) it >links again on i386 (get_clock vs. get_cycles), >furthermore some sparse warning about non-ANSI fn decls, and 0 vs. NULL. >Also make pp_client_exch_dest and pp_server_connect static. > >Please consider. > >Signed-off-by: Bernhard Fischer <[EMAIL PROTECTED]> s/\.com/\.at/ whatever. ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] perftest-03: put C code in .c file
On Wed, May 11, 2005 at 08:11:05AM -0700, Grant Grundler wrote: >On Wed, May 11, 2005 at 03:00:04PM +0200, Bernhard Fischer wrote: >> >- gcc $(CFLAGS) -o rdma_lat rdma_lat.c -libverbs >> >> please replace that gcc with $(CC) > >yes, good idea. Patch below (perftest-04) removes the line. >Applies on top of perftest-03. With the attached (let's call it perftest-05 as it's on top of -04) it links again on i386 (get_clock vs. get_cycles), furthermore some sparse warning about non-ANSI fn decls, and 0 vs. NULL. Also make pp_client_exch_dest and pp_server_connect static. Please consider. Signed-off-by: Bernhard Fischer <[EMAIL PROTECTED]> > >thanks, >grant diff -rup perftest-04.oorig/get_clock.h perftest/get_clock.h --- perftest-04.oorig/get_clock.h 2005-05-11 16:05:48.0 + +++ perftest/get_clock.h2005-05-11 16:09:39.0 + @@ -34,7 +34,7 @@ #if defined (__x86_64__) || defined(__i386__) typedef unsigned long long cycles_t; -static inline cycles_t get_clock() +static inline cycles_t get_cycles(void) { unsigned low, high; unsigned long long val; @@ -45,7 +45,7 @@ static inline cycles_t get_clock() } #elif defined(__PPC64__) typedef unsigned long long cycles_t; -static inline cycles_t get_cycles() +static inline cycles_t get_cycles(void) { cycles_t ret; @@ -54,7 +54,7 @@ static inline cycles_t get_cycles() } #elif defined(__ia64__) typedef unsigned long long cycles_t; -static inline cycles_t get_cycles() +static inline cycles_t get_cycles(void) { cycles_t ret; diff -rup perftest-04.oorig/rdma_lat.c perftest/rdma_lat.c --- perftest-04.oorig/rdma_lat.c2005-05-11 10:59:37.0 + +++ perftest/rdma_lat.c 2005-05-11 16:10:42.0 + @@ -163,7 +163,7 @@ static int pp_client_connect(const char return sockfd; } -struct pingpong_dest * pp_client_exch_dest(int sockfd, +static struct pingpong_dest * pp_client_exch_dest(int sockfd, const struct pingpong_dest *my_dest) { struct pingpong_dest *rem_dest = NULL; @@ -202,7 +202,7 @@ out: return rem_dest; } -int pp_server_connect(int port) +static int pp_server_connect(int port) { struct addrinfo *res, *t; struct addrinfo hints = { @@ -244,7 +244,7 @@ int pp_server_connect(int port) } listen(sockfd, 1); - connfd = accept(sockfd, NULL, 0); + connfd = accept(sockfd, NULL, NULL); if (connfd < 0) { perror("server accept"); fprintf(stderr, "accept() failed\n"); ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] perftest-03: put C code in .c file
Hi, On Tue, May 10, 2005 at 06:40:14PM -0700, Grant Grundler wrote: >Hi Michael, >Patch below fixes up the Makefile and getclock.h so C code is in C files >--- Makefile (revision 2296) >+++ Makefile (working copy) >@@ -1,6 +1,9 @@ > CFLAGS = -Wall -O2 -g -D_GNU_SOURCE + CC = gcc > >-rdma_lat: rdma_lat.c get_clock.h >- gcc $(CFLAGS) -o rdma_lat rdma_lat.c -libverbs please replace that gcc with $(CC) >+RDMA_LAT_OBJ = rdma_lat.o get_clock.o >+ >+rdma_lat: $(RDMA_LAT_OBJ) >+ gcc -o rdma_lat $(RDMA_LAT_OBJ) -libverbs ditto >+ > clean: >- rm -f rdma_lat >+ rm -f rdma_lat $(RDMA_LAT_OBJ) thank you ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] [KDAPL] module initialization order
On Tue, May 10, 2005 at 02:13:47PM -0700, Roland Dreier wrote: >James> That situation could be viewed as a configuration error. I >James> think we can use the kconfig language to prevent this. You could select y the parent-provider of a module, yes. This doesn't take care of removal when all parts are modules, though. > >How do you deal with > >modprobe kdapl >modprobe nfs_rdma >modprobe mthca > >Or how about: > >modprobe mthca >modprobe kdapl >modprobe nfs_rdma > If i understand the question correctly, isn't hotplug (i.e. userspace) taking care of this? locally for e.g. ipoib, i have something along the lines of: tst11:~ # modprobe -v ib0 install /sbin/modprobe -v ib_mthca && /sbin/modprobe -v ib_ipoib insmod /lib/modules/2.6.12-rc4/kernel/drivers/infiniband/core/ib_core.ko insmod /lib/modules/2.6.12-rc4/kernel/drivers/infiniband/core/ib_mad.ko insmod /lib/modules/2.6.12-rc4/kernel/drivers/infiniband/hw/mthca/ib_mthca.ko insmod /lib/modules/2.6.12-rc4/kernel/drivers/infiniband/core/ib_sa.ko insmod /lib/modules/2.6.12-rc4/kernel/drivers/infiniband/ulp/ipoib/ib_ipoib.ko tst11:~ # now hotplug would call the 'remove' line for e.g. ipoib et al^A^K^J tst11:~ # modprobe -v -r ib_ipoib remove /sbin/rmmod ib_ipoib ; /sbin/rmmod ib_sa ; /sbin/rmmod ib_mthca ; /sbin/rmmod ib_mad ; /sbin/rmmod ib_core The pci hotplug should have taken care to decrement the use count, so if you pulled the last user, userspace hotplug would unload the unused (if configured to do so) parts of the driver chain. NFS users in the above example would just sit there and block, i think. ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[PATCH] Re: [openib-general] 0 op factor
On Thu, May 05, 2005 at 05:21:22PM -0700, Libor Michalek wrote: > No point other then the obvious, changing the sign of the variables >value. It's just a style convention that I have no problem changing. - remove '0 operator factor' statements. - a bit of whitespace removal. - remove return at end of void functions. - remove expect from _sdp_cm_path_complete(). - comment out empty if(var&FLAG){}else{} in _sdp_inet_release(). untested. hw/mthca/mthca_cmd.c |2 - ulp/sdp/sdp_actv.c | 12 +++--- ulp/sdp/sdp_inet.c | 58 ++- ulp/sdp/sdp_rcvd.c |3 -- ulp/sdp/sdp_wall.c |4 +-- 5 files changed, 34 insertions(+), 45 deletions(-) diff -rup infiniband.oorig/hw/mthca/mthca_cmd.c infiniband/hw/mthca/mthca_cmd.c --- infiniband.oorig/hw/mthca/mthca_cmd.c 2005-01-19 15:53:15.0 +0100 +++ infiniband/hw/mthca/mthca_cmd.c 2005-05-07 19:28:34.924018992 +0200 @@ -219,7 +219,7 @@ static int mthca_cmd_post(struct mthca_d * (and some architectures such as ia64 implement memcpy_toio * in terms of writeb). */ - __raw_writel(cpu_to_be32(in_param >> 32), dev->hcr + 0 * 4); + __raw_writel(cpu_to_be32(in_param >> 32), dev->hcr/*+0 * 4*/); __raw_writel(cpu_to_be32(in_param & 0xul), dev->hcr + 1 * 4); __raw_writel(cpu_to_be32(in_modifier), dev->hcr + 2 * 4); __raw_writel(cpu_to_be32(out_param >> 32), dev->hcr + 3 * 4); diff -rup infiniband.oorig/ulp/sdp/sdp_actv.c infiniband/ulp/sdp/sdp_actv.c --- infiniband.oorig/ulp/sdp/sdp_actv.c 2005-02-19 14:33:28.0 +0100 +++ infiniband/ulp/sdp/sdp_actv.c 2005-05-07 19:04:41.244971336 +0200 @@ -385,7 +385,6 @@ static void _sdp_cm_path_complete(u64 id struct sdp_opt *conn = (struct sdp_opt *) arg; struct sdpc_buff *buff; int result = 0; - int expect; /* * lock the socket */ @@ -491,8 +490,7 @@ static void _sdp_cm_path_complete(u64 id sdp_dbg_warn(conn, "Error <%d> buffering hello msg.", result); - expect = sdp_buff_pool_put(buff); - SDP_EXPECT(!(0 > expect)); + sdp_buff_pool_put(buff); goto failed; } @@ -560,20 +558,18 @@ static void _sdp_cm_path_complete(u64 id goto done; failed: - result = sdp_wall_recv_reject(conn, (0 - status)); + result = sdp_wall_recv_reject(conn, -status); if (0 > result) { sdp_dbg_warn(conn, "Error <%d> rejecting connection", result); - - expect = sdp_wall_recv_drop(conn); - SDP_EXPECT(!(0 > expect)); + + sdp_wall_recv_drop(conn); } /* if */ done: SDP_CONN_UNLOCK(conn); SDP_CONN_PUT(conn); - return; } /* _sdp_cm_path_complete */ /* diff -rup infiniband.oorig/ulp/sdp/sdp_inet.c infiniband/ulp/sdp/sdp_inet.c --- infiniband.oorig/ulp/sdp/sdp_inet.c 2005-02-19 14:33:28.0 +0100 +++ infiniband/ulp/sdp/sdp_inet.c 2005-05-07 19:21:34.928867960 +0200 @@ -103,11 +103,11 @@ void sdp_inet_wake_send(struct sock *sk) { struct sdp_opt *conn; - if (NULL == sk || + if (NULL == sk || NULL == (conn = SDP_GET_CONN(sk))) { return; - } + } if (NULL != sk->sk_socket && 0 < test_bit(SOCK_NOSPACE, &sk->sk_socket->flags) && @@ -128,7 +128,6 @@ void sdp_inet_wake_send(struct sock *sk) read_unlock(&sk->sk_callback_lock); } - return; } /* sdp_inet_wake_send */ /* @@ -149,7 +148,6 @@ void sdp_inet_wake_generic(struct sock * read_unlock(&sk->sk_callback_lock); } - return; } /* sdp_inet_wake_generic */ /* @@ -161,7 +159,7 @@ void sdp_inet_wake_recv(struct sock *sk, read_lock(&sk->sk_callback_lock); if (NULL != sk->sk_sleep) { - + wake_up_interruptible(sk->sk_sleep); } @@ -169,7 +167,6 @@ void sdp_inet_wake_recv(struct sock *sk, read_unlock(&sk->sk_callback_lock); } - return; } /* sdp_inet_wake_recv */ /* @@ -181,7 +178,7 @@ void sdp_inet_wake_error(struct sock *sk read_lock(&sk->sk_callback_lock); if (NULL != sk->sk_sleep) { - + wake_up_interruptible(sk->sk_sleep); } @@ -189,7 +186,6 @@ void sdp_inet_wake_error(struct sock *sk read_unlock(&sk->sk_callback_lock); } - return; } /* sdp_inet_wake_error */ /* @@ -205,7 +201,6 @@ void sdp_inet_wake_urg(struct sock *sk) sk_send_sigurg(sk); } - return; } /* sdp_inet_wake_urg */ /* @@ -384,7 +379,7 @@ static int _sdp_inet_release(struct sock */ result =
Re: [openib-general] sdp_pass.c: sdp_cm_listen_lookup()
On Fri, May 06, 2005 at 04:56:33PM -0700, Tom Duffy wrote: >On Sat, 2005-05-07 at 00:35 +0200, Bernhard Fischer wrote: >> Hi, >> >> I get compile errors in sdp_cm_listen_lookup for these being unknown: >> >> sk_debug, sk_localroute, sk_rcvtstamp > >I posted a patch to fix these to the list. It should be applied to the >tree once 2.6.12 is out. In the mean time, the tree should compile >cleanly against 2.6.11. Libor hinted at another problem all together >and that is the new socket slab allocator. I posted a patch to update >SDP to use the new mechanism as well. If you cannot find the patches in >the archives, I will resend. Thanks for the hint. I'll find them in the archives. ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] sdp_pass.c: sdp_cm_listen_lookup()
On Fri, May 06, 2005 at 03:57:50PM -0700, Libor Michalek wrote: >On Sat, May 07, 2005 at 12:35:06AM +0200, Bernhard Fischer wrote: >> Hi, >> >> I get compile errors in sdp_cm_listen_lookup for these being unknown: >> >> sk_debug, sk_localroute, sk_rcvtstamp > > Which version of the kernel are you using, and is this the only error >that you get? 2.6.12-rc, gen2 rev. 2270. I did not see any other obvious compile errors. I did not find any other occurance of the abovementioned three, is this correct? > >-Libor > ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] [patch] misc trivia nitpicks 04
On Fri, May 06, 2005 at 02:08:07PM -0700, Grant Grundler wrote: >On Fri, May 06, 2005 at 10:25:51PM +0200, Bernhard Fischer wrote: >> I've also inlined double_compar > >Eh? That's pointless. erm, right. Other notes: getopt 'i' fills in ib_port but checks port, it seems. Didn't look closely. The segfault i was seeing did came from dlist_start(). iters is off by one (stops at 998 when asked for 1000). And the timings look odd towards the end: 987, 4.944600 988, 4.977098 989, 5.158333 990, 5.482057 991, 6.955688 992, 7.680629 993, 8.210586 994, 8.759292 995, 9.820456 996, 26.436613 997, 72.905357 998, 704572144.206260 ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] sdp_pass.c: sdp_cm_listen_lookup()
Hi, I get compile errors in sdp_cm_listen_lookup for these being unknown: sk_debug, sk_localroute, sk_rcvtstamp ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] [patch] misc trivia nitpicks 04
On Fri, May 06, 2005 at 02:08:07PM -0700, Grant Grundler wrote: >On Fri, May 06, 2005 at 10:25:51PM +0200, Bernhard Fischer wrote: >Hrm...can you post output how far it gets? > >You verified your libibverbs matches your kernel modules? Ah, yes it's an old kernel. thanks, ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] [patch] misc trivia nitpicks 04
On Fri, May 06, 2005 at 12:43:30PM -0700, Grant Grundler wrote: >On Fri, May 06, 2005 at 07:41:04PM +0200, Bernhard Fischer wrote: >> 02 was get_clock.h: if mhz!=m set mhz=0.0 and break so the file is >> closed proper. I haven't sent that yet. > >I've fixed now too. > >I've added an option for the program to print out the results in cycles. > >And get_clock_khz() is silly - it should be called get_clock_mhz since >the usage is: > double mhz = get_cpu_khz()/1000; > >and all the arches I have report: > ia64 reports "cpu MHz: 1500.00" > x86 reports "cpu MHz : 1000.814" > parisc reports "cpu MHz : 650.00" > >I also fixed a typo "meadian" (vs median). > >I'll post a -02 version shortly at: > http://gsyprf3.external.hp.com/openib/perftest-02.tgz > > >> Should i send a patch later or will MST fix this? > >I'll keep pestering MST with this bit. :^) I've also inlined double_compar and inserted an else befor the second if to have a better if elif else chain. Cosmetic since it's fed into qsort, i assume. Even the original version just segfaults on me. I must be missing something.. Still looking > >thanks, >grant > ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] [patch] misc trivia nitpicks 04
On Fri, May 06, 2005 at 11:13:51AM -0400, Hal Rosenstock wrote: >On Thu, 2005-05-05 at 20:39, Bernhard Fischer wrote: >> Hi, >> >> - s/incomm/incom/g > >Patch applied (osm/include/iba/ib_types.h) Thank you. > >BTW, I saw niopicks 01, 03, and 04. Did I miss 02 ? 02 was get_clock.h: if mhz!=m set mhz=0.0 and break so the file is closed proper. I haven't sent that yet. Should i send a patch later or will MST fix this? ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] Re: [openib-commits] r2239 - gen2/trunk/src/linux-kernel/infiniband/include
On Thu, May 05, 2005 at 04:33:23PM -0400, Hal Rosenstock wrote: >On Thu, 2005-05-05 at 16:16, Bernhard Fischer wrote: >> Furthermore it would be really, really nice if anyone who touches files >> would trim any whitespace as that is extremely annoying if you happen to >> be forced to look at. Also, whitespace just eats my precious >> (disc-)space for no good. > >Are OpenSM files guilty of this ? Just looking at the headers (which are probably most likely to be installed) and within those, superfluous noise before eol. $ cat ~/huh4 find ./ -iname "*h" -exec grep -HinE "[[:space:]]$" {} \; gives the attached (file:line as of gen2 rev. 2257) I'd say yes. >From my POV, it would be good to trim at least those that might get propagated to installed images. YMMV. feeding /usr/local/src/obj/gen2/trunk/src/userspace into /home/cow/huh4 ./libibcm/include/infiniband/cm_abi.h:53: ./libibcm/include/infiniband/cm.h:267: * IB_CM_REQ_RECEIVED and IB_CM_SIDR_REQ_RECEIVED communication events ./libibcm/include/infiniband/cm.h:268: * generated as a result of listen requests result in the allocation of a ./libibcm/include/infiniband/cm.h:269: * new @cm_id. ./libibcm/include/infiniband/cm.h:328:int ib_cm_attr_id(uint32_t cm_id, ./libmthca/config/ltmain.sh:110:: ${IFS=" ./mstflint/mtcr.h:146: unsigned long long ./mstflint/mtcr.h:192: Binary file ./libibverbs/examples/.libs/ibv_asyncwatch matches ./libibverbs/config/ltmain.sh:110:: ${IFS=" ./libibverbs/include/infiniband/verbs.h:378:struct ibv_pd *pd; ./libibverbs/include/infiniband/verbs.h:522: ./libibverbs/include/infiniband/driver.h:63: ./management/libibcommon/include/infiniband/common.h:187:get_build_version(void) ./management/libibmad/include/infiniband/mad.h:242: IB_GS_DATA_F, ./management/osm/include/vendor/osm_vendor_mlx_sender.h:63: * ./management/osm/include/vendor/osm_vendor_mlx_sender.h:118: * NAME osmv_rmpp_snd_error ./management/osm/include/vendor/osm_vendor_mtl.h:91:* ./management/osm/include/vendor/osm_vendor_mtl.h:153:* Handle returned by MTL open call (ib_open_al). ./management/osm/include/vendor/osm_vendor_mtl.h:242: * Given the vendor obj and a guid ./management/osm/include/vendor/osm_vendor_mtl.h:282: * Fill in the array of port_attr with all available ports on ALL the ./management/osm/include/vendor/osm_vendor_mtl.h:284: * ALSO - ./management/osm/include/vendor/osm_vendor_mlx.h:61: ./management/osm/include/vendor/osm_vendor_mlx.h:68:uint32_t resp_timeout; ./management/osm/include/vendor/osm_vendor_mlx.h:69: ./management/osm/include/vendor/osm_vendor_mlx.h:75: ./management/osm/include/vendor/osm_vendor_mlx_transport.h:40: * ./management/osm/include/vendor/osm_vendor_mlx_transport.h:43: * that allows sending a single MAD/receiving a callback ./management/osm/include/vendor/osm_vendor_mlx_transport.h:44: * when a single MAD is received. ./management/osm/include/vendor/osm_vendor_mlx_inout.h:56:#ifndef OSM_VENDOR_INTF_ANAFA ./management/osm/include/vendor/osm_vendor_ts.h:139: * Points to the MAD pool used by OSM ./management/osm/include/vendor/osm_vendor_ts.h:160: * ./management/osm/include/vendor/osm_vendor_ts.h:220: * Handle returned by TS open call . ./management/osm/include/vendor/osm_vendor_ts.h:314: * Given the vendor obj and a guid ./management/osm/include/vendor/osm_vendor_ts.h:354: * Fill in the array of port_attr with all available ports on ALL the ./management/osm/include/vendor/osm_vendor_ts.h:356: * ALSO - ./management/osm/include/vendor/osm_vendor_mtl_hca_guid.h:40: * Provides interface over VAPI for obtaining the local ports guids or from guid ./management/osm/include/vendor/osm_vendor_mtl_hca_guid.h:134: * Given the vendor obj and a guid ./management/osm/include/vendor/osm_vendor_mtl_hca_guid.h:173: * Fill in the array of port_attr with all available ports on ALL the ./management/osm/include/vendor/osm_vendor_mtl_hca_guid.h:175: * ALSO - ./management/osm/include/vendor/osm_vendor_umadt.h:75:* ./management/osm/include/vendor/osm_vendor_umadt.h:104:*Umadt APIs. ./management/osm/include/vendor/osm_vendor_umadt.h:131:*Umadt mad structure to identify a mad. ./management/osm/include/vendor/osm_vendor_umadt.h:140: ./management/osm/include/vendor/osm_vendor_ibumad.h:68:* ./management/osm/include/vendor/osm_vendor_mlx_sar.h:54:void* p_arbt_mad; ./management/osm/include/vendor/osm_vendor_mlx_sar.h:58:uint32_t data_sz; /*typical data sz for this kind of mad (sa or regular */ ./management/osm/include/vendor/osm_vendor_mlx_sar.h:59: ./management/osm/include/vendor/osm_vendor_mlx_sar.h:71: * ./management/osm/include/vendor/osm_vendor_mlx_sar.h:74:osmv_rmpp_sar_init(osmv_rmpp_sar_t* p_sar, voi
Re: [openib-general] [patch] misc trivia nitpicks 01
On Thu, May 05, 2005 at 06:08:28PM -0700, Libor Michalek wrote: >On Fri, May 06, 2005 at 02:49:05AM +0200, Bernhard Fischer wrote: >> On Thu, May 05, 2005 at 05:32:57PM -0700, Libor Michalek wrote: >> >On Thu, May 05, 2005 at 07:09:56PM +0200, Bernhard Fischer wrote: >> >> Hi, >> >> >> >> please consider attached. >> >> >> >> diffstat ./gen2.2257.cow-20050505-1902.diff >> >> linux-kernel/infiniband/ulp/sdp/sdp_actv.c | 27 +-- >> >> userspace/libmthca/src/mthca.c |4 >> >> userspace/management/osm/include/opensm/osm_subnet.h | 79 +- >> >> userspace/management/osm/include/vendor/osm_vendor_mtl_transaction_mgr.h >> >> | 43 ++--- >> >> userspace/management/osm/opensm/osm_sa_service_record.c |8 - >> >> userspace/management/osm/opensm/osm_sm_state_mgr.c | 38 ++-- >> >> userspace/management/osm/opensm/osm_sw_info_rcv.c |9 - >> >> userspace/management/osm/opensm/osm_trap_rcv.c | 62 +++ >> >> 8 files changed, 134 insertions(+), 136 deletions(-) [] >> At any rate, i'll try to send them as separate patches per vendor in the >> future.. [] > I was just suggesting breaking it up by module/program which would >avoid a loaded value judgment like this yet still achieves the desired Yes, i see. Thanks in advance for applying the parts you feel comfortable with :) cheers, >result. The only reason I suggested it was that I saw SDP and OpenSM >in the same patch and I haven't even built OpenSM in a while, so I >didn't want to check in changes to it. > >-Libor > > ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] [patch] misc trivia nitpicks 01
On Thu, May 05, 2005 at 05:32:57PM -0700, Libor Michalek wrote: >On Thu, May 05, 2005 at 07:09:56PM +0200, Bernhard Fischer wrote: >> Hi, >> >> please consider attached. >> >> diffstat ./gen2.2257.cow-20050505-1902.diff >> linux-kernel/infiniband/ulp/sdp/sdp_actv.c | 27 +-- >> userspace/libmthca/src/mthca.c |4 >> userspace/management/osm/include/opensm/osm_subnet.h | 79 +- >> userspace/management/osm/include/vendor/osm_vendor_mtl_transaction_mgr.h | >> 43 ++--- >> userspace/management/osm/opensm/osm_sa_service_record.c |8 - >> userspace/management/osm/opensm/osm_sm_state_mgr.c | 38 ++-- >> userspace/management/osm/opensm/osm_sw_info_rcv.c |9 - >> userspace/management/osm/opensm/osm_trap_rcv.c | 62 +++ >> 8 files changed, 134 insertions(+), 136 deletions(-) > >Bernhard, > > I have no problem with the content, but I think this would be easier if >you submitted a patch per module/program, so each maintainer could treat >this independently. I tend to disagree. I'm an independant nitpicker. I'll try to separate those, but if a patch -- especially a relatively trivial one, which mostly touches obvious typos in the documentation -- comes in, i'd think that you in the sense of payed maintainers should be able to weight if you can apply them, regardless of who pays the 3 minutes to read through it. At any rate, i'll try to send them as separate patches per vendor in the future.. PS: cisco is core, mthca, sdp, and voltaire is osm, yes? Or would a split per individual be better? I really don't know. > >-Libor > > ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] [patch] misc trivia nitpicks 04
Hi, - s/incomm/incom/g diff -X excl -rduNp gen2.2257.oorig/trunk/src/linux-kernel/infiniband/core/ucm.c gen2.2257/trunk/src/linux-kernel/infiniband/core/ucm.c --- gen2.2257.oorig/trunk/src/linux-kernel/infiniband/core/ucm.c 2005-04-26 20:58:50.0 +0200 +++ gen2.2257/trunk/src/linux-kernel/infiniband/core/ucm.c 2005-05-06 02:46:18.055867696 +0200 @@ -99,7 +99,7 @@ static void ib_ucm_ctx_put(struct ib_ucm list_del(&uevent->file_list); list_del(&uevent->ctx_list); - /* clear incomming connections. */ + /* clear incoming connections. */ if (uevent->cm_id) (void)ib_destroy_cm_id(uevent->cm_id); diff -X excl -rduNp gen2.2257.oorig/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_inet.c gen2.2257/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_inet.c --- gen2.2257.oorig/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_inet.c 2005-04-26 20:58:46.0 +0200 +++ gen2.2257/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_inet.c 2005-05-06 02:42:17.965367000 +0200 @@ -37,7 +37,7 @@ #include "sdp_main.h" /* - * list of connections waiting for an incomming connection + * list of connections waiting for an incoming connection */ static int proto_family= AF_INET_SDP; static int buff_min= SDP_BUFF_POOL_COUNT_MIN; @@ -690,7 +690,7 @@ done: } /* - * sdp_inet_listen - listen on a socket for incomming addresses + * sdp_inet_listen - listen on a socket for incoming addresses */ static int sdp_inet_listen(struct socket *sock, int backlog) { diff -X excl -rduNp gen2.2257.oorig/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_pass.c gen2.2257/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_pass.c --- gen2.2257.oorig/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_pass.c 2005-04-26 20:58:46.0 +0200 +++ gen2.2257/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_pass.c 2005-05-06 02:45:48.961290744 +0200 @@ -35,7 +35,7 @@ #include "sdp_main.h" /* - * handle incomming passive connection error. (REJ) + * handle incoming passive connection error. (REJ) */ void sdp_cm_pass_error(struct sdp_opt *conn, int error) { @@ -67,7 +67,7 @@ void sdp_cm_pass_error(struct sdp_opt *c } /* - * handle incomming passive connection establishment. (RTU) + * handle incoming passive connection establishment. (RTU) */ int sdp_cm_pass_establish(struct sdp_opt *conn) { @@ -295,7 +295,7 @@ static int sdp_cm_listen_lookup(struct s struct sock *sk; int result; /* -* match a listener with an incomming conn, and generate +* match a listener with an incoming conn, and generate * accept message and state on success. */ sdp_dbg_ctrl(conn, diff -X excl -rduNp gen2.2257.oorig/trunk/src/userspace/examples/aio/README gen2.2257/trunk/src/userspace/examples/aio/README --- gen2.2257.oorig/trunk/src/userspace/examples/aio/README 2005-05-05 10:49:55.0 +0200 +++ gen2.2257/trunk/src/userspace/examples/aio/README 2005-05-06 02:46:34.885309232 +0200 @@ -17,7 +17,7 @@ ttcp.aio.c: gcc -I../../../linux-kernel/infiniband/ulp/sdp ttcp.aio.c \ -o ttcp.aio.x -laio - The program runs in two modes, a server which listens for incomming + The program runs in two modes, a server which listens for incoming connections and acts as a data sink, and a client which connects to a given address/port and acts as a data source. A decent help exists for available parameters, but here are some reasonable defaults: diff -X excl -rduNp gen2.2257.oorig/trunk/src/userspace/management/osm/include/iba/ib_types.h gen2.2257/trunk/src/userspace/management/osm/include/iba/ib_types.h --- gen2.2257.oorig/trunk/src/userspace/management/osm/include/iba/ib_types.h 2005-04-26 20:58:28.0 +0200 +++ gen2.2257/trunk/src/userspace/management/osm/include/iba/ib_types.h 2005-05-06 02:47:15.938068264 +0200 @@ -7403,7 +7403,7 @@ typedef struct _ib_ca_attr * * max_qp_resp_res * max_eec_resp_res -* Maximum limit on number of responder resources for incomming RDMA +* Maximum limit on number of responder resources for incoming RDMA * operations, on QPs and EEC's respectively. * * max_resp_res ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] [SDP] missing MODULE_PARM_DESC
Hi, in src/linux-kernel/infiniband/ulp/sdp/sdp_inet.c module_param seem to be missing MODULE_PARM_DESC please fix. ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] Re: [openib-commits] r2239 - gen2/trunk/src/linux-kernel/infiniband/include
On Thu, May 05, 2005 at 04:29:45PM -0400, Hal Rosenstock wrote: >On Thu, 2005-05-05 at 16:29, Bernhard Fischer wrote: > >> >> Please, if you happen to find those, grep the whole tree for occurances. >> >> >> >> My current -ENOPARSE list looks like: >> >> >> >> conenctions >> >> esablished >> >> etablish >> >> defintions >> >> defaul\ >> >> a\ single\ buffers >> >> retreival >> >> recieve >> >> explict >> >> implict >> >> responsab >> >> listning >> >> conntension >> >> consistancy >> >> the\ the >> >> intia >> > >> >Libor is the maintainer for SDP so this is a request to him. >> >> Libor is maintaining >> trunk/src/userspace/management/osm/opensm/osm_state_mgr.c ? >> >> Guess i'm looking at an old copy (i've got revision 2265), am i? > >OK. Are others in this list in OpenSM as well ? Haven't specifically grepped for occurances in osm, no. I'm not encouraged nor payed to take care of those at all ;) Nonetheless thanks for fixing those you're forced to be concerned about. kind regards, ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] Re: [openib-commits] r2239 - gen2/trunk/src/linux-kernel/infiniband/include
On Thu, May 05, 2005 at 04:16:37PM -0400, Hal Rosenstock wrote: >On Thu, 2005-05-05 at 16:16, Bernhard Fischer wrote: >> Hal, >> >> - Fix commentary typo: s/intialized/initialized/g >> >> Would you mind also fixing >> src/userspace/libsdp/src/port.c >> >> Please, if you happen to find those, grep the whole tree for occurances. >> >> My current -ENOPARSE list looks like: >> >> conenctions >> esablished >> etablish >> defintions >> defaul\ >> a\ single\ buffers >> retreival >> recieve >> explict >> implict >> responsab >> listning >> conntension >> consistancy >> the\ the >> intia > >Libor is the maintainer for SDP so this is a request to him. Libor is maintaining trunk/src/userspace/management/osm/opensm/osm_state_mgr.c ? Guess i'm looking at an old copy (i've got revision 2265), am i? > >-- Hal > >> Furthermore it would be really, really nice if anyone who touches files >> would trim any whitespace as that is extremely annoying if you happen to >> be forced to look at. Also, whitespace just eats my precious >> (disc-)space for no good. >> >> TIA, ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] HCA Firmware
Hi, I may have not searched in the proper places, so may miss the obvious "standard" sink of vendors for firmware.. Still, where can i get current, 'blessed' (fwiw) firmware for infiniband HCA's? PS: On hardware manufactured by an unnamed, recently sold vendor, i get: Mellanox Tavor Device Driver is creating device "InfiniHost0" (bus=0a, devfn=00) THH(1): TAVOR_IF_CMD_STAT_BAD_OP - Operation/command not supported or opcode modifier not supported THH(1): infiniband/hw/mellanox-hca/thh/thh_hob.c[3793]: THH_hob_create: CMD_error in THH_cmd_SYS_EN (2) THH(1): infiniband/hw/mellanox-hca/thh/thh_mod_obj.c[292]: Failed creating THH_hob for InfiniHost0 THH(1): THH_init_hh_all_tavor: For all 1 Tavor devices initialization was not successful THH(1): infiniband/hw/mellanox-hca/thh/thh_mod_obj.c[329]: Failed initialization of all available InfiniHost devices where current firmware is: # cat /proc/infiniband/core/ca1/info name: mthca provider: mthca node GUID: 0005:ad00:0002:312c ports: 2 vendor ID: 0x5ad device ID: 0x5a44 HW revision: 0x0 FW revision: 0x30002 which let's me suspect that that revision needs upgrading, no? Thanks in advance for any hint in this respect. ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] Re: [openib-commits] r2239 - gen2/trunk/src/linux-kernel/infiniband/include
Hal, - Fix commentary typo: s/intialized/initialized/g Would you mind also fixing src/userspace/libsdp/src/port.c Please, if you happen to find those, grep the whole tree for occurances. My current -ENOPARSE list looks like: conenctions esablished etablish defintions defaul\ a\ single\ buffers retreival recieve explict implict responsab listning conntension consistancy the\ the intia Furthermore it would be really, really nice if anyone who touches files would trim any whitespace as that is extremely annoying if you happen to be forced to look at. Also, whitespace just eats my precious (disc-)space for no good. TIA, ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] [patch] misc trivia nitpicks 03
Hi, What's the status of SDP_EXPECT ? will it be submitted upstream or cleaned up beforehand? Just curious. - s/conenct/connect/g - s/conect/connect/g - trim whitespace - fold sdp_conn_internal_lock into sdp_conn_lock. - fold sdp_conn_internal_unlock into sdp_conn_unlock. PS: those folds are completely untested.. Libor, anyone else? thank you, diff -X excl -rduNp gen2.2257.oorig/trunk/src/userspace/management/doc/diagtools.txt gen2.2257/trunk/src/userspace/management/doc/diagtools.txt --- gen2.2257.oorig/trunk/src/userspace/management/doc/diagtools.txt 2005-04-26 20:58:41.0 +0200 +++ gen2.2257/trunk/src/userspace/management/doc/diagtools.txt 2005-05-05 20:33:53.572745456 +0200 @@ -335,7 +335,7 @@ smpquery, smpquery output format, ibaddr 16. ibnetchecknode Description: -Check conectivity and do some simple sanity checks for the specified node. +Check connectivity and do some simple sanity checks for the specified node. Port address is lid unless -G option is used to specify a GUID address. Syntax: diff -X excl -rduNp gen2.2257.oorig/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_conn.c gen2.2257/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_conn.c --- gen2.2257.oorig/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_conn.c 2005-04-26 20:58:46.0 +0200 +++ gen2.2257/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_conn.c 2005-05-05 20:32:42.415562992 +0200 @@ -595,8 +595,8 @@ done: return conn; } -/* - * Functions to cancel IOCB requests in a conenctions queues. +/* + * Functions to cancel IOCB requests in a connections queues. */ static int sdp_desc_q_cancel_lookup_func(struct sdpc_desc *element, void *arg) { diff -X excl -rduNp gen2.2257.oorig/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_conn.c gen2.2257/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_conn.c --- gen2.2257.oorig/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_conn.c 2005-04-26 20:58:46.0 +0200 +++ gen2.2257/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_conn.c 2005-05-05 21:24:27.895458288 +0200 @@ -352,13 +352,13 @@ int sdp_inet_port_get(struct sdp_opt *co INADDR_ANY == look->src_addr || conn->src_addr == look->src_addr) { - sdp_dbg_warn(conn, + sdp_dbg_warn(conn, "port rejected. <%04x><%d:%d><%d:%d><%04x><%u:%u>", port, sk->sk_bound_dev_if, srch->sk_bound_dev_if, sk->sk_reuse, -srch->sk_reuse, +srch->sk_reuse, look->state, conn->src_addr, look->src_addr); @@ -670,6 +670,7 @@ void sdp_conn_destruct(struct sdp_opt *c result = sdp_conn_stat_dump(conn); SDP_EXPECT(result >= 0); + /* * really there shouldn't be anything in these tables, but it's * really bad if we leave a dangling reference here. @@ -748,7 +749,7 @@ void sdp_conn_destruct(struct sdp_opt *c result); } /* -* check consistancy +* check consistency */ if (atomic_read(&conn->refcnt) < 0) sdp_dbg_warn(conn, "destruct low ref count <%04x>", @@ -812,7 +789,7 @@ void sdp_conn_relock(struct sdp_opt *con if (1 == result_r) { result = sdp_cq_event_locked(&entry, conn); if (result < 0) - sdp_dbg_warn(conn, + sdp_dbg_warn(conn, "Error <%d> from event handler.", result); @@ -823,7 +800,7 @@ void sdp_conn_relock(struct sdp_opt *con if (1 == result_s) { result = sdp_cq_event_locked(&entry, conn); if (result < 0) - sdp_dbg_warn(conn, + sdp_dbg_warn(conn, "Error <%d> from event handler.", result); rearm = 1; @@ -836,17 +813,17 @@ void sdp_conn_relock(struct sdp_opt *con result = ib_req_notify_cq(conn->recv_cq, IB_CQ_NEXT_COMP); if (result) - sdp_dbg_warn(co
[openib-general] [patch] misc trivia nitpicks 01
Hi, please consider attached. diffstat ./gen2.2257.cow-20050505-1902.diff linux-kernel/infiniband/ulp/sdp/sdp_actv.c | 27 +-- userspace/libmthca/src/mthca.c |4 userspace/management/osm/include/opensm/osm_subnet.h | 79 +- userspace/management/osm/include/vendor/osm_vendor_mtl_transaction_mgr.h | 43 ++--- userspace/management/osm/opensm/osm_sa_service_record.c |8 - userspace/management/osm/opensm/osm_sm_state_mgr.c | 38 ++-- userspace/management/osm/opensm/osm_sw_info_rcv.c |9 - userspace/management/osm/opensm/osm_trap_rcv.c | 62 +++ 8 files changed, 134 insertions(+), 136 deletions(-) Thank you, - s/reciev/receiv/g; trim superfluous whitespace while touching those. - src/linux-kernel/infiniband/ulp/sdp/sdp_actv.c (sdp_cm_actv_error): trim return off void(). - src/userspace/libmthca/src/mthca.c (openib_driver_init): vendor is unsigned. diff -X excl -rduNp gen2.2257.oorig/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_actv.c gen2.2257/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_actv.c --- gen2.2257.oorig/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_actv.c 2005-04-26 20:58:46.0 +0200 +++ gen2.2257/trunk/src/linux-kernel/infiniband/ulp/sdp/sdp_actv.c 2005-05-05 17:20:25.968369544 +0200 @@ -43,7 +43,7 @@ void sdp_cm_actv_error(struct sdp_opt *c /* * error value is positive error. * -* Handle errors within active connections stream. +* Handle errors within active connections stream. * First generate appropriate response, REJ, DREQ or nothing. * Second the socket must be notified of the error. */ @@ -59,14 +59,14 @@ void sdp_cm_actv_error(struct sdp_opt *c break; case SDP_CONN_ST_ERROR_STRM: /* socket has been destroyed. */ error = ECONNRESET; -case SDP_CONN_ST_REQ_SENT: +case SDP_CONN_ST_REQ_SENT: case SDP_CONN_ST_REP_RECV: case SDP_CONN_ST_RTU_SENT: /* * All four states we have gotten a REP and are now in * one of these states. */ - result = ib_send_cm_rej(conn->cm_id, + result = ib_send_cm_rej(conn->cm_id, IB_CM_REJ_CONSUMER_DEFINED, NULL, 0, NULL, 0); @@ -78,14 +78,14 @@ void sdp_cm_actv_error(struct sdp_opt *c break; case SDP_CONN_ST_ESTABLISHED: /* -* Made it all the way to esablished, need to initiate a +* Made it all the way to established, need to initiate a * full disconnect. */ result = ib_send_cm_dreq(conn->cm_id, NULL, 0); if (result < 0) - sdp_dbg_warn(NULL, "Error <%d> sending CM DREQ", + sdp_dbg_warn(NULL, "Error <%d> sending CM DREQ", result); - + conn->state = SDP_CONN_ST_TIME_WAIT_1; break; } @@ -100,7 +100,6 @@ void sdp_cm_actv_error(struct sdp_opt *c sdp_iocb_q_cancel_all(conn, (0 - error)); sdp_inet_wake_error(conn->sk); - return; } /* @@ -113,8 +112,8 @@ static int sdp_cm_actv_establish(struct struct sock *sk; int result; - sdp_dbg_ctrl(conn, "active etablish. src <%08x:%04x> dst <%08x:%04x>", -conn->src_addr, conn->src_port, + sdp_dbg_ctrl(conn, "active establish. src <%08x:%04x> dst <%08x:%04x>", +conn->src_addr, conn->src_port, conn->dst_addr, conn->dst_port); sk = conn->sk; @@ -134,7 +133,7 @@ static int sdp_cm_actv_establish(struct } qp_attr->rq_psn= conn->rq_psn; - + attr_mask |= IB_QP_RQ_PSN; result = ib_modify_qp(conn->qp, qp_attr, attr_mask); @@ -190,7 +189,7 @@ static int sdp_cm_actv_establish(struct * release disconnects. */ conn->flags &= ~SDP_CONN_F_DIS_HOLD; - + inet_sk(sk)->saddr = htonl(conn->src_addr); inet_sk(sk)->rcv_saddr = htonl(conn->src_addr); @@ -204,7 +203,7 @@ static int sdp_cm_actv_establish(struct */ sdp_inet_wake_send(sk); sdp_inet_wake_recv(sk, 0); - + result = 0; done: kfree(qp_attr); @@ -273,7 +272,7 @@ int sdp_cm_rep_handler(struct ib_cm_id * { struct sdp_msg_hello_ack *hello_ack; int result = -EPROTO; - + if (cm_id != conn->cm_id) { sdp_dbg_warn(conn, "REP comm ID mismatch. <%08x:%08x>", conn->cm_id->local_id, cm_id->local_id); @@ -310,7 +309,7 @@ int sdp_cm_rep_handler(struct ib_cm_id * conn->d_qpn = event->param.rep_rcvd.remote_qpn; /* * The maximum amount of data that can be sent to the remote -* p
[openib-general] 0 op factor
Hi, I forgot, but what's the purpose of statements like: $ egrep -ri "[^[:digit:]|\.]0[[:space:]]*[-+*/]" *|egrep -v "(\.(svn-base|txt):|/\*)" infiniband/ulp/sdp/sdp_inet.c: * flag: 0 - recv shutdown infiniband/ulp/sdp/sdp_actv.c: sdp_iocb_q_cancel_all(conn, (0 - error)); infiniband/ulp/sdp/sdp_actv.c: sdp_cm_actv_error(conn, (0 - result)); infiniband/ulp/sdp/sdp_actv.c: sdp_cm_actv_error(conn, (0 - status)); infiniband/ulp/sdp/sdp_pass.c: sdp_iocb_q_cancel_all(conn, (0 - error)); infiniband/ulp/sdp/sdp_pass.c: sdp_cm_pass_error(conn, (0 - result)); infiniband/ulp/sdp/sdp_rcvd.c: 0 - ((SDP_SRC_AVAIL_RECV > size) ? size : 0); infiniband/hw/mthca/mthca_cmd.c: __raw_writel(cpu_to_be32(in_param >> 32), dev->hcr + 0 * 4); Could somebody please elaborate? thanks ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] Re: [PATCH][RFC][0/4] InfiniBand userspace verbs implementation
On Mon, Apr 18, 2005 at 09:40:40PM +0200, Arjan van de Ven wrote: >On Mon, 2005-04-18 at 11:25 -0500, Timur Tabi wrote: >> Arjan van de Ven wrote: >> >> > this is a myth; linux is free to move the page about in physical memory >> > even if it's mlock()ed!! darn, yes, this is true. I know people who introduced #define VM_RESERVED 0x0008 /* Don't unmap it from swap_out */ to vm_flags just because of this. I'll just hold my breath and won't delve further. >> >> Then Linux has a very odd definition of the word "locked". >> >> > And even then, the user can munlock the memory from another thread etc >> > etc. Not a good idea. >> >> Well, that's okay, because then the app is doing something stupid, so we >> don't worry about >> that. > >you should since that physical page can be reused, say by a root >process, and you'd be majorly screwed ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
[openib-general] Re: [openib-commits] r1821
Hi, Minor commentary changes. Thank you, diff -X ../excl -rdup upstream/gen2/trunk/src/userspace/management/osm/opensm/osm_sa_mcmember_record.c gen2/trunk/src/userspace/management/osm/opensm/osm_sa_mcmember_record.c --- upstream/gen2/trunk/src/userspace/management/osm/opensm/osm_sa_mcmember_record.c 2005-02-19 14:33:27.0 +0100 +++ gen2/trunk/src/userspace/management/osm/opensm/osm_sa_mcmember_record.c 2005-02-19 14:45:40.0 +0100 @@ -490,7 +490,7 @@ __validate_more_comp_fields( if(mtu_mgrp < mtu_required) return FALSE; break; -case 1: /* Less than than MTU specified */ +case 1: /* Less than MTU specified */ if(mtu_mgrp > mtu_required) return FALSE; break; @@ -516,7 +516,7 @@ __validate_more_comp_fields( if(rate_mgrp < rate_required) return FALSE; break; -case 1: /* Less than than RATE specified */ +case 1: /* Less than RATE specified */ if(rate_mgrp > rate_required) return FALSE; break; ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Re: [openib-general] opensm segfaults after CL_INSUFFICIENT_MEMORY
On Fri, Feb 11, 2005 at 10:49:08AM -0500, Hal Rosenstock wrote: >On Fri, 2005-02-11 at 10:31, Bernhard Fischer wrote: >> On Thu, Feb 10, 2005 at 03:26:45PM -0500, Hal Rosenstock wrote: >> >Initialization continues (not sure it should) and attempting to clear a >> >> I'd say it should not but exit. >> >> >static object fails. Even if initialization didn't continue, opensm >> >would not run properly. >> > >> >The first problem is why the malloc fails for the timer provider. The >> >second problem is pretty strange as the system callback object is local >> >and should not be an address that causes a segmentation violation. >> > >> >Is this built with the autotools version ? >> >> Disregard the second problem, pathological compiler settings (wrong >> regparm for that system). >> >> Please do, however, fix the path for ENOMEM. > >I presume you mean exiting on failures during initialization (of the >component library) ? Yes. > >BTW, are you now past these issues ? opensm does work now. I'm still getting ib_mthca :04:00.0: CQ overrun on CQN 0082 on the sender after having transmitted 5MB to 50MB. I was only playing with opensm while waiting for an answer to the CQ overrun... ___ openib-general mailing list openib-general@openib.org http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general