Quoting r. Roland Dreier [EMAIL PROTECTED]:
Subject: Re: [openib-general] scaling issues, was: uDAPL cma: add support for
address and route retries, call disconnect when recving dreq
With 256 nodes all to all already gives you 65000 requests
which is the same order of magnitude as the
Quoting r. Roland Dreier [EMAIL PROTECTED]:
Subject: Re: [PATCH 1/7] IB/core - Add DMA mapping functions to allow device
drivers to interpose
However, this means that the API must give the HCA the choice of
what to keep inside the mapping. This could mean, for example, returning
a
Quoting r. Or Gerlitz [EMAIL PROTECTED]:
what about enhancing xxx_destory_id() to sense that it was called from
this id callback context so the xxx module code defers the destory_id()
execution to run after the callback is over. This can be done by
writing at the id the pid of the thread
Quoting r. Jeff Squyres [EMAIL PROTECTED]:
Subject: Static linking with libibverbs
Do you like large executables?
Do you eschew system-installed libc's?
Were you ever frustrated that linking against libibverbs would
prevent using -static?
Well, now you too can produce 100% statically
Quoting r. Or Gerlitz [EMAIL PROTECTED]:
Subject: Re: [PATCH] for 2-6-19 rdma/addr: use client registration to fix
module unload race
Sean Hefty wrote:
Require registration with ib_addr module to prevent caller from unloading
while a callback is in progress.
Sean,
Is there a
Quoting r. Sean Hefty [EMAIL PROTECTED]:
Subject: [PATCH] for 2-6-19 rdma/addr: use client registration to fix module
unload race
Require registration with ib_addr module to prevent caller from unloading
while a callback is in progress.
Signed-off-by: Sean Hefty [EMAIL PROTECTED]
ib_cm
Quoting r. Or Gerlitz [EMAIL PROTECTED]:
Require registration with ib_addr module to prevent caller from unloading
while a callback is in progress.
ib_cm has this bug as well. Shouldn't we patch it for 2.6.19 too?
I know there was a similar discussion which i was not tracking... re
Quoting r. Or Gerlitz [EMAIL PROTECTED]:
Subject: Re: [openib-general] [PATCH] for 2-6-19 rdma/addr: use client
registration to fix module unload race
Michael S. Tsirkin wrote:
Quoting r. Or Gerlitz [EMAIL PROTECTED]:
Require registration with ib_addr module to prevent caller from
Quoting r. Or Gerlitz [EMAIL PROTECTED]:
there must be away to avoid the race within the ID provider module,
There's no other way - go over the archives.
so at least the api can be saved...
See Documentation/stable_api_nonsense.txt
--
MST
___
Quoting r. Sasha Khapyorsky [EMAIL PROTECTED]:
Look at this for example:
http://sourceware.org/ml/bug-glibc/2005-02/msg00123.html
Hmm, yea. Do you understand the answer there?
Does not make sense to me ...
I less care about the answer, but more about valgrind output (btw
cannot
Quoting r. Michael S. Tsirkin [EMAIL PROTECTED]:
Subject: [PATCH repost] IB/srp: destroy/recreate qp/cq at reconnect
From: Ishai Rabinovitz [EMAIL PROTECTED]
This makes SRP more robust in presence of hardware errors
and is closer to behaviour suggested by IB spec,
reducing chance of stale
Quoting r. Roland Dreier [EMAIL PROTECTED]:
Subject: Re: Ordering between PCI config space writes and MMIO reads?
The discussion fizzled out without really reaching a definitive
answer, so I'm going to apply the original patch (below), since I
pretty much convinced myself that only the
Quoting r. Sasha Khapyorsky [EMAIL PROTECTED]:
Subject: [PATCH] diags/saquery: fix node_desc.description as string usages
node_desc.description buffer is received from the network and should
not be NULL-terminated. In such cases using it as regular string in
functions like strcmp() or
-by: Michael S. Tsirkin [EMAIL PROTECTED]
---
Roland, the offsets look wrong. Am I missing something?
Please review vs the spec, and if correct consider for 2.6.19.
diff --git a/drivers/infiniband/hw/mthca/mthca_cmd.c
b/drivers/infiniband/hw/mthca/mthca_cmd.c
index 99a94d7..768df72 100644
Quoting r. Roland Dreier [EMAIL PROTECTED]:
Subject: Re: [PATCH] IB/mthca: fix MAD extended header format
Roland, the offsets look wrong. Am I missing something?
Please review vs the spec, and if correct consider for 2.6.19.
I just applied your original patch without checking it
Quoting r. Roland Dreier [EMAIL PROTECTED]:
Subject: Re: [PATCH/RFC 2/2] IPoIB: Convert to NAPI
#of sockets : 124 8 16
NONAPI BW :276 297 304 306 303
NAPI BW :302 311 349 408 367
OK, the NAPI numbers are than the NONAPI numbers, which I guess is
Quoting r. Roland Dreier [EMAIL PROTECTED]:
Subject: Re: creating releases for the libraries you own
How about allowing web server serve static pages from
/pub/html and ~user/html? Then anyone with an account
can easily upload files by scp, and post a link.
I think that solves the
Quoting r. Roland Dreier [EMAIL PROTECTED]:
Subject: Re: convention for git directories on the new git OFA server
1. Development trees: each one will place the development tree under
his/her home:
~username/scm/topic.git (e.g. ~mst/scm/sdp.git)
2. Stable/release
Quoting r. Sasha Khapyorsky [EMAIL PROTECTED]:
Subject: Re: [PATCH] diags/saquery: fix node_desc.description as string usages
On 13:44 Mon 30 Oct , Michael S. Tsirkin wrote:
Quoting r. Sasha Khapyorsky [EMAIL PROTECTED]:
Subject: [PATCH] diags/saquery: fix node_desc.description
Quoting r. Roland Dreier [EMAIL PROTECTED]:
Subject: Re: creating releases for the libraries you own
Well, with ~user/html I can easily cook up a perl script to generate
a MD5
checksums or sign stuff and just stick them in the same directory as
original files.
But you
Quoting r. Roland Dreier [EMAIL PROTECTED]:
Subject: Re: convention for git directories on the new git OFA server
OK, good idea. Same for OFED I guess - we can have e.g. ~ofed/ if
several people need to share the same ofed trees.
So it seems it is best to stick with ~user/
Quoting r. Sasha Khapyorsky [EMAIL PROTECTED]:
Subject: Re: [PATCH] diags/saquery: fix node_desc.description as string usages
On 06:30 Tue 31 Oct , Michael S. Tsirkin wrote:
Quoting r. Sasha Khapyorsky [EMAIL PROTECTED]:
Subject: Re: [PATCH] diags/saquery: fix node_desc.description
Quoting r. Roland Dreier [EMAIL PROTECTED]:
Subject: Re: convention for git directories on the new git OFA server
So, what do you suggest for openfabrics? Could you write it up?
Actually, ~user is probably OK. If we want to do ~ofed then we just
have to create an ofed user, and a group
Quoting r. Michael S. Tsirkin [EMAIL PROTECTED]:
Subject: Re: convention for git directories on the new git OFA server
Quoting r. Roland Dreier [EMAIL PROTECTED]:
Subject: Re: convention for git directories on the new git OFA server
So, what do you suggest for openfabrics? Could you
Quoting r. Roland Dreier [EMAIL PROTECTED]:
Subject: Re: [PATCH/RFC 2/2] IPoIB: Convert to NAPI
Convert the IP-over-InfiniBand network device driver over to using
NAPI to handle all completions (both receive and send).
Signed-off-by: Roland Dreier [EMAIL PROTECTED]
OK, I finally found
Quoting r. Matt Leininger [EMAIL PROTECTED]:
Subject: Re: creating releases for the libraries you own
On Sun, 2006-10-29 at 12:52 +0200, Tziporet Koren wrote:
Roland Dreier wrote:
I want to suggest that you will create releases to the libraries you
own
To make this simpler,
Quoting r. Sasha Khapyorsky [EMAIL PROTECTED]:
I even think we should replace --base-path with --user-path=scm.
As it is, we have e.g.
/pub/scm/management.git - /home/sashak/management.git
which is just confusing.
It is accessable just as git://staging.openfabrics.org/management ,
Quoting r. Sasha Khapyorsky [EMAIL PROTECTED]:
I think it is best to be running with --user-path=scm so that I can
put git trees under scm sub-directory and have that part exported.
'--user-path=scm' in addition to '--base-path'? Looks good for me.
Could this be done please?
I even think
OK, I tested this a bit.
The machine and the inet connection seem very fast, thanks for that!
An issue I've run into:
[EMAIL PROTECTED]:~$ git clone -n
'git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git'
fatal: read error (Connection reset by peer)
fetch-pack from
FYI.
Of interest to openib is the change fixing backwards-compatibility -
this might help people running old git clients.
--
MST
---BeginMessage---
The latest maintenance release GIT 1.4.3.2 is available at the
usual places:
http://www.kernel.org/pub/software/scm/git/
Quoting r. Michael S. Tsirkin [EMAIL PROTECTED]:
Subject: Re: new server up and running
OK, I tested this a bit.
The machine and the inet connection seem very fast, thanks for that!
An issue I've run into:
[EMAIL PROTECTED]:~$ git clone -n
'git://git.kernel.org/pub/scm/linux/kernel/git
Quoting r. Matthew Wilcox [EMAIL PROTECTED]:
Subject: Re: Ordering between PCI config space writes and MMIO reads?
On Wed, Oct 25, 2006 at 12:30:22AM -0600, Grant Grundler wrote:
Can someone provide a quote of the PCI Local bus spec that allows this?
(Or at least a reference to a spec
Quoting r. Sean Hefty [EMAIL PROTECTED]:
Subject: [PATCH] librdmacm: updated librdmacm to work with proposed 2.6.20
kernel CMA
Updates the librdmacm to work with ABI version 3, which is the proposed
kernel changes for inclusion in 2.6.20.
Test programs are also updated.
Signed-off-by:
Quoting r. Jason Gunthorpe [EMAIL PROTECTED]:
After thinking about it some more, there are more cases in the kernel
than just the one in mthca. For instance there are paths in the pci
core than manipulate BAR registers without a barrier. It would be bad
to change a bar via config write and
Quoting vishal [EMAIL PROTECTED]:
Thread model: posix
gcc version 4.1.0 (SUSE Linux)
configure:2471: $? = 0
configure:2473: gcc -V /dev/null 5
gcc: '-V' option must have argument
configure:2476: $? = 1
configure:2499: checking for C compiler default output file name
configure:2502: gcc
Quoting r. Parks Fields [EMAIL PROTECTED]:
Subject: Kernel for IPoIB Question
Sincethe to[pic of kernel affecting the performance, which kernel is
now giving the best performance. We are trying to choose the next
kernel for our clusters and wanted to choose one that does well with
Quoting r. Scott Weitzenkamp (sweitzen) [EMAIL PROTECTED]:
Netperf (www.netperf.org) is a good tool to measure IPoIB performance.
Of special note is the -T flag which often lets you get more consistent results
by pinning the test to a single CPU.
Another useful tool is iperf, which has a -P
Quoting r. Sean Hubbell [EMAIL PROTECTED]:
Subject: Re: IPoIB Question
Scott,
Thanks for the reply again. The third party api that we use leverages a
combination of UDP and TCP socket conntections for speed. Is there
something for UCP as well?
iperf supports UDP as well. Again, check
Quoting r. Sean Hubbell [EMAIL PROTECTED]:
I am looking at libsdp for the TCP funcationality and wanted to know if
libsdp supports UDP as well
AFAIK, SDP can only emulate TCP sockets.
--
MST
___
openib-general mailing list
Quoting r. Bala [EMAIL PROTECTED]:
Subject: Is there a way to recover Mellanox HCA card from wrong Firmware
Hi All,
while updating the firmware in a set of
nodes, accidentlly updated one of the Mellanox
PCI-X card with
fw-25218-5_1_400-MHEL-CF128-T-MEMFREE.bin.zip
(PCI-Express
Quoting r. Jeff Squyres [EMAIL PROTECTED]:
Subject: Re: [openib-general] OFED 1.1 - Official Release
Tziporet --
Is OFED 1.1 going to be listed on www.openib.org? I see that the
Downloads section still lists all the OFED 1.0 stuff.
I don't think we have access to www.openib.org.
That's
Quoting r. Shirley Ma [EMAIL PROTECTED]:
Subject: Re: [PATCH/RFC 1/2] IB: Return maybe_missed_event hint from
ib_req_notify_cq()
Michael S. Tsirkin [EMAIL PROTECTED] wrote on 10/19/2006 01:21:45 PM:
Please also note that due to factors such as TCP window limits, TX on a
single
Quoting r. Sean Hefty [EMAIL PROTECTED]:
* Added client register/unregister routines.
BTW, we want this stuff for CM an CMA as well, don't we?
Could be nice to have that module unloading race in CM and CMA closed
for 2.6.19.
--
MST
___
Quoting r. Shirley Ma [EMAIL PROTECTED]:
Subject: Re: [openfabrics-ewg] [openib-general] RHEL5 and OFED ...
Roland Dreier [EMAIL PROTECTED] wrote on 10/19/2006 09:19:14 AM:
Shirley How can RHEL5 pick up this particular patch? Applications
Shirley with fork() depend on this patch.
Quoting r. Roland Dreier [EMAIL PROTECTED]:
If all you care about is system()/fork()+exec(), then be aware that
RHEL5 has a new enough kernel so that will work without doing anything.
As was pointed out, its really system/fork+wait.
--
MST
___
Quoting r. Doug Ledford [EMAIL PROTECTED]:
Subject: Re: [openfabrics-ewg] RHEL5 and OFED ...
On Thu, 2006-10-19 at 07:09 +0200, Michael S. Tsirkin wrote:
Quoting r. Doug Ledford [EMAIL PROTECTED]:
Subject: Re: [openfabrics-ewg] RHEL5 and OFED ...
On Wed, 2006-10-18 at 09:29 +0200
Quoting r. Doug Ledford [EMAIL PROTECTED]:
Subject: Re: [openfabrics-ewg] RHEL5 and OFED ...
On Thu, 2006-10-19 at 07:09 +0200, Michael S. Tsirkin wrote:
Yeah, I'll look over the diff today.
How does it look?
Not too far in yet, but the srp_topspin patch in the
kernel_patches
From: Ishai Rabinovitz [EMAIL PROTECTED]
This makes SRP more robust in presence of hardware errors
and is closer to behaviour suggested by IB spec,
reducing chance of stale packets.
Signed-off-by: Ishai Rabinovitz [EMAIL PROTECTED]
Signed-off-by: Michael S. Tsirkin [EMAIL PROTECTED]
---
Hello
Quoting r. Shirley Ma [EMAIL PROTECTED]:
Subject: Re: [PATCH/RFC 1/2] IB: Return maybe_missed_event hint from
ib_req_notify_cq()
Roland Dreier [EMAIL PROTECTED] wrote on 10/18/2006 01:55:13 PM:
I would like to understand why there's a throughput difference with
scaling turned off, since
Quoting r. Or Gerlitz [EMAIL PROTECTED]:
Subject: Re: OFED-1.1-pre1 is ready
Tziporet Koren wrote:
OFED 1.1-pre1 is available:
URL:
https://openib.org/svn/gen2/branches/1.1/ofed/releases/OFED-1.1-pre1.tgz
Release details:
BUILD_ID:
OFED-1.1-pre1
openib-1.1 (REV=9854)
#
Quoting r. Doug Ledford [EMAIL PROTECTED]:
Subject: Re: [openfabrics-ewg] RHEL5 and OFED ...
On Sun, 2006-10-15 at 12:13 -0400, Doug Ledford wrote:
Now for userspace - does RHEL5 include at least libibverbs-1.0?
This has been released a while back, and Roland makes regular bugfix
Quoting r. Doug Ledford [EMAIL PROTECTED]:
From our dicussion, it seems we should be able to just push the
small number of missing bits into RHEL5 directly. That would be
nicer of course.
It depends. If there's lots of individual changes, it might be easier
to push the OFED 1.1 change.
Quoting r. Or Gerlitz [EMAIL PROTECTED]:
Try something like
git log v2.6.18-rc6..936b9fc0bd1411b52826213a5d89e2ceb4f52a78
to get the list of OFED changes against v2.6.18-rc6.
thanks for all the info, however i think the OFED docs must state on
what upstream version are the OFED kernel
Quoting r. Doug Ledford [EMAIL PROTECTED]:
Hmm, no, I really want to take a srpm from amd64 and get a 32 bit
gcc executable that will build 64 bit binaries that match these
built on native amd64 system exectly.
Between just i386 and x86_64, you might be able to do that.
I guess what I
Quoting r. Jeff Squyres [EMAIL PROTECTED]:
However, this is a larger audience than was on the call. Is there a
significant movement here from the developers to move to 100% git?
Life would be somewhat easier for me with 100% git.
--
MST
___
Quoting r. Jeff Squyres [EMAIL PROTECTED]:
Does anyone have any sysadmin cycles to do this kind of stuff? I
would expect it to be a flurry of activity here in the beginning
followed by short bursts of activity separated by long periods of
nothing.
FWIW, I can help out keeping the git
Quoting r. Hoang-Nam Nguyen [EMAIL PROTECTED]:
Subject: [PATCH OFED-1.1-rc7] libehca configure: fix missing check of libsysfs
Hello,
here is the patch of configure in libehca as a result of the patch
libehca configure.in and config.h.in. It is generated by autogen.sh
and pretty lengthy.
Quoting r. Ira Weiny [EMAIL PROTECTED]:
Subject: Catastrophic error detected.
I got the following error running with OFED 1.1 on a modified 2.6.9 RHEL4
kernel. Hal mentioned that there might be a catastrophic error recovery patch
submitted since then? I can't find a mention of that in the
Quoting r. Roland Dreier [EMAIL PROTECTED]:
could have each plugin drop a file in /etc/libibverbs.conf.d/ with the
name -- something like
OK, feature request time :)
Hopefully that is under prefix:
$prefix/etc/libibverbs.conf.d/
and I think an environment with a list of additional directories
Quoting r. Roland Dreier [EMAIL PROTECTED]:
Hopefully that is under prefix:
$prefix/etc/libibverbs.conf.d/
Well, $sysconfdir/libibverbs.conf.d
Ugh, is that a problem if I want to build and run as non-root?
I'm used to be able to set --prefix on config line for all libs
to some directory,
Quoting r. Roland Dreier [EMAIL PROTECTED]:
Subject: Re: [PATCH] use mmiowb after doorbell ring
Hopefully that is under prefix:
$prefix/etc/libibverbs.conf.d/
Well, $sysconfdir/libibverbs.conf.d
Ugh, is that a problem if I want to build and run as non-root?
I'm used
Quoting r. Roland Dreier [EMAIL PROTECTED]:
libraries don't stick anything in home directories -- I'm just
suggesting $HOME/.libibverbs.conf as a place to stick extra configs
that users might want to add.
I'm kind of thinking that we might want other config options beyond
just driver names
Quoting r. Roland Dreier [EMAIL PROTECTED]:
Subject: Re: [PATCH] use mmiowb after doorbell ring
I dunno what's better. Maybe separate environment variables for
user-specific configs are just as good -- eg that's what ld.so does.
Hmm.
I guess what I'm trying to say is - let's
Quoting Roland Dreier [EMAIL PROTECTED]:
I think the real question is whether we expect to have complex config
options that would be hard to stick in an environment variable. At
this point I suspect not, so I guess I'll go with $sysconfdir/libibverbs.d
and $IBV_DRIVERS.
OK, both make sense.
Quoting r. Shirley Ma [EMAIL PROTECTED]:
/* can be added later once ipoib support sg
.get_sg = ethtool_op_get_sg,
.set_sg = ethtool_op_set_sg,
*/
The difficulty here is that sg currently requires checksum offloading in
netdevice.
--
MST
___
Quoting r. Doug Ledford [EMAIL PROTECTED]:
Subject: Re: RHEL5 and OFED ...
On Sat, 2006-10-14 at 22:14 +0200, Michael S. Tsirkin wrote:
Quoting r. Doug Ledford [EMAIL PROTECTED]:
Sorry. RHEL5 Beta1 has been out for a while, but OFED 1.1 still isn't
done yet. Obviously, I wasn't able
Quoting Jeff Squyres [EMAIL PROTECTED]:
- Some version of git and svn are installed on the new server, but
that's about it
The tool versions installed on openib are ancient.
Can site admins please install latest svn and git versions from source?
--
MST
Quoting r. Jeff Squyres [EMAIL PROTECTED]:
It seems like trac can integrate with both SVN and git and would also
provide us with integrated wiki capabilities.
One feature that bugzilla has (and that seems to be disabled in openib bugzilla
:() is mail integration, where I can Cc bugzilla and
Quoting Michael Newton [EMAIL PROTECTED]:
Also, these counts do not wrap: they peg at all 1s.
At infiniband speeds, these counts can peg out very quickly indeed,
to the point they can really only be of use if they can be reset each time
there read.
I mostly use them to verify network health.
Quoting r. [EMAIL PROTECTED] [EMAIL PROTECTED]:
From the OFED release notes:
3. Fork support from kernel 2.6.12 and above is available provided that
applications do not use threads. Only system() or fork() and immediate
exec() are supported.
system() works because parent calls
Quoting r. Roland Dreier [EMAIL PROTECTED]:
Subject: Re: [openib-general] Tools for development
Michael The tool versions installed on openib are ancient. Can
Michael site admins please install latest svn and git versions
Michael from source?
What distro is on the new
Quoting r. Doug Ledford [EMAIL PROTECTED]:
Subject: Re: RHEL5 and OFED ...
On Tue, 2006-10-17 at 15:35 +0200, Michael S. Tsirkin wrote:
Quoting r. Doug Ledford [EMAIL PROTECTED]:
Subject: Re: RHEL5 and OFED ...
On Sat, 2006-10-14 at 22:14 +0200, Michael S. Tsirkin wrote
Quoting r. Roland Dreier [EMAIL PROTECTED]:
Subject: Re: [PATCH] use mmiowb after doorbell ring
OK, here's what I actually put in my tree. Can you eyeball this and
maybe give it a quick test? If it looks good to you, I'll send it on
to the stable team for 2.6.18.x.
BTW, something like
Quoting r. Shirley Ma [EMAIL PROTECTED]:
/* can be added later once ipoib support sg
.get_sg = ethtool_op_get_sg,
.set_sg = ethtool_op_set_sg,
*/
The difficulty here is that sg currently requires checksum offloading in
netdevice.
I read the discussion in net-dev.
Hmm, any
Quoting r. Matt Leininger [EMAIL PROTECTED]:
Developers had requested git 1.4, but Ubuntu had an older version. We
went ahead and installed git from source. I'd prefer to stick to Ubuntu
packages if possible.
We have much to gain from newer versions - just look at gitweb change log.
But my
Quoting r. [EMAIL PROTECTED] [EMAIL PROTECTED]:
Subject: Re: [openfabrics-ewg] OFED 1.1 RC7 fork() issue.
On Tue, Oct 17, 2006 at 10:34:23AM -0500, Tang, Changqing wrote:
3. Fork support from kernel 2.6.12 and above is available
provided that applications do not use threads. The fork()
Quoting r. Sean Hefty [EMAIL PROTECTED]:
Subject: Re: OFED 1.1 release schedule
Tziporet Koren wrote:
I checked it and saw that the patch is applied, but since in the patch
Sean put the cm_issue_drep as a static, thus nm does not show it.
from the patch: +static int cm_issue_drep(struct
Quoting r. Roland Dreier [EMAIL PROTECTED]:
Subject: Re: [PATCH] use mmiowb after doorbell ring
Michael BTW, something like this will be needed for userspace too?
Ugh, I forgot about that.
I don't think an mmiowb() equivalent is available from userspace.
Isn't this just an asm()
Quoting r. Roland Dreier [EMAIL PROTECTED]:
No, it's never a good idea to turn off TCP or IP checksums. That
leads to possibilities of silent data corruption too easily.
never is probably too strong a word - hardware checksum offloading
turns off checksumming in software, moving that to
Quoting r. Roland Dreier [EMAIL PROTECTED]:
Subject: Re: [PATCH] use mmiowb after doorbell ring
I don't think an mmiowb() equivalent is available from userspace.
Isn't this just an asm() command?
Nope, look at the kernel source, specifically arch/ia64/sn/kernel/iomv.c
BTW, I
Quoting r. Roland Dreier [EMAIL PROTECTED]:
Subject: Re: [PATCH] use mmiowb after doorbell ring
Michael kernel code does rmb rather than mb there.
OK, but that's an optimization rather than a correctness issue: mb is
stronger than rmb.
Very strange. Let's consider amd64: libibverbs
Quoting Doug Ledford [EMAIL PROTECTED]:
Evidently, I was mistaken and rhn is still populated with the beta1
rpms. So, I've made the latest kernel available on my web page as
referenced below (amongst other rpms as well). However, it may still be
a while before the rpms are fully populated as
On a tangent, is there a way to set up a cross-build environment that will
build kernel modules for e.g. RHEL amd64 kernel on a 32 bit machine?
I'm doing this now with gcc and kernel.org kernel I built myself from source.
I guess I mostly need to get gcc and binutils SRPMs to generate
Quoting r. Roland Dreier [EMAIL PROTECTED]:
But of course not all x86 processors support lfence/mfence
True, but I dont think anyone us still running libibverbs on processors that
don't. What happens if an older processors when you call lfence?
--
MST
Quoting r. Roland Dreier [EMAIL PROTECTED]:
Another confusing thing is that asm-i386 defines mb() and rmb() just
to be compiler barriers,
I see:
#define rmb() alternative(lock; addl $0,0(%%esp), lfence, X86_FEATURE_XMM2)
as for mb() - I don't thnk our kernel code uses that so I think userspace
Quoting r. Roland Dreier [EMAIL PROTECTED]:
But of course not all x86 processors
support lfence/mfence which leads to some ugly issues of how to handle
this
lfence seems to be part of SSE2,
and I don't think we really need sfence/mfence.
We can just require SSE2 support:
Quoting r. Doug Ledford [EMAIL PROTECTED]:
Subject: Re: RHEL5 and OFED ...
On Tue, 2006-10-17 at 22:28 +0200, Michael S. Tsirkin wrote:
On a tangent, is there a way to set up a cross-build environment that will
build kernel modules for e.g. RHEL amd64 kernel on a 32 bit machine?
I'm
Quoting r. Roland Dreier [EMAIL PROTECTED]:
as for mb() - I don't thnk our kernel code uses that so I think userspace
should switch to wmb as well. wmb isjust a compiler barrier on most
arhitectures.
I'm not sure it's worth the trouble to split up the two cases at this
point.
Quoting r. Roland Dreier [EMAIL PROTECTED]:
Detecting SSE2 is easy -- we could just do the cpuid ourselves if we
wanted to. The problem is what do you do when you see that the CPU
does or doesn't have the instruction? The runtime patching that the
kernel does is way too complicated, and if
Quoting r. Doug Ledford [EMAIL PROTECTED]:
Far easier would be to go the other way around,
run on x86_64 and build for i386, in which case gcc supports that out of
the box.
All that's left is to convince Lenovo there's a market for x86_64
thinkpads.
--
MST
Quoting r. Roland Dreier [EMAIL PROTECTED]:
Subject: Re: [PATCH] use mmiowb after doorbell ring
Roland For now I just used lock; addl %0 to implement rmb on
Roland i386. I'm really not comfortable making libmthca depend
Roland on sse2, and I don't see a good way to detect and
Quoting r. Jason Gunthorpe [EMAIL PROTECTED]:
Subject: Re: [PATCH] use mmiowb after doorbell ring
On Tue, Oct 17, 2006 at 08:44:34PM -0700, Roland Dreier wrote:
Jason I think the typical way this is done would be to use
Jason ld.so's 'hwcap' handling and stick an optimized library
Quoting r. Doug Ledford [EMAIL PROTECTED]:
Subject: Re: RHEL5 and OFED ...
On Wed, 2006-10-18 at 06:01 +0200, Michael S. Tsirkin wrote:
Quoting r. Doug Ledford [EMAIL PROTECTED]:
Far easier would be to go the other way around,
run on x86_64 and build for i386, in which case gcc
Quoting r. Maestas, Christopher Daniel [EMAIL PROTECTED]:
Subject: Re: [openib-general] RHEL5 and OFED ...
Now for userspace - does RHEL5 include at least libibverbs-1.0?
This has been released a while back, and Roland makes regular bugfix
releases.
Here's what I see on a rhel4 u4
Quoting r. Roland Dreier [EMAIL PROTECTED]:
Subject: Re: RHEL5 and OFED ...
Michael 1. The first thing would be to list fixes between 2.6.18
Michael and 2.6.19-rc1 and backport these. Some of them are in
Michael OFED.
If you want to do this, I think it would be great to also
Quoting r. Doug Ledford [EMAIL PROTECTED]:
1. The first thing would be to list fixes between 2.6.18 and 2.6.19-rc1 and
backport these. Some of them are in OFED.
That would be helpful. Since 2.6.19-rc looks to have integrated the
iWARP merge, the fixes are no doubt mixed in with a bunch
Quoting r. Roland Dreier [EMAIL PROTECTED]:
There are further changes I would like to add on top of that, but
comments on the two patches there would be appreciated.
A small optimization:
if (missed_event netif_rx_reschedule(dev, 0))
should be, I think
if
Quoting r. Doug Ledford [EMAIL PROTECTED]:
Dough, would it be possible to update this + libmthca?
Possibly. What's the justification? What's in 1.0.4 that is the
primary reason for wanting to update from 1.0.3?
Hmm, Roland, I went to look into Changelog and I note that you don't
label svn
Quoting r. Krishna Kumar [EMAIL PROTECTED]:
Subject: [PATCH] rdma_bind_addr() leaks a cma_dev reference count
rdma_bind_addr leaks a cma_dev reference count in failure
case.
Signed-off-by: Krishna Kumar [EMAIL PROTECTED]
diff -ruNp org/drivers/infiniband/core/cma.c
Quoting r. Roland Dreier [EMAIL PROTECTED]:
I do put the releases into the changelog. There has not been a
libibverbs 1.0.4 release yet, but you can find the entry
2006-05-02 Roland Dreier [EMAIL PROTECTED]
* Release version 1.0.3.
in the libibverbs ChangeLog. Everything after
401 - 500 of 2913 matches
Mail list logo