tag:old.nabble.com,2006:forum-3273Nabble - OpenSSI Clusters for Linux2009-11-20T10:39:29ZThe OpenSSI project is a comprehensive clustering solution offering a full, highly available SSI environment for Linux. Goals for OpenSSI Clusters include availability, scalability and manageability, built from standard servers. OpenSSI Clusters for Linux home is <a href="http://sourceforge.net/projects/ssic-linux/" target="_top" rel="nofollow">here</a>.tag:old.nabble.com,2006:post-26448346OpenSSI on CentOS 5.42009-11-20T10:39:29Z2009-11-20T10:39:29ZDaniel BahenaHi, I was looking around the wiki for OpenSSI and didn't find any
<br>information about somebody installing OpenSSI on CentOS 5.4
<br><br>I downloaded openssi-fc3-1.9.2.i686.tar.gz and when I run the install
<br>script I get a bunch of errors about missing some packages that I
<br>already have on a higher version. Any of you has some info on how to
<br>get around this?
<br><br>This is what I get
<br><br>[root@centos openssi-fc3-1.9.2.i686]# ./install
<br>This version of OpenSSI 1.9 is intended for a system
<br>installed with Fedora Core 3. You should run the install
<br>script on a fresh Fedora Core 3 installation, or download
<br>the correct version of OpenSSI 1.9 for your distribution.
<br><br>Versions for Debian and SuSE are also available.
<br><br>Would you like to install this version of OpenSSI anyway (y/n) [n]: y
<br><br>Checking for base packages...
<br> checking
<br> checking
<br> checking
<br> checking
<br> checking
<br> checking
<br> checking
<br> checking
<br> checking
<br> checking
<br> checking
<br> checking
<br> checking
<br> checking
<br> checking
<br> checking
<br> checking
<br><br>Checking for OpenSSI packages...
<br> checking
<br> checking
<br> checking
<br> checking
<br> checking
<br> checking
<br> checking
<br> checking
<br> checking
<br> checking
<br> checking
<br> checking
<br> checking
<br> checking
<br> checking
<br> checking
<br> checking
<br> checking
<br> checking
<br> checking
<br><br>warning: RPMS/base/dhcp-3.0.1-44_FC3.i386.rpm: Header V3 DSA
<br>signature: NOKEY, key ID 4f2a6fd2
<br>error: Failed dependencies:
<br> /usr/lib/python2.3 is needed by libxml2-python-2.6.16-3.i386
<br> beecrypt >= 3.0.1 is needed by rpm-4.3.2-21.i386
<br> libbeecrypt.so.6 is needed by rpm-4.3.2-21.i386
<br> librpm-4.3.so is needed by rpm-4.3.2-21.i386
<br> librpmbuild-4.3.so is needed by rpm-4.3.2-21.i386
<br> librpmdb-4.3.so is needed by rpm-4.3.2-21.i386
<br> librpmio-4.3.so is needed by rpm-4.3.2-21.i386
<br> popt = 1.9.1 is needed by rpm-4.3.2-21.i386
<br> initscripts >= 8.11.1-1 is needed by kernel-2.6.18-164.el5.i686
<br> initscripts < 8.40-1 conflicts with kudzu-1.2.57.1.21-1.el5.centos.i386
<br> nfs-utils < 1.0.7-12 conflicts with kernel-2.6.18-164.el5.i686
<br> libnewt.so.0.51 is needed by ntsysv-1.3.11.1-1_ssi_1devel.i386
<br> db4 = 4.2.52 is needed by pam-0.77-65_ssi_1devel.i386
<br> pam < 0.78 conflicts with setuptool-1.19.2-1.el5.centos.i386
<br> procps < 3.2.5-6.3 conflicts with kernel-2.6.18-164.el5.i686
<br> hotplug is needed by udev-039-10.FC3.7_ssi_1devel.i386
<br> udev < 063-6 conflicts with kernel-2.6.18-164.el5.i686
<br> libxml2 >= 2.6.25 is needed by (installed) libxslt-1.1.17-2.el5_2.2.i386
<br> rpm = 4.4.2.3-18.el5 is needed by (installed) rpm-libs-4.4.2.3-18.el5.i386
<br> rpm = 4.4.2.3-18.el5 is needed by (installed) rpm-python-4.4.2.3-18.el5.i386
<br> rpm >= 0:4.4.2 is needed by (installed) yum-3.2.22-20.el5.centos.noarch
<br> rpm = 4.4.2.3-18.el5 is needed by (installed) rpm-build-4.4.2.3-18.el5.i386
<br> chkconfig >= 1.3.26 is needed by (installed) dbus-glib-0.73-8.el5.i386
<br> chkconfig >= 1.3.26 is needed by (installed) dbus-1.1.2-12.el5.i386
<br> initscripts >= 8.11.1-1 is needed by (installed) kernel-2.6.18-164.el5.i686
<br> initscripts >= 8.04-1 is needed by (installed) hal-0.5.8.1-52.el5.i386
<br> mkinitrd >= 4.2.21-1 is needed by (installed) kernel-2.6.18-164.el5.i686
<br> libpam.so.0(LIBPAM_1.0) is needed by (installed) coreutils-5.97-23.el5.i386
<br> libpam.so.0(LIBPAM_1.0) is needed by (installed) ppp-2.4.4-2.el5.i386
<br> libpam.so.0(LIBPAM_1.0) is needed by (installed) libuser-0.54.7-2.el5.5.i386
<br> libpam.so.0(LIBPAM_1.0) is needed by (installed) passwd-0.73-1.i386
<br> libpam.so.0(LIBPAM_1.0) is needed by (installed) usermode-1.88-3.el5.2.i386
<br> libpam.so.0(LIBPAM_1.0) is needed by (installed) at-3.1.8-82.fc6.i386
<br> libpam.so.0(LIBPAM_1.0) is needed by (installed) vixie-cron-4.1-76.el5.i386
<br> libpam.so.0(LIBPAM_1.0) is needed by (installed) cyrus-sasl-2.1.22-5.el5.i386
<br> libpam.so.0(LIBPAM_1.0) is needed by (installed)
<br>krb5-workstation-1.6.1-36.el5.i386
<br> libpam.so.0(LIBPAM_1.0) is needed by (installed) nss_ldap-253-21.el5.i386
<br> libpam.so.0(LIBPAM_1.0) is needed by (installed) pam_pkcs11-0.5.3-23.i386
<br> libpam.so.0(LIBPAM_1.0) is needed by (installed) pam_ccreds-3-5.i386
<br> libpam.so.0(LIBPAM_1.0) is needed by (installed) pam_krb5-2.2.14-10.i386
<br> libpam.so.0(LIBPAM_1.0) is needed by (installed)
<br>policycoreutils-1.33.12-14.6.el5.i386
<br> libpam.so.0(LIBPAM_1.0) is needed by (installed) cups-1.3.7-11.el5.i386
<br> libpam.so.0(LIBPAM_1.0) is needed by (installed)
<br>openssh-server-4.3p2-36.el5.i386
<br> libpam.so.0(LIBPAM_1.0) is needed by (installed) cvs-1.11.22-7.el5.i386
<br> libpam.so.0(LIBPAM_1.0) is needed by (installed) screen-4.0.3-1.el5.i386
<br> libpam.so.0(LIBPAM_1.0) is needed by (installed) squid-2.6.STABLE21-3.el5.i386
<br> libpam.so.0(LIBPAM_EXTENSION_1.0) is needed by (installed)
<br>pam_pkcs11-0.5.3-23.i386
<br> libpam_misc.so.0(LIBPAM_MISC_1.0) is needed by (installed)
<br>coreutils-5.97-23.el5.i386
<br> libpam_misc.so.0(LIBPAM_MISC_1.0) is needed by (installed)
<br>libuser-0.54.7-2.el5.5.i386
<br> libpam_misc.so.0(LIBPAM_MISC_1.0) is needed by (installed) passwd-0.73-1.i386
<br> libpam_misc.so.0(LIBPAM_MISC_1.0) is needed by (installed)
<br>usermode-1.88-3.el5.2.i386
<br> libpam_misc.so.0(LIBPAM_MISC_1.0) is needed by (installed)
<br>policycoreutils-1.33.12-14.6.el5.i386
<br> pam >= 0.78 is needed by (installed) passwd-0.73-1.i386
<br> pam >= 0.99.6.2-3.27 is needed by (installed) authconfig-5.3.21-6.el5.i386
<br> pam >= 0.77-66.8 is needed by (installed) vixie-cron-4.1-76.el5.i386
<br> /lib/security/pam_loginuid.so is needed by (installed)
<br>openssh-server-4.3p2-36.el5.i386
<br> udev >= 062 is needed by (installed) pcmciautils-014-5.i386
<br> udev >= 039-10.14.EL4 is needed by (installed) isdn4k-utils-3.2-56.el5.i386
<br> udev >= 089-1 is needed by (installed) hal-0.5.8.1-52.el5.i386
<br>Failed to install packages
<br>[root@centos openssi-fc3-1.9.2.i686]#
<br><br>Any help will be greatly appreciated.
<br><br>Best regards,
<br>--
<br>Dan Bahena
<br><a href="http://twitter.com/informatux" target="_top" rel="nofollow">http://twitter.com/informatux</a><br><a href="http://xti.informatux.net" target="_top" rel="nofollow">http://xti.informatux.net</a><br><br>------------------------------------------------------------------------------
<br>Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
<br>trial. Simplify your report design, integration and deployment - and focus on
<br>what you do best, core application coding. Discover what's new with
<br>Crystal Reports now. <a href="http://p.sf.net/sfu/bobj-july" target="_top" rel="nofollow">http://p.sf.net/sfu/bobj-july</a><br>_______________________________________________
<br>Ssic-linux-users mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26448346&i=0" target="_top" rel="nofollow">Ssic-linux-users@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-users" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-users</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-users-f3276.html" embed="fixTarget[3276]" target="_top" >ssic-linux-users</a></p>tag:old.nabble.com,2006:post-26290095Re: OpenSSI 1.9.6 in the making2009-11-10T11:49:10Z2009-11-10T11:49:10ZBrian J. WatsonRoger, John, and the other active OpenSSI developers,
<br><br>Congratulations on all your hard work this year!
<br><br>Best regards,
<br>Brian
<br><div class='shrinkable-quote'><br>> -----Original Message-----
<br>> From: Roger Tsang [mailto:<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26290095&i=0" target="_top" rel="nofollow">roger.tsang@...</a>]
<br>> Sent: Tuesday, November 10, 2009 12:06 AM
<br>> To: ssic-linux-devel
<br>> Subject: [SSI-devel] OpenSSI 1.9.6 in the making
<br>>
<br>> Hi,
<br>>
<br>> These are just some thoughts I'd like to share.
<br>>
<br>> As you know the original plan was to release 2.0.0pre3 as 1.9.6 around
<br>> first half of year 2009. Early in the year we decided to rename the
<br>> release to 1.9.6 due to the amount of changes to the core. However
<br>> one criteria I had in mind for making this release was improved
<br>> stability from the previous release. At the time I did not feel we've
<br>> met this criteria. At the same time some people felt that porting to
<br>> 2.6.18 is more urgent, but I strongly believe the core should be
<br>> stabilized first to ensure we won't have to deal with outstanding core
<br>> bugs that could impact our porting effort.
<br>>
<br>> So we ended up focusing our effort fixing core bugs. At the same time
<br>> I took the opportunity to increase the performance of OpenSSI. Thanks
<br>> to John Hughes we also identified some useful regression tests. Today
<br>> most of the bugs in our tracker are closed or fixed. We also
<br>> completed more than a few OpenSSI enhancements / feature requests in
<br>> our tracker.
<br>>
<br>> During the course of the year we've seen almost 20,000 lines of code
<br>> (LoC) changes at the core. The net change is less but this is still
<br>> 5-10x more than the previous year. I haven't really looked up the
<br>> numbers; that's my impression. Worthy of mention in addition to
<br>> system latency performance improvements is the time and effort taken
<br>> to address the more obscure bugs only seen on multi-core / SMP
<br>> machines and preemptive kernels.
<br>>
<br>> Though rarely mentioned I believe one of the major technical obstacles
<br>> to reaching stable was OpenSSI-1.9.x had been suffering the same class
<br>> of problems the openMosix project (now defunct) had long struggled
<br>> with and that is SMP bugs. If I'm not wrong the developers who have
<br>> forked openMosix are still dealing with these problems. So I am
<br>> thankful that most of our OpenSSI developers and designers kept this
<br>> in mind when porting to Linux and am grateful that we are able to
<br>> overcome the remaining issues considering our OpenSSI code base is
<br>> much larger than openMosix and considering we've integrated some
<br>> openMosix code (eek!).
<br>>
<br>>
<br>> Regards,
<br>> Roger
<br>>
<br>> -----------------------------------------------------------------------
<br>> -------
<br>> Let Crystal Reports handle the reporting - Free Crystal Reports 2008
<br>> 30-Day
<br>> trial. Simplify your report design, integration and deployment - and
<br>> focus on
<br>> what you do best, core application coding. Discover what's new with
<br>> Crystal Reports now. <a href="http://p.sf.net/sfu/bobj-july" target="_top" rel="nofollow">http://p.sf.net/sfu/bobj-july</a><br>> _______________________________________________
<br>> ssic-linux-devel mailing list
<br>> <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26290095&i=1" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br>> <a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a></div><br>------------------------------------------------------------------------------
<br>Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
<br>trial. Simplify your report design, integration and deployment - and focus on
<br>what you do best, core application coding. Discover what's new with
<br>Crystal Reports now. <a href="http://p.sf.net/sfu/bobj-july" target="_top" rel="nofollow">http://p.sf.net/sfu/bobj-july</a><br>_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26290095&i=2" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-26279682OpenSSI 1.9.6 in the making2009-11-10T00:05:45Z2009-11-10T00:05:45ZRoger TsangHi,
<br><br>These are just some thoughts I'd like to share.
<br><br>As you know the original plan was to release 2.0.0pre3 as 1.9.6 around
<br>first half of year 2009. Early in the year we decided to rename the
<br>release to 1.9.6 due to the amount of changes to the core. However
<br>one criteria I had in mind for making this release was improved
<br>stability from the previous release. At the time I did not feel we've
<br>met this criteria. At the same time some people felt that porting to
<br>2.6.18 is more urgent, but I strongly believe the core should be
<br>stabilized first to ensure we won't have to deal with outstanding core
<br>bugs that could impact our porting effort.
<br><br>So we ended up focusing our effort fixing core bugs. At the same time
<br>I took the opportunity to increase the performance of OpenSSI. Thanks
<br>to John Hughes we also identified some useful regression tests. Today
<br>most of the bugs in our tracker are closed or fixed. We also
<br>completed more than a few OpenSSI enhancements / feature requests in
<br>our tracker.
<br><br>During the course of the year we've seen almost 20,000 lines of code
<br>(LoC) changes at the core. The net change is less but this is still
<br>5-10x more than the previous year. I haven't really looked up the
<br>numbers; that's my impression. Worthy of mention in addition to
<br>system latency performance improvements is the time and effort taken
<br>to address the more obscure bugs only seen on multi-core / SMP
<br>machines and preemptive kernels.
<br><br>Though rarely mentioned I believe one of the major technical obstacles
<br>to reaching stable was OpenSSI-1.9.x had been suffering the same class
<br>of problems the openMosix project (now defunct) had long struggled
<br>with and that is SMP bugs. If I'm not wrong the developers who have
<br>forked openMosix are still dealing with these problems. So I am
<br>thankful that most of our OpenSSI developers and designers kept this
<br>in mind when porting to Linux and am grateful that we are able to
<br>overcome the remaining issues considering our OpenSSI code base is
<br>much larger than openMosix and considering we've integrated some
<br>openMosix code (eek!).
<br><br><br>Regards,
<br>Roger
<br><br>------------------------------------------------------------------------------
<br>Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
<br>trial. Simplify your report design, integration and deployment - and focus on
<br>what you do best, core application coding. Discover what's new with
<br>Crystal Reports now. <a href="http://p.sf.net/sfu/bobj-july" target="_top" rel="nofollow">http://p.sf.net/sfu/bobj-july</a><br>_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26279682&i=0" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-26217171[PATCH] for 2.6.32-rc6 make ses enclosure links even if diagnostic page7 is absent2009-11-05T07:47:01Z2009-11-05T07:47:01ZJohn HughesUse the diagnostic page 10 information from SES enclosure devices even
<br>if page 7 (slot names) is missing. This allows creation of the
<br>enclosure/slot/device and device/enclosure links even if the enclosure
<br>doesn't name its slots.
<br><br>Signed-off-by: John Hughes <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26217171&i=0" target="_top" rel="nofollow">john@...</a>>
<br><br>diff --git a/drivers/scsi/ses.c b/drivers/scsi/ses.c
<br>index 55b034b..725a095 100644
<br>--- a/drivers/scsi/ses.c
<br>+++ b/drivers/scsi/ses.c
<br>@@ -389,9 +389,9 @@ static void ses_enclosure_data_process(struct enclosure_devi
<br> len = (desc_ptr[2] << 8) + desc_ptr[3];
<br> /* skip past overall descriptor */
<br> desc_ptr += len + 4;
<br>- if (ses_dev->page10)
<br>- addl_desc_ptr = ses_dev->page10 + 8;
<br> }
<br>+ if (ses_dev->page10)
<br>+ addl_desc_ptr = ses_dev->page10 + 8;
<br> type_ptr = ses_dev->page1 + 12 + ses_dev->page1[11];
<br> components = 0;
<br> for (i = 0; i < types; i++, type_ptr += 4) {
<br><br><br><br>------------------------------------------------------------------------------
<br>Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
<br>trial. Simplify your report design, integration and deployment - and focus on
<br>what you do best, core application coding. Discover what's new with
<br>Crystal Reports now. <a href="http://p.sf.net/sfu/bobj-july" target="_top" rel="nofollow">http://p.sf.net/sfu/bobj-july</a><br>_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26217171&i=1" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-26210503Re: VProc child-sibling list race2009-11-04T23:54:08Z2009-11-04T23:54:08ZRoger TsangIn the latest CVS there is a regression in switch_exec_pids() that would eventually cause an oops decrementing reference count on an already released vproc at vpop_wait() while running glibc tst-exec4 from ssic-linux-bugs-2000692. I checked in the fix.<br>
<br><div class="gmail_quote">On Wed, Nov 4, 2009 at 6:18 AM, Roger Tsang <span dir="ltr"><<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26210503&i=0" target="_top" rel="nofollow">roger.tsang@...</a>></span> wrote:<br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
Hi,<br><br>This bug is either staring right at me or I am just having a bad hair day. In vpop_wait() we see VPROC LOCK is used to protect its parent-child-sibling list. However the ND LIST lock doesn't protect PVP(vo)->pvp_childl from other threads. Also if I am not wrong the for-loop in vpop_wait() is not SMP-safe since PVP(vc)->pvp_childl is neither protected by VPROC LOCK nor ND LIST lock.<br>
<br>-Roger<br><br><snip><br>
VPROC_LOCK_EXCL(vp, "vpop_wait");<br><snip><br> for (vo = NULL, vc = pvp->pvp_head_childl;<br> vc != NULL;<br> vo = vc, vc = PVP(vc)->pvp_childl) {<br>
<snip><br> VPROC_LOCK_ND_LIST_EXCL(vp, "vpop_wait");<br> if (vo == NULL)<br> pvp->pvp_head_childl = pvc->pvp_childl;<br>
else<br> PVP(vo)->pvp_childl = pvc->pvp_childl;<br> rmb();<br> pvc->pvp_childl = NULL;<br>
VPROC_UNLOCK_ND_LIST_EXCL(vp, "vpop_wait");<br><snip><br>VPROC_UNLOCK_EXCL(vp, "vpop_wait");<br>
</blockquote></div><br>
<br />------------------------------------------------------------------------------
<br>Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
<br>trial. Simplify your report design, integration and deployment - and focus on
<br>what you do best, core application coding. Discover what's new with
<br>Crystal Reports now. <a href="http://p.sf.net/sfu/bobj-july" target="_top" rel="nofollow">http://p.sf.net/sfu/bobj-july</a><br />_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26210503&i=1" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-26191561VProc child-sibling list race2009-11-03T22:18:46Z2009-11-03T22:18:46ZRoger TsangHi,<br><br>This bug is either staring right at me or I am just having a bad hair day. In vpop_wait() we see VPROC LOCK is used to protect its parent-child-sibling list. However the ND LIST lock doesn't protect PVP(vo)->pvp_childl from other threads. Also if I am not wrong the for-loop in vpop_wait() is not SMP-safe since PVP(vc)->pvp_childl is neither protected by VPROC LOCK nor ND LIST lock.<br>
<br>-Roger<br><br><snip><br>
VPROC_LOCK_EXCL(vp, "vpop_wait");<br><snip><br> for (vo = NULL, vc = pvp->pvp_head_childl;<br> vc != NULL;<br> vo = vc, vc = PVP(vc)->pvp_childl) {<br>
<snip><br> VPROC_LOCK_ND_LIST_EXCL(vp, "vpop_wait");<br> if (vo == NULL)<br> pvp->pvp_head_childl = pvc->pvp_childl;<br>
else<br> PVP(vo)->pvp_childl = pvc->pvp_childl;<br> rmb();<br> pvc->pvp_childl = NULL;<br>
VPROC_UNLOCK_ND_LIST_EXCL(vp, "vpop_wait");<br><snip><br>VPROC_UNLOCK_EXCL(vp, "vpop_wait");<br>
<br />------------------------------------------------------------------------------
<br>Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
<br>trial. Simplify your report design, integration and deployment - and focus on
<br>what you do best, core application coding. Discover what's new with
<br>Crystal Reports now. <a href="http://p.sf.net/sfu/bobj-july" target="_top" rel="nofollow">http://p.sf.net/sfu/bobj-july</a><br />_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26191561&i=0" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-26071909[ ssic-linux-Bugs-2838006 ] Corrupt SHM_LOCK_DEST flag2009-10-26T21:11:14Z2009-10-26T21:11:14ZSourceForge.netBugs item #2838006, was opened at 2009-08-14 22:43
<br>Message generated for change (Settings changed) made by rogertsang
<br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2838006&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2838006&group_id=32541</a><br><br>Please note that this message will contain a full copy of the comment thread,
<br>including the initial issue submission, for this request,
<br>not just the latest update.
<br>Category: IPC
<br>Group: default
<br>Status: Open
<br>>Resolution: Accepted
<br>Priority: 5
<br>Private: No
<br>Submitted By: Roger Tsang (rogertsang)
<br>Assigned to: Roger Tsang (rogertsang)
<br>Summary: Corrupt SHM_LOCK_DEST flag
<br><br>Initial Comment:
<br>Reported by John Hughes.
<br><a href="http://article.gmane.org/gmane.linux.cluster.ssic.devel/6037" target="_top" rel="nofollow">http://article.gmane.org/gmane.linux.cluster.ssic.devel/6037</a><br><br>Current definition of SHM_LOCK_DEST flag is 023420 (octal)
<br><br>----------------------------------------------------------------------
<br><br>>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-10-27 00:11
<br><br>Message:
<br>please ignore previous comment. bug is not fixed.
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-10-26 23:47
<br><br>Message:
<br>checked-in
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-08-14 22:53
<br><br>Message:
<br>Further due to this flag the SHM node down code assumes shm_ids semaphore
<br>is held, but in fact is not always the case and could race causing memory
<br>corruption.
<br><br>----------------------------------------------------------------------
<br><br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2838006&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2838006&group_id=32541</a><br><br>------------------------------------------------------------------------------
<br>Come build with us! The BlackBerry(R) Developer Conference in SF, CA
<br>is the only developer event you need to attend this year. Jumpstart your
<br>developing skills, take BlackBerry mobile applications to market and stay
<br>ahead of the curve. Join us from November 9 - 12, 2009. Register now!
<br><a href="http://p.sf.net/sfu/devconference" target="_top" rel="nofollow">http://p.sf.net/sfu/devconference</a><br>_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26071909&i=0" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-26071828[ ssic-linux-Bugs-1925545 ] Processes stuck in I/O after transparent CFS failover2009-10-26T20:57:56Z2009-10-26T20:57:56ZSourceForge.netBugs item #1925545, was opened at 2008-03-25 17:35
<br>Message generated for change (Settings changed) made by rogertsang
<br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405834&aid=1925545&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405834&aid=1925545&group_id=32541</a><br><br>Please note that this message will contain a full copy of the comment thread,
<br>including the initial issue submission, for this request,
<br>not just the latest update.
<br>Category: Filesystem
<br>Group: v1.9.1
<br>Status: Open
<br>>Resolution: Fixed
<br>Priority: 5
<br>>Private: No
<br>Submitted By: Roger Tsang (rogertsang)
<br>Assigned to: Roger Tsang (rogertsang)
<br>Summary: Processes stuck in I/O after transparent CFS failover
<br><br>Initial Comment:
<br>Processes or threads doing I/O during transparent CFS failover get stuck waiting in I/O. They cannot be interrupted. If they are part of a thread group zombies can also appear.
<br><br>How to reproduce: Start reading a large file on a CFS hard mount on the surviving OpenSSI node and force CFS to transparently failover. The application on the surviving OpenSSI node is expected to continue uninterrupted and finish reading the entire file, but instead gets stuck waiting in I/O.
<br><br>----------------------------------------------------------------------
<br><br>>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-10-26 23:57
<br><br>Message:
<br>checked-in final fix
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2008-07-03 23:21
<br><br>Message:
<br>Logged In: YES
<br>user_id=1246761
<br>Originator: YES
<br><br>Fixed but not yet in code repository.
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2008-03-25 18:04
<br><br>Message:
<br>Logged In: YES
<br>user_id=1246761
<br>Originator: YES
<br><br>NB: This bug might not manifest in the earlier 1.9.x releases because the
<br>CFS rebuild thread is RT before (around) 1.9.3.
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2008-03-25 17:53
<br><br>Message:
<br>Logged In: YES
<br>user_id=1246761
<br>Originator: YES
<br><br>This is a CFS super block rebuild flush race with CFS async code where
<br>down requests being flushed are asynchronously pushed back into the list of
<br>down requests. These requests have the page lock.
<br><br>Fixed in 2.0.0pre3.
<br><br>----------------------------------------------------------------------
<br><br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405834&aid=1925545&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405834&aid=1925545&group_id=32541</a><br><br>------------------------------------------------------------------------------
<br>Come build with us! The BlackBerry(R) Developer Conference in SF, CA
<br>is the only developer event you need to attend this year. Jumpstart your
<br>developing skills, take BlackBerry mobile applications to market and stay
<br>ahead of the curve. Join us from November 9 - 12, 2009. Register now!
<br><a href="http://p.sf.net/sfu/devconference" target="_top" rel="nofollow">http://p.sf.net/sfu/devconference</a><br>_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26071828&i=0" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-26071809[ ssic-linux-Bugs-1718329 ] Problem with migration of process using tcp socket2009-10-26T20:53:50Z2009-10-26T20:53:50ZSourceForge.netBugs item #1718329, was opened at 2007-05-13 17:25
<br>Message generated for change (Comment added) made by rogertsang
<br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405834&aid=1718329&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405834&aid=1718329&group_id=32541</a><br><br>Please note that this message will contain a full copy of the comment thread,
<br>including the initial issue submission, for this request,
<br>not just the latest update.
<br>Category: Process Management
<br>Group: v1.9.1
<br>Status: Open
<br>>Resolution: Fixed
<br>Priority: 5
<br>Private: No
<br>Submitted By: ch0hlik (ch0hlik)
<br>Assigned to: Roger Tsang (rogertsang)
<br>Summary: Problem with migration of process using tcp socket
<br><br>Initial Comment:
<br>For example I use this little tcp server:
<br><a href="http://www.cs.rpi.edu/courses/sysprog/sockets/server.c" target="_top" rel="nofollow">http://www.cs.rpi.edu/courses/sysprog/sockets/server.c</a><br>$ gcc -o server server.c
<br><br>On the first node 10.0.0.1:
<br>$ ./server 8888
<br><br>$ migrate 2 server_pid
<br><br>On the second node 10.0.0.2:
<br>$ telnet 10.0.0.1 8888
<br>Trying 10.0.0.1...
<br>Connected to openone (10.0.0.1).
<br>Escape character is '^]'.
<br>Connection closed by foreign host.
<br><br>After this server show:
<br>ERROR reading from socket: Invalid argument
<br><br>In 1.2.2 I'm the same steps and it works fine.
<br>In 1.9.3 problem looks just like in 1.9.2 or worse. (1.9.3 from here: <a href="http://radian.org/~roger/OPENSSI-FC/kernel/2.6.11/" target="_top" rel="nofollow">http://radian.org/~roger/OPENSSI-FC/kernel/2.6.11/</a>)
<br><br>----------------------------------------------------------------------
<br><br>>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-10-26 23:53
<br><br>Message:
<br>checked-in fix
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-07-04 17:11
<br><br>Message:
<br>Finally got around to this... kernel-2.6 socket_file_ops no longer provide
<br>read() file operation that the existing OpenSSI code depends on.
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2007-12-06 02:37
<br><br>Message:
<br>Logged In: YES
<br>user_id=1246761
<br>Originator: NO
<br><br>I don't have documentation. Try code inspection... ghost sockets get
<br>created during migration and they inherit ssi socket ops used mostly to
<br>ship syscalls to origin node.
<br><br>----------------------------------------------------------------------
<br><br>Comment By: ch0hlik (ch0hlik)
<br>Date: 2007-12-05 10:36
<br><br>Message:
<br>Logged In: YES
<br>user_id=1771064
<br>Originator: YES
<br><br>Roger,
<br>Is there is any possibility to recive documentation for code used to
<br>modify sockets inside kernel?
<br>Thank you and regards,
<br>RJ
<br><br>----------------------------------------------------------------------
<br><br>Comment By: ch0hlik (ch0hlik)
<br>Date: 2007-12-05 10:33
<br><br>Message:
<br>Logged In: YES
<br>user_id=1771064
<br>Originator: YES
<br><br>Roger,
<br>Is there is any possibility to recive documentation for code used to
<br>modify sockets inside kernel?
<br>Thank you and regards,
<br>RJ
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2007-10-29 01:36
<br><br>Message:
<br>Logged In: YES
<br>user_id=1246761
<br>Originator: NO
<br><br>VPROC recreates the socket for SOCKET_I() during process migration. No
<br>-ERFB_TRYAGAIN errors were encountered, so the problem is somewhere in
<br>rmtsock.
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2007-08-14 02:24
<br><br>Message:
<br>Logged In: YES
<br>user_id=1246761
<br>Originator: NO
<br><br>This is looking more like a revisit of the rmtfb_getsvr_handle() SSI-1.9.1
<br>bug for which rmtfb_tryagain_handle() was written in SSI-1.9.2 to fix
<br>remote tty but maybe should be extended for rmtsock if it is a case of
<br>client/server ino mismatch.
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2007-08-13 03:05
<br><br>Message:
<br>Logged In: YES
<br>user_id=1246761
<br>Originator: NO
<br><br>Looking at the code again... In SSI-1.9 the base kernel-2.6 inode
<br>structure no longer directly references its socket structure unlike in
<br>SSI-1.2. Unless the contents of this socket structure is exported to the
<br>destination node along with the inode SSI-1.9 rmtsocksvr_* operations that
<br>use SOCKET_I() are still working with bugus information from the local
<br>node's sock_inode_cache. This explains why TCP sockets did not get
<br>recreated properly on the destination migration node.
<br><br>----------------------------------------------------------------------
<br><br>Comment By: ch0hlik (ch0hlik)
<br>Date: 2007-06-06 15:18
<br><br>Message:
<br>Logged In: YES
<br>user_id=1771064
<br>Originator: YES
<br><br>Is there is any possibilities that socket migration will be supported in
<br>1.9.2 od 1.9.3?
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2007-05-14 03:10
<br><br>Message:
<br>Logged In: YES
<br>user_id=1246761
<br>Originator: NO
<br><br>According to reop_import_pathinit() the current OpenSSI implementation
<br>does not support socket migration. However for some reason existing
<br>outbound connections continue even after process migration. The migration
<br>code looks pretty much the same in SSI-1.2 and SSI-1.9.
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2007-05-13 23:59
<br><br>Message:
<br>Logged In: YES
<br>user_id=1246761
<br>Originator: NO
<br><br>In 2.6.11-ssi TCP socket did not get recreated properly at the destination
<br>node.
<br><br>At origin node:
<br>server 236031 root 3u IPv4 121308792 TCP *:8888
<br>(LISTEN)
<br><br>At destination node:
<br>server 236031 root 3u sock 0,2 121308792 can't identify
<br>protocol
<br><br>When migrating the process back to the origin node, the listen socket
<br>reappears.
<br><br>----------------------------------------------------------------------
<br><br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405834&aid=1718329&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405834&aid=1718329&group_id=32541</a><br><br>------------------------------------------------------------------------------
<br>Come build with us! The BlackBerry(R) Developer Conference in SF, CA
<br>is the only developer event you need to attend this year. Jumpstart your
<br>developing skills, take BlackBerry mobile applications to market and stay
<br>ahead of the curve. Join us from November 9 - 12, 2009. Register now!
<br><a href="http://p.sf.net/sfu/devconference" target="_top" rel="nofollow">http://p.sf.net/sfu/devconference</a><br>_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26071809&i=0" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-26071805[ ssic-linux-Bugs-1811510 ] deadlock on loop mounted fs2009-10-26T20:53:23Z2009-10-26T20:53:23ZSourceForge.netBugs item #1811510, was opened at 2007-10-11 08:22
<br>Message generated for change (Settings changed) made by rogertsang
<br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405834&aid=1811510&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405834&aid=1811510&group_id=32541</a><br><br>Please note that this message will contain a full copy of the comment thread,
<br>including the initial issue submission, for this request,
<br>not just the latest update.
<br>Category: Filesystem
<br>Group: default
<br>Status: Open
<br>>Resolution: Fixed
<br>Priority: 5
<br>Private: No
<br>Submitted By: John Hughes (hughesj)
<br>Assigned to: Roger Tsang (rogertsang)
<br>Summary: deadlock on loop mounted fs
<br><br>Initial Comment:
<br>1. Make a sparse file
<br><br> perl -e 'open BIGFILE, ">BIGFILE"; seek BIGFILE, 1024 * 1024 * 1024, 0; print BIGFILE "big"'
<br><br>2. make a filesystem on it
<br><br> losetup /dev/loop/0 BIGFILE
<br> mkfs -t ext3 /dev/loop/0
<br><br>3. mount it
<br><br> mount -t ext3 /dev/loop/0 /mnt
<br><br>4. write a lot of files to it
<br><br> cd /mnt
<br> dump 0f - / | restore rf -
<br><br>eventualy the node where we are writing to the loopback mounted fs gets deadlocked. It's still up as far as the cluster is concerned, but any attempt to start a process on it blocks.
<br><br><br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-03-24 01:06
<br><br>Message:
<br>checked-in but needs verification
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-03-22 17:35
<br><br>Message:
<br>Testing new code to associate separate BDI per mount. This should allow us
<br>to support recursively stacked CFS.
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2008-10-10 10:52
<br><br>Message:
<br>checked-in
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2008-06-19 02:32
<br><br>Message:
<br>Logged In: YES
<br>user_id=1246761
<br>Originator: NO
<br><br>Try the attached patch.
<br><br>More work would need to be done to pass a flag to kernel space for CFS to
<br>use a different congestion bit in the case of CFS on loopback. However the
<br>proposed solution only works if you are not going to CFS mount another
<br>loopback on top of a CFS mount on loopback on CFS. So the simple fix would
<br>be this patch. Loopback becomes a standard mount.
<br>File Added: util-linux.1811510.patch
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2008-03-16 20:35
<br><br>Message:
<br>Logged In: YES
<br>user_id=1246761
<br>Originator: NO
<br><br>Should be fixed in 2.0.0pre3...
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2007-10-20 21:58
<br><br>Message:
<br>Logged In: YES
<br>user_id=1246761
<br>Originator: NO
<br><br>It looks like CFS ran out of memory. Try the latest checkin of
<br>kernel/cluster/ssi/cfs code that re-enables commit for soft mounts.
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2007-10-20 14:42
<br><br>Message:
<br>Logged In: YES
<br>user_id=1246761
<br>Originator: NO
<br><br>Does 2.6.10-ssi run into this bug?
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Nobody/Anonymous (nobody)
<br>Date: 2007-10-16 10:33
<br><br>Message:
<br>Logged In: NO
<br><br>Still looks the same as the old bug... This time it is stacked
<br>generic_file_writev().
<br><br>cfs_async (has i_sem)
<br> loop0
<br> pdflush
<br> kjournald
<br> cfs_async (waiting for i_sem)
<br><br>----------------------------------------------------------------------
<br><br>Comment By: John Hughes (hughesj)
<br>Date: 2007-10-12 07:36
<br><br>Message:
<br>Logged In: YES
<br>user_id=166336
<br>Originator: YES
<br><br>Here's some debugging. I've got to the point where the "restore" process
<br>on node 1 seems hung. On node 2 I try an "onnode 1 pwd". It hangs.
<br><br>One node 1:
<br><br>Entering kdb (current=0xc0502bc0, pid 0) on processor 0 due to Keyboard
<br>Entry
<br>[0]kdb> ps
<br>1 idle process (state I) and 50 sleeping system daemon (state M) processes
<br>suppressed
<br>Task Addr Pid Parent [*] cpu State Thread Command
<br><br>0xcf82a5d0 5 2 0 0 R 0xcf82a7b0 events/0
<br>0xcf68b990 117 11 0 0 D 0xcf68bb70 pdflush
<br>0xcf68a310 121 2 0 0 D 0xcf68a4f0 cfs_async
<br>0xcf6b99b0 122 2 0 0 D 0xcf6b9b90 cfs_async
<br>0xcf6b9410 123 2 0 0 D 0xcf6b95f0 cfs_async
<br>0xcf6b8e70 124 2 0 0 D 0xcf6b9050 cfs_async
<br>0xcf6b88d0 125 2 0 0 D 0xcf6b8ab0 cfs_async
<br>0xcf6b8330 126 2 0 0 D 0xcf6b8510 cfs_async
<br>0xcf6c99d0 127 2 0 0 D 0xcf6c9bb0 cfs_async
<br>0xcf6c9430 128 2 0 0 D 0xcf6c9610 cfs_async
<br>0xce92b730 1 0 0 0 D 0xce92b910 init
<br>[...]
<br>0xce90b150 67763 2 0 0 D 0xce90b330 loop0
<br>0xce90d170 67820 2 0 0 D 0xce90d350 kjournald
<br>0xce8f96d0 67822 67636 0 0 S 0xce8f98b0 dump
<br>0xce8f8b90 67823 67636 0 0 D 0xce8f8d70 restore
<br>0xcf13f970 67824 67822 0 0 S 0xcf13fb50 dump
<br>0xcf13f3d0 67825 67824 0 0 S 0xcf13f5b0 dump
<br>0xcf7861f0 67826 67824 0 0 S 0xcf7863d0 dump
<br>0xcf786790 67827 67824 0 0 S 0xcf786970 dump
<br>0xcf47d9b0 132773 2 0 0 D 0xcf47db90 onnode
<br>[0]kdb> btp 132773
<br>Stack traceback for pid 132773
<br>0xcf47d9b0 132773 2 0 0 D 0xcf47db90 onnode
<br>EBP EIP Function (args)
<br>0xce879ba8 0xc046c2e6 schedule+0x3a6 (0xce879c10)
<br>0xce879bb4 0xc046d348 io_schedule+0x28 (0xc1271c70)
<br>0xce879bc0 0xc014aed5 sync_page+0x45 (0xc10c37f8, 0x0, 0xc014ae90,
<br>0xcf47d9b0, 0xce879c10)
<br>0xce879be0 0xc046d6fe __wait_on_bit_lock+0x5e (0x2, 0xc10c37f8,
<br>0xc10c37f8, 0x0, 0x0)
<br>0xce879c3c 0xc014b744 __lock_page+0x84 (0xc049efb5, 0xa7, 0xce7c31a0, 0x0,
<br>0x1)
<br>0xce879cc4 0xc014beeb do_generic_mapping_read+0x3db (0xce88ca00,
<br>0xce7c31f0, 0xce7c31a0, 0xce879e00, 0xce879d00)
<br>0xce879d1c 0xc014c3ed __generic_file_aio_read+0x1ed (0xce879dc4,
<br>0xce879d34, 0x1, 0xce879e00, 0xcf06d600)
<br>0xce879d48 0xc014c473 generic_file_aio_read+0x53 (0xce879dc4, 0xcf06d600,
<br>0x80, 0x0, 0x0)
<br>0xce879d84 0xc028375a __cfs_file_read+0xaa (0xce879dc4, 0x0, 0xcf06d600,
<br>0x80, 0xce879da0)
<br>0xce879da8 0xc0283828 cfs_file_aio_read+0x38 (0xce879dc4, 0xcf06d600,
<br>0x80, 0x0, 0x0)
<br>0xce879e50 0xc016c3b3 do_sync_read+0xa3 (0xce7c31a0, 0xcf06d600, 0x80,
<br>0xce879e8c, 0xce879000)
<br>0xce879e74 0xc016c490 vfs_read+0xb0 (0xce7c31a0, 0xcf06d600, 0x80,
<br>0xce879e8c, 0x0)
<br>0xce879e9c 0xc017895a kernel_read+0x4a (0xce7c31a0, 0x0, 0xcf06d600, 0x80,
<br>0xcf06d600)
<br>0xce879ec0 0xc017946a prepare_binprm+0xca (0xcf06d600, 0x7fff, 0xc13b4080,
<br>0x0, 0x0)
<br>0xce879eec 0xc0179a16 ssi_do_execve+0x1a6 (0xcf012920, 0xce6f8800,
<br>0xcf6aa400, 0xce879fa0, 0x0)
<br>0xce879f78 0xc0245c3a rexecve_server+0xea (0xcf50e000, 0xcf47d9b0,
<br>0xcf012920, 0xce6f8800, 0xcf6aa400)
<br>0xce879fec 0xc02454f5 rexecve_server_setup+0x55
<br> 0xc01023a5 kernel_thread_helper+0x5
<br>[0]kdb> btp 67823
<br>Stack traceback for pid 67823
<br>0xce8f8b90 67823 67636 0 0 D 0xce8f8d70 restore
<br>EBP EIP Function (args)
<br>0xca2fcea0 0xc046c2e6 schedule+0x3a6 (0x0, 0xce8f8b90, 0xc013f0a0,
<br>0xca2fced4, 0xca2fced4)
<br>0xca2fcef4 0xc029f3ba cfs_wait_on_request+0x7a (0xc9a8c200, 0xca2fcf14,
<br>0x0, 0x1, 0x0)
<br>0xca2fcf24 0xc0285a9e cfs_wait_on_requests+0x8e (0xccb63be4, 0x0, 0x0,
<br>0x0, 0xce7c3600)
<br>0xca2fcf48 0xc0286f66 cfs_sync_inode+0x76 (0xccb63be4, 0x0, 0x0, 0x2,
<br>0x0)
<br>0xca2fcf80 0xc0283653 cfs_file_flush+0x93 (0xce7c3600, 0x81a4, 0xccdef200,
<br>0x5, 0xccdef204)
<br>0xca2fcf9c 0xc016bb3c filp_close+0x6c (0xce7c3600, 0xccdef200, 0xce7c3600,
<br>0x5, 0x0)
<br>0xca2fcfbc 0xc016bbce sys_close+0x6e
<br> 0xc0105a3b syscall_call+0x7
<br>[0]kdb>
<br>[0]kdb> btp 67763
<br>Stack traceback for pid 67763
<br>0xce90b150 67763 2 0 0 D 0xce90b330 loop0
<br>EBP EIP Function (args)
<br>0xca488db8 0xc046c2e6 schedule+0x3a6 (0xca488e20)
<br>0xca488dc4 0xc046d348 io_schedule+0x28 (0xc12711e0)
<br>0xca488dd0 0xc014aed5 sync_page+0x45 (0xc11d6be0, 0x0, 0xc014ae90,
<br>0xce90b150, 0xca488e20)
<br>0xca488df0 0xc046d6fe __wait_on_bit_lock+0x5e (0x2, 0xc11d6be0,
<br>0xc11d6be0, 0x0, 0x0)
<br>0xca488e4c 0xc014b744 __lock_page+0x84 (0xc049efb5, 0xa7, 0xcd6ca600,
<br>0x38002, 0x1)
<br>0xca488ed4 0xc014beeb do_generic_mapping_read+0x3db (0xcb632f40,
<br>0xcd6ca650, 0xcd6ca600, 0xca488f58, 0xca488ef4)
<br>0xca488f04 0xc014c61b generic_file_sendfile+0x5b (0xcd6ca600, 0xca488f58,
<br>0x1000, 0xd08f15d0, 0xca488f60)
<br>0xca488f3c 0xc02838bd cfs_file_sendfile+0x8d (0xcd6ca600, 0xca488f58,
<br>0x1000, 0xd08f15d0, 0xca488f60)
<br>0xca488f74 0xd08f16fc [loop]do_lo_receive+0x5c (0xc9353000, 0xc4279630,
<br>0x1000, 0x38002000, 0x0)
<br>0xca488fa4 0xd08f176e [loop]lo_receive+0x5e (0xc9353000, 0xc1ed33e0,
<br>0x1000, 0x38002000, 0x0)
<br>0xca488fc8 0xd08f17eb [loop]do_bio_filebacked+0x4b (0xc9353000,
<br>0xc1ed33e0, 0x0, 0xc9353138, 0xd08f1a60)
<br>0xca488fec 0xd08f1b3b [loop]loop_thread+0xdb
<br> 0xc01023a5 kernel_thread_helper+0x5
<br><br><br><br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2007-10-11 22:21
<br><br>Message:
<br>Logged In: YES
<br>user_id=1246761
<br>Originator: NO
<br><br>Sounds like [ 686748 ] Filesystem stacking deadlock.
<br><br>----------------------------------------------------------------------
<br><br>Comment By: John Hughes (hughesj)
<br>Date: 2007-10-11 08:22
<br><br>Message:
<br>Logged In: YES
<br>user_id=166336
<br>Originator: YES
<br><br>This is with the 2.6.11 kernel
<br><br>----------------------------------------------------------------------
<br><br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405834&aid=1811510&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405834&aid=1811510&group_id=32541</a><br><br>------------------------------------------------------------------------------
<br>Come build with us! The BlackBerry(R) Developer Conference in SF, CA
<br>is the only developer event you need to attend this year. Jumpstart your
<br>developing skills, take BlackBerry mobile applications to market and stay
<br>ahead of the curve. Join us from November 9 - 12, 2009. Register now!
<br><a href="http://p.sf.net/sfu/devconference" target="_top" rel="nofollow">http://p.sf.net/sfu/devconference</a><br>_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26071805&i=0" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-26071800[ ssic-linux-Bugs-2095071 ] From time to time /proc deadlocks2009-10-26T20:52:41Z2009-10-26T20:52:41ZSourceForge.netBugs item #2095071, was opened at 2008-09-05 10:07
<br>Message generated for change (Comment added) made by rogertsang
<br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2095071&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2095071&group_id=32541</a><br><br>Please note that this message will contain a full copy of the comment thread,
<br>including the initial issue submission, for this request,
<br>not just the latest update.
<br>Category: Process Management
<br>Group: None
<br>Status: Open
<br>Resolution: Accepted
<br>Priority: 5
<br>Private: No
<br>Submitted By: John Hughes (hughesj)
<br>Assigned to: Nobody/Anonymous (nobody)
<br>Summary: From time to time /proc deadlocks
<br><br>Initial Comment:
<br>For quite some time (maybe even on the old 2.4 based kernel) we've occasionally seen a problem where the /proc filesystem seems to be deadlocked - any attempt to read /proc hangs.
<br><br>When this happens rebooting one the nodes (not any node, it has to be the "right" one) will free up the system and things will continue as normal.
<br><br>Today I just noticed that when I rebooted the node that was "causing" the problem I had the following messages on the init node:
<br><br>Node 6 has gone down!!!
<br>Assertion failed! origin_lock != ((void *)0), cluster/ssi/vproc/dvp_pvpsops.c, pvpsop_get_execnode, line=376
<br>nm_add_node: Node 6 added
<br><br>Is this a clue?
<br>
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-10-26 23:52
<br><br>Message:
<br>bug fixed?
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-03-28 17:37
<br><br>Message:
<br>please try latest CVS (March 24th)
<br><br>----------------------------------------------------------------------
<br><br>Comment By: John Hughes (hughesj)
<br>Date: 2008-12-03 10:59
<br><br>Message:
<br>Well. I finally found a (crazy) way to duplicate this - launch a windows
<br>app with wine and hit control-c before it gets going. Eventually it will
<br>provoke the hang.
<br><br>A "bta A" trace of the running processes is attached.
<br><br>----------------------------------------------------------------------
<br><br>Comment By: John Hughes (hughesj)
<br>Date: 2008-11-20 05:14
<br><br>Message:
<br>Sorry, wasn't clear above - this is not straight CVS, it's my port of
<br>current CVS to 2.6.12. However I'm pretty sure this part of the port is
<br>good.
<br><br><br>----------------------------------------------------------------------
<br><br>Comment By: John Hughes (hughesj)
<br>Date: 2008-11-20 05:13
<br><br>Message:
<br>With current CVS (20/11/2008) I still see this bug, on coming in to work I
<br>found my (non-init) node stuck, apparently in the screensaver, and when I
<br>tried to see what was going on from the initnode each time I did a stat on
<br>"/proc/1" it hung. stat on other things in /proc was working - stat
<br>/proc/self or stat /proc/$$ for example.
<br><br>When I turned off the node that was stuck the hung "stat" operations on
<br>the initnode sprang back to life and I see messages like this in the log:
<br><br>Node 6 has gone down!!!
<br>Assertion failed! origin_lock != ((void *)0),
<br>cluster/ssi/vproc/dvp_pvpsops.c, pvpsop_get_execnode, line=379
<br>Assertion failed! origin_lock != ((void *)0),
<br>cluster/ssi/vproc/dvp_pvpsops.c, pvpsop_get_execnode, line=379
<br>Assertion failed! origin_lock != ((void *)0),
<br>cluster/ssi/vproc/dvp_pvpsops.c, pvpsop_get_execnode, line=379
<br>nm_add_node: Node 6 added
<br><br><br><br><br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2008-10-02 06:24
<br><br>Message:
<br>a fix is going into CVS
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2008-09-23 22:07
<br><br>Message:
<br>Related to this origin_lock assertion is a possible race in
<br>vproc_origin_list traversal supposedly fixed by pragma #ifdef VOD_HLIST
<br>since SSI-1.9.x, but the fix introduced a possible deadlock bug and should
<br>be fixed in 1.9.6.
<br><br>Lock ordering pre-1.9.6 (with #ifdef VOD_HLIST):
<br> -> vproc_origin_cleanup (down_read origin list)
<br> -> vproc_origin_fgpgrp_cleanup
<br> -> pvpop_getctty
<br> -> rpvpop_start_op
<br> -> pvpopsop_get_execnode
<br> -> vproc_lock_origin_node
<br> -> vproc_origin_find (down_read origin list)
<br><br>----------------------------------------------------------------------
<br><br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2095071&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2095071&group_id=32541</a><br><br>------------------------------------------------------------------------------
<br>Come build with us! The BlackBerry(R) Developer Conference in SF, CA
<br>is the only developer event you need to attend this year. Jumpstart your
<br>developing skills, take BlackBerry mobile applications to market and stay
<br>ahead of the curve. Join us from November 9 - 12, 2009. Register now!
<br><a href="http://p.sf.net/sfu/devconference" target="_top" rel="nofollow">http://p.sf.net/sfu/devconference</a><br>_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26071800&i=0" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-26071769[ ssic-linux-Bugs-2719590 ] size-32 slab leak2009-10-26T20:49:18Z2009-10-26T20:49:18ZSourceForge.netBugs item #2719590, was opened at 2009-03-28 16:36
<br>Message generated for change (Comment added) made by rogertsang
<br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2719590&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2719590&group_id=32541</a><br><br>Please note that this message will contain a full copy of the comment thread,
<br>including the initial issue submission, for this request,
<br>not just the latest update.
<br>Category: Virtual Memory
<br>Group: None
<br>Status: Open
<br>>Resolution: Fixed
<br>Priority: 5
<br>Private: No
<br>Submitted By: Roger Tsang (rogertsang)
<br>Assigned to: Roger Tsang (rogertsang)
<br>Summary: size-32 slab leak
<br><br>Initial Comment:
<br>I have not investigated whether this bug is present in previous releases, but in the latest CVS as of this writing there is a slow slab leak on the init node.
<br>size-32 969791 969850
<br>size-32 976961 976990
<br>size-32 979727 979727
<br>size-32 1015844 1015903
<br><br>The failover init node does not seem to have this problem.
<br>size-32 29247 33082
<br>size-32 29328 33082
<br>size-32 29262 33082
<br><br>----------------------------------------------------------------------
<br><br>>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-10-26 23:49
<br><br>Message:
<br>checked-in fix
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-04-03 10:00
<br><br>Message:
<br>Found more than a few places leaking memory. Preliminary tests look good.
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-03-30 22:54
<br><br>Message:
<br>correction: problem did not go away with upgrade to drbd-8.3.1
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-03-29 12:38
<br><br>Message:
<br>seems to affect only SMP nodes
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-03-28 20:43
<br><br>Message:
<br>problem seems to have disappeared after rolling upgrade from drbd-8.3.0 to
<br>drbd-8.3.1
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-03-28 16:58
<br><br>Message:
<br>Still seeing the same symptoms after applying patches from Bug 2719607.
<br><br>----------------------------------------------------------------------
<br><br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2719590&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2719590&group_id=32541</a><br><br>------------------------------------------------------------------------------
<br>Come build with us! The BlackBerry(R) Developer Conference in SF, CA
<br>is the only developer event you need to attend this year. Jumpstart your
<br>developing skills, take BlackBerry mobile applications to market and stay
<br>ahead of the curve. Join us from November 9 - 12, 2009. Register now!
<br><a href="http://p.sf.net/sfu/devconference" target="_top" rel="nofollow">http://p.sf.net/sfu/devconference</a><br>_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26071769&i=0" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-26071767[ ssic-linux-Bugs-2719607 ] memory leak in ipcname_gettotal() path2009-10-26T20:49:01Z2009-10-26T20:49:01ZSourceForge.netBugs item #2719607, was opened at 2009-03-28 16:47
<br>Message generated for change (Comment added) made by rogertsang
<br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2719607&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2719607&group_id=32541</a><br><br>Please note that this message will contain a full copy of the comment thread,
<br>including the initial issue submission, for this request,
<br>not just the latest update.
<br>Category: IPC
<br>Group: default
<br>Status: Open
<br>>Resolution: Fixed
<br>Priority: 5
<br>Private: No
<br>Submitted By: Roger Tsang (rogertsang)
<br>Assigned to: John Hughes (hughesj)
<br>Summary: memory leak in ipcname_gettotal() path
<br><br>Initial Comment:
<br><a href="https://sourceforge.net/mailarchive/forum.php?thread_name=49C8C9D9.6000204%40Calva.COM&forum_name=ssic-linux-devel" target="_top" rel="nofollow">https://sourceforge.net/mailarchive/forum.php?thread_name=49C8C9D9.6000204%40Calva.COM&forum_name=ssic-linux-devel</a><br><br>----------------------------------------------------------------------
<br><br>>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-10-26 23:49
<br><br>Message:
<br>checked-in
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-04-28 09:56
<br><br>Message:
<br>node_id_pairs is a vector
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-04-28 09:54
<br><br>Message:
<br>There is also a memory corruption bug in this path.
<br>When doing RPC the length of the struct node_id_pairs buffer is
<br>incorrect.
<br><br>--- linux.orig/cluster/ssi/ipc/namesvr_clnt.c
<br>+++ linux/cluster/ssi/ipc/namesvr_clnt.c
<br>@@ -188,7 +188,7 @@ again:
<br> {
<br> clusternode_t server_node;
<br>- int status, rval, len = 0;
<br>+ int rval;
<br> extern clms_key_svc_t ipc_key_service;
<br> again:
<br> server_node = name_server_node;
<br>@@ -203,13 +203,12 @@ again:
<br> rval = ipcname_gettotal(service, *node_id_pairs, sz);
<br> } else {
<br> ssi_procstate_t pstate;
<br>+ int status, count = *sz;
<br><br>- if (*sz > 0)
<br>- len = *sz * sizeof(struct ssi_nodeid_pair);
<br> ssi_procstate_get(&pstate);
<br><br>- node_id_pairs, &len, sz);
<br>+ node_id_pairs, &count,
<br>sz);
<br> if (status == -EAGAIN || status == -EREMOTE) {
<br> /* Server is doing failover,dying, or isn't ready
<br>yet.*/
<br> clms_waitfor_key_service(0);
<br>Index: linux/cluster/ssi/ipc/namesvr_svr.c
<br>===================================================================
<br>@@ -88,34 +88,33 @@ ripc_ipcname_getid(clusternode_t *node,
<br> */
<br> void
<br>- int *len, int *sz)
<br>+ ssi_procstate_t *pstate,
<br>+ struct ssi_nodeid_pair **node_id_pairs, int *count,
<br>+ int *sz)
<br> {
<br> ssi_procstate_t save_pstate;
<br>- int count = *sz;
<br><br> if (ipcname_failover_flag) {
<br> *rval = -EAGAIN;
<br> return;
<br> }
<br><br>- *len = 0;
<br> ssi_procstate_get(&save_pstate);
<br> ssi_procstate_set(pstate);
<br>- if (count > 0) {
<br>- if (*node_id_pairs == NULL) {
<br>- *sz = 0;
<br>+ if (*count > 0) {
<br>+ if (!*node_id_pairs) {
<br>+ *count = *sz = 0;
<br> goto done;
<br> }
<br> }
<br><br> *rval = ipcname_gettotal(service, *node_id_pairs, sz);
<br>- if (count > 0) {
<br>- if (count > *sz)
<br>- count = *sz;
<br>- *len = count * sizeof(**node_id_pairs);
<br>- }
<br>+ if (!*rval) {
<br>+ if (*count && *sz < *count)
<br>+ *count = *sz;
<br>+ } else
<br>+ *count = 0;
<br> done:
<br> ssi_procstate_set(&save_pstate);
<br> }
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-03-28 16:49
<br><br>Message:
<br>fix attached
<br><br>----------------------------------------------------------------------
<br><br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2719607&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2719607&group_id=32541</a><br><br>------------------------------------------------------------------------------
<br>Come build with us! The BlackBerry(R) Developer Conference in SF, CA
<br>is the only developer event you need to attend this year. Jumpstart your
<br>developing skills, take BlackBerry mobile applications to market and stay
<br>ahead of the curve. Join us from November 9 - 12, 2009. Register now!
<br><a href="http://p.sf.net/sfu/devconference" target="_top" rel="nofollow">http://p.sf.net/sfu/devconference</a><br>_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26071767&i=0" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-26071765[ ssic-linux-Bugs-2719679 ] migration stuck in move_data_wait_for_state_change()2009-10-26T20:48:42Z2009-10-26T20:48:42ZSourceForge.netBugs item #2719679, was opened at 2009-03-28 17:28
<br>Message generated for change (Comment added) made by rogertsang
<br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2719679&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2719679&group_id=32541</a><br><br>Please note that this message will contain a full copy of the comment thread,
<br>including the initial issue submission, for this request,
<br>not just the latest update.
<br>Category: Process Management
<br>Group: None
<br>Status: Open
<br>>Resolution: Fixed
<br>Priority: 5
<br>Private: No
<br>Submitted By: Roger Tsang (rogertsang)
<br>Assigned to: Roger Tsang (rogertsang)
<br>Summary: migration stuck in move_data_wait_for_state_change()
<br><br>Initial Comment:
<br>>From John Hughes:
<br><br>With current (March 17th) CVS I'm seeing a problem migrating processes from time to time.
<br><br>Running my zz-migrate-cruelty script (attached) it sometimes seems to get stuck in move_data_wait_for_state_change
<br><br>Stack trace shows:
<br>schedule
<br>move_data_wait_for_state_change
<br>ssi_quiesce_movement
<br>do_signal
<br>do_notify_resume
<br>work_notifysig
<br><br>----------------------------------------------------------------------
<br><br>>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-10-26 23:48
<br><br>Message:
<br>checked-in
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-05-10 21:48
<br><br>Message:
<br>Looks like an issue with SMP synchronization (or lack of)...
<br><br>With my latest changes (not yet in CVS) the zz-migrate-cruelty script no
<br>longer gets stuck. I tested successfully for 10+ minutes on two-node
<br>cluster.
<br><br>----------------------------------------------------------------------
<br><br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2719679&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2719679&group_id=32541</a><br><br>------------------------------------------------------------------------------
<br>Come build with us! The BlackBerry(R) Developer Conference in SF, CA
<br>is the only developer event you need to attend this year. Jumpstart your
<br>developing skills, take BlackBerry mobile applications to market and stay
<br>ahead of the curve. Join us from November 9 - 12, 2009. Register now!
<br><a href="http://p.sf.net/sfu/devconference" target="_top" rel="nofollow">http://p.sf.net/sfu/devconference</a><br>_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26071765&i=0" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-26071758[ ssic-linux-Bugs-2719752 ] stuck while tracing ssi-ksync2009-10-26T20:48:03Z2009-10-26T20:48:03ZSourceForge.netBugs item #2719752, was opened at 2009-03-28 18:12
<br>Message generated for change (Settings changed) made by rogertsang
<br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2719752&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2719752&group_id=32541</a><br><br>Please note that this message will contain a full copy of the comment thread,
<br>including the initial issue submission, for this request,
<br>not just the latest update.
<br>Category: Process Management
<br>Group: v2.0.0pre2
<br>Status: Open
<br>>Resolution: Fixed
<br>Priority: 5
<br>Private: No
<br>Submitted By: Roger Tsang (rogertsang)
<br>Assigned to: Roger Tsang (rogertsang)
<br>Summary: stuck while tracing ssi-ksync
<br><br>Initial Comment:
<br>>From John Hughes (March 16):
<br><br># strace -f -o zz /sbin/ssi-ksync
<br>...
<br>pvpop_rmv_child_from_parent: waiting for parent lock 68272
<br>pvpop_rmv_child_from_parent: waiting for parent lock 68272
<br>pvpop_rmv_child_from_parent: waiting for parent lock 68272
<br>pvpop_rmv_child_from_parent: waiting for parent lock 68272
<br>...
<br><br>----------------------------------------------------------------------
<br><br>>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-10-26 23:48
<br><br>Message:
<br>checked-in
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-07-04 16:54
<br><br>Message:
<br>This pvpop_rmv_child_from_parent() was remotely called from
<br>__ptrace_unlink() path and is stuck waiting for lock held by strace process
<br>which is waiting for remote child to get reaped.
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-03-28 18:16
<br><br>Message:
<br>>From John Hughes (March 16):
<br><br>Further investigation shows that the problem is happening in
<br>mkdhcpd.conf:
<br><br># strace -f -o zz /sbin/mkdhcpd.conf
<br>pvpop_rmv_child_from_parent: waiting for parent lock xxxxx
<br><br><br>If I power-off the node that's waiting for the parent lock then the other
<br>node crashes:
<br><br>Taking over master from node 2.
<br>Node 2 has gone down!!!
<br>passed the first scan in ipcname_pull_data
<br>num_objects[MSG] = 0
<br>num_objects[SEM] = 0
<br>num_objects[SHM] = 0
<br>ipcnameserver ready completed
<br>write handler down off 2400000 len 4096
<br><br>EXT3-fs: INFO: recovery required on readonly filesystem.
<br>EXT3-fs: write access will be enabled during recovery.
<br>kjournald starting. Commit interval 5 seconds
<br>EXT3-fs: recovery complete.
<br>EXT3-fs: mounted filesystem with ordered data mode.
<br>fsck 1.37 (21-Mar-2005)
<br>EXT3 FS on hdb1, internal journal
<br>/etc/init.d/rc.sysrecover running
<br>ptrace_unlink: vpop_reclaim failed Unable to handle kernel NULL pointer
<br>dereference at virtual address 00000008
<br>printing eip:
<br>c0241911
<br>*pde = 00000000
<br>Oops: 0000 [#1]
<br>SMP Modules linked in: parport_pc parport floppy uhci_hcd ohci_hcd
<br>ehci_hcd i2c_piix4 i2c_core ide_scsi scsi_mod ext3 jbd ne2k_pci 8390
<br>CPU: 0
<br>EIP: 0060:[<c0241911>] Not tainted VLI
<br>EFLAGS: 00000282 (2.6.11-ssi-686-smp) EIP is at
<br>pvpop_reassign_child+0x281/0x920
<br>eax: 00000000 ebx: 0002074b ecx: 00000000 edx: cea561f0
<br>esi: 0002074c edi: cf6ae000 ebp: cf6afebc esp: cf6afe1c
<br>ds: 007b es: 007b ss: 0068
<br>Process VPROC Slave Dae (pid: 67247, threadinfo=cf6ae000 task=cf980b10)
<br>Stack: 0002074b 00000001 c04cffc1 00000001 cf6afe4c c01540a2 00000014
<br>00000000 00000000 cf532040 cf532000 00000000 cea561f0 c0155036
<br>cf6afe78 c025c57f 00000001 000008d0 cf9ef700 00000010 cea56cc0
<br>00000001 00000001 cf6afe94 Call Trace:
<br>[<c010679f>] show_stack+0x7f/0xa0
<br>[<c0106944>] show_registers+0x164/0x230
<br>[<c0106cf4>] die+0xf4/0x1c0
<br>[<c011f39d>] do_page_fault+0x46d/0x669
<br>[<c0106403>] error_code+0x2b/0x30
<br>[<c025965a>] vproc_child_lost_parent+0x7a/0x160
<br>[<c0259ead>] vproc_lost_relations+0xad/0x218
<br>[<c024d554>] pvpsop_lost_relations+0x94/0xa0
<br>[<c025956e>] vproc_carelist_cleanup+0x7e/0xf0
<br>[<c0256933>] vproc_origin_node_cleanup+0x63/0x2b0
<br>[<c0256dde>] vproc_origin_cleanup+0xbe/0x190
<br>[<c0257fc1>] vproc_slave_daemon+0x1a1/0x260
<br>[<c01023a5>] kernel_thread_helper+0x5/0x10
<br>Code: 8b 4d 8c 89 0c 24 e8 6f a9 01 00 ba 01 00 00 00 89 54 24 04 8b 55 90
<br>8b 42 2c 89 04 24 e8 88 a6 01 00 89 45 8c 8b 4d 8c 8b 55 90 <8b> 41 08 39
<br>42 28 0f 84 12 03 00 00 8b 85 7c ff ff ff 85 c0 75
<br><br>----------------------------------------------------------------------
<br><br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2719752&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2719752&group_id=32541</a><br><br>------------------------------------------------------------------------------
<br>Come build with us! The BlackBerry(R) Developer Conference in SF, CA
<br>is the only developer event you need to attend this year. Jumpstart your
<br>developing skills, take BlackBerry mobile applications to market and stay
<br>ahead of the curve. Join us from November 9 - 12, 2009. Register now!
<br><a href="http://p.sf.net/sfu/devconference" target="_top" rel="nofollow">http://p.sf.net/sfu/devconference</a><br>_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26071758&i=0" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-26071757[ ssic-linux-Bugs-2782858 ] thread group fails to migrate2009-10-26T20:47:42Z2009-10-26T20:47:42ZSourceForge.netBugs item #2782858, was opened at 2009-04-28 09:39
<br>Message generated for change (Settings changed) made by rogertsang
<br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2782858&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2782858&group_id=32541</a><br><br>Please note that this message will contain a full copy of the comment thread,
<br>including the initial issue submission, for this request,
<br>not just the latest update.
<br>Category: Process Management
<br>Group: v1.9.1
<br>Status: Open
<br>>Resolution: Fixed
<br>Priority: 5
<br>Private: No
<br>Submitted By: Roger Tsang (rogertsang)
<br>Assigned to: Roger Tsang (rogertsang)
<br>Summary: thread group fails to migrate
<br><br>Initial Comment:
<br>[created from discussion <a href="http://article.gmane.org/gmane.linux.cluster.ssic.devel/6010" target="_top" rel="nofollow">http://article.gmane.org/gmane.linux.cluster.ssic.devel/6010</a>]
<br><br>When running the attached test program the origin node displays the following messages:
<br><br>add_thread_group:Cannot move process 438181 (calc) - cannot find all shares
<br>add_thread_group: group(438181)/share imbalance 4/1
<br><br>To build attached test program
<br># gcc -o calc pthread-test.c -pthread -lm
<br><br>----------------------------------------------------------------------
<br><br>>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-10-26 23:47
<br><br>Message:
<br>checked-in
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-05-06 01:28
<br><br>Message:
<br>apply patch against CVS
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-05-01 21:10
<br><br>Message:
<br>now looking good; will CVS checkin my fixes soon
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-04-29 23:50
<br><br>Message:
<br>ran into other bugs while fixing this...
<br>1) fails to migrate due to shared SysV sem
<br>2) sending signal after migration results in zombie thread group when the
<br>parent of the test program reside on a different node
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-04-28 09:42
<br><br>Message:
<br>from discussion
<br><a href="http://article.gmane.org/gmane.linux.cluster.ssic.devel/6011" target="_top" rel="nofollow">http://article.gmane.org/gmane.linux.cluster.ssic.devel/6011</a><br><br>John Huges wrote:
<br>I've also seen another error when trying to migrate the process:
<br><br>$ migrate 1 395359
<br>migrate: process 395359 has exited/is exiting
<br>$ dmesg | tail -1
<br>add_thread_group:Cannot move process 394990 (gdm) has different tgid
<br>394990 (expected 395359)
<br><br>It seems confused about which threads belong to which processes.
<br><br>I get the same behaviours with a 2.6.11 kernel based on CVS as of 16
<br>March.
<br><br>----------------------------------------------------------------------
<br><br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2782858&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2782858&group_id=32541</a><br><br>------------------------------------------------------------------------------
<br>Come build with us! The BlackBerry(R) Developer Conference in SF, CA
<br>is the only developer event you need to attend this year. Jumpstart your
<br>developing skills, take BlackBerry mobile applications to market and stay
<br>ahead of the curve. Join us from November 9 - 12, 2009. Register now!
<br><a href="http://p.sf.net/sfu/devconference" target="_top" rel="nofollow">http://p.sf.net/sfu/devconference</a><br>_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26071757&i=0" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-26071754[ ssic-linux-Bugs-2838006 ] Corrupt SHM_LOCK_DEST flag2009-10-26T20:47:19Z2009-10-26T20:47:19ZSourceForge.netBugs item #2838006, was opened at 2009-08-14 22:43
<br>Message generated for change (Comment added) made by rogertsang
<br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2838006&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2838006&group_id=32541</a><br><br>Please note that this message will contain a full copy of the comment thread,
<br>including the initial issue submission, for this request,
<br>not just the latest update.
<br>Category: IPC
<br>Group: default
<br>Status: Open
<br>>Resolution: Fixed
<br>Priority: 5
<br>Private: No
<br>Submitted By: Roger Tsang (rogertsang)
<br>Assigned to: Roger Tsang (rogertsang)
<br>Summary: Corrupt SHM_LOCK_DEST flag
<br><br>Initial Comment:
<br>Reported by John Hughes.
<br><a href="http://article.gmane.org/gmane.linux.cluster.ssic.devel/6037" target="_top" rel="nofollow">http://article.gmane.org/gmane.linux.cluster.ssic.devel/6037</a><br><br>Current definition of SHM_LOCK_DEST flag is 023420 (octal)
<br><br>----------------------------------------------------------------------
<br><br>>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-10-26 23:47
<br><br>Message:
<br>checked-in
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-08-14 22:53
<br><br>Message:
<br>Further due to this flag the SHM node down code assumes shm_ids semaphore
<br>is held, but in fact is not always the case and could race causing memory
<br>corruption.
<br><br>----------------------------------------------------------------------
<br><br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2838006&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2838006&group_id=32541</a><br><br>------------------------------------------------------------------------------
<br>Come build with us! The BlackBerry(R) Developer Conference in SF, CA
<br>is the only developer event you need to attend this year. Jumpstart your
<br>developing skills, take BlackBerry mobile applications to market and stay
<br>ahead of the curve. Join us from November 9 - 12, 2009. Register now!
<br><a href="http://p.sf.net/sfu/devconference" target="_top" rel="nofollow">http://p.sf.net/sfu/devconference</a><br>_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26071754&i=0" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-26071749[ ssic-linux-Feature Requests-1213029 ] clusterwide flock2009-10-26T20:46:16Z2009-10-26T20:46:16ZSourceForge.netFeature Requests item #1213029, was opened at 2005-06-01 16:33
<br>Message generated for change (Comment added) made by rogertsang
<br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405837&aid=1213029&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405837&aid=1213029&group_id=32541</a><br><br>Please note that this message will contain a full copy of the comment thread,
<br>including the initial issue submission, for this request,
<br>not just the latest update.
<br>Category: Filesystem
<br>Group: 1.9
<br>Status: Open
<br>>Resolution: Fixed
<br>Priority: 3
<br>Private: No
<br>Submitted By: Jakob Goldbach (goldbach)
<br>Assigned to: Roger Tsang (rogertsang)
<br>Summary: clusterwide flock
<br><br>Initial Comment:
<br>I would like that BSD style flock is made clusterwide.
<br><br>Many webapps in php is using flock.
<br><br>Also, php itself is using flock for locking session
<br>files. (although, you can use somthing else than files
<br>for session data).
<br><br><br><br><br><br><br>----------------------------------------------------------------------
<br><br>>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-10-26 23:46
<br><br>Message:
<br>checked-in
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-05-28 22:49
<br><br>Message:
<br>Try attached patch. Apply against CVS.
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-05-24 18:35
<br><br>Message:
<br>clusterwide flock() in CVS still needs a bit of work, but should be ready
<br>by 1.9.6 release.
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-03-24 01:08
<br><br>Message:
<br>checked-in
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-03-22 17:50
<br><br>Message:
<br>Adding new CFS code to convert flock to POSIX style lock.
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2007-04-24 23:27
<br><br>Message:
<br>Logged In: YES
<br>user_id=1246761
<br>Originator: NO
<br><br>Take that back flock is not clusterwide in SSI-1.9. flock needs to be
<br>converted to posix style lock to be clusterwide. For now use fcntl or
<br>another PHP session.save_handler.
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2007-04-24 02:13
<br><br>Message:
<br>Logged In: YES
<br>user_id=1246761
<br>Originator: NO
<br><br>Supported in SSI-1.9
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Nobody/Anonymous (nobody)
<br>Date: 2005-08-25 07:49
<br><br>Message:
<br>Logged In: NO
<br><br>OpenSSI-1.9 may already have this, not sure yet.
<br><br>----------------------------------------------------------------------
<br><br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405837&aid=1213029&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405837&aid=1213029&group_id=32541</a><br><br>------------------------------------------------------------------------------
<br>Come build with us! The BlackBerry(R) Developer Conference in SF, CA
<br>is the only developer event you need to attend this year. Jumpstart your
<br>developing skills, take BlackBerry mobile applications to market and stay
<br>ahead of the curve. Join us from November 9 - 12, 2009. Register now!
<br><a href="http://p.sf.net/sfu/devconference" target="_top" rel="nofollow">http://p.sf.net/sfu/devconference</a><br>_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26071749&i=0" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-26071746[ ssic-linux-Feature Requests-1750747 ] CFS ioctl2009-10-26T20:45:42Z2009-10-26T20:45:42ZSourceForge.netFeature Requests item #1750747, was opened at 2007-07-09 15:35
<br>Message generated for change (Comment added) made by rogertsang
<br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405837&aid=1750747&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405837&aid=1750747&group_id=32541</a><br><br>Please note that this message will contain a full copy of the comment thread,
<br>including the initial issue submission, for this request,
<br>not just the latest update.
<br>Category: Filesystem
<br>Group: 1.9
<br>Status: Open
<br>>Resolution: Fixed
<br>Priority: 5
<br>Private: No
<br>Submitted By: Roger Tsang (rogertsang)
<br>Assigned to: Roger Tsang (rogertsang)
<br>Summary: CFS ioctl
<br><br>Initial Comment:
<br>Add ioctl support so xfs can be manipulated while CFS stacked.
<br><br>----------------------------------------------------------------------
<br><br>>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-10-26 23:45
<br><br>Message:
<br>checked-in
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-09-08 23:14
<br><br>Message:
<br>ioctl() works from any node even when CFS client not at server
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-07-14 02:06
<br><br>Message:
<br>So far xfs_io seems to work on CFS client at server.
<br><br>----------------------------------------------------------------------
<br><br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405837&aid=1750747&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405837&aid=1750747&group_id=32541</a><br><br>------------------------------------------------------------------------------
<br>Come build with us! The BlackBerry(R) Developer Conference in SF, CA
<br>is the only developer event you need to attend this year. Jumpstart your
<br>developing skills, take BlackBerry mobile applications to market and stay
<br>ahead of the curve. Join us from November 9 - 12, 2009. Register now!
<br><a href="http://p.sf.net/sfu/devconference" target="_top" rel="nofollow">http://p.sf.net/sfu/devconference</a><br>_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26071746&i=0" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-26071743[ ssic-linux-Feature Requests-1925610 ] NFS export CFS stacked PFS with depreciated iget() support2009-10-26T20:45:23Z2009-10-26T20:45:23ZSourceForge.netFeature Requests item #1925610, was opened at 2008-03-25 18:54
<br>Message generated for change (Settings changed) made by rogertsang
<br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405837&aid=1925610&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405837&aid=1925610&group_id=32541</a><br><br>Please note that this message will contain a full copy of the comment thread,
<br>including the initial issue submission, for this request,
<br>not just the latest update.
<br>Category: Filesystem
<br>Group: 1.9
<br>Status: Open
<br>>Resolution: Fixed
<br>Priority: 5
<br>Private: No
<br>Submitted By: Roger Tsang (rogertsang)
<br>Assigned to: Roger Tsang (rogertsang)
<br>Summary: NFS export CFS stacked PFS with depreciated iget() support
<br><br>Initial Comment:
<br>Some file systems in kernel-2.6 no longer support iget(). For example XFS. To NFS export these CFS stack mounts, CFS export operations code path will have to phase out PFS iget() calls.
<br><br>Though I haven't attempted to trigger any oops we expect to see fh_verify() in the NFS stack trace when the file system exported by CFS is not supported.
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2008-04-20 13:50
<br><br>Message:
<br>Logged In: YES
<br>user_id=1246761
<br>Originator: YES
<br><br>Checked into CVS.
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2008-04-06 13:18
<br><br>Message:
<br>Logged In: YES
<br>user_id=1246761
<br>Originator: YES
<br><br>CFS no longer using iget() in 2.0.0pre3
<br><br>----------------------------------------------------------------------
<br><br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405837&aid=1925610&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405837&aid=1925610&group_id=32541</a><br><br>------------------------------------------------------------------------------
<br>Come build with us! The BlackBerry(R) Developer Conference in SF, CA
<br>is the only developer event you need to attend this year. Jumpstart your
<br>developing skills, take BlackBerry mobile applications to market and stay
<br>ahead of the curve. Join us from November 9 - 12, 2009. Register now!
<br><a href="http://p.sf.net/sfu/devconference" target="_top" rel="nofollow">http://p.sf.net/sfu/devconference</a><br>_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26071743&i=0" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-26071739[ ssic-linux-Feature Requests-1947313 ] CFS only zap caches on revalidate2009-10-26T20:45:08Z2009-10-26T20:45:08ZSourceForge.netFeature Requests item #1947313, was opened at 2008-04-20 13:53
<br>Message generated for change (Settings changed) made by rogertsang
<br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405837&aid=1947313&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405837&aid=1947313&group_id=32541</a><br><br>Please note that this message will contain a full copy of the comment thread,
<br>including the initial issue submission, for this request,
<br>not just the latest update.
<br>Category: Filesystem
<br>Group: 1.9
<br>Status: Open
<br>>Resolution: Fixed
<br>Priority: 5
<br>Private: No
<br>Submitted By: Roger Tsang (rogertsang)
<br>Assigned to: Roger Tsang (rogertsang)
<br>Summary: CFS only zap caches on revalidate
<br><br>Initial Comment:
<br>for CFS performance
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2008-04-20 13:53
<br><br>Message:
<br>Logged In: YES
<br>user_id=1246761
<br>Originator: YES
<br><br>checked into CVS
<br><br>----------------------------------------------------------------------
<br><br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405837&aid=1947313&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405837&aid=1947313&group_id=32541</a><br><br>------------------------------------------------------------------------------
<br>Come build with us! The BlackBerry(R) Developer Conference in SF, CA
<br>is the only developer event you need to attend this year. Jumpstart your
<br>developing skills, take BlackBerry mobile applications to market and stay
<br>ahead of the curve. Join us from November 9 - 12, 2009. Register now!
<br><a href="http://p.sf.net/sfu/devconference" target="_top" rel="nofollow">http://p.sf.net/sfu/devconference</a><br>_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26071739&i=0" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-26071742[ ssic-linux-Feature Requests-1947441 ] process migration memory balancing2009-10-26T20:44:51Z2009-10-26T20:44:51ZSourceForge.netFeature Requests item #1947441, was opened at 2008-04-20 18:27
<br>Message generated for change (Settings changed) made by rogertsang
<br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405837&aid=1947441&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405837&aid=1947441&group_id=32541</a><br><br>Please note that this message will contain a full copy of the comment thread,
<br>including the initial issue submission, for this request,
<br>not just the latest update.
<br>Category: Load Balancing
<br>Group: 1.9
<br>Status: Open
<br>>Resolution: Fixed
<br>Priority: 5
<br>Private: No
<br>Submitted By: Roger Tsang (rogertsang)
<br>Assigned to: Roger Tsang (rogertsang)
<br>Summary: process migration memory balancing
<br><br>Initial Comment:
<br>Finish porting this feature and enable in 1.9
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2008-04-20 18:27
<br><br>Message:
<br>Logged In: YES
<br>user_id=1246761
<br>Originator: YES
<br><br>checked into CVS
<br><br>----------------------------------------------------------------------
<br><br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405837&aid=1947441&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405837&aid=1947441&group_id=32541</a><br><br>------------------------------------------------------------------------------
<br>Come build with us! The BlackBerry(R) Developer Conference in SF, CA
<br>is the only developer event you need to attend this year. Jumpstart your
<br>developing skills, take BlackBerry mobile applications to market and stay
<br>ahead of the curve. Join us from November 9 - 12, 2009. Register now!
<br><a href="http://p.sf.net/sfu/devconference" target="_top" rel="nofollow">http://p.sf.net/sfu/devconference</a><br>_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26071742&i=0" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-26071741[ ssic-linux-Feature Requests-1947318 ] CFS async increase priority for wb_priority / swapfile2009-10-26T20:44:39Z2009-10-26T20:44:39ZSourceForge.netFeature Requests item #1947318, was opened at 2008-04-20 14:03
<br>Message generated for change (Settings changed) made by rogertsang
<br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405837&aid=1947318&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405837&aid=1947318&group_id=32541</a><br><br>Please note that this message will contain a full copy of the comment thread,
<br>including the initial issue submission, for this request,
<br>not just the latest update.
<br>Category: Filesystem
<br>Group: 1.9
<br>Status: Open
<br>>Resolution: Fixed
<br>Priority: 5
<br>Private: No
<br>Submitted By: Roger Tsang (rogertsang)
<br>Assigned to: Roger Tsang (rogertsang)
<br>Summary: CFS async increase priority for wb_priority / swapfile
<br><br>Initial Comment:
<br>Async respect memory priority
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2008-04-20 14:04
<br><br>Message:
<br>Logged In: YES
<br>user_id=1246761
<br>Originator: YES
<br><br>checked into CVS
<br><br>----------------------------------------------------------------------
<br><br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405837&aid=1947318&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405837&aid=1947318&group_id=32541</a><br><br>------------------------------------------------------------------------------
<br>Come build with us! The BlackBerry(R) Developer Conference in SF, CA
<br>is the only developer event you need to attend this year. Jumpstart your
<br>developing skills, take BlackBerry mobile applications to market and stay
<br>ahead of the curve. Join us from November 9 - 12, 2009. Register now!
<br><a href="http://p.sf.net/sfu/devconference" target="_top" rel="nofollow">http://p.sf.net/sfu/devconference</a><br>_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26071741&i=0" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-26071729[ ssic-linux-Feature Requests-2525489 ] remote unix_dgram_sendmsg() support >1 data block2009-10-26T20:43:49Z2009-10-26T20:43:49ZSourceForge.netFeature Requests item #2525489, was opened at 2009-01-20 21:30
<br>Message generated for change (Comment added) made by rogertsang
<br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2525489&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2525489&group_id=32541</a><br><br>Please note that this message will contain a full copy of the comment thread,
<br>including the initial issue submission, for this request,
<br>not just the latest update.
<br>Category: IPC
<br>Group: 1.9
<br>Status: Open
<br>>Resolution: Fixed
<br>Priority: 5
<br>Private: No
<br>Submitted By: Roger Tsang (rogertsang)
<br>Assigned to: Roger Tsang (rogertsang)
<br>Summary: remote unix_dgram_sendmsg() support >1 data block
<br><br>Initial Comment:
<br>Remote unix_dgram_sendmsg() add support for more than one data block
<br><br>----------------------------------------------------------------------
<br><br>>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-10-26 23:43
<br><br>Message:
<br>checked-in
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-01-20 21:30
<br><br>Message:
<br>going in 1.9.6
<br><br>----------------------------------------------------------------------
<br><br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2525489&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2525489&group_id=32541</a><br><br>------------------------------------------------------------------------------
<br>Come build with us! The BlackBerry(R) Developer Conference in SF, CA
<br>is the only developer event you need to attend this year. Jumpstart your
<br>developing skills, take BlackBerry mobile applications to market and stay
<br>ahead of the curve. Join us from November 9 - 12, 2009. Register now!
<br><a href="http://p.sf.net/sfu/devconference" target="_top" rel="nofollow">http://p.sf.net/sfu/devconference</a><br>_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26071729&i=0" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-26071725[ ssic-linux-Feature Requests-2559191 ] Interruptible CFS token request2009-10-26T20:43:26Z2009-10-26T20:43:26ZSourceForge.netFeature Requests item #2559191, was opened at 2009-02-02 21:50
<br>Message generated for change (Comment added) made by rogertsang
<br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2559191&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2559191&group_id=32541</a><br><br>Please note that this message will contain a full copy of the comment thread,
<br>including the initial issue submission, for this request,
<br>not just the latest update.
<br>Category: System Architecture
<br>Group: 1.9
<br>Status: Open
<br>>Resolution: Fixed
<br>Priority: 5
<br>Private: No
<br>Submitted By: Roger Tsang (rogertsang)
<br>Assigned to: Roger Tsang (rogertsang)
<br>Summary: Interruptible CFS token request
<br><br>Initial Comment:
<br>Appropriate SMP-safe handling is added so that blocking calls to DLM can be interrupted. This feature allows better interactivity.
<br><br>----------------------------------------------------------------------
<br><br>>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-10-26 23:43
<br><br>Message:
<br>checked-in
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-07-14 02:08
<br><br>Message:
<br>testing fix for "extra tok_release" warnings
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-03-28 17:43
<br><br>Message:
<br>postponed due to getting extra tok_release warnings with CONFIG_PREEMPT
<br><br>----------------------------------------------------------------------
<br><br>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-02-02 21:51
<br><br>Message:
<br>going in 1.9.6
<br><br>----------------------------------------------------------------------
<br><br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2559191&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2559191&group_id=32541</a><br><br>------------------------------------------------------------------------------
<br>Come build with us! The BlackBerry(R) Developer Conference in SF, CA
<br>is the only developer event you need to attend this year. Jumpstart your
<br>developing skills, take BlackBerry mobile applications to market and stay
<br>ahead of the curve. Join us from November 9 - 12, 2009. Register now!
<br><a href="http://p.sf.net/sfu/devconference" target="_top" rel="nofollow">http://p.sf.net/sfu/devconference</a><br>_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26071725&i=0" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-26071724[ ssic-linux-Feature Requests-2881146 ] CFS to utilize TCP sendpage support in ICS2009-10-26T20:43:07Z2009-10-26T20:43:07ZSourceForge.netFeature Requests item #2881146, was opened at 2009-10-17 22:47
<br>Message generated for change (Comment added) made by rogertsang
<br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2881146&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2881146&group_id=32541</a><br><br>Please note that this message will contain a full copy of the comment thread,
<br>including the initial issue submission, for this request,
<br>not just the latest update.
<br>Category: Filesystem
<br>Group: 1.9
<br>Status: Open
<br>>Resolution: Fixed
<br>Priority: 5
<br>Private: No
<br>Submitted By: Roger Tsang (rogertsang)
<br>Assigned to: Roger Tsang (rogertsang)
<br>Summary: CFS to utilize TCP sendpage support in ICS
<br><br>Initial Comment:
<br>In [ ssic-linux-Feature 1738728 ] CFS is optimized to send coalesced / multiple requests per operation via an array of iovec's instead of one request per operation. With the new sendpage support in [ ci-linux-Feature 2881140 ] we take this one step further and push the pages (the coalesced requests) directly on the wire and eliminate various overhead associated with iovec's at multiple layers.
<br><br>----------------------------------------------------------------------
<br><br>>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-10-26 23:43
<br><br>Message:
<br>checked-in
<br><br>----------------------------------------------------------------------
<br><br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2881146&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2881146&group_id=32541</a><br><br>------------------------------------------------------------------------------
<br>Come build with us! The BlackBerry(R) Developer Conference in SF, CA
<br>is the only developer event you need to attend this year. Jumpstart your
<br>developing skills, take BlackBerry mobile applications to market and stay
<br>ahead of the curve. Join us from November 9 - 12, 2009. Register now!
<br><a href="http://p.sf.net/sfu/devconference" target="_top" rel="nofollow">http://p.sf.net/sfu/devconference</a><br>_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26071724&i=0" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-26071718[ ssic-linux-Feature Requests-2881147 ] CFS to support ~1MB file IO buffer2009-10-26T20:42:30Z2009-10-26T20:42:30ZSourceForge.netFeature Requests item #2881147, was opened at 2009-10-17 22:55
<br>Message generated for change (Settings changed) made by rogertsang
<br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2881147&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2881147&group_id=32541</a><br><br>Please note that this message will contain a full copy of the comment thread,
<br>including the initial issue submission, for this request,
<br>not just the latest update.
<br>Category: Filesystem
<br>Group: 1.9
<br>Status: Open
<br>>Resolution: Fixed
<br>Priority: 5
<br>Private: No
<br>Submitted By: Roger Tsang (rogertsang)
<br>Assigned to: Roger Tsang (rogertsang)
<br>Summary: CFS to support ~1MB file IO buffer
<br><br>Initial Comment:
<br>Increase the maximum size of each CFS file I/O request by more than one factor or to approx. one megabyte.
<br>Depends on [ ssic-linux-Feature 2881146 ]
<br><br>----------------------------------------------------------------------
<br><br>>Comment By: Roger Tsang (rogertsang)
<br>Date: 2009-10-26 23:42
<br><br>Message:
<br>checked in
<br><br>----------------------------------------------------------------------
<br><br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2881147&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2881147&group_id=32541</a><br><br>------------------------------------------------------------------------------
<br>Come build with us! The BlackBerry(R) Developer Conference in SF, CA
<br>is the only developer event you need to attend this year. Jumpstart your
<br>developing skills, take BlackBerry mobile applications to market and stay
<br>ahead of the curve. Join us from November 9 - 12, 2009. Register now!
<br><a href="http://p.sf.net/sfu/devconference" target="_top" rel="nofollow">http://p.sf.net/sfu/devconference</a><br>_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26071718&i=0" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-25976499Re: OpenSSI under Xen HVM2009-10-20T07:48:21Z2009-10-20T07:48:21ZJohn HughesJohn Hughes wrote:
<br>> Here's my /etc/xen/scripts/openssi-bridge script to set up two bridges:
<br>>
<br>> #!/bin/sh
<br>> # network-xen-custom
<br>> # Exit if anything goes wrong
<br>> set -e
<br>>
<br>> interfaces="calvaedi openssi"
<br>>
<br>^ obviously should be
<br> interfaces="eth0 eth1"
<br>> dir=`dirname $0`
<br>> script=$dir/network-bridge
<br>>
<br>> for i in ${interfaces}
<br>> do
<br>> $script $1 netdev=$i
<br>> done
<br>>
<br><br><br>------------------------------------------------------------------------------
<br>Come build with us! The BlackBerry(R) Developer Conference in SF, CA
<br>is the only developer event you need to attend this year. Jumpstart your
<br>developing skills, take BlackBerry mobile applications to market and stay
<br>ahead of the curve. Join us from November 9 - 12, 2009. Register now!
<br><a href="http://p.sf.net/sfu/devconference" target="_top" rel="nofollow">http://p.sf.net/sfu/devconference</a><br>_______________________________________________
<br>Ssic-linux-users mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=25976499&i=0" target="_top" rel="nofollow">Ssic-linux-users@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-users" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-users</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-users-f3276.html" embed="fixTarget[3276]" target="_top" >ssic-linux-users</a></p>tag:old.nabble.com,2006:post-25973111OpenSSI under Xen HVM2009-10-20T03:52:54Z2009-10-20T03:52:54ZJohn HughesAnyone wanting to play with OpenSSI but doesn't want to dedicate a
<br>machine or machines, just a little notice that it seesm to work OK under
<br>Xen HVM (hardware virtual machine, on modern hardware that has HVM
<br>instructions, e.g. Intel Core or recent AMD's). To run in PVM mode the
<br>old OpenSSI/Xen patches would need resurrection.
<br><br>Example Xen config:
<br><br>kernel = '/usr/lib/xen-3.2-1/boot/hvmloader'
<br>builder = 'hvm'
<br>memory = '256'
<br>vcpus = 2
<br>device_model='/usr/lib/xen-3.2-1/bin/qemu-dm'
<br><br># Disks
<br>disk = [
<br> 'file:/home/xen/node1.img,hda,w',
<br> 'phy:/dev/some_vg/openssi.img,hdb,w' ]
<br><br># Hostname
<br>name = 'node1'
<br><br># Networking, one RTL 8139+ for external network and one for
<br># interconnect.
<br># Need specialised bridge script.
<br>vif = ['type=ioemu, bridge=eth0, mac=00:16:3E:52:c1:9B',
<br> 'type=ioemu, bridge=eth1, mac=00:16:3E:52:d1:9B' ]
<br><br># Behaviour
<br>boot='c'
<br><br># console
<br>sdl=0
<br><br>vnc=1
<br>vnclisten='0.0.0.0'
<br><br><br>Here's my /etc/xen/scripts/openssi-bridge script to set up two bridges:
<br><br>#!/bin/sh
<br># network-xen-custom
<br># Exit if anything goes wrong
<br>set -e
<br><br>interfaces="calvaedi openssi"
<br>dir=`dirname $0`
<br>script=$dir/network-bridge
<br><br>for i in ${interfaces}
<br>do
<br> $script $1 netdev=$i
<br>done
<br><br><br><br><br><br>------------------------------------------------------------------------------
<br>Come build with us! The BlackBerry(R) Developer Conference in SF, CA
<br>is the only developer event you need to attend this year. Jumpstart your
<br>developing skills, take BlackBerry mobile applications to market and stay
<br>ahead of the curve. Join us from November 9 - 12, 2009. Register now!
<br><a href="http://p.sf.net/sfu/devconference" target="_top" rel="nofollow">http://p.sf.net/sfu/devconference</a><br>_______________________________________________
<br>Ssic-linux-users mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=25973111&i=0" target="_top" rel="nofollow">Ssic-linux-users@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-users" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-users</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-users-f3276.html" embed="fixTarget[3276]" target="_top" >ssic-linux-users</a></p>tag:old.nabble.com,2006:post-25943657[ ssic-linux-Feature Requests-2881147 ] CFS to support ~1MB file IO buffer2009-10-17T20:08:36Z2009-10-17T20:08:36ZSourceForge.netFeature Requests item #2881147, was opened at 2009-10-17 22:55
<br>Message generated for change (Settings changed) made by rogertsang
<br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2881147&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2881147&group_id=32541</a><br><br>Please note that this message will contain a full copy of the comment thread,
<br>including the initial issue submission, for this request,
<br>not just the latest update.
<br>Category: Filesystem
<br>Group: 1.9
<br>Status: Open
<br>>Resolution: Accepted
<br>Priority: 5
<br>Private: No
<br>Submitted By: Roger Tsang (rogertsang)
<br>Assigned to: Roger Tsang (rogertsang)
<br>Summary: CFS to support ~1MB file IO buffer
<br><br>Initial Comment:
<br>Increase the maximum size of each CFS file I/O request by more than one factor or to approx. one megabyte.
<br>Depends on [ ssic-linux-Feature 2881146 ]
<br><br>----------------------------------------------------------------------
<br><br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2881147&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2881147&group_id=32541</a><br><br>------------------------------------------------------------------------------
<br>Come build with us! The BlackBerry(R) Developer Conference in SF, CA
<br>is the only developer event you need to attend this year. Jumpstart your
<br>developing skills, take BlackBerry mobile applications to market and stay
<br>ahead of the curve. Join us from November 9 - 12, 2009. Register now!
<br><a href="http://p.sf.net/sfu/devconference" target="_top" rel="nofollow">http://p.sf.net/sfu/devconference</a><br>_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=25943657&i=0" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-25943598[ ssic-linux-Feature Requests-2881146 ] CFS to utilize TCP sendpage support in ICS2009-10-17T19:56:00Z2009-10-17T19:56:00ZSourceForge.netFeature Requests item #2881146, was opened at 2009-10-17 22:47
<br>Message generated for change (Settings changed) made by rogertsang
<br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2881146&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2881146&group_id=32541</a><br><br>Please note that this message will contain a full copy of the comment thread,
<br>including the initial issue submission, for this request,
<br>not just the latest update.
<br>Category: Filesystem
<br>Group: 1.9
<br>Status: Open
<br>>Resolution: Accepted
<br>Priority: 5
<br>Private: No
<br>Submitted By: Roger Tsang (rogertsang)
<br>Assigned to: Roger Tsang (rogertsang)
<br>Summary: CFS to utilize TCP sendpage support in ICS
<br><br>Initial Comment:
<br>In [ ssic-linux-Feature 1738728 ] CFS is optimized to send coalesced / multiple requests per operation via an array of iovec's instead of one request per operation. With the new sendpage support in [ ci-linux-Feature 2881140 ] we take this one step further and push the pages (the coalesced requests) directly on the wire and eliminate various overhead associated with iovec's at multiple layers.
<br><br>----------------------------------------------------------------------
<br><br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2881146&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2881146&group_id=32541</a><br><br>------------------------------------------------------------------------------
<br>Come build with us! The BlackBerry(R) Developer Conference in SF, CA
<br>is the only developer event you need to attend this year. Jumpstart your
<br>developing skills, take BlackBerry mobile applications to market and stay
<br>ahead of the curve. Join us from November 9 - 12, 2009. Register now!
<br><a href="http://p.sf.net/sfu/devconference" target="_top" rel="nofollow">http://p.sf.net/sfu/devconference</a><br>_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=25943598&i=0" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-25943593[ ssic-linux-Feature Requests-2881147 ] CFS to support ~1MB file IO buffer2009-10-17T19:55:24Z2009-10-17T19:55:24ZSourceForge.netFeature Requests item #2881147, was opened at 2009-10-17 22:55
<br>Message generated for change (Tracker Item Submitted) made by rogertsang
<br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2881147&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2881147&group_id=32541</a><br><br>Please note that this message will contain a full copy of the comment thread,
<br>including the initial issue submission, for this request,
<br>not just the latest update.
<br>Category: Filesystem
<br>Group: 1.9
<br>Status: Open
<br>Resolution: None
<br>Priority: 5
<br>Private: No
<br>Submitted By: Roger Tsang (rogertsang)
<br>Assigned to: Roger Tsang (rogertsang)
<br>Summary: CFS to support ~1MB file IO buffer
<br><br>Initial Comment:
<br>Increase the maximum size of each CFS file I/O request by more than one factor or to approx. one megabyte.
<br>Depends on [ ssic-linux-Feature 2881146 ]
<br><br>----------------------------------------------------------------------
<br><br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2881147&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2881147&group_id=32541</a><br><br>------------------------------------------------------------------------------
<br>Come build with us! The BlackBerry(R) Developer Conference in SF, CA
<br>is the only developer event you need to attend this year. Jumpstart your
<br>developing skills, take BlackBerry mobile applications to market and stay
<br>ahead of the curve. Join us from November 9 - 12, 2009. Register now!
<br><a href="http://p.sf.net/sfu/devconference" target="_top" rel="nofollow">http://p.sf.net/sfu/devconference</a><br>_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=25943593&i=0" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-25943571[ ssic-linux-Feature Requests-2881146 ] CFS to utilize TCP sendpage support in ICS2009-10-17T19:47:22Z2009-10-17T19:47:22ZSourceForge.netFeature Requests item #2881146, was opened at 2009-10-17 22:47
<br>Message generated for change (Tracker Item Submitted) made by rogertsang
<br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2881146&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2881146&group_id=32541</a><br><br>Please note that this message will contain a full copy of the comment thread,
<br>including the initial issue submission, for this request,
<br>not just the latest update.
<br>Category: Filesystem
<br>Group: 1.9
<br>Status: Open
<br>Resolution: None
<br>Priority: 5
<br>Private: No
<br>Submitted By: Roger Tsang (rogertsang)
<br>Assigned to: Roger Tsang (rogertsang)
<br>Summary: CFS to utilize TCP sendpage support in ICS
<br><br>Initial Comment:
<br>In [ ssic-linux-Feature 1738728 ] CFS is optimized to send coalesced / multiple requests per operation via an array of iovec's instead of one request per operation. With the new sendpage support in [ ci-linux-Feature 2881140 ] we take this one step further and push the pages (the coalesced requests) directly on the wire and eliminate various overhead associated with iovec's at multiple layers.
<br><br>----------------------------------------------------------------------
<br><br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2881146&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405837&aid=2881146&group_id=32541</a><br><br>------------------------------------------------------------------------------
<br>Come build with us! The BlackBerry(R) Developer Conference in SF, CA
<br>is the only developer event you need to attend this year. Jumpstart your
<br>developing skills, take BlackBerry mobile applications to market and stay
<br>ahead of the curve. Join us from November 9 - 12, 2009. Register now!
<br><a href="http://p.sf.net/sfu/devconference" target="_top" rel="nofollow">http://p.sf.net/sfu/devconference</a><br>_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=25943571&i=0" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-25937523Re: What I did on my holidays.2009-10-17T04:28:33Z2009-10-17T04:28:33ZJohn HughesJohn Hughes wrote:
<div class='shrinkable-quote'><br>> Since I had some free time on my holidays I've been attempting to
<br>> forward port current OpenSSI.
<br>>
<br>> Due to the poor working conditions (the only internet access I was able
<br>> to get going was via my cellphone, about 1 EUR per megabyte at around
<br>> 8Kbps) I didn't get as far as I wanted, but I've got a 2.6.15 based
<br>> kernel that boots and forms a cluster.
<br>>
<br>> So far it has one major bug (ptrace oopses) and one regression (exec
<br>> from a thread never finishes).
<br>>
<br>> I'll try and fix those problems then arrange to host the source somewhere.
<br>>
</div>I've fixed the strace crash, although cross node strace is not working
<br>at the moment.
<br><br>You can get a copy of what I'm working on using git from
<br><br><a href="http://perso.calvaedi.com/~john/openssi-future.git" target="_top" rel="nofollow">http://perso.calvaedi.com/~john/openssi-future.git</a><br><br>------------------------------------------------------------------------------
<br>Come build with us! The BlackBerry(R) Developer Conference in SF, CA
<br>is the only developer event you need to attend this year. Jumpstart your
<br>developing skills, take BlackBerry mobile applications to market and stay
<br>ahead of the curve. Join us from November 9 - 12, 2009. Register now!
<br><a href="http://p.sf.net/sfu/devconference" target="_top" rel="nofollow">http://p.sf.net/sfu/devconference</a><br>_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=25937523&i=0" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>tag:old.nabble.com,2006:post-25891408[ ssic-linux-Bugs-2878871 ] ha-lvs sometimes starts before interfaces2009-10-14T06:45:26Z2009-10-14T06:45:26ZSourceForge.netBugs item #2878871, was opened at 2009-10-14 15:44
<br>Message generated for change (Settings changed) made by hughesj
<br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2878871&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2878871&group_id=32541</a><br><br>Please note that this message will contain a full copy of the comment thread,
<br>including the initial issue submission, for this request,
<br>not just the latest update.
<br>>Category: Booting / init
<br>Group: None
<br>Status: Open
<br>Resolution: None
<br>Priority: 5
<br>Private: No
<br>Submitted By: John Hughes (hughesj)
<br>Assigned to: Nobody/Anonymous (nobody)
<br>Summary: ha-lvs sometimes starts before interfaces
<br><br>Initial Comment:
<br>Under some currently unknown circumstances ha-lvs starts before the network interfaces, preventing it from adding the CVIP address. Stopping and restarting ha-lvs "fixes" the problem.
<br><br>Log:
<br>EXT3 FS on hda1, internal journal
<br>EXT3-fs: mounted filesystem with ordered data mode.
<br>IPVS: sync thread started: state = MASTER, mcast_ifn = eth1, syncid = 0
<br>eth0: link up, 100Mbps, full-duplex, lpa 0x05E1
<br>eth1: no IPv6 routers present
<br>eth0: no IPv6 routers present
<br><br>Seen running OpenSSI kernels 2.6.12 and 2.6.14 in a XEN hvm (hardware virtual machine).
<br><br><br>----------------------------------------------------------------------
<br><br>You can respond by visiting:
<br><a href="https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2878871&group_id=32541" target="_top" rel="nofollow">https://sourceforge.net/tracker/?func=detail&atid=405834&aid=2878871&group_id=32541</a><br><br>------------------------------------------------------------------------------
<br>Come build with us! The BlackBerry(R) Developer Conference in SF, CA
<br>is the only developer event you need to attend this year. Jumpstart your
<br>developing skills, take BlackBerry mobile applications to market and stay
<br>ahead of the curve. Join us from November 9 - 12, 2009. Register now!
<br><a href="http://p.sf.net/sfu/devconference" target="_top" rel="nofollow">http://p.sf.net/sfu/devconference</a><br>_______________________________________________
<br>ssic-linux-devel mailing list
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=25891408&i=0" target="_top" rel="nofollow">ssic-linux-devel@...</a>
<br><a href="https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel</a><br><p>From forum: <a href="http://old.nabble.com/ssic-linux-devel-f3275.html" embed="fixTarget[3275]" target="_top" >ssic-linux-devel</a></p>