<?xml version="1.0"?>
<?xml-stylesheet type="text/xsl" href="/rss.xsl.xml"?>
<rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/">
<channel>
    <title>Changes in dlm.h</title>
    <description></description>
    <language>en</language>
    <copyright>Copyright 2015</copyright>
    <generator>Java</generator><item>
        <title>1131f339 - dlm: remove lkb from callback tracepoints</title>
        <link>http://172.16.0.5:8080/history/linux-6.15/include/trace/events/dlm.h#1131f339</link>
        <description>dlm: remove lkb from callback tracepointsStop using lkb structs in the callback tracepoints so that lkbreferences are not needed. This prepares for separating lkbstructs from callbacks.Signed-off-by: Alexander Aring &lt;aahringo@redhat.com&gt;Signed-off-by: David Teigland &lt;teigland@redhat.com&gt;

            List of files:
            /linux-6.15/include/trace/events/dlm.h</description>
        <pubDate>Thu, 28 Mar 2024 15:48:37 +0000</pubDate>
        <dc:creator>Alexander Aring &lt;aahringo@redhat.com&gt;</dc:creator>
    </item>
<item>
        <title>8c95006d - fs: dlm: add plock dev tracepoints</title>
        <link>http://172.16.0.5:8080/history/linux-6.15/include/trace/events/dlm.h#8c95006d</link>
        <description>fs: dlm: add plock dev tracepointsI currently debug nfs plock handling and introduce those two tracepointsfor getting more information about what is happening there if the userspace reads plock operations from kernel and writing the result back.Signed-off-by: Alexander Aring &lt;aahringo@redhat.com&gt;Signed-off-by: David Teigland &lt;teigland@redhat.com&gt;

            List of files:
            /linux-6.15/include/trace/events/dlm.h</description>
        <pubDate>Tue, 01 Aug 2023 18:09:42 +0000</pubDate>
        <dc:creator>Alexander Aring &lt;aahringo@redhat.com&gt;</dc:creator>
    </item>
<item>
        <title>8a39dcd9 - fs: dlm: change dflags to use atomic bits</title>
        <link>http://172.16.0.5:8080/history/linux-6.15/include/trace/events/dlm.h#8a39dcd9</link>
        <description>fs: dlm: change dflags to use atomic bitsCurrently manipulating lkb_dflags assumes to held the rsb lock assignedto the lkb. This is held by dlm message processing after certaintime to lookup the right rsb from the received lkb message id. For userspace locks flags, which is currently the only use case for lkb_dflags,flags are also being set during dlm character device handling withoutholding the rsb lock. To minimize the risk that bit operations aregetting corrupted we switch to atomic bit operations. This patch willalso introduce helpers to snapshot atomic bit values in an non atomicway. There might be still issues with the flag handling e.g. running incase of manipulating bit ops and snapshot them at the same time, but thispatch minimize them and will start to use atomic bit operations.Signed-off-by: Alexander Aring &lt;aahringo@redhat.com&gt;Signed-off-by: David Teigland &lt;teigland@redhat.com&gt;

            List of files:
            /linux-6.15/include/trace/events/dlm.h</description>
        <pubDate>Mon, 06 Mar 2023 20:48:15 +0000</pubDate>
        <dc:creator>Alexander Aring &lt;aahringo@redhat.com&gt;</dc:creator>
    </item>
<item>
        <title>8c11ba64 - fs: dlm: store lkb distributed flags into own value</title>
        <link>http://172.16.0.5:8080/history/linux-6.15/include/trace/events/dlm.h#8c11ba64</link>
        <description>fs: dlm: store lkb distributed flags into own valueThis patch stores lkb distributed flags value in an separate valueinstead of sharing internal and distributed flags in lkb-&gt;lkb_flags value.This has the advantage to not mask/write back flag values inreceive_flags() functionality. The dlm debug_fs does not provide thedistributed flags anymore, those can be added in future.Signed-off-by: Alexander Aring &lt;aahringo@redhat.com&gt;Signed-off-by: David Teigland &lt;teigland@redhat.com&gt;

            List of files:
            /linux-6.15/include/trace/events/dlm.h</description>
        <pubDate>Mon, 06 Mar 2023 20:48:14 +0000</pubDate>
        <dc:creator>Alexander Aring &lt;aahringo@redhat.com&gt;</dc:creator>
    </item>
<item>
        <title>9f48eead - fs: dlm: remove DLM_IFL_LOCAL_MS flag</title>
        <link>http://172.16.0.5:8080/history/linux-6.15/include/trace/events/dlm.h#9f48eead</link>
        <description>fs: dlm: remove DLM_IFL_LOCAL_MS flagThe DLM_IFL_LOCAL_MS flag is an internal non shared flag but used inm_flags of dlm messages. It is not shared because it is only used forlocal messaging. Instead using DLM_IFL_LOCAL_MS in dlm messages we pass aparameter around to signal local messaging or not. This patch is addingthe local parameter to signal local messaging.Signed-off-by: Alexander Aring &lt;aahringo@redhat.com&gt;Signed-off-by: David Teigland &lt;teigland@redhat.com&gt;

            List of files:
            /linux-6.15/include/trace/events/dlm.h</description>
        <pubDate>Mon, 06 Mar 2023 20:48:13 +0000</pubDate>
        <dc:creator>Alexander Aring &lt;aahringo@redhat.com&gt;</dc:creator>
    </item>
<item>
        <title>a7e7ffac - fs: dlm: rename stub to local message flag</title>
        <link>http://172.16.0.5:8080/history/linux-6.15/include/trace/events/dlm.h#a7e7ffac</link>
        <description>fs: dlm: rename stub to local message flagThis patch renames DLM_IFL_STUB_MS to DLM_IFL_LOCAL_MS flag. TheDLM_IFL_STUB_MS flag is somewhat misnamed, it means the dlm message isused for local message transfer only. It is used by recovery to resolvelock states if a node got fenced.Signed-off-by: Alexander Aring &lt;aahringo@redhat.com&gt;Signed-off-by: David Teigland &lt;teigland@redhat.com&gt;

            List of files:
            /linux-6.15/include/trace/events/dlm.h</description>
        <pubDate>Mon, 06 Mar 2023 20:48:12 +0000</pubDate>
        <dc:creator>Alexander Aring &lt;aahringo@redhat.com&gt;</dc:creator>
    </item>
<item>
        <title>17827754 - fs: dlm: add dst nodeid for msg tracing</title>
        <link>http://172.16.0.5:8080/history/linux-6.15/include/trace/events/dlm.h#17827754</link>
        <description>fs: dlm: add dst nodeid for msg tracingIn DLM when we send a dlm message it is easy to add the lock resourcename, but additional lookup is required when to trace the receivemessage side. The idea here is to move the lookup work to the user byusing a lookup to find the right send message with recv message. As noteDLM can&apos;t drop any message which is guaranteed by a special sessionlayer.For doing the lookup a 3 tupel is required as an unique identificationwhich is dst nodeid, src nodeid and sequence number. This patch adds thedestination nodeid to the dlm message trace points. The source nodeid isgiven by the h_nodeid field inside the header.Signed-off-by: Alexander Aring &lt;aahringo@redhat.com&gt;Signed-off-by: David Teigland &lt;teigland@redhat.com&gt;

            List of files:
            /linux-6.15/include/trace/events/dlm.h</description>
        <pubDate>Thu, 17 Nov 2022 22:11:45 +0000</pubDate>
        <dc:creator>Alexander Aring &lt;aahringo@redhat.com&gt;</dc:creator>
    </item>
<item>
        <title>81889255 - fs: dlm: rename seq to h_seq for msg tracing</title>
        <link>http://172.16.0.5:8080/history/linux-6.15/include/trace/events/dlm.h#81889255</link>
        <description>fs: dlm: rename seq to h_seq for msg tracingThis patch renames seq to h_seq as it is named in the dlm headerstructure.Signed-off-by: Alexander Aring &lt;aahringo@redhat.com&gt;Signed-off-by: David Teigland &lt;teigland@redhat.com&gt;

            List of files:
            /linux-6.15/include/trace/events/dlm.h</description>
        <pubDate>Thu, 17 Nov 2022 22:11:44 +0000</pubDate>
        <dc:creator>Alexander Aring &lt;aahringo@redhat.com&gt;</dc:creator>
    </item>
<item>
        <title>e01c4b7b - fd: dlm: trace send/recv of dlm message and rcom</title>
        <link>http://172.16.0.5:8080/history/linux-6.15/include/trace/events/dlm.h#e01c4b7b</link>
        <description>fd: dlm: trace send/recv of dlm message and rcomThis patch adds tracepoints for send and recv cases of dlm messages anddlm rcom messages. In case of send and dlm message we add the dlm rsbresource name this dlm messages belongs to. This has the advantage tofollow dlm messages on a per lock basis. In case of recv message theresource name can be extracted by follow the send message sequencenumber.The dlm message DLM_MSG_PURGE doesn&apos;t belong to a lock request and willnot set the resource name in a dlm_message trace. The same for all rcommessages.There is additional handling required for this debugging functionalitywhich is tried to be small as possible. Also the midcomms layer getsaware of lock resource names, for now this is required to make aconnection between sequence number and lock resource names. It is fordebugging purpose only.Signed-off-by: Alexander Aring &lt;aahringo@redhat.com&gt;Signed-off-by: David Teigland &lt;teigland@redhat.com&gt;

            List of files:
            /linux-6.15/include/trace/events/dlm.h</description>
        <pubDate>Thu, 27 Oct 2022 20:45:15 +0000</pubDate>
        <dc:creator>Alexander Aring &lt;aahringo@redhat.com&gt;</dc:creator>
    </item>
<item>
        <title>56171e0d - fs: dlm: const void resource name parameter</title>
        <link>http://172.16.0.5:8080/history/linux-6.15/include/trace/events/dlm.h#56171e0d</link>
        <description>fs: dlm: const void resource name parameterThe resource name parameter should never be changed by DLM so we declareit as const. At some point it is handled as a char pointer, a resourcename can be a non printable ascii string as well. This patch change itto handle it as void pointer as it is offered by DLM API.Signed-off-by: Alexander Aring &lt;aahringo@redhat.com&gt;Signed-off-by: David Teigland &lt;teigland@redhat.com&gt;

            List of files:
            /linux-6.15/include/trace/events/dlm.h</description>
        <pubDate>Mon, 15 Aug 2022 19:43:27 +0000</pubDate>
        <dc:creator>Alexander Aring &lt;aahringo@redhat.com&gt;</dc:creator>
    </item>
<item>
        <title>7a3de732 - fs: dlm: trace user space callbacks</title>
        <link>http://172.16.0.5:8080/history/linux-6.15/include/trace/events/dlm.h#7a3de732</link>
        <description>fs: dlm: trace user space callbacksThis patch adds trace callbacks for user locks. Unfortenately user locksare handled in a different way than kernel locks in some cases. Userlocks never call the dlm_lock()/dlm_unlock() kernel API and use the nextstep internal API of dlm. Adding those traces from user API callersshould make it possible for dlm trace system to see lock handling foruser locks as well.Signed-off-by: Alexander Aring &lt;aahringo@redhat.com&gt;Signed-off-by: David Teigland &lt;teigland@redhat.com&gt;

            List of files:
            /linux-6.15/include/trace/events/dlm.h</description>
        <pubDate>Mon, 15 Aug 2022 19:43:24 +0000</pubDate>
        <dc:creator>Alexander Aring &lt;aahringo@redhat.com&gt;</dc:creator>
    </item>
<item>
        <title>5d92a30e - fs: dlm: add resource name to tracepoints</title>
        <link>http://172.16.0.5:8080/history/linux-6.15/include/trace/events/dlm.h#5d92a30e</link>
        <description>fs: dlm: add resource name to tracepointsThis patch adds the resource name to dlm tracepoints.  The nameusually comes through the lkb_resource, but in some cases a resourcemay not yet be associated with an lkb, in which case the name andnamelen parameters are used.It should be okay to access the lkb_resource and the res_name field atthe time when the tracepoint is invoked. The resource is assigned to alkb and it&apos;s reference is being held during the tracepoint call. Duringthis time the resource cannot be freed. Also a lkb will never switchits assigned resource. The name of a dlm_rsb is assigned at creationtime and should never be changed during runtime as well.The TP_printk() call uses always a hexadecimal string arrayrepresentation for the resource name (which is not necessarily ascii.)Signed-off-by: Alexander Aring &lt;aahringo@redhat.com&gt;Signed-off-by: David Teigland &lt;teigland@redhat.com&gt;

            List of files:
            /linux-6.15/include/trace/events/dlm.h</description>
        <pubDate>Wed, 22 Jun 2022 18:45:12 +0000</pubDate>
        <dc:creator>Alexander Aring &lt;aahringo@redhat.com&gt;</dc:creator>
    </item>
<item>
        <title>0c4c516f - fs: dlm: remove additional dereference of lksb</title>
        <link>http://172.16.0.5:8080/history/linux-6.15/include/trace/events/dlm.h#0c4c516f</link>
        <description>fs: dlm: remove additional dereference of lksbThis patch removes a dereference of lksb of lkb when calling asttracepoint. First it reduces additional overhead, even if tracesare not active. Second we can deference it in TP_fast_assign fromthe existing lkb parameter.Signed-off-by: Alexander Aring &lt;aahringo@redhat.com&gt;Signed-off-by: David Teigland &lt;teigland@redhat.com&gt;

            List of files:
            /linux-6.15/include/trace/events/dlm.h</description>
        <pubDate>Wed, 22 Jun 2022 18:45:11 +0000</pubDate>
        <dc:creator>Alexander Aring &lt;aahringo@redhat.com&gt;</dc:creator>
    </item>
<item>
        <title>92732376 - fs: dlm: trace socket handling</title>
        <link>http://172.16.0.5:8080/history/linux-6.15/include/trace/events/dlm.h#92732376</link>
        <description>fs: dlm: trace socket handlingThis patch adds tracepoints for dlm socket receive and sendfunctionality. We can use it to track how much data was send or receivedto or from a specific nodeid.Signed-off-by: Alexander Aring &lt;aahringo@redhat.com&gt;Signed-off-by: David Teigland &lt;teigland@redhat.com&gt;

            List of files:
            /linux-6.15/include/trace/events/dlm.h</description>
        <pubDate>Tue, 02 Nov 2021 19:17:16 +0000</pubDate>
        <dc:creator>Alexander Aring &lt;aahringo@redhat.com&gt;</dc:creator>
    </item>
<item>
        <title>f1d3b8f9 - fs: dlm: initial support for tracepoints</title>
        <link>http://172.16.0.5:8080/history/linux-6.15/include/trace/events/dlm.h#f1d3b8f9</link>
        <description>fs: dlm: initial support for tracepointsThis patch adds initial support for dlm tracepoints. It will introducetracepoints to dlm main functionality dlm_lock()/dlm_unlock() and theircomplete ast() callback or blocking bast() callback.The lock/unlock functionality has a start and end tracepoint, this isbecause there exists a race in case if would have a tracepoint at theend position only the complete/blocking callbacks could occur before. Towork with eBPF tracing and using their lookup hash functionality therecould be problems that an entry was not inserted yet. However use thestart functionality for hash insert and check again in end functionalityif there was an dlm internal error so there is no ast callback. In furtherit might also that locks with local masters will occur those callbacksimmediately so we must have such functionality.I did not make everything accessible yet, although it seems eBPF can beused to access a lot of internal datastructures if it&apos;s aware of thestruct definitions of the running kernel instance. We still can changeit, if you do eBPF experiments e.g. time measurements between lock andcallback functionality you can simple use the local lkb_id field as hashvalue in combination with the lockspace id if you have multiplelockspaces. Otherwise you can simple use trace-cmd for some functionality,e.g. `trace-cmd record -e dlm` and `trace-cmd report` afterwards.Signed-off-by: Alexander Aring &lt;aahringo@redhat.com&gt;Signed-off-by: David Teigland &lt;teigland@redhat.com&gt;

            List of files:
            /linux-6.15/include/trace/events/dlm.h</description>
        <pubDate>Tue, 02 Nov 2021 19:17:15 +0000</pubDate>
        <dc:creator>Alexander Aring &lt;aahringo@redhat.com&gt;</dc:creator>
    </item>
</channel>
</rss>
