public inbox for pve-devel@lists.proxmox.com
 help / color / mirror / Atom feed
From: Shannon Sterz <s.sterz@proxmox.com>
To: pve-devel@lists.proxmox.com
Subject: [PATCH cluster 3/4] pmxcfs: dfsm: wake waiters in dfsm_send_message_sync() on error mode
Date: Mon, 16 Nov 2026 17:01:30 +0100	[thread overview]
Message-ID: <20261116160131.344857-4-s.sterz@proxmox.com> (raw)
In-Reply-To: <20261116160131.344857-1-s.sterz@proxmox.com>

if the dfsm switches to an abnormal state (`>= DFSM_ERROR_MODE_START`)
and a thread is waiting in `dfsm_send_message_sync()` for a message to
return, it will wait indefinitely. this is because only two functions
will broadcast on the condition `&dfsm->sync_cond` that is waited for:

- `dfsm_record_local_result()`: only called by
  `dfsm_cpg_deliver_callback()` for local messages and only if mode is
  `< DFSM_ERROR_MODE_START`.
- `dfsm_send_sync_message_abort()`: only called on `dfsm_finalize()`
  or `dfsm_initialize()`. which are only called when the dfsm service
  is finalized or initialized; not on error modes.

further a call to `dcdb_deliver()` can call `dfsm_set_errormode()`
during a local delivery. then right after the abort
`dfsm_record_local_result()` will set the `msgcount_rcvd` back to that
message's number. a waiter in `dfsm_send_message_sync()` will be
woken, but the `msgcount_rcvd` will not be increased by default.
meaning the thread goes back to waiting. so a check on whether an
error mode here is set is necessary.

this patch does three things:

- wake any waiter of `&dfsm->sync_cond` upon entering an abnormal mode
- return an error on `dfsm_send_message_sync()` when in an error mode
- check whether dfsm is in an error mode when awoken in
  `dfsm_send_message_sync()`

Suggested-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
Signed-off-by: Shannon Sterz <s.sterz@proxmox.com>
---
 src/pmxcfs/dfsm.c | 15 ++++++++++++++-
 1 file changed, 14 insertions(+), 1 deletion(-)

diff --git a/src/pmxcfs/dfsm.c b/src/pmxcfs/dfsm.c
index 741db54..a5f241c 100644
--- a/src/pmxcfs/dfsm.c
+++ b/src/pmxcfs/dfsm.c
@@ -145,6 +145,7 @@ struct dfsm {
 
 static gboolean dfsm_deliver_queue(dfsm_t *dfsm);
 static gboolean dfsm_deliver_sync_queue(dfsm_t *dfsm);
+static dfsm_mode_t dfsm_get_mode(dfsm_t *dfsm);
 
 gboolean dfsm_nodeid_is_local(dfsm_t *dfsm, uint32_t nodeid, uint32_t pid) {
     g_return_val_if_fail(dfsm != NULL, FALSE);
@@ -262,6 +263,13 @@ cs_error_t dfsm_send_message_sync(
     g_return_val_if_fail(!len || iov != NULL, CS_ERR_INVALID_PARAM);
 
     g_mutex_lock(&dfsm->sync_mutex);
+    /* abort sending if we are in an abnormal state */
+    if (dfsm_get_mode(dfsm) >= DFSM_ERROR_MODE_START) {
+        cfs_dom_debug(dfsm->log_domain, "error mode - refusing to send message");
+        g_mutex_unlock(&dfsm->sync_mutex);
+        return CS_ERR_NOT_EXIST;
+    }
+
     /* note: hold lock until message is sent - to guarantee ordering */
     uint64_t msgcount = ++dfsm->msgcount;
     if (rp) {
@@ -303,7 +311,7 @@ cs_error_t dfsm_send_message_sync(
     if (rp) {
         g_mutex_lock(&dfsm->sync_mutex);
 
-        while (dfsm->msgcount_rcvd < msgcount) {
+        while (dfsm->msgcount_rcvd < msgcount && dfsm_get_mode(dfsm) < DFSM_ERROR_MODE_START) {
             g_cond_wait(&dfsm->sync_cond, &dfsm->sync_mutex);
         }
 
@@ -476,6 +484,11 @@ static void dfsm_set_mode(dfsm_t *dfsm, dfsm_mode_t new_mode) {
         return;
     }
 
+    if (new_mode >= DFSM_ERROR_MODE_START) {
+        /* abort waiting for sync messages to arrive if we are in an abnormal state */
+        dfsm_send_sync_message_abort(dfsm);
+    }
+
     if (new_mode == DFSM_MODE_START) {
         cfs_dom_message(dfsm->log_domain, "start cluster connection");
     } else if (new_mode == DFSM_MODE_START_SYNC) {
-- 
2.47.3





  parent reply	other threads:[~2026-11-16 16:01 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-11-16 16:01 [PATCH cluster 0/4] Fix 8325 and minor pmxcfs Clean-Up Shannon Sterz
2026-11-16 16:01 ` [PATCH cluster 1/4] tree wide: update copyright notice to current year Shannon Sterz
2026-11-16 16:01 ` [PATCH cluster 2/4] fix #8325: pmxcfs: dfsm: abort send if dfsm is not joined via a handle Shannon Sterz
2026-11-16 16:01 ` Shannon Sterz [this message]
2026-11-16 16:01 ` [PATCH cluster 4/4] pmxcfs: dfsm: make sure error results are processed correctly Shannon Sterz
2026-11-17  1:32 ` applied: [PATCH cluster 0/4] Fix 8325 and minor pmxcfs Clean-Up Thomas Lamprecht

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20261116160131.344857-4-s.sterz@proxmox.com \
    --to=s.sterz@proxmox.com \
    --cc=pve-devel@lists.proxmox.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
Service provided by Proxmox Server Solutions GmbH | Privacy | Legal