On Thu, Apr 18, 2024 at 01:48 PM -0500, Jonathon Jongsma <jjongsma@redhat.com> wrote:
On 4/12/24 8:36 AM, Marc Hartmayer wrote:
Since @driver->privateData is modified take the lock.
[…snip…]
* signal if that event never comes */ - scheduleMdevctlUpdate(priv, (event_type == G_FILE_MONITOR_EVENT_CHANGES_DONE_HINT)); + VIR_WITH_OBJECT_LOCK_GUARD(priv) { + scheduleMdevctlUpdate(priv, (event_type == G_FILE_MONITOR_EVENT_CHANGES_DONE_HINT)); + } }
I don't see any cases where we would not want to take the lock (e.g. because it has already been taken), so maybe it would be better simply to lock the object within the scheduleMdevctlUpdate() function rather than requiring each caller to lock it?
Hmm, with patch 15 in mind this would make the code more complicated.
Reviewed-by: Jonathon Jongsma <jjongsma@redhat.com>
-- Kind regards / Beste Grüße Marc Hartmayer IBM Deutschland Research & Development GmbH Vorsitzender des Aufsichtsrats: Wolfgang Wendt Geschäftsführung: David Faller Sitz der Gesellschaft: Böblingen Registergericht: Amtsgericht Stuttgart, HRB 243294