For a long time now, libvirt has pre-created the monitor connection
socket and passed the pre-opened FD into QEMU during startup. Thus
libvirt does not have any timeouts waiting for the monitor socket
to appear, it is immediately connected.

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
---
 system/vl.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/system/vl.c b/system/vl.c
index 6b87bfa32c..1d1508e28f 100644
--- a/system/vl.c
+++ b/system/vl.c
@@ -1911,7 +1911,6 @@ static bool object_create_early(const char *type)
      * Allocation of large amounts of memory may delay
      * chardev initialization for too long, and trigger timeouts
      * on software that waits for a monitor socket to be created
-     * (e.g. libvirt).
      */
     if (g_str_has_prefix(type, "memory-backend-")) {
         return false;
-- 
2.43.0

On Wed, Jan 03, 2024 at 12:30:42PM +0000, Daniel P. Berrangé wrote:
> Date: Wed,  3 Jan 2024 12:30:42 +0000
> From: "Daniel P. Berrangé" <berrange@redhat.com>
> Subject: [PATCH] softmmu: remove obsolete comment about libvirt timeouts
> 
> For a long time now, libvirt has pre-created the monitor connection
> socket and passed the pre-opened FD into QEMU during startup. Thus
> libvirt does not have any timeouts waiting for the monitor socket
> to appear, it is immediately connected.
> 
> Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
> ---
>  system/vl.c | 1 -
>  1 file changed, 1 deletion(-)
> 
> diff --git a/system/vl.c b/system/vl.c
> index 6b87bfa32c..1d1508e28f 100644
> --- a/system/vl.c
> +++ b/system/vl.c
> @@ -1911,7 +1911,6 @@ static bool object_create_early(const char *type)
>       * Allocation of large amounts of memory may delay
>       * chardev initialization for too long, and trigger timeouts
>       * on software that waits for a monitor socket to be created
> -     * (e.g. libvirt).

From the commit message of 6546d0dba6c2 ("vl: Delay initialization of
memory backends"), and related bugzilla, I understand the only software
Eduardo wanted to describe is libvirt.

Do you know of any other software that has the similar timeout mechanism?
If there is no other software, the description of "trigger timeouts on
software ..." in the comment could be deleted as well.

Otherwise,
Reviewed-by: Zhao Liu <zhao1.liu@intel.com>

Regards,
Zhao

>       */
>      if (g_str_has_prefix(type, "memory-backend-")) {
>          return false;
> -- 
> 2.43.0
> 
> 

On Wed, Jan 03, 2024 at 10:49:20PM +0800, Zhao Liu wrote:
> On Wed, Jan 03, 2024 at 12:30:42PM +0000, Daniel P. Berrangé wrote:
> > Date: Wed,  3 Jan 2024 12:30:42 +0000
> > From: "Daniel P. Berrangé" <berrange@redhat.com>
> > Subject: [PATCH] softmmu: remove obsolete comment about libvirt timeouts
> > 
> > For a long time now, libvirt has pre-created the monitor connection
> > socket and passed the pre-opened FD into QEMU during startup. Thus
> > libvirt does not have any timeouts waiting for the monitor socket
> > to appear, it is immediately connected.
> > 
> > Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
> > ---
> >  system/vl.c | 1 -
> >  1 file changed, 1 deletion(-)
> > 
> > diff --git a/system/vl.c b/system/vl.c
> > index 6b87bfa32c..1d1508e28f 100644
> > --- a/system/vl.c
> > +++ b/system/vl.c
> > @@ -1911,7 +1911,6 @@ static bool object_create_early(const char *type)
> >       * Allocation of large amounts of memory may delay
> >       * chardev initialization for too long, and trigger timeouts
> >       * on software that waits for a monitor socket to be created
> > -     * (e.g. libvirt).
> 
> From the commit message of 6546d0dba6c2 ("vl: Delay initialization of
> memory backends"), and related bugzilla, I understand the only software
> Eduardo wanted to describe is libvirt.
> 
> Do you know of any other software that has the similar timeout mechanism?
> If there is no other software, the description of "trigger timeouts on
> software ..." in the comment could be deleted as well.

I haven't checked any other users recently, but waiting for the monitor
socket with a timeout has been a common usage pattern, so I expect there
are other users still besides libvirt which do this.

> 
> Otherwise,
> Reviewed-by: Zhao Liu <zhao1.liu@intel.com>
> 
> Regards,
> Zhao
> 
> >       */
> >      if (g_str_has_prefix(type, "memory-backend-")) {
> >          return false;
> > -- 
> > 2.43.0
> > 
> > 
> 

With regards,
Daniel
-- 
|: https://berrange.com      -o-    https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org         -o-            https://fstop138.berrange.com :|
|: https://entangle-photo.org    -o-    https://www.instagram.com/dberrange :|