diff mbox series

ui/gtk: Fix regression in gd_refresh_rate_milihz

Message ID 3ZSPEQ.MH51LARSPXFN2@gmail.com
State New
Headers show
Series ui/gtk: Fix regression in gd_refresh_rate_milihz | expand

Commit Message

Nikola Pavlica Aug. 7, 2020, 10:25 p.m. UTC 
From: Nikola Pavlica <pavlica.nikola@gmail.com 
<mailto:pavlica.nikola@gmail.com>>

In January (I think), I suggested a fix for the constant refresh rate 
issue in QEMU despite VMs being able to render graphics at a framerate 
much higher than that.

Here's the link to the patch in question:
<https://lists.nongnu.org/archive/html/qemu-devel/2020-01/msg01343.html>

However, I've noticed that the fix for older systems that run a older 
system of GTK fail to compile with my version of the patch. Which was 
fixed here: 
<https://lists.nongnu.org/archive/html/qemu-devel/2020-01/msg03386.html>, 
but improperly.

The issue lies if the target system (that the package is build on) has 
a higher version of GTK, let's say 3.24, it fails the #ifdef macro and 
thus does not include the changes that I've made. I've noticed this 
issue under the ArchLinux build system, which can be fixed by just 
using the included macro for checking GTK versions.

Also being that this is a regression and not a plain bugfix, is this 
mergable into stable?

Signed-off-by: Nikola Pavlica <pavlica.nikola@gmail.com 
<mailto:pavlica.nikola@gmail.com>>
     if (win) {

Comments

Nikola Pavlica Aug. 8, 2020, 1 a.m. UTC | #1 
Apologies, false alarm.

I found out that my testing methodology was flawed. Close this.

On Sat, Aug 8, 2020 at 00:25, Nikola Pavlica <pavlica.nikola@gmail.com> 
wrote:
> From: Nikola Pavlica <pavlica.nikola@gmail.com 
> <mailto:pavlica.nikola@gmail.com>>
> 
> In January (I think), I suggested a fix for the constant refresh rate 
> issue in QEMU despite VMs being able to render graphics at a 
> framerate much higher than that.
> 
> Here's the link to the patch in question:
> <https://lists.nongnu.org/archive/html/qemu-devel/2020-01/msg01343.html>
> 
> However, I've noticed that the fix for older systems that run a older 
> system of GTK fail to compile with my version of the patch. Which was 
> fixed here: 
> <https://lists.nongnu.org/archive/html/qemu-devel/2020-01/msg03386.html>, 
> but improperly.
> 
> The issue lies if the target system (that the package is build on) 
> has a higher version of GTK, let's say 3.24, it fails the #ifdef 
> macro and thus does not include the changes that I've made. I've 
> noticed this issue under the ArchLinux build system, which can be 
> fixed by just using the included macro for checking GTK versions.
> 
> Also being that this is a regression and not a plain bugfix, is this 
> mergable into stable?
> 
> Signed-off-by: Nikola Pavlica <pavlica.nikola@gmail.com 
> <mailto:pavlica.nikola@gmail.com>>
> --- a/ui/gtk.c
> +++ b/ui/gtk.c
> @@ -1968,7 +1968,7 @@ static GtkWidget 
> *gd_create_menu_machine(GtkDisplayState *s)
>   */
>  static int gd_refresh_rate_millihz(GtkWidget *window)
>  {
> -#ifdef GDK_VERSION_3_22
> +#if GTK_CHECK_VERSION(3, 22, 0)
>      GdkWindow *win = gtk_widget_get_window(window);
> 
>      if (win) {
> --
> 2.24.1
diff mbox series

Patch

--- a/ui/gtk.c
+++ b/ui/gtk.c
@@ -1968,7 +1968,7 @@  static GtkWidget 
*gd_create_menu_machine(GtkDisplayState *s)
  */
 static int gd_refresh_rate_millihz(GtkWidget *window)
 {
-#ifdef GDK_VERSION_3_22
+#if GTK_CHECK_VERSION(3, 22, 0)
     GdkWindow *win = gtk_widget_get_window(window);