Attachment 59098 Details for Bug 44397 – Patch to catch loops of length > 1

[patch] Patch to catch loops of length > 1

prevent-filter-loops.patch (text/plain), 1.73 KB, created by Arun Raghavan on 2012-03-26 20:19:05 UTC

(hide)

Description:

Filename:

MIME Type:

Creator: Arun Raghavan

Created: 2012-03-26 20:19:05 UTC

Size: 1.73 KB

patch

obsolete

>diff --git a/src/modules/module-remap-sink.c b/src/modules/module-remap-sink.c
>index 2822a7f..0c216d4 100644
>--- a/src/modules/module-remap-sink.c
>+++ b/src/modules/module-remap-sink.c
>@@ -291,7 +291,7 @@ static pa_bool_t sink_input_may_move_to_cb(pa_sink_input *i, pa_sink *dest) {
>     pa_sink_input_assert_ref(i);
>     pa_assert_se(u = i->userdata);
> 
>-    return u->sink != dest;
>+    return (u->sink != dest) && pa_sink_check_loop(dest, u->module);
> }
> 
> /* Called from main context */
>diff --git a/src/pulsecore/sink.c b/src/pulsecore/sink.c
>index 2d214cf..9c7e2bb 100644
>--- a/src/pulsecore/sink.c
>+++ b/src/pulsecore/sink.c
>@@ -1456,6 +1456,17 @@ pa_sink *pa_sink_get_master(pa_sink *s) {
>     return s;
> }
> 
>+/* Called from the main thread */
>+/* Checks if "dest" is loaded by another instance of "m". This is meant to be
>+ * used by filter modules to verify that they are not causing a loop. */
>+pa_bool_t pa_sink_check_loop(pa_sink *dest, pa_module *m) {
>+    pa_assert(dest->module);
>+    pa_assert(dest->module->name);
>+    pa_assert(m->name);
>+
>+    return pa_streq(dest->module->name, m->name);
>+}
>+
> /* Called from main context */
> pa_bool_t pa_sink_is_passthrough(pa_sink *s) {
>     pa_sink_input *alt_i;
>diff --git a/src/pulsecore/sink.h b/src/pulsecore/sink.h
>index 56fa735..812105f 100644
>--- a/src/pulsecore/sink.h
>+++ b/src/pulsecore/sink.h
>@@ -421,6 +421,8 @@ pa_bool_t pa_sink_flat_volume_enabled(pa_sink *s);
> 
> /* Get the master sink when sharing volumes */
> pa_sink *pa_sink_get_master(pa_sink *s);
>+/* Make sure sink dest isn't loaded by an instance of module m */
>+pa_bool_t pa_sink_check_loop(pa_sink *dest, pa_module *m);
> 
> /* Is the sink in passthrough mode? (that is, is there a passthrough sink input
>  * connected to this sink? */

Actions: View | Splinter Review

Attachments on bug 44397: 55057 | 55942 | 55943 | 58920 | 58966 | 58967 | 58968 | 59067 | 59079 | 59098 | 59107

Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.