diff --git a/src/Magnum/GL/Implementation/FramebufferState.cpp b/src/Magnum/GL/Implementation/FramebufferState.cpp
index f28cde3a9..0aec3fc09 100644
--- a/src/Magnum/GL/Implementation/FramebufferState.cpp
+++ b/src/Magnum/GL/Implementation/FramebufferState.cpp
@@ -183,8 +183,8 @@ FramebufferState::FramebufferState(Context& context, Containers::StaticArrayView
 
     /* DSA/non-DSA implementation for framebuffer clearing. Yes, it's because
        Intel Windows drivers are shit, and Mesa 24 seems to be stealing bugs
-       from the wrong driver. Doesn't happen on 23 and will hopefully be fixed
-       on 25? */
+       from the wrong driver. Doesn't happen on 23, doesn't seem to happen on
+       25 anymore either. */
     #ifndef MAGNUM_TARGET_GLES
     if(context.isExtensionSupported<Extensions::ARB::direct_state_access>()
         && (!context.versionString().contains("Mesa 24.") ||
diff --git a/src/Magnum/GL/Implementation/driverSpecific.cpp b/src/Magnum/GL/Implementation/driverSpecific.cpp
index 3c4a3d15e..8f25717e4 100644
--- a/src/Magnum/GL/Implementation/driverSpecific.cpp
+++ b/src/Magnum/GL/Implementation/driverSpecific.cpp
@@ -433,8 +433,10 @@ constexpr Containers::StringView KnownWorkarounds[]{
    any relevant changelog entry and unfortunately the previous version I had
    was only 23.3.5, so it could be anything in between. My hunch is that it's
    due to some new code that deals with framebuffer compression and which was
-   only correctly cleaned up in the non-DSA code path. Or something. See also
-   "intel-windows-broken-dsa-framebuffer-clear" above. */
+   only correctly cleaned up in the non-DSA code path. Or something. Does not
+   seem to happen on version 25 anymore. It might have been fixed earlier
+   already but enabling it for all version 24 patch versions because that's
+   easiest. See also "intel-windows-broken-dsa-framebuffer-clear" above. */
 "mesa-broken-dsa-framebuffer-clear"_s,
 #endif