summaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authorNaseer Ahmed <naseer@codeaurora.org>2013-07-09 16:36:28 -0400
committerIliyan Malchev <malchev@google.com>2013-07-09 18:01:51 -0700
commite3de50824f3fe007d62e5b7efd87f5b7ed93efcf (patch)
tree23147bf662909a153dcf52a7d988aada04a27c7e
parentfbc4be199f1c2a50562ea12ae14aa439fd73557c (diff)
downloadplatform_hardware_qcom_display-jb-mr2-dev.tar.gz
platform_hardware_qcom_display-jb-mr2-dev.tar.bz2
platform_hardware_qcom_display-jb-mr2-dev.zip
hwc: Don't lock eventControl for vsync eventsjb-mr2-dev
Locking here can lead to a delay in vsync disable/enable where prepare/set is not yet completed. The HWC spec says that eventControl needs to take effect immediately. This lock was earlier present when we waited for SF to unblank primary at startup and the vsync control ioctls could fail. This is no longer necessary as we unblank primary on startup now. Bug: 9669813 Change-Id: I2bf18ae9235f8dd428a495a9a23af7f00f34ff1d Signed-off-by: Iliyan Malchev <malchev@google.com>
-rw-r--r--libhwcomposer/hwc.cpp1
1 files changed, 0 insertions, 1 deletions
diff --git a/libhwcomposer/hwc.cpp b/libhwcomposer/hwc.cpp
index 03efe1a7..3ee320c0 100644
--- a/libhwcomposer/hwc.cpp
+++ b/libhwcomposer/hwc.cpp
@@ -245,7 +245,6 @@ static int hwc_eventControl(struct hwc_composer_device_1* dev, int dpy,
{
int ret = 0;
hwc_context_t* ctx = (hwc_context_t*)(dev);
- Locker::Autolock _l(ctx->mBlankLock);
if(!ctx->dpyAttr[dpy].isActive) {
ALOGE("Display is blanked - Cannot %s vsync",
enable ? "enable" : "disable");