[Mesa-dev] [PATCH] Break instead of loop all paths once dri driver is gotten successfully
Matt Turner
mattst88 at gmail.com
Thu Dec 19 09:30:38 PST 2013
On Thu, Dec 19, 2013 at 8:34 AM, Brian Paul <brianp at vmware.com> wrote:
> On 12/18/2013 10:11 PM, Quanxian Wang wrote:
>>
>> Originally it will continue all paths. The next search will overwrite
>> previous
>> result even if the previous is successful. In some case it will cause
>> error.
>>
>> Signed-off-by: Quanxian Wang <quanxian.wang at intel.com>
>> Reviewed-By: Gong, Zhigang <zhigang.gong at gmail.com>
>> ---
>> src/egl/drivers/dri2/egl_dri2.c | 3 +++
>> src/gbm/backends/dri/gbm_dri.c | 5 +++--
>> 2 files changed, 6 insertions(+), 2 deletions(-)
>>
>> diff --git a/src/egl/drivers/dri2/egl_dri2.c
>> b/src/egl/drivers/dri2/egl_dri2.c
>> index e703f28..892f1f4 100644
>> --- a/src/egl/drivers/dri2/egl_dri2.c
>> +++ b/src/egl/drivers/dri2/egl_dri2.c
>> @@ -403,6 +403,9 @@ dri2_open_driver(_EGLDisplay *disp)
>> if (dri2_dpy->driver == NULL)
>> _eglLog(_EGL_DEBUG, "failed to open %s: %s\n", path,
>> dlerror());
>> }
>> + /* not need continue to loop all paths once the driver is found */
>> + if (dri2_dpy->driver != NULL)
>> + break;
>> }
>>
>> if (dri2_dpy->driver == NULL) {
>> diff --git a/src/gbm/backends/dri/gbm_dri.c
>> b/src/gbm/backends/dri/gbm_dri.c
>> index b4b97ac..cd31221 100644
>> --- a/src/gbm/backends/dri/gbm_dri.c
>> +++ b/src/gbm/backends/dri/gbm_dri.c
>> @@ -224,9 +224,10 @@ dri_load_driver(struct gbm_dri_device *dri)
>> snprintf(path, sizeof path,
>> "%.*s/%s_dri.so", len, p, dri->base.driver_name);
>> dri->driver = dlopen(path, RTLD_NOW | RTLD_GLOBAL);
>> - if (dri->driver == NULL)
>> - fprintf(stderr, "failed to open %s: %s\n", path, dlerror());
>> }
>> + /* not need continue to loop all paths once the driver is found */
>> + if (dri->driver != NULL)
>> + break;
>> }
>>
>> if (dri->driver == NULL) {
>>
>
> Reviewed-by: Brian Paul <brianp at vmware.com>
>
> Do you need someone to push this for you?
If you do push it, would you mind fixing up the comments to sound a bit better?
More information about the mesa-dev
mailing list