[Mesa-dev] [PATCH] i965: Make the driver compile until a proper libdrm can be released.

[Kristian Høgsberg]

On Fri, Nov 8, 2013 at 2:25 PM, Ian Romanick <idr at freedesktop.org> wrote:
>
> On 11/08/2013 10:49 AM, Eric Anholt wrote:
>> No depending on unreleased code.
>> ---
>>  src/mesa/drivers/dri/i965/brw_context.c | 8 +++-----
>>  src/mesa/drivers/dri/i965/brw_reset.c   | 7 +++++++
>>  2 files changed, 10 insertions(+), 5 deletions(-)
>>
>> diff --git a/src/mesa/drivers/dri/i965/brw_context.c b/src/mesa/drivers/dri/i965/brw_context.c
>> index a33e993..bee98e3 100644
>> --- a/src/mesa/drivers/dri/i965/brw_context.c
>> +++ b/src/mesa/drivers/dri/i965/brw_context.c
>> @@ -714,12 +714,10 @@ brwCreateContext(gl_api api,
>>     }
>>
>>     /* Notification of GPU resets requires hardware contexts and a kernel new
>> -    * enough to support DRM_IOCTL_I915_GET_RESET_STATS.
>> +    * enough to support DRM_IOCTL_I915_GET_RESET_STATS, which isn't upstream
>> +    * yet.
>>      */
>> -   if (notify_reset &&
>> -       (brw->hw_ctx == NULL
>> -        || drm_intel_get_reset_stats(brw->hw_ctx, &brw->reset_count, NULL,
>> -                                     NULL))) {
>> +   if (notify_reset) {
>
> If anything, this should #if around DRM_IOCTL_I915_GET_RESET_STATS being
> defined.  If that's not defined, then the DRI extension shouldn't be
> enabled.
>
>       if (notify_reset
> #ifdef DRM_IOCTL_I915_GET_RESET_STATS
>            && (brw->hw_ctx == NULL
>                || drm_intel_get_reset_stats(brw->hw_ctx,
>                                             &brw->reset_count, NULL,
>                                             NULL))
> #endif
>          ) {
>
> Then we don't have to revert this patch when libdrm bits (re-)land.

We don't want the #ifdef in there long term.  Once it lands in libdrm
and we depend on that version, the code will always be enabled.

I committed Erics patch as is, we need a driver that compiles while we
work this out.

Kristian

> And add #ifdef DRM_IOCTL_I915_GET_RESET_STATS around the dri2Robustness
> stuff in intel_screen.c.
x>
>>        /* This is the wrong error code, but the correct error code (one that
>>         * will cause EGL to generate EGL_BAD_MATCH) doesn't seem to exist.
>>         */
>> diff --git a/src/mesa/drivers/dri/i965/brw_reset.c b/src/mesa/drivers/dri/i965/brw_reset.c
>> index 7eca1bc..e93b2e2 100644
>> --- a/src/mesa/drivers/dri/i965/brw_reset.c
>> +++ b/src/mesa/drivers/dri/i965/brw_reset.c
>> @@ -42,10 +42,17 @@ brw_get_graphics_reset_status(struct gl_context *ctx)
>>      */
>>     assert(brw->hw_ctx != NULL);
>>
>> +#if 0
>
> #ifdef DRM_IOCTL_I915_GET_RESET_STATS
>
>> +   /* This is waiting until the kernel code can be merged and a new libdrm
>> +    * actually released.
>> +    */
>>     err = drm_intel_get_reset_stats(brw->hw_ctx, &reset_count, &active,
>>                                     &pending);
>>     if (err)
>>        return GL_NO_ERROR;
>> +#else
>> +   return GL_NO_ERROR;
>> +#endif
>>
>>     /* A reset was observed while a batch from this context was executing.
>>      * Assume that this context was at fault.
>>
>
> _______________________________________________
> mesa-dev mailing list
> mesa-dev at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/mesa-dev