[Intel-gfx] [PATCH] drm/i915: fix the wrong latency value while computing wm0
Jesse Barnes
jbarnes at virtuousgeek.org
Thu Dec 23 17:22:42 CET 2010
On Thu, 23 Dec 2010 16:35:40 +0800
Yuanhan Liu <yuanhan.liu at linux.intel.com> wrote:
> On Ironlake, the LP0 latency is hardcoded and in ns unit, while on
> Sandybridge, it comes from a register and with unit 0.1 us. So, fix
> the wrong latency value while computing wm0 on Ironlake and Sandybridge.
>
> Signed-off-by: Yuanhan Liu <yuanhan.liu at linux.intel.com>
> ---
> drivers/gpu/drm/i915/intel_display.c | 10 +++++-----
> 1 files changed, 5 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
> index 9645140..c784506 100644
> --- a/drivers/gpu/drm/i915/intel_display.c
> +++ b/drivers/gpu/drm/i915/intel_display.c
> @@ -3418,9 +3418,9 @@ static void i830_update_wm(struct drm_device *dev, int planea_clock, int unused,
> static bool ironlake_compute_wm0(struct drm_device *dev,
> int pipe,
> const struct intel_watermark_params *display,
> - int display_latency,
> + int display_latency_ns,
> const struct intel_watermark_params *cursor,
> - int cursor_latency,
> + int cursor_latency_ns,
> int *plane_wm,
> int *cursor_wm)
> {
> @@ -3438,7 +3438,7 @@ static bool ironlake_compute_wm0(struct drm_device *dev,
> pixel_size = crtc->fb->bits_per_pixel / 8;
>
> /* Use the small buffer method to calculate plane watermark */
> - entries = ((clock * pixel_size / 1000) * display_latency * 100) / 1000;
> + entries = ((clock * pixel_size / 1000) * display_latency_ns) / 1000;
> entries = DIV_ROUND_UP(entries, display->cacheline_size);
> *plane_wm = entries + display->guard_size;
> if (*plane_wm > (int)display->max_wm)
> @@ -3446,7 +3446,7 @@ static bool ironlake_compute_wm0(struct drm_device *dev,
>
> /* Use the large buffer method to calculate cursor watermark */
> line_time_us = ((htotal * 1000) / clock);
> - line_count = (cursor_latency * 100 / line_time_us + 1000) / 1000;
> + line_count = (cursor_latency_ns / line_time_us + 1000) / 1000;
> entries = line_count * 64 * pixel_size;
> entries = DIV_ROUND_UP(entries, cursor->cacheline_size);
> *cursor_wm = entries + cursor->guard_size;
> @@ -3652,7 +3652,7 @@ static void sandybridge_update_wm(struct drm_device *dev,
> int pixel_size)
> {
> struct drm_i915_private *dev_priv = dev->dev_private;
> - int latency = SNB_READ_WM0_LATENCY();
> + int latency = SNB_READ_WM0_LATENCY() * 100; /* In unit 0.1 */
> int fbc_wm, plane_wm, cursor_wm, enabled;
> int clock;
>
Good catch, thanks for the fix.
Acked-by: Jesse Barnes <jbarnes at virtuousgeek.org>
--
Jesse Barnes, Intel Open Source Technology Center
More information about the Intel-gfx
mailing list