On 3/13/2014 11:46 PM, Ville Syrjälä wrote:
On Thu, Mar 13, 2014 at 09:30:18PM +0530, deepak.s@xxxxxxxxxxxxxxx wrote:
From: Deepak S <deepak.s@xxxxxxxxx>
We are adding a module paramter to control rps boost. By default, we
enable the boost for better performace. Based on the need (perf/power)
we can either enable/disable.
v2: Addressed rps default comment (Jani)
Signed-off-by: Deepak S <deepak.s@xxxxxxxxxxxxxxx>
---
drivers/gpu/drm/i915/i915_drv.h | 1 +
drivers/gpu/drm/i915/i915_gem.c | 16 +++++++++++++++-
drivers/gpu/drm/i915/i915_params.c | 5 +++++
3 files changed, 21 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 607042b..7808319 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -2106,6 +2106,7 @@ struct i915_params {
int panel_use_ssc;
int vbt_sdvo_panel_type;
int enable_rc6;
+ int enable_rps_boost;
Should be bool like Jani said. And then it should be thrown somewhere
somewhere at the end of the structure next to the other bools.
I will address this.
int enable_fbc;
int enable_ppgtt;
int enable_psr;
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 92b0b41..23a4700 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -1002,6 +1002,17 @@ static bool can_wait_boost(struct drm_i915_file_private *file_priv)
return !atomic_xchg(&file_priv->rps_wait_boost, true);
}
+static int intel_enable_rps_boost(struct drm_device *dev)
+{
+ /* No RPS Boost before Ironlake */
This comment is still wrong. I'd just drop it, everyone should know what
the gen check below means.
Ok
+ if (INTEL_INFO(dev)->gen < 6)
+ return 0;
+
+ /* Respect the kernel parameter if it is set */
This comment too seems rather obvious. I'd drop it as well.
Ok
+ return i915.enable_rps_boost;
+
+}
This function is still just a wrapper for i915.enable_rps_boost since
__wait_seqno() already does the gen check. You could just check
i915.enable_rps_boost directly in __wait_seqno(). The other option is
to just drop the gen check from __wait_seqno() and just let this
function take care of it. Hmm. Yeah that might be the nicest choice in
fact.
Agreed. Does not make sense to have multiple platform check's.
+
/**
* __wait_seqno - wait until execution of seqno has finished
* @ring: the ring expected to report seqno
@@ -1042,8 +1053,11 @@ static int __wait_seqno(struct intel_ring_buffer *ring, u32 seqno,
timeout_expire = timeout ? jiffies + timespec_to_jiffies_timeout(timeout) : 0;
- if (INTEL_INFO(dev)->gen >= 6 && can_wait_boost(file_priv)) {
+ if (INTEL_INFO(dev)->gen >= 6 && can_wait_boost(file_priv) &&
+ intel_enable_rps_boost(ring->dev)) {
Indentation is quite wrong. There's also trailing whitespace around
these parts. Please run patches through checkpatch.pl before submitting.
+
gen6_rps_boost(dev_priv);
+
if (file_priv)
mod_delayed_work(dev_priv->wq,
&file_priv->mm.idle_work,
diff --git a/drivers/gpu/drm/i915/i915_params.c b/drivers/gpu/drm/i915/i915_params.c
index a66ffb6..2d207e3 100644
--- a/drivers/gpu/drm/i915/i915_params.c
+++ b/drivers/gpu/drm/i915/i915_params.c
@@ -34,6 +34,7 @@ struct i915_params i915 __read_mostly = {
.panel_use_ssc = -1,
.vbt_sdvo_panel_type = -1,
.enable_rc6 = -1,
+ .enable_rps_boost = 1,
true
.enable_fbc = -1,
.enable_hangcheck = true,
.enable_ppgtt = -1,
@@ -78,6 +79,10 @@ MODULE_PARM_DESC(enable_rc6,
"For example, 3 would enable rc6 and deep rc6, and 7 would enable everything. "
"default: -1 (use per-chip default)");
+module_param_named(enable_rps_boost, i915.enable_rps_boost, int, 0600);
bool
+MODULE_PARM_DESC(enable_rps_boost,
+ "Enable/Disable boost RPS frequency (default: enabled (1))");
default: true
+
module_param_named(enable_fbc, i915.enable_fbc, int, 0600);
MODULE_PARM_DESC(enable_fbc,
"Enable frame buffer compression for power savings "
--
1.8.4.2
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
