Message ID | 1418055747-22702-1-git-send-email-damien.lespiau@intel.com (mailing list archive) |
---|---|
State | New, archived |
Headers | show |
On Mon, Dec 08, 2014 at 04:22:27PM +0000, Damien Lespiau wrote: > I was playing with clang and oh surprise! a warning trigerred by > -Wshift-overflow (gcc doesn't have this one): > > WA_SET_BIT_MASKED(GEN7_GT_MODE, > GEN6_WIZ_HASHING_MASK | GEN6_WIZ_HASHING_16x4); > > drivers/gpu/drm/i915/intel_ringbuffer.c:786:2: warning: signed shift result > (0x28002000000) requires 43 bits to represent, but 'int' only has 32 bits > [-Wshift-overflow] > WA_SET_BIT_MASKED(GEN7_GT_MODE, > ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > drivers/gpu/drm/i915/intel_ringbuffer.c:737:15: note: expanded from macro > 'WA_SET_BIT_MASKED' > WA_REG(addr, _MASKED_BIT_ENABLE(mask), (mask) & 0xffff) > > Turned out GEN6_WIZ_HASHING_MASK was already shifted by 16, and we were > trying to shift it a bit more. > > The other thing is that it's not the usual case of setting WA bits here, we > need to have separate mask and value. > > To fix this, I've introduced a new _MASKED_FIELD() macro that takes both the > (unshifted) mask and the desired value and the rest of the patch ripples > through from it. > > This bug was introduced when reworking the WA emission in: > > Commit 7225342ab501befdb64bcec76ded41f5897c0855 > Author: Mika Kuoppala <mika.kuoppala@linux.intel.com> > Date: Tue Oct 7 17:21:26 2014 +0300 > > drm/i915: Build workaround list in ring initialization > > v2: Invert the order of the mask and value arguments (Daniel Vetter) > Rewrite _MASKED_BIT_ENABLE() and _MASKED_BIT_DISABLE() with > _MASKED_FIELD() (Jani Nikula) > Make sure we only evaluate 'a' once in _MASKED_BIT_ENABLE() (Dave Gordon) > Add check to ensure the value is within the mask boundaries (Chris Wilson) > > Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> > Cc: Arun Siluvery <arun.siluvery@linux.intel.com> > Signed-off-by: Damien Lespiau <damien.lespiau@intel.com> > --- > drivers/gpu/drm/i915/i915_reg.h | 13 ++++++++++--- > drivers/gpu/drm/i915/intel_pm.c | 6 +++--- > drivers/gpu/drm/i915/intel_ringbuffer.c | 8 ++++++-- > 3 files changed, 19 insertions(+), 8 deletions(-) > > diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h > index dc03fac..e0cd461 100644 > --- a/drivers/gpu/drm/i915/i915_reg.h > +++ b/drivers/gpu/drm/i915/i915_reg.h > @@ -34,8 +34,15 @@ > #define _PORT3(port, a, b, c) ((port) == PORT_A ? (a) : \ > (port) == PORT_B ? (b) : (c)) > > -#define _MASKED_BIT_ENABLE(a) (((a) << 16) | (a)) > -#define _MASKED_BIT_DISABLE(a) ((a) << 16) > +#define _MASKED_FIELD(mask, value) ({ \ > + if (__builtin_constant_p(mask) && __builtin_constant_p(value)) \ > + BUILD_BUG_ON_MSG((value) & ~(mask), \ > + "Incorrect value for mask"); \ Imo just use WARN_ON, with my patch it'll do the right thing. And maybe we'll use this in some context once where it's not a compile-time constant and then the runtime check would be better. With that addressed it's Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch> I just realized that WA_REG has the order the other way round too, maybe throw a follow-up on top? -Daniel > + (mask) << 16 | (value); }) > +#define _MASKED_BIT_ENABLE(a) ({ typeof(a) _a = (a); _MASKED_FIELD(_a, _a); }) > +#define _MASKED_BIT_DISABLE(a) (_MASKED_FIELD((a), 0)) > + > + > > /* PCI config space */ > > @@ -1284,7 +1291,7 @@ enum punit_power_well { > #define GEN6_WIZ_HASHING_8x8 GEN6_WIZ_HASHING(0, 0) > #define GEN6_WIZ_HASHING_8x4 GEN6_WIZ_HASHING(0, 1) > #define GEN6_WIZ_HASHING_16x4 GEN6_WIZ_HASHING(1, 0) > -#define GEN6_WIZ_HASHING_MASK (GEN6_WIZ_HASHING(1, 1) << 16) > +#define GEN6_WIZ_HASHING_MASK GEN6_WIZ_HASHING(1, 1) > #define GEN6_TD_FOUR_ROW_DISPATCH_DISABLE (1 << 5) > > #define GFX_MODE 0x02520 > diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c > index 78911e2..0f2febd 100644 > --- a/drivers/gpu/drm/i915/intel_pm.c > +++ b/drivers/gpu/drm/i915/intel_pm.c > @@ -6389,7 +6389,7 @@ static void gen6_init_clock_gating(struct drm_device *dev) > * to keep in mind (see 3DSTATE_PS and 3DSTATE_WM). > */ > I915_WRITE(GEN6_GT_MODE, > - GEN6_WIZ_HASHING_MASK | GEN6_WIZ_HASHING_16x4); > + _MASKED_FIELD(GEN6_WIZ_HASHING_MASK, GEN6_WIZ_HASHING_16x4)); > > ilk_init_lp_watermarks(dev); > > @@ -6587,7 +6587,7 @@ static void haswell_init_clock_gating(struct drm_device *dev) > * to keep in mind (see 3DSTATE_PS and 3DSTATE_WM). > */ > I915_WRITE(GEN7_GT_MODE, > - GEN6_WIZ_HASHING_MASK | GEN6_WIZ_HASHING_16x4); > + _MASKED_FIELD(GEN6_WIZ_HASHING_MASK, GEN6_WIZ_HASHING_16x4)); > > /* WaSwitchSolVfFArbitrationPriority:hsw */ > I915_WRITE(GAM_ECOCHK, I915_READ(GAM_ECOCHK) | HSW_ECOCHK_ARB_PRIO_SOL); > @@ -6684,7 +6684,7 @@ static void ivybridge_init_clock_gating(struct drm_device *dev) > * to keep in mind (see 3DSTATE_PS and 3DSTATE_WM). > */ > I915_WRITE(GEN7_GT_MODE, > - GEN6_WIZ_HASHING_MASK | GEN6_WIZ_HASHING_16x4); > + _MASKED_FIELD(GEN6_WIZ_HASHING_MASK, GEN6_WIZ_HASHING_16x4)); > > snpcr = I915_READ(GEN6_MBCUNIT_SNPCR); > snpcr &= ~GEN6_MBC_SNPCR_MASK; > diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c > index 79b4ca5..9deb152 100644 > --- a/drivers/gpu/drm/i915/intel_ringbuffer.c > +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c > @@ -739,6 +739,9 @@ static int wa_add(struct drm_i915_private *dev_priv, > #define WA_CLR_BIT_MASKED(addr, mask) \ > WA_REG(addr, _MASKED_BIT_DISABLE(mask), (mask) & 0xffff) > > +#define WA_SET_FIELD_MASKED(addr, mask, value) \ > + WA_REG(addr, _MASKED_FIELD(mask, value), mask) > + > #define WA_SET_BIT(addr, mask) WA_REG(addr, I915_READ(addr) | (mask), mask) > #define WA_CLR_BIT(addr, mask) WA_REG(addr, I915_READ(addr) & ~(mask), mask) > > @@ -783,8 +786,9 @@ static int bdw_init_workarounds(struct intel_engine_cs *ring) > * disable bit, which we don't touch here, but it's good > * to keep in mind (see 3DSTATE_PS and 3DSTATE_WM). > */ > - WA_SET_BIT_MASKED(GEN7_GT_MODE, > - GEN6_WIZ_HASHING_MASK | GEN6_WIZ_HASHING_16x4); > + WA_SET_FIELD_MASKED(GEN7_GT_MODE, > + GEN6_WIZ_HASHING_MASK, > + GEN6_WIZ_HASHING_16x4); > > return 0; > } > -- > 1.8.3.1 > > _______________________________________________ > Intel-gfx mailing list > Intel-gfx@lists.freedesktop.org > http://lists.freedesktop.org/mailman/listinfo/intel-gfx
On 08/12/14 16:27, Daniel Vetter wrote: > On Mon, Dec 08, 2014 at 04:22:27PM +0000, Damien Lespiau wrote: >> I was playing with clang and oh surprise! a warning trigerred by >> -Wshift-overflow (gcc doesn't have this one): [snip] >> diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c >> index 79b4ca5..9deb152 100644 >> --- a/drivers/gpu/drm/i915/intel_ringbuffer.c >> +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c >> @@ -739,6 +739,9 @@ static int wa_add(struct drm_i915_private *dev_priv, >> #define WA_CLR_BIT_MASKED(addr, mask) \ >> WA_REG(addr, _MASKED_BIT_DISABLE(mask), (mask) & 0xffff) >> >> +#define WA_SET_FIELD_MASKED(addr, mask, value) \ >> + WA_REG(addr, _MASKED_FIELD(mask, value), mask) >> + >> #define WA_SET_BIT(addr, mask) WA_REG(addr, I915_READ(addr) | (mask), mask) >> #define WA_CLR_BIT(addr, mask) WA_REG(addr, I915_READ(addr) & ~(mask), mask) Not your changes, but: * WA_{SET,CLR}_BIT() above look dubious and don't seem to be used anyway * dev_priv->workarounds.reg[idx].mask = mask; The mask field is set but not used in intel_ring_workarounds_emit() or intel_logical_ring_workarounds_emit(), only in debugfs printout. And it's redundant since the 'value' incorporates the bit(field) mask and the new target value into one parameter, hence 3rd parameter of WA_REG() is surplus and calculating it in WA_{SET,CLR_BIT_MASKED() is also redundant. Unless I've missed something? .Dave.
On Mon, Dec 08, 2014 at 04:50:13PM +0000, Dave Gordon wrote: > On 08/12/14 16:27, Daniel Vetter wrote: > > On Mon, Dec 08, 2014 at 04:22:27PM +0000, Damien Lespiau wrote: > >> I was playing with clang and oh surprise! a warning trigerred by > >> -Wshift-overflow (gcc doesn't have this one): > > [snip] > > >> diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c > >> index 79b4ca5..9deb152 100644 > >> --- a/drivers/gpu/drm/i915/intel_ringbuffer.c > >> +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c > >> @@ -739,6 +739,9 @@ static int wa_add(struct drm_i915_private *dev_priv, > >> #define WA_CLR_BIT_MASKED(addr, mask) \ > >> WA_REG(addr, _MASKED_BIT_DISABLE(mask), (mask) & 0xffff) > >> > >> +#define WA_SET_FIELD_MASKED(addr, mask, value) \ > >> + WA_REG(addr, _MASKED_FIELD(mask, value), mask) > >> + > >> #define WA_SET_BIT(addr, mask) WA_REG(addr, I915_READ(addr) | (mask), mask) > >> #define WA_CLR_BIT(addr, mask) WA_REG(addr, I915_READ(addr) & ~(mask), mask) > > Not your changes, but: > > * WA_{SET,CLR}_BIT() above look dubious and don't seem to be used anyway > > * dev_priv->workarounds.reg[idx].mask = mask; > > The mask field is set but not used in intel_ring_workarounds_emit() or > intel_logical_ring_workarounds_emit(), only in debugfs printout. > And it's redundant since the 'value' incorporates the bit(field) mask > and the new target value into one parameter, hence 3rd parameter of > WA_REG() is surplus and calculating it in WA_{SET,CLR_BIT_MASKED() is > also redundant. > > Unless I've missed something? The mask is used to test that we correctly set/clear W/A values in i-g-t tests. Imagine the W/A being "clear bit 2", we have a generic (value, mask) to check that we do indeed do that.
On 08/12/14 16:22, Damien Lespiau wrote: > I was playing with clang and oh surprise! a warning trigerred by > -Wshift-overflow (gcc doesn't have this one): > > WA_SET_BIT_MASKED(GEN7_GT_MODE, > GEN6_WIZ_HASHING_MASK | GEN6_WIZ_HASHING_16x4); > > drivers/gpu/drm/i915/intel_ringbuffer.c:786:2: warning: signed shift result > (0x28002000000) requires 43 bits to represent, but 'int' only has 32 bits > [-Wshift-overflow] > WA_SET_BIT_MASKED(GEN7_GT_MODE, > ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > drivers/gpu/drm/i915/intel_ringbuffer.c:737:15: note: expanded from macro > 'WA_SET_BIT_MASKED' > WA_REG(addr, _MASKED_BIT_ENABLE(mask), (mask) & 0xffff) > > Turned out GEN6_WIZ_HASHING_MASK was already shifted by 16, and we were > trying to shift it a bit more. > > The other thing is that it's not the usual case of setting WA bits here, we > need to have separate mask and value. > > To fix this, I've introduced a new _MASKED_FIELD() macro that takes both the > (unshifted) mask and the desired value and the rest of the patch ripples > through from it. > > This bug was introduced when reworking the WA emission in: > > Commit 7225342ab501befdb64bcec76ded41f5897c0855 > Author: Mika Kuoppala <mika.kuoppala@linux.intel.com> > Date: Tue Oct 7 17:21:26 2014 +0300 > > drm/i915: Build workaround list in ring initialization > > v2: Invert the order of the mask and value arguments (Daniel Vetter) > Rewrite _MASKED_BIT_ENABLE() and _MASKED_BIT_DISABLE() with > _MASKED_FIELD() (Jani Nikula) > Make sure we only evaluate 'a' once in _MASKED_BIT_ENABLE() (Dave Gordon) > Add check to ensure the value is within the mask boundaries (Chris Wilson) > > Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> > Cc: Arun Siluvery <arun.siluvery@linux.intel.com> > Signed-off-by: Damien Lespiau <damien.lespiau@intel.com> > --- > drivers/gpu/drm/i915/i915_reg.h | 13 ++++++++++--- > drivers/gpu/drm/i915/intel_pm.c | 6 +++--- > drivers/gpu/drm/i915/intel_ringbuffer.c | 8 ++++++-- > 3 files changed, 19 insertions(+), 8 deletions(-) > > diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h > index dc03fac..e0cd461 100644 > --- a/drivers/gpu/drm/i915/i915_reg.h > +++ b/drivers/gpu/drm/i915/i915_reg.h > @@ -34,8 +34,15 @@ > #define _PORT3(port, a, b, c) ((port) == PORT_A ? (a) : \ > (port) == PORT_B ? (b) : (c)) > > -#define _MASKED_BIT_ENABLE(a) (((a) << 16) | (a)) > -#define _MASKED_BIT_DISABLE(a) ((a) << 16) > +#define _MASKED_FIELD(mask, value) ({ \ > + if (__builtin_constant_p(mask) && __builtin_constant_p(value)) \ > + BUILD_BUG_ON_MSG((value) & ~(mask), \ > + "Incorrect value for mask"); \ > + (mask) << 16 | (value); }) For even more compile- and run-time robustness we could check that 'mask' and 'value' each fit in 16 bits, as we have an explicit '16' in there already. .Dave.
Tested-By: PRC QA PRTS (Patch Regression Test System Contact: shuang.he@intel.com)
-------------------------------------Summary-------------------------------------
Platform Delta drm-intel-nightly Series Applied
PNV 364/364 364/364
ILK +1 364/366 365/366
SNB 448/450 448/450
IVB 497/498 497/498
BYT 289/289 289/289
HSW 563/564 563/564
BDW 417/417 417/417
-------------------------------------Detailed-------------------------------------
Platform Test drm-intel-nightly Series Applied
ILK igt_kms_flip_wf_vblank-ts-check DMESG_WARN(1, M26)PASS(7, M26M37) PASS(1, M37)
Note: You need to pay more attention to line start with '*'
diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h index dc03fac..e0cd461 100644 --- a/drivers/gpu/drm/i915/i915_reg.h +++ b/drivers/gpu/drm/i915/i915_reg.h @@ -34,8 +34,15 @@ #define _PORT3(port, a, b, c) ((port) == PORT_A ? (a) : \ (port) == PORT_B ? (b) : (c)) -#define _MASKED_BIT_ENABLE(a) (((a) << 16) | (a)) -#define _MASKED_BIT_DISABLE(a) ((a) << 16) +#define _MASKED_FIELD(mask, value) ({ \ + if (__builtin_constant_p(mask) && __builtin_constant_p(value)) \ + BUILD_BUG_ON_MSG((value) & ~(mask), \ + "Incorrect value for mask"); \ + (mask) << 16 | (value); }) +#define _MASKED_BIT_ENABLE(a) ({ typeof(a) _a = (a); _MASKED_FIELD(_a, _a); }) +#define _MASKED_BIT_DISABLE(a) (_MASKED_FIELD((a), 0)) + + /* PCI config space */ @@ -1284,7 +1291,7 @@ enum punit_power_well { #define GEN6_WIZ_HASHING_8x8 GEN6_WIZ_HASHING(0, 0) #define GEN6_WIZ_HASHING_8x4 GEN6_WIZ_HASHING(0, 1) #define GEN6_WIZ_HASHING_16x4 GEN6_WIZ_HASHING(1, 0) -#define GEN6_WIZ_HASHING_MASK (GEN6_WIZ_HASHING(1, 1) << 16) +#define GEN6_WIZ_HASHING_MASK GEN6_WIZ_HASHING(1, 1) #define GEN6_TD_FOUR_ROW_DISPATCH_DISABLE (1 << 5) #define GFX_MODE 0x02520 diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 78911e2..0f2febd 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -6389,7 +6389,7 @@ static void gen6_init_clock_gating(struct drm_device *dev) * to keep in mind (see 3DSTATE_PS and 3DSTATE_WM). */ I915_WRITE(GEN6_GT_MODE, - GEN6_WIZ_HASHING_MASK | GEN6_WIZ_HASHING_16x4); + _MASKED_FIELD(GEN6_WIZ_HASHING_MASK, GEN6_WIZ_HASHING_16x4)); ilk_init_lp_watermarks(dev); @@ -6587,7 +6587,7 @@ static void haswell_init_clock_gating(struct drm_device *dev) * to keep in mind (see 3DSTATE_PS and 3DSTATE_WM). */ I915_WRITE(GEN7_GT_MODE, - GEN6_WIZ_HASHING_MASK | GEN6_WIZ_HASHING_16x4); + _MASKED_FIELD(GEN6_WIZ_HASHING_MASK, GEN6_WIZ_HASHING_16x4)); /* WaSwitchSolVfFArbitrationPriority:hsw */ I915_WRITE(GAM_ECOCHK, I915_READ(GAM_ECOCHK) | HSW_ECOCHK_ARB_PRIO_SOL); @@ -6684,7 +6684,7 @@ static void ivybridge_init_clock_gating(struct drm_device *dev) * to keep in mind (see 3DSTATE_PS and 3DSTATE_WM). */ I915_WRITE(GEN7_GT_MODE, - GEN6_WIZ_HASHING_MASK | GEN6_WIZ_HASHING_16x4); + _MASKED_FIELD(GEN6_WIZ_HASHING_MASK, GEN6_WIZ_HASHING_16x4)); snpcr = I915_READ(GEN6_MBCUNIT_SNPCR); snpcr &= ~GEN6_MBC_SNPCR_MASK; diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c index 79b4ca5..9deb152 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.c +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c @@ -739,6 +739,9 @@ static int wa_add(struct drm_i915_private *dev_priv, #define WA_CLR_BIT_MASKED(addr, mask) \ WA_REG(addr, _MASKED_BIT_DISABLE(mask), (mask) & 0xffff) +#define WA_SET_FIELD_MASKED(addr, mask, value) \ + WA_REG(addr, _MASKED_FIELD(mask, value), mask) + #define WA_SET_BIT(addr, mask) WA_REG(addr, I915_READ(addr) | (mask), mask) #define WA_CLR_BIT(addr, mask) WA_REG(addr, I915_READ(addr) & ~(mask), mask) @@ -783,8 +786,9 @@ static int bdw_init_workarounds(struct intel_engine_cs *ring) * disable bit, which we don't touch here, but it's good * to keep in mind (see 3DSTATE_PS and 3DSTATE_WM). */ - WA_SET_BIT_MASKED(GEN7_GT_MODE, - GEN6_WIZ_HASHING_MASK | GEN6_WIZ_HASHING_16x4); + WA_SET_FIELD_MASKED(GEN7_GT_MODE, + GEN6_WIZ_HASHING_MASK, + GEN6_WIZ_HASHING_16x4); return 0; }
I was playing with clang and oh surprise! a warning trigerred by -Wshift-overflow (gcc doesn't have this one): WA_SET_BIT_MASKED(GEN7_GT_MODE, GEN6_WIZ_HASHING_MASK | GEN6_WIZ_HASHING_16x4); drivers/gpu/drm/i915/intel_ringbuffer.c:786:2: warning: signed shift result (0x28002000000) requires 43 bits to represent, but 'int' only has 32 bits [-Wshift-overflow] WA_SET_BIT_MASKED(GEN7_GT_MODE, ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ drivers/gpu/drm/i915/intel_ringbuffer.c:737:15: note: expanded from macro 'WA_SET_BIT_MASKED' WA_REG(addr, _MASKED_BIT_ENABLE(mask), (mask) & 0xffff) Turned out GEN6_WIZ_HASHING_MASK was already shifted by 16, and we were trying to shift it a bit more. The other thing is that it's not the usual case of setting WA bits here, we need to have separate mask and value. To fix this, I've introduced a new _MASKED_FIELD() macro that takes both the (unshifted) mask and the desired value and the rest of the patch ripples through from it. This bug was introduced when reworking the WA emission in: Commit 7225342ab501befdb64bcec76ded41f5897c0855 Author: Mika Kuoppala <mika.kuoppala@linux.intel.com> Date: Tue Oct 7 17:21:26 2014 +0300 drm/i915: Build workaround list in ring initialization v2: Invert the order of the mask and value arguments (Daniel Vetter) Rewrite _MASKED_BIT_ENABLE() and _MASKED_BIT_DISABLE() with _MASKED_FIELD() (Jani Nikula) Make sure we only evaluate 'a' once in _MASKED_BIT_ENABLE() (Dave Gordon) Add check to ensure the value is within the mask boundaries (Chris Wilson) Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> Cc: Arun Siluvery <arun.siluvery@linux.intel.com> Signed-off-by: Damien Lespiau <damien.lespiau@intel.com> --- drivers/gpu/drm/i915/i915_reg.h | 13 ++++++++++--- drivers/gpu/drm/i915/intel_pm.c | 6 +++--- drivers/gpu/drm/i915/intel_ringbuffer.c | 8 ++++++-- 3 files changed, 19 insertions(+), 8 deletions(-)