mesa.git
9 years agometa: Fix the y offset for 1D_ARRAY in _mesa_meta_pbo_TexSubImage
Neil Roberts [Wed, 25 Feb 2015 15:33:08 +0000 (15:33 +0000)]
meta: Fix the y offset for 1D_ARRAY in _mesa_meta_pbo_TexSubImage

The yoffset needs to be interpreted as a slice offset for 1D array
textures. This patch implements that by moving the yoffset into
zoffset similar to how it moves the height into depth.

Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Cc: "10.5" <mesa-stable@lists.freedesktop.org>
9 years agometa: Allow GL_UN/PACK_IMAGE_HEIGHT in _mesa_meta_pbo_Get/TexSubImage
Neil Roberts [Thu, 26 Feb 2015 12:53:50 +0000 (12:53 +0000)]
meta: Allow GL_UN/PACK_IMAGE_HEIGHT in _mesa_meta_pbo_Get/TexSubImage

Now that a layered source PBO is interpreted as a single tall 2D image
it's quite easy to accept the image height packing option by just
creating an image that is tall enough to include the image padding.

I'm not sure whether the image height property should affect 1D_ARRAY
textures. My intuition and interpretation of the GL spec (which is a
bit vague) would be that it shouldn't. However the software fallback
path in Mesa uses the property for packing but not for unpacking. The
binary NVidia driver uses it for both. This patch doesn't use it for
either case so it is different from the software fallback. There is
some discussion about this here:

http://lists.freedesktop.org/archives/mesa-dev/2015-February/077925.html

This is tested by the texsubimage Piglit test with the array and pbo
arguments. Previously this test was skipping this code path because it
always sets the image height.

I've also tested it by modifying the getteximage-targets test. It
wasn't using this code path before because it was using the default
texture object so this code couldn't successfully create a frame
buffer. I also modified it to add some image padding with the image
height in the PBO.

Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Cc: "10.5" <mesa-stable@lists.freedesktop.org>
9 years agoRevert "common: Fix PBOs for 1D_ARRAY."
Neil Roberts [Thu, 26 Feb 2015 12:12:15 +0000 (12:12 +0000)]
Revert "common: Fix PBOs for 1D_ARRAY."

This reverts commit 546aba143d13ba3f993ead4cc30b2404abfc0202.

I think the changes to the calls to glBlitFramebuffer from this patch
are no different to what it was doing previously because it used to
set height to 1 before doing the blits. However it was introducing
some problems with the blit for layer 0 because this was no longer
special cased. It didn't fix problems with the yoffset which needs to
be interpreted as a slice offset. I think a better solution would be
to modify the original if statement to cope with the yoffset.

Conflicts:
src/mesa/drivers/common/meta_tex_subimage.c

Cc: "10.5" <mesa-stable@lists.freedesktop.org>
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
9 years agoglsl: Fix GCC unused-variable warning in release build.
Vinson Lee [Wed, 4 Mar 2015 02:46:13 +0000 (18:46 -0800)]
glsl: Fix GCC unused-variable warning in release build.

  CXX      ast_array_index.lo
ast_array_index.cpp: In function ‘void update_max_array_access(ir_rvalue*, int, YYLTYPE*, _mesa_glsl_parse_state*)’:
ast_array_index.cpp:86:30: warning: unused variable ‘interface_type’ [-Wunused-variable]
             const glsl_type *interface_type =
                              ^

Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Reviewed-by: Timothy Arceri <t_arceri@yahoo.com.au>
9 years agoilo: improve WA handling in rectlist path
Chia-I Wu [Wed, 4 Mar 2015 21:38:50 +0000 (14:38 -0700)]
ilo: improve WA handling in rectlist path

Add wrappers for 3DPRIMITIVE to make sure we clear current_pipe_control_dw1
and deferred_pipe_control_dw1 after it.  Add missing
gen7_wa_post_ps_and_later().

9 years agoilo: clean up Gen7.5 WAs
Chia-I Wu [Wed, 4 Mar 2015 21:09:26 +0000 (14:09 -0700)]
ilo: clean up Gen7.5 WAs

These WAs

  gen7_wa_post_3dstate_push_constant_alloc_ps()
  gen7_wa_pre_vs()
  gen7_wa_pre_3dstate_sf_depth_bias()
  first half of gen7_wa_pre_depth()
  gen7_wa_post_ps_and_later()

are Gen7-specific.  Update copy-and-pasted gen8_wa_pre_depth() also.

9 years agoclover: Fix build since llvm r231270
Tom Stellard [Wed, 4 Mar 2015 21:09:50 +0000 (13:09 -0800)]
clover: Fix build since llvm r231270

9 years agoilo: add ILO_DEBUG=hang
Chia-I Wu [Wed, 4 Mar 2015 20:07:55 +0000 (13:07 -0700)]
ilo: add ILO_DEBUG=hang

When set, detect and dump the hanging batch bufffer.

9 years agoilo: add some more winsys functions
Chia-I Wu [Wed, 4 Mar 2015 19:02:12 +0000 (12:02 -0700)]
ilo: add some more winsys functions

Add intel_winsys_get_reset_stats(), intel_winsys_import_userptr(), and
intel_bo_map_async().  The latter two are stubs, but we are not going to use
them immediately either.

9 years agoi965/fs: Don't propagate cmod to inst with different type.
Matt Turner [Fri, 27 Feb 2015 18:22:21 +0000 (10:22 -0800)]
i965/fs: Don't propagate cmod to inst with different type.

Cc: 10.5 <mesa-stable@lists.freedesktop.org>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=89317
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
9 years agor300g: Check return value of snprintf().
Matt Turner [Wed, 4 Mar 2015 00:09:58 +0000 (16:09 -0800)]
r300g: Check return value of snprintf().

Would have at least prevented the crash the previous patch fixed.

Cc: 10.4, 10.5 <mesa-stable@lists.freedesktop.org>
Bugzilla: https://bugs.gentoo.org/show_bug.cgi?id=540970
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
9 years agor300g: Use PATH_MAX instead of limiting ourselves to 100 chars.
Matt Turner [Wed, 4 Mar 2015 00:02:40 +0000 (16:02 -0800)]
r300g: Use PATH_MAX instead of limiting ourselves to 100 chars.

When built with Gentoo's package manager, the Mesa source directory
exists seven directories deep. The path to the .test file is too long
and is silently truncated, leading to a crash. Just use PATH_MAX.

Cc: 10.4, 10.5 <mesa-stable@lists.freedesktop.org>
Bugzilla: https://bugs.gentoo.org/show_bug.cgi?id=540970
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
9 years agoglx/tests: add -I src/ to fix make check
Brian Paul [Wed, 4 Mar 2015 17:03:09 +0000 (10:03 -0700)]
glx/tests: add -I src/ to fix make check

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
9 years agoi965: Fix uint64_t overflow in intel_client_wait_sync()
Kristian Høgsberg [Tue, 3 Mar 2015 00:19:52 +0000 (16:19 -0800)]
i965: Fix uint64_t overflow in intel_client_wait_sync()

DRM_IOCTL_I915_GEM_WAIT takes an int64_t for the timeout value but
GL_ARB_sync takes an uint64_t.  Further, the ioctl used to wait
indefinitely when passed a negative timeout, but it's been broken and
now returns immediately in that case.  Thus, if an application passes
UINT64_MAX to wait forever, we overflow to -1LL and return immediately.
Work around this mess by clamping the wait timeout to INT64_MAX.

Signed-off-by: Kristian Høgsberg <krh@bitplanet.net>
Reviewed-by: Chad Versace <chad.versace@intel.com>
9 years agoegl: Take alpha bits into account when selecting GBM formats
Daniel Stone [Mon, 2 Mar 2015 13:52:59 +0000 (13:52 +0000)]
egl: Take alpha bits into account when selecting GBM formats

This fixes piglit when using PIGLIT_PLATFORM=gbm

Tom Stellard:
  - Fix ARGB2101010 format

Cc: "10.4 10.5" <mesa-stable@lists.freedesktop.org>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Chad Versace <chad.versace@intel.com>
9 years agofreedreno/ir3: fix old compiler after f6b2e8af742
Rob Clark [Wed, 4 Mar 2015 16:36:32 +0000 (11:36 -0500)]
freedreno/ir3: fix old compiler after f6b2e8af742

If first_driver_param is left as zero (calloc'd struct), the result is
c0 getting clobbered.

Signed-off-by: Rob Clark <robclark@freedesktop.org>
9 years agogallivm: init MM = NULL to silence warning
Brian Paul [Tue, 3 Mar 2015 20:20:56 +0000 (13:20 -0700)]
gallivm: init MM = NULL to silence warning

Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
9 years agomapi: remove u_compiler.h
Brian Paul [Wed, 4 Mar 2015 00:15:05 +0000 (17:15 -0700)]
mapi: remove u_compiler.h

Just include c99_compat.h or util/macros.h where needed.

Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
9 years agomapi: use util/macros.h instead of locally defined macros
Brian Paul [Tue, 3 Mar 2015 16:11:35 +0000 (09:11 -0700)]
mapi: use util/macros.h instead of locally defined macros

The next step is to get rid of u_compiler.h completely.

Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
9 years agomapi: replace INLINE with inline
Brian Paul [Tue, 3 Mar 2015 16:08:22 +0000 (09:08 -0700)]
mapi: replace INLINE with inline

Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
9 years agomesa: consolidate PUBLIC macro definition
Brian Paul [Tue, 3 Mar 2015 16:01:03 +0000 (09:01 -0700)]
mesa: consolidate PUBLIC macro definition

Define the macro in src/util/macros.h rather than in two different
places.  Note that USED isn't actually used anywhere at this time.

Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
9 years agost/xlib: include p_compiler.h to get PUBLIC definition
Brian Paul [Tue, 3 Mar 2015 16:18:36 +0000 (09:18 -0700)]
st/xlib: include p_compiler.h to get PUBLIC definition

To prevent build break with following changes.

Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
9 years agomapi: remove unneeded ARRAY_SIZE #define
Brian Paul [Tue, 3 Mar 2015 16:40:40 +0000 (09:40 -0700)]
mapi: remove unneeded ARRAY_SIZE #define

include util/macros.h instead.

Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
9 years agoglx: use ARRAY_SIZE from macros.h
Brian Paul [Tue, 3 Mar 2015 16:51:19 +0000 (09:51 -0700)]
glx: use ARRAY_SIZE from macros.h

Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
9 years agoscons: Update for the fact that we require GCC 4.2
Jose Fonseca [Wed, 4 Mar 2015 13:56:35 +0000 (13:56 +0000)]
scons: Update for the fact that we require GCC 4.2

Reviewed-by: Brian Paul <brianp@vmware.com>
9 years agosvga: Set MSVC2013 compat flags.
Jose Fonseca [Wed, 4 Mar 2015 14:25:54 +0000 (14:25 +0000)]
svga: Set MSVC2013 compat flags.

Reviewed-by: Brian Paul <brianp@vmware.com>
9 years agosoftpipe,trace: Set MSVC 2008 compat flags.
Jose Fonseca [Wed, 4 Mar 2015 14:25:39 +0000 (14:25 +0000)]
softpipe,trace: Set MSVC 2008 compat flags.

Although we don't deploy these, we need to use them for debugging.

Reviewed-by: Brian Paul <brianp@vmware.com>
9 years agoscons: Use -Werror MSVC compatibility flags per-directory.
Jose Fonseca [Wed, 4 Mar 2015 14:23:52 +0000 (14:23 +0000)]
scons: Use -Werror MSVC compatibility flags per-directory.

Matching what we already do with autotools builds.

Reviewed-by: Brian Paul <brianp@vmware.com>
9 years agost/vega: Remove.
Jose Fonseca [Tue, 3 Mar 2015 16:58:21 +0000 (16:58 +0000)]
st/vega: Remove.

OpenVG API seems to have dwindled away.  The code
would still be interesting if we wanted to implement NV_path_rendering
but given the trend of the next gen graphics APIs, it seems
unlikely that this becomes ARB or core.

v2: Remove a few "openvg" references left, per Emil Velikov.

Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
v3: Update release notes.

9 years agost/egl: Remove.
Jose Fonseca [Tue, 3 Mar 2015 16:01:22 +0000 (16:01 +0000)]
st/egl: Remove.

Largely superseeded by src/egl, and
WGL/GLX_EXT_create_context_es_profile extensions.

Note this will break Android.mk with gallium drivers -- somebody
familiar with that build infrastructure will need to update it to use
gallium drivers through egl_dri2.

v2: Remove the _EGL_BUILT_IN_DRIVER_GALLIUM define from
src/egl/main/Android.mk; and update the src/egl/main/Sconscript to
create a SharedLibrary, add versioning, create symlink - copy the bits
from egl-static, per Emil Velikov.

Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
v3: Disallow undefined symbols in libEGL.so.  Update release notes

9 years agowindows/gdi: Remove.
Jose Fonseca [Tue, 3 Mar 2015 14:52:15 +0000 (14:52 +0000)]
windows/gdi: Remove.

This classic driver is so far behind Gallium softpipe/llvmpipe based
one, that's hard to imagine ever being useful.

v2: Drop drivers/windows from src/mesa/Makefile.am:EXTRA_DIST per Emil
Velikov.

Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
v3: Update release notes.

9 years agonir: Use helper macros for dealing with VLAs.
Jose Fonseca [Fri, 27 Feb 2015 15:32:24 +0000 (15:32 +0000)]
nir: Use helper macros for dealing with VLAs.

v2:
- Single statement, by using memset return value as suggested by Ian
Romanick.
- No internal declaration, as suggested by Jason Ekstrand.
- Move macros to a header.

Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
9 years agogallium/auxiliary/indices: fix start param
Marc-Andre Lureau [Fri, 27 Feb 2015 18:40:19 +0000 (19:40 +0100)]
gallium/auxiliary/indices: fix start param

Since commit 28f3f8d, indices generator take a start parameter. However, some
index values have been left to start at 0.

This fixes the glean/fbo test with the virgl driver, and copytexsubimage
with freedreno.

Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Cc: "10.4 10.5" <mesa-stable@lists.freedesktop.org>
9 years agoscons: Define _DEFAULT_SOURCE.
Vinson Lee [Sun, 1 Mar 2015 08:41:48 +0000 (00:41 -0800)]
scons: Define _DEFAULT_SOURCE.

Fix GCC cpp warnings with glibc >= 2.19.

/usr/include/features.h:148:3: warning: #warning "_BSD_SOURCE and _SVID_SOURCE are deprecated, use _DEFAULT_SOURCE" [-Wcpp]
 # warning "_BSD_SOURCE and _SVID_SOURCE are deprecated, use _DEFAULT_SOURCE"
   ^

Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Acked-by: Emil Velikov <emil.l.velikov@gmail.com>
9 years agointel: fix EGLImage renderbuffer _BaseFormat
Frank Henigman [Thu, 19 Feb 2015 23:09:20 +0000 (15:09 -0800)]
intel: fix EGLImage renderbuffer _BaseFormat

Correctly set _BaseFormat field when creating a gl_renderbuffer
with EGLImage storage.

Change-Id: I8c9f7302d18b617f54fa68304d8ffee087ed8a77
Signed-off-by: Frank Henigman <fjhenigman@google.com>
Reviewed-by: Stéphane Marchesin <marcheu@chromium.org>
Reviewed-by: Chad Versace <chad.versace@intel.com>
9 years agofreedreno/a4xx: re-enable int (conditional on glsl130)
Rob Clark [Fri, 27 Feb 2015 00:37:31 +0000 (19:37 -0500)]
freedreno/a4xx: re-enable int (conditional on glsl130)

Re-enable integer, now that we can handle flat varyings.  Still, ofc,
conditional on FD_MESA_DEBUG=glsl130, until we can deprecate _old
compiler..

Signed-off-by: Rob Clark <robclark@freedesktop.org>
9 years agofreedreno/ir3: handle flat bypass for a4xx
Rob Clark [Wed, 25 Feb 2015 18:54:25 +0000 (13:54 -0500)]
freedreno/ir3: handle flat bypass for a4xx

We may not need this for later a4xx patchlevels, but we do at least need
this for patchlevel 0.  Bypass bary.f for fetching varyings when flat
shading is needed (rather than configure via cmdstream).  This requires
a special dummy bary.f w/ (ei) flag to signal to scheduler when all
varyings are consumed.  And requires shader variants based on rasterizer
flatshade state to handle TGSI_INTERPOLATE_COLOR.

Signed-off-by: Rob Clark <robclark@freedesktop.org>
9 years agofreedreno/ir3: add support for memory (cat6) instructions
Rob Clark [Thu, 26 Feb 2015 20:13:10 +0000 (15:13 -0500)]
freedreno/ir3: add support for memory (cat6) instructions

Scheduled basically the same as texture (cat5) instructions, using (sy)
flag for synchronization.

Signed-off-by: Rob Clark <robclark@freedesktop.org>
9 years agofreedreno/ir3: fix up cat6 instruction encodings
Rob Clark [Thu, 26 Feb 2015 18:35:31 +0000 (13:35 -0500)]
freedreno/ir3: fix up cat6 instruction encodings

I think there is at least one more sub-encoding, but these two should be
enough to cover the common load/store instructions.

Signed-off-by: Rob Clark <robclark@freedesktop.org>
9 years agotgsi/lowering: don't forget interp for BCOLOR inputs
Rob Clark [Fri, 27 Feb 2015 14:02:48 +0000 (09:02 -0500)]
tgsi/lowering: don't forget interp for BCOLOR inputs

To lower two sided color, tgsi_lowering creates additional BCOLOR inputs
(matching up to the BCOLOR outputs on the vert shader).  These inputs
should copy the interpolation state of their matching COLOR input.

Signed-off-by: Rob Clark <robclark@freedesktop.org>
9 years agofreedreno/a3xx,a4xx: silence some warnings
Rob Clark [Fri, 27 Feb 2015 02:06:02 +0000 (21:06 -0500)]
freedreno/a3xx,a4xx: silence some warnings

  fd3_emit.c: In function ‘fd3_emit_vertex_bufs’:
  fd3_emit.c:377:11: warning: unused variable ‘semantic’ [-Wunused-variable]
     uint8_t semantic = sem2name(vp->inputs[i].semantic);

and

  fd4_emit.c: In function ‘fd4_emit_vertex_bufs’:
  fd4_emit.c:304:11: warning: unused variable ‘semantic’ [-Wunused-variable]
     uint8_t semantic = sem2name(vp->inputs[i].semantic);

Signed-off-by: Rob Clark <robclark@freedesktop.org>
9 years agoc99_alloca.h: add case for __sun
Brian Paul [Tue, 3 Mar 2015 15:09:21 +0000 (08:09 -0700)]
c99_alloca.h: add case for __sun

Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com>
9 years agoconfigure: Leverage gcc warn options to enable safe use of C99 features where possible.
Jose Fonseca [Thu, 26 Feb 2015 16:46:48 +0000 (16:46 +0000)]
configure: Leverage gcc warn options to enable safe use of C99 features where possible.

The main objective of this change is to enable Linux developers to use
more of C99 throughout Mesa, with confidence that the portions that need
to be built with MSVC -- and only those portions --, stay portable.

This is achieved by using the appropriate -Werror= options only on the
places they need to be used.

Unfortunately we still need MSVC 2008 on a few portions of the code
(namely llvmpipe and its dependencies).  I hope to eventually eliminate
this so that we can use C99 everywhere, but there are technical/logistic
challenges (specifically, newer Windows SDKs no longer bundle MSVC,
instead require a full installation of Visual Studio, and that has
hindered adoption of newer MSVC versions on our build processes.)
Thankfully we have more directy control over our OpenGL driver, which is
why we're now able to migrate to MSVC 2013 for most of the tree.

Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
9 years agoi965: Fix assertion in brw_reg_type_letters
Ben Widawsky [Fri, 28 Nov 2014 22:28:19 +0000 (14:28 -0800)]
i965: Fix assertion in brw_reg_type_letters

While using various debugging features (optimization debug, instruction dumping,
etc) this function is called in order to get a readable letter for the type of
unit.

On GEN8, two new units were added, the Qword and the Unsigned Qword (Q, and UQ
respectively). The existing assertion tries to determine that the argument
passed in is within the correct boundary, however, it was using UQ as the upper
limit instead of Q.

To my knowledge you can only hit this case with the branch I am currently
working on, so it doesn't fix any known issues.

Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Reviewed-by: Matt Turner <mattst88@gmail.com>
9 years agoi965: Rename some PIPE_CONTROL flags
Ben Widawsky [Fri, 27 Feb 2015 07:01:33 +0000 (23:01 -0800)]
i965: Rename some PIPE_CONTROL flags

I'm not really sure of the origins of the existing flag names. Modern docs have
some slightly different names. Having the correct names makes it easier to
determine if existing PIPE_CONTROL flag settings are correct, as well as making
adding new PIPE_CONTROLs easier.

This originally came up while I was trying to implement workarounds and spotted
some things called, "flush" which should have been called "invalidate."

Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
9 years agoi965/fs: Don't use backend_visitor::instructions after creating the CFG.
Matt Turner [Tue, 13 Jan 2015 23:35:57 +0000 (15:35 -0800)]
i965/fs: Don't use backend_visitor::instructions after creating the CFG.

This is a fix for a regression introduced in commit a9f8296d ("i965/fs:
Preserve the CFG in a few more places.").

The errata this code works around is described in a comment before the function:

   "[DevBW, DevCL] Errata: A destination register from a send can not be
    used as a destination register until after it has been sourced by an
    instruction with a different destination register.

The framebuffer write's sources must be in message registers, which SEND
instructions cannot have as a destination. There's no way for this
errata to affect anything at the end of the program. Just remove the
code.

Cc: 10.4, 10.5 <mesa-stable@lists.freedesktop.org>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=84613
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
9 years agomain/base_tex_format: Properly handle STENCIL_INDEX1/4/16
Jason Ekstrand [Fri, 27 Feb 2015 23:53:11 +0000 (18:53 -0500)]
main/base_tex_format: Properly handle STENCIL_INDEX1/4/16

This takes "fbo-stencil blit GL_STENCIL_INDEX1/4/16" from crash to pass on
BDW.

Cc: 10.5 <mesa-stable@lists.freedesktop.org>
Reviewed-by: Matt Turner <mattst88@gmail.com>
9 years agometa/TexSubImage: Stash everything other than PIXEL_TRANSFER/store in meta_begin
Jason Ekstrand [Fri, 27 Feb 2015 20:29:03 +0000 (12:29 -0800)]
meta/TexSubImage: Stash everything other than PIXEL_TRANSFER/store in meta_begin

Previously, there were bugs where if the app set a scissor it could affect
the area of the texture that was downloaded.  There was also potential that
the framebuffer SRGB state could affect downloads.  This ensures that those
will get saved/restored and can't affect the texture download.

Cc: 10.5 <mesa-stable@lists.freedesktop.org>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=89292
Reviewed-by: Neil Roberts <neil@linux.intel.com>
9 years agoi915: Remove hand-rolled memcpy implementation.
Matt Turner [Sat, 21 Feb 2015 04:32:14 +0000 (20:32 -0800)]
i915: Remove hand-rolled memcpy implementation.

Reviewed-by: Eric Anholt <eric@anholt.net>
9 years agoi965: Remove hand-rolled memcpy implementation.
Matt Turner [Sat, 21 Feb 2015 04:28:52 +0000 (20:28 -0800)]
i965: Remove hand-rolled memcpy implementation.

Reviewed-by: Eric Anholt <eric@anholt.net>
9 years agoi965: Consider scratch writes to have side effects.
Matt Turner [Sat, 28 Feb 2015 21:36:21 +0000 (13:36 -0800)]
i965: Consider scratch writes to have side effects.

We could do better by tracking scratch reads and writes.

Cc: 10.5 <mesa-stable@lists.freedesktop.org>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=88793
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
9 years agomesa: Correct backwards NULL check.
Matt Turner [Sat, 28 Feb 2015 19:14:02 +0000 (11:14 -0800)]
mesa: Correct backwards NULL check.

Cc: "10.4, 10.5" <mesa-stable@lists.freedesktop.org>
Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
9 years agomesa: Free memory allocated for luminance in readpixels.
Matt Turner [Sat, 28 Feb 2015 19:08:17 +0000 (11:08 -0800)]
mesa: Free memory allocated for luminance in readpixels.

Cc: "10.4, 10.5" <mesa-stable@lists.freedesktop.org>
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
9 years agomesa: Indent break statements and add a missing one.
Matt Turner [Sat, 28 Feb 2015 19:00:51 +0000 (11:00 -0800)]
mesa: Indent break statements and add a missing one.

Always indenting break statements makes spotting missing ones easier.

Cc: "10.4, 10.5" <mesa-stable@lists.freedesktop.org>
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
9 years agoc99_alloca.h: Include stdlib.h on all non-Windows.
Vinson Lee [Sun, 1 Mar 2015 20:52:00 +0000 (13:52 -0700)]
c99_alloca.h: Include stdlib.h on all non-Windows.

Fix build on FreeBSD.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=89364
Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Tested-by: Brian Paul <brianp@vmware.com>
9 years agomesa: remove extra definition of ARRAY_SIZE in src/mesa/main/macros.h
Brian Paul [Sat, 28 Feb 2015 20:33:11 +0000 (13:33 -0700)]
mesa: remove extra definition of ARRAY_SIZE in src/mesa/main/macros.h

Already defined in src/util/macros.h

Reviewed-by: Matt Turner <mattst88@gmail.com>
9 years agomesa: remove the Elements() macro definition
Brian Paul [Sat, 28 Feb 2015 16:11:43 +0000 (09:11 -0700)]
mesa: remove the Elements() macro definition

No longer used.

Acked-by: Ilia Mirkin <imirkin@alum.mit.edu>
9 years agoutil: replace Elements() with ARRAY_SIZE()
Brian Paul [Sat, 28 Feb 2015 20:28:06 +0000 (13:28 -0700)]
util: replace Elements() with ARRAY_SIZE()

Reviewed-by: Matt Turner <mattst88@gmail.com>
9 years agoradeon: replace Elements() with ARRAY_SIZE()
Brian Paul [Sat, 28 Feb 2015 16:05:54 +0000 (09:05 -0700)]
radeon: replace Elements() with ARRAY_SIZE()

Acked-by: Ilia Mirkin <imirkin@alum.mit.edu>
9 years agor200: replace Elements() with ARRAY_SIZE()
Brian Paul [Sat, 28 Feb 2015 16:05:48 +0000 (09:05 -0700)]
r200: replace Elements() with ARRAY_SIZE()

Acked-by: Ilia Mirkin <imirkin@alum.mit.edu>
9 years agonouveau: replace Elements() with ARRAY_SIZE()
Brian Paul [Sat, 28 Feb 2015 16:05:40 +0000 (09:05 -0700)]
nouveau: replace Elements() with ARRAY_SIZE()

Acked-by: Ilia Mirkin <imirkin@alum.mit.edu>
9 years agoi965: replace Elements() with ARRAY_SIZE()
Brian Paul [Sat, 28 Feb 2015 16:05:29 +0000 (09:05 -0700)]
i965: replace Elements() with ARRAY_SIZE()

Acked-by: Ilia Mirkin <imirkin@alum.mit.edu>
9 years agoi915: replace Elements() with ARRAY_SIZE()
Brian Paul [Sat, 28 Feb 2015 16:05:24 +0000 (09:05 -0700)]
i915: replace Elements() with ARRAY_SIZE()

Acked-by: Ilia Mirkin <imirkin@alum.mit.edu>
9 years agomapi: replace Elements() with ARRAY_SIZE()
Brian Paul [Sat, 28 Feb 2015 16:11:29 +0000 (09:11 -0700)]
mapi: replace Elements() with ARRAY_SIZE()

Acked-by: Ilia Mirkin <imirkin@alum.mit.edu>
9 years agoglsl: replace Elements() with ARRAY_SIZE()
Brian Paul [Sat, 28 Feb 2015 16:11:23 +0000 (09:11 -0700)]
glsl: replace Elements() with ARRAY_SIZE()

Acked-by: Ilia Mirkin <imirkin@alum.mit.edu>
9 years agost/dri: replace Elements() with ARRAY_SIZE()
Brian Paul [Sat, 28 Feb 2015 16:11:13 +0000 (09:11 -0700)]
st/dri: replace Elements() with ARRAY_SIZE()

Acked-by: Ilia Mirkin <imirkin@alum.mit.edu>
9 years agost/mesa: replace Elements() with ARRAY_SIZE()
Brian Paul [Sat, 28 Feb 2015 16:01:33 +0000 (09:01 -0700)]
st/mesa: replace Elements() with ARRAY_SIZE()

Acked-by: Ilia Mirkin <imirkin@alum.mit.edu>
9 years agomesa/program: replace Elements() with ARRAY_SIZE()
Brian Paul [Sat, 28 Feb 2015 15:57:27 +0000 (08:57 -0700)]
mesa/program: replace Elements() with ARRAY_SIZE()

Acked-by: Ilia Mirkin <imirkin@alum.mit.edu>
9 years agomesa/swrast: replace Elements() with ARRAY_SIZE()
Brian Paul [Sat, 28 Feb 2015 15:57:20 +0000 (08:57 -0700)]
mesa/swrast: replace Elements() with ARRAY_SIZE()

Acked-by: Ilia Mirkin <imirkin@alum.mit.edu>
9 years agomesa/vbo: replace Elements() with ARRAY_SIZE()
Brian Paul [Sat, 28 Feb 2015 15:57:15 +0000 (08:57 -0700)]
mesa/vbo: replace Elements() with ARRAY_SIZE()

Acked-by: Ilia Mirkin <imirkin@alum.mit.edu>
9 years agomesa/main: replace Elements() with ARRAY_SIZE()
Brian Paul [Sat, 28 Feb 2015 15:57:11 +0000 (08:57 -0700)]
mesa/main: replace Elements() with ARRAY_SIZE()

We've been using a mix of these two macros for a while now.  Let's
just use the later everywhere.  It seems to be the convention used
by other open-source projects.

Acked-by: Ilia Mirkin <imirkin@alum.mit.edu>
9 years agomesa: trim down #includes in api_loopback.h
Brian Paul [Thu, 26 Feb 2015 20:06:57 +0000 (13:06 -0700)]
mesa: trim down #includes in api_loopback.h

Acked-by: Matt Turner <mattst88@gmail.com>
9 years agomesa: trim down includes of compiler.h
Brian Paul [Thu, 26 Feb 2015 20:03:05 +0000 (13:03 -0700)]
mesa: trim down includes of compiler.h

In some cases, glheader.h is the right #include.
Also remove some instances of struct _glapi_table declarations.

Acked-by: Matt Turner <mattst88@gmail.com>
9 years agoscons: Fix HAVE___* definition.
Jose Fonseca [Fri, 27 Feb 2015 15:42:23 +0000 (15:42 +0000)]
scons: Fix HAVE___* definition.

These definitions must be moved before `cppdefines` is used to have effect.

Trivial.

9 years agoidentity: Remove.
Jose Fonseca [Fri, 27 Feb 2015 14:43:51 +0000 (14:43 +0000)]
identity: Remove.

It's unmaintained, and most likely broken: I use trace driver every now
and then, and everytime I do I need to fix it up.

It's also unused: identity_screen_create is never called.

Above all, it's dead weight: if identity driver had the infrastructure
for other pass-through drivers (like trace and rbug), then it would make
sense on its own right.  But as it is implemmented, it's just another
driver to (forget) to update whenever there is a gallium interface
change.

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
9 years agoi965: Remove the create_raw_surface vtbl hook.
Francisco Jerez [Tue, 10 Feb 2015 13:54:55 +0000 (15:54 +0200)]
i965: Remove the create_raw_surface vtbl hook.

It's a wrapper around emit_buffer_surface_state with format=RAW, pitch=1,
rw=true and the remaining arguments ordered differently.  There's no point in
having a separate vtbl pointer for that.

Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
9 years agoi965: Add missing defines for render cache messages.
Francisco Jerez [Sat, 23 Nov 2013 00:18:43 +0000 (16:18 -0800)]
i965: Add missing defines for render cache messages.

And remove duplicated definition of OWORD_DUAL_BLOCK_WRITE.

Reviewed-by: Paul Berry <stereotype441@gmail.com>
9 years agoi965/skl: Lay out a 1D miptree horizontally
Neil Roberts [Thu, 12 Feb 2015 14:39:07 +0000 (14:39 +0000)]
i965/skl: Lay out a 1D miptree horizontally

On Gen9+ the 1D miptree is laid out with all of the mipmap levels in a
horizontal line.

Reviewed-by: Ben Widawsky <ben@bwidawsk.net>
9 years agoi965/skl: Lay out 3D textures the same as array textures
Neil Roberts [Wed, 18 Feb 2015 18:27:45 +0000 (18:27 +0000)]
i965/skl: Lay out 3D textures the same as array textures

On Gen9+ the 3D textures use the same mipmap layout as 2D array
textures.

Reviewed-by: Ben Widawsky <ben@bwidawsk.net>
9 years agoi965/skl: Fix the maximum thread count format for the PS
Neil Roberts [Fri, 27 Feb 2015 16:57:22 +0000 (16:57 +0000)]
i965/skl: Fix the maximum thread count format for the PS

According to the bspec for some reason the format of the maximum
number of threads field has changed from U8-2 to U8-1 for the PS.

Reviewed-by: Anuj Phogat <anuj.phogat@gmail.com>
9 years agodraw: fix division-by-zero for empty geometry shaders
Marek Olšák [Sun, 1 Mar 2015 10:18:42 +0000 (11:18 +0100)]
draw: fix division-by-zero for empty geometry shaders

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=89372

Reviewed-by: Dave Airlie <airlied@redhat.com>
9 years agoi965/gs: Check newly-generated GS-out VUE map against correct stage
Chris Forbes [Sat, 28 Feb 2015 06:57:20 +0000 (19:57 +1300)]
i965/gs: Check newly-generated GS-out VUE map against correct stage

Previously, we compared our new GS-out VUE map to the existing *VS*-out
VUE map, which is bogus.

This would mostly manifest as redundant dirty flagging where the GS is
in use but the VS and GS output layouts differ; but there is a scary
case where we would fail to flag a GS-out layout change if it happened
to match the VS-out layout.

Signed-off-by: Chris Forbes <chrisf@ijw.co.nz>
Cc: "10.5, 10.4" <mesa-stable@lists.freedesktop.org>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=88885

9 years agoi965: add GLSL_TYPE_DOUBLE switch case to silence warning
Brian Paul [Sat, 28 Feb 2015 15:43:16 +0000 (08:43 -0700)]
i965: add GLSL_TYPE_DOUBLE switch case to silence warning

Reviewed-by: Matt Turner <mattst88@gmail.com>
9 years agomesa: include macros.h in stencil.h
Brian Paul [Thu, 26 Feb 2015 20:06:31 +0000 (13:06 -0700)]
mesa: include macros.h in stencil.h

Since it uses the CLAMP macro.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
9 years agomesa: move finite macro to imports.h
Brian Paul [Thu, 26 Feb 2015 20:22:36 +0000 (13:22 -0700)]
mesa: move finite macro to imports.h

Move it to the only place it's used.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
9 years agomesa: remove _NORMAPI, _NORMAPIP macros
Brian Paul [Thu, 26 Feb 2015 19:26:09 +0000 (12:26 -0700)]
mesa: remove _NORMAPI, _NORMAPIP macros

Was only used in one place.  Use equivalent _XFORMAPIP there instead.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
9 years agomesa: move FLT_MAX_EXP to c99_math.h
Brian Paul [Thu, 26 Feb 2015 18:35:10 +0000 (11:35 -0700)]
mesa: move FLT_MAX_EXP to c99_math.h

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
9 years agomesa: move ONE_DIV_SQRT_LN2 to prog_statevars.c
Brian Paul [Thu, 26 Feb 2015 18:34:23 +0000 (11:34 -0700)]
mesa: move ONE_DIV_SQRT_LN2 to prog_statevars.c

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
9 years agomesa: remove unused uninitialized_var() macro
Brian Paul [Sat, 28 Feb 2015 20:38:11 +0000 (13:38 -0700)]
mesa: remove unused uninitialized_var() macro

Reviewed-by: Matt Turner <mattst88@gmail.com>
9 years agomesa: Check return value of __get_cpuid().
Matt Turner [Sat, 28 Feb 2015 18:06:26 +0000 (10:06 -0800)]
mesa: Check return value of __get_cpuid().

The use of the uninitialized_var() macro was to silence an uninitialized
variable warning that I assumed stemmed from gcc being unable to see
inside __get_cpuid() or understand its inline assembly.

In fact, it was because the __get_cpuid() function can fail, and not
initialize its arguments. Instead, check for failure and return early.

Reviewed-by: Brian Paul <brianp@vmware.com>
9 years agoi965/fs/nir: Mark fallthrough.
Matt Turner [Sat, 28 Feb 2015 18:46:33 +0000 (10:46 -0800)]
i965/fs/nir: Mark fallthrough.

9 years agoi965/fs/nir: Mark fallthrough.
Matt Turner [Sat, 28 Feb 2015 18:37:50 +0000 (10:37 -0800)]
i965/fs/nir: Mark fallthrough.

9 years agoi965: Avoid applying negate to wrong MAD source.
Matt Turner [Fri, 27 Feb 2015 18:59:17 +0000 (10:59 -0800)]
i965: Avoid applying negate to wrong MAD source.

For some given GLSL IR like (+ (neg x) (* 1.2 x)), the try_emit_mad
function would see that one of the +'s sources was a negate expression
and set mul_negate = true without confirming that it was actually a
multiply.

Cc: 10.5 <mesa-stable@lists.freedesktop.org>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=89315
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=89095
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
9 years agoi965/vec4: Fix implementation of i2b.
Matt Turner [Fri, 27 Feb 2015 06:49:47 +0000 (22:49 -0800)]
i965/vec4: Fix implementation of i2b.

I broke this in commit 2881b123d. I must have misread i2b as b2i.

Cc: 10.5 <mesa-stable@lists.freedesktop.org>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=88246
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
9 years agoi965/fs/nir: Use emit_math for nir_op_fpow
Ian Romanick [Fri, 27 Feb 2015 22:17:50 +0000 (14:17 -0800)]
i965/fs/nir: Use emit_math for nir_op_fpow

It appears that all the other instructions that need it already use it.
This one just got missed.

Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Cc: "10.5" <mesa-stable@lists.freedesktop.org>
9 years agomapi: Don't rely on GNU void pointer arithmetic.
Matt Turner [Fri, 27 Feb 2015 19:42:43 +0000 (11:42 -0800)]
mapi: Don't rely on GNU void pointer arithmetic.

Commit 79daa510c added -Werror=pointer-arith to CFLAGS, which makes
arithmetic on void pointers an error.

See https://gcc.gnu.org/onlinedocs/gcc/Pointer-Arith.html

Reviewed-by: Brian Paul <brianp@vmware.com>
9 years agoRevert "configure: Leverage gcc warn options to enable safe use of C99 features where...
Kenneth Graunke [Fri, 27 Feb 2015 23:48:06 +0000 (15:48 -0800)]
Revert "configure: Leverage gcc warn options to enable safe use of C99 features where possible."

This reverts commit 79daa510c7a871a33797308a2ccb4b83a067ffbe.

I apparently hadn't done a clean build when testing this; it broke the
build for Tom, Ben, and myself.  We like the idea; let's try a v2.

9 years agoauxilary/os: correct sysctl use in os_get_total_physical_memory()
Jonathan Gray [Sun, 22 Feb 2015 08:19:25 +0000 (19:19 +1100)]
auxilary/os: correct sysctl use in os_get_total_physical_memory()

The length argument passed to sysctl was the size of the pointer
not the type.  The result of this is sysctl calls would fail on
32 bit BSD/Mac OS X.

Additionally the wrong pointer was passed as an argument to store
the result of the sysctl call.

Cc: "10.4, 10.5" <mesa-stable@lists.freedesktop.org>
Signed-off-by: Jonathan Gray <jsg@jsg.id.au>
Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
9 years agoglsl: silence uninitialized var warning on MinGW
Brian Paul [Fri, 27 Feb 2015 20:06:06 +0000 (13:06 -0700)]
glsl: silence uninitialized var warning on MinGW

Reviewed-by: Anuj Phogat <anuj.phogat@gmail.com>
9 years agomesa: silence unused var warning in get_tex_rgba_uncompressed()
Brian Paul [Fri, 27 Feb 2015 20:05:42 +0000 (13:05 -0700)]
mesa: silence unused var warning in get_tex_rgba_uncompressed()

Reviewed-by: Anuj Phogat <anuj.phogat@gmail.com>