- Nov 24, 2021
-
-
Dylan Baker authored
-
Dylan Baker authored
-
spirv_to_nir sometimes wraps derefs in vec2 or mov instructions as part of its texture handling. These get in the way of nir_rematerialize_derefs_in_use_blocks_impl. Running copy propagation should get rid of the extra move instructions and get us back to intact deref chains for everything except variable pointer use-cases. fossil-db (Sienna Cichlid): Totals from 6 (0.00% of 134572) affected shaders: CodeSize: 92656 -> 93088 (+0.47%) Instrs: 17060 -> 17138 (+0.46%) Latency: 224408 -> 227539 (+1.40%) InvThroughput: 37402 -> 37924 (+1.40%) VClause: 408 -> 402 (-1.47%) Copies: 1065 -> 1107 (+3.94%) Signed-off-by: Rhys Perry <pendingchaos02@gmail.com> Reviewed-by: Jason Ekstrand <jason@jlekstrand.net> Closes: #5668 Fixes: 14a12b77 ("spirv: Rework our handling of images and samplers") Part-of: <!13924> (cherry picked from commit b4251007)
-
The function need_temp_reg_initialization looks suspecious. It will only ever return true if we get past this if: if (!(emit->info.indirect_files && (1u << TGSI_FILE_TEMPORARY)) ... Using the logical && means the intended initialization done based on the result of this check is not performed. This code was both introduced and altered in MR 5317. ccb4ea5a introduces the function. ba37d408 is a collection of performance improvements and misc fixes. This altered the if from using bitwise to logical and. This commit changes it back to bitwise. Spotted from a compile warning. Fixes: ba37d408 ("svga: Performance fixes") Reviewed-by: Zoltán Böszörményi <zboszor@gmail.com> Part-of: <!12157> (cherry picked from commit 64292c0f)
-
Android build system may use different internal variables to specify cflags/cppflags. Small change in product confguration may force Android to use diffrent set of variables, therefore we should keep all of them attached to the make rule's target. Closes: mesa/mesa#5549 Fixes: 8621bd8d ("android: Add scripts to build using meson") Signed-off-by: Roman Stratiienko <r.stratiienko@gmail.com> Acked-by: Jesse Natalie <jenatali@microsoft.com> Reviewed-by: Marijn Suijten <marijn.suijten@somainline.org> Part-of: <mesa/mesa!13914> (cherry picked from commit 32ec0fff)
-
If a predecessor has only spilled constants (no temporaries), spills_exit will be empty. fossil-db (Sienna Cichlid): Totals from 2 (0.00% of 128647) affected shaders: Latency: 139106 -> 139104 (-0.00%) Signed-off-by: Rhys Perry <pendingchaos02@gmail.com> Reviewed-by: Daniel Schürmann <daniel@schuermann.dev> Closes: #5633 Cc: mesa-stable Part-of: <!13821> (cherry picked from commit cc289434)
-
Now that we removed the intel intrinsic and just use the generic one, we can skip it in the intel call lowering pass and just deal with it in the intel rt intrinsic lowering. v2: rewrite with nir_shader_instructions_pass() (Jason) v3: handle everything in switch (Jason) Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com> Fixes: 423c47de ("nir: drop the btd_resume_intel intrinsic") Reviewed-by: Jason Ekstrand <jason@jlekstrand.net> Part-of: <mesa/mesa!12113> (cherry picked from commit c5a42e40)
-
If the source and destination were within the same full register, like hr90.x and hr90.y (which both map to r45.x), then we'd perform the swap/copy with the wrong register. This broke dEQP-VK.ssbo.phys.layout.random.16bit.scalar.35 once BDA is enabled. Fixes: 0ffcb19b ("ir3: Rewrite register allocation") Part-of: <!13818> (cherry picked from commit c98adc56)
-
The immediate shouldn't be half-reg because the other source isn't. Fixes an assertion failure with dEQP-VK.ssbo.phys.layout.random.16bit.scalar.35. Fixes: 0ffcb19b ("ir3: Rewrite register allocation") Part-of: <mesa/mesa!13818> (cherry picked from commit 65da866a)
-
Fixes: 82d9a37a ("glsl/nir: Add a shared helper for building float64 shaders") Closes: #5556 Reviewed-by: Matt Turner <mattst88@gmail.com> Reviewed-by: Tapani Pälli <tapani.palli@intel.com> Part-of: <mesa/mesa!13828> (cherry picked from commit 04f5c543)
-
Fixes: 3be42f9c ("mesa: rewrite glPushAttrib/glPopAttrib to get rid of malloc") Closes: #5621 Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu> Part-of: <!13773> (cherry picked from commit 9d9de15a)
-
VK CTS just added some new tests to write to a compressed image from a compute shader, which was overrunning memory. The image width/height need to be sized according to the block sizes to avoid overwriting memory. dEQP-VK.image.sample_texture.*bit_compressed* Cc: mesa-stable Reviewed-by: Roland Scheidegger <sroland@vmware.com> Part-of: <mesa/mesa!13618> (cherry picked from commit 27903abb)
-
This fixes some crashes in VK-GL-CTS where it doesn't deal with these. Cc: mesa-stable Reviewed-by: Roland Scheidegger <sroland@vmware.com> Part-of: <mesa/mesa!13617> (cherry picked from commit 53a8faaf)
-
p_is_helper doesn't have any operands, so ACO's value numbering and/or the pre-RA optimizer could incorrectly recognize two such instructions as the same. This patch adds exec as an operand to p_is_helper in order to achieve correct behavior. Cc: mesa-stable Closes: mesa/mesa#5570 Signed-off-by: Timur Kristóf <timur.kristof@gmail.com> Reviewed-by: Daniel Schürmann <daniel@schuermann.dev> Part-of: <mesa/mesa!13577> (cherry picked from commit d80c7f34)
-
Dylan Baker authored
-
Future LLVM header leads to __declspec(__restrict), which is invalid. Just undefine the restrict macro to keep __declspec(restrict). Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl> (cherry picked from commit e0de7aa4) # Conflicts: # src/amd/compiler/aco_print_asm.cpp Part-of: <mesa/mesa!13813>
-
zero is for success, nonzero is failure. Fixes: 0b60d6a2 ("d3d12: Support Linux eventfds for fences") Reviewed-by: Jason Ekstrand <jason@jlekstrand.net> Part-of: <mesa/mesa!12268> (cherry picked from commit accd8326)
-
When the client calls vkMapMemory(), we have to align the requested offset down to the nearest page or else the map will fail. On platforms where we have DRM_IOCTL_I915_GEM_MMAP_OFFSET, we always map the whole buffer. In either case, the original map may start before the requested offset and we need to take that into account when we clflush. Cc: mesa-stable@lists.freedesktop.org Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com> Part-of: <mesa/mesa!13610> (cherry picked from commit 90ac06e5)
-
This is already done on lines 475-480, resulting in them appearing twice in the summary. Fixes: 47946855 ("meson: allow egl_native_platform to be specified") Signed-off-by: Eric Engestrom <eric@engestrom.ch> Reviewed-by: Adam Jackson <ajax@redhat.com> Part-of: <mesa/mesa!13278> (cherry picked from commit 9ad375bd)
-
Dylan Baker authored
-
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com> Fixes: d399c3e8 ("intel/dev: Add device info for ADL-S") Reviewed-by: Jordan Justen <jordan.l.justen@intel.com> Part-of: <mesa/mesa!13476> (cherry picked from commit 3b1a5b8f)
-
Dylan Baker authored
-
A bit difficult to find what commit introduced the issue because of all the renaming, but it was my bug :) Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com> Cc: mesa-stable Reviewed-by: Francisco Jerez <currojerez@riseup.net> Part-of: <mesa/mesa!10015> (cherry picked from commit 349bfb72)
-
v2: Fix comment change (Marcin) Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com> Cc: <mesa-stable@lists.freedesktop.org> Reviewed-by: Francisco Jerez <currojerez@riseup.net> Part-of: <mesa/mesa!10015> (cherry picked from commit 67619d81)
-
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com> Cc: <mesa-stable@lists.freedesktop.org> Reviewed-by: Francisco Jerez <currojerez@riseup.net> Part-of: <mesa/mesa!10015> (cherry picked from commit d7c6a90c)
-
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com> Cc: <mesa-stable@lists.freedesktop.org> Reviewed-by: Francisco Jerez <currojerez@riseup.net> Part-of: <mesa/mesa!10015> (cherry picked from commit d1db5d56)
-
10c75ae4 moved handling of this state to the functions that depend on ctx->_ImageTransferState. So we can't depend on _NEW_PIXEL being set to call this function, since it'll be always clear earlier by _mesa_update_state_locked. Example sequence that would trigger the issue: glPixelTransferi(...) glClear(...) glTexSubImage2D(...) <-- won't use the new value set by glPixelTransferi because glClear caused _NEW_PIXEL to be cleared. _NEW_PIXEL itself is kept because st_update_pixel_transfer depends on it. Fixes: 10c75ae4 ("mesa: move _mesa_update_pixel out of _mesa_update_state") Closes: mesa/mesa#5273 Reviewed-by: Marek Olšák <marek.olsak@amd.com> Part-of: <mesa/mesa!13596> (cherry picked from commit 1ee3fbd7)
-
Fixes: 9defe8ac - radeonsi: implement fast Z/S clears using clear_buffer on HTILE Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com> Part-of: <mesa/mesa!13603> (cherry picked from commit 74adf22a)
-
- Unset depth_cleared_level_mask for non-clear blits. Set the flag after the clear, so that we don't have to check blitter_running. - Set depth_cleared_level_mask only when we set depth_clear_value. Fixes: ff8a930c - radeonsi: add _once suffix to depth_cleared_level_mask Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com> Part-of: <mesa/mesa!13603> (cherry picked from commit 5d3aea49)
-
vertex shader stages that can produce xfb must have their input size clamped to the compiler define MAX_VARYING to successfully be able to export an xfb output for each input fixes KHR-GL46.geometry_shader.limits.max_input_components cc: mesa-stable Reviewed-by: Dave Airlie <airlied@redhat.com> Part-of: <mesa/mesa!13632> (cherry picked from commit 5d1b81d8)
-
Cc: mesa-stable Reviewed-by: Marek Olšák <marek.olsak@amd.com> Part-of: <mesa/mesa!13550> (cherry picked from commit dbf602a6)
-
This is used by variable indexing of constant arrays, to build code like this: s_add_u32 s6, s6, const_data@rel32@lo+4 s_addc_u32 s7, s7, const_data@rel32@hi+12 [...] global_load_dword v4, v4, s[6:7 Closes: mesa/mesa#5118 Fixes: 82888829 ("radeonsi: set MEM_ORDERED optimally") Reviewed-by: Marek Olšák <marek.olsak@amd.com> Part-of: <mesa/mesa!13550> (cherry picked from commit dc56301f)
-
if no restart indices are found, this draw must be discarded to avoid crashing later on Fixes: 58307074 ("util/primconvert: handle rewriting of prim-restart draws with unsupported primtype") Reviewed-by: Jesse Natalie <jenatali@microsoft.com> Part-of: <mesa/mesa!13630> (cherry picked from commit bc345281)
-
Dylan Baker authored
-
Dylan Baker authored
-
- Nov 03, 2021
-
-
Fixes: e532a47f ("util/hash_table: do not leak u64 struct key") Reviewed-by: Emma Anholt <emma@anholt.net> Part-of: <mesa/mesa!13390> (cherry picked from commit ffd4157b)
-
Fixes dEQP-VK.reconvergence.*nesting* tests. There are cases when cmod is set to an instruction that cannot have conditional modifier. E.g. following: find_live_channel(32) vgrf166:UD, NoMask cmp.z.f0.0(32) null:D, vgrf166+0.0<0>:D, 0d is optimized to: find_live_channel.z.f0.0(32) vgrf166:UD, NoMask v2: - Add unit test to check cmod is not set to 'find_live_channel' (Matt Turner) - Update flag_subreg when conditonal_mod is updated (Ian Romanick) Signed-off-by: Vadym Shovkoplias <vadym.shovkoplias@globallogic.com> Closes: mesa/mesa#5431 Fixes: 32b7ba66 ("intel/compiler: fix cmod propagation optimisations") Reviewed-by: Matt Turner <mattst88@gmail.com> Part-of: <mesa/mesa!13268> (cherry picked from commit 2dbb6699)
-
Dylan Baker authored
-
- Nov 01, 2021
-
-
these aren't exact matches in spirv, so set the expected result type to float where necessary cc: mesa-stable fixes #5567 Reviewed-by: Dave Airlie <airlied@redhat.com> Part-of: <mesa/mesa!13562> (cherry picked from commit 73af6788)
-