Issue metadata
Sign in to add a comment
|
Chrome is not rendering some PNG images correctly
Reported by
rodgersn...@gmail.com,
Jul 26
|
|||||||||||||||||||||||||||||||
Issue descriptionUserAgent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/68.0.3440.75 Safari/537.36 Example URL: https://techzim.co.zw Steps to reproduce the problem: 1. Visit https://techzim.co.zw 2. Observe the "techzim" logo image in the top left corner at time of reporting What is the expected behavior? Png images should display without a green border. The same behavior was observed on other sites which are not linked because they are under development. What went wrong? There is a png image that displays with a green border. I have attached a screenshot of the bug, and the original png image that is not rendering correctly Does it occur on multiple sites: Yes Is it a problem with a plugin? N/A Did this work before? Yes Don't know my last version Does this work in other browsers? Yes Chrome version: 68.0.3440.75 Channel: stable OS Version: Flash Version:
,
Jul 26
,
Jul 27
rodgersnigel@ Thanks for the issue. Able to reproduce this issue on Ubuntu 16.04 on the latest Stable 68.0.3440.75 and Dev 69.0.3497.12 as per the original comment. Note: Issue is not reproducible on Windows 10 and Mac OS 10.13.3. This issue looks similar to issue 862908 and the bisect information is also same. Bisect information: =================== Good build :68.0.3402.0 Bad build : 68.0.3403.0 Hence assigning the issue to brianosman@ - owner of the issue 862908 and requesting to look into the issue. Adding ReleaseBlock-Stable for M-68 as this is a recent regression. Please feel free to remove if it is not applicable. Thanks..
,
Jul 27
,
Jul 27
The following revision refers to this bug: https://skia.googlesource.com/skcms/+/23e7777f421d11992e8245771641141aa273556b commit 23e7777f421d11992e8245771641141aa273556b Author: Mike Klein <mtklein@google.com> Date: Fri Jul 27 14:42:26 2018 clamp before premul When alpha is <1, it's possible to fit r,g,b >1 into premul unorms, and that subverts the expectations users who consume the output of skcms_Transform(). When working with unorm pixels, downstream consumers will assume the r,g,b values are also within [0,1], i.e. are less than the alpha channel. Put another way, they assume the colors are in gamut. So clamp earlier, before premul. (The reordering has no effect on force_opaque or swap_rb.) Bug: chromium:867813 Bug: chromium:862908 Change-Id: I553b42df06482a4e3cd3a9377c2adddb8ff3d1bc Reviewed-on: https://skia-review.googlesource.com/143710 Reviewed-by: Brian Osman <brianosman@google.com> Commit-Queue: Brian Osman <brianosman@google.com> Auto-Submit: Mike Klein <mtklein@google.com> [modify] https://crrev.com/23e7777f421d11992e8245771641141aa273556b/skcms.cc
,
Jul 27
This is the same as issue 862908 , and the above fix is going to roll into chromium ToT soon. For this to manifest, it requires: 1) SW rasterization (that's why we're only seeing it on Linux). 2) A color profile installed with a gamut that's smaller than sRGB. This is very rare. Most users have sRGB profiles, and most users that don't have sRGB have larger gamut profiles (like AdobeRGB). Given this, there is a workaround, which happens to be the same workaround we're currently advising for issue 847024 : Go to chrome://flags#force-color-profile, and set it to sRGB. Given these factors, I'm not sure if we should plan to cherry-pick this into m68?
,
Jul 27
,
Jul 30
,
Jul 30
The following revision refers to this bug: https://skia.googlesource.com/skcms/+/6b7ef53d3298a7ab7e3410ba3ffe5623539f025e commit 6b7ef53d3298a7ab7e3410ba3ffe5623539f025e Author: Mike Klein <mtklein@google.com> Date: Mon Jul 30 13:04:23 2018 add unit test for clamp-before-premul Bug: chromium:867813 Bug: chromium:862908 Change-Id: I79a82b6637c8a707844642b9bce4444bb85828b5 Reviewed-on: https://skia-review.googlesource.com/144295 Auto-Submit: Mike Klein <mtklein@google.com> Commit-Queue: Brian Osman <brianosman@google.com> Reviewed-by: Brian Osman <brianosman@google.com> [modify] https://crrev.com/6b7ef53d3298a7ab7e3410ba3ffe5623539f025e/skcms.cc [modify] https://crrev.com/6b7ef53d3298a7ab7e3410ba3ffe5623539f025e/tests.c
,
Jul 31
Issue 862908 has been merged into this issue.
,
Jul 31
https://skia-review.googlesource.com/c/skia/+/143713 should port this back to M69 if needed.
,
Aug 1
Your change meets the bar and is auto-approved for M69. Please go ahead and merge the CL to branch 3497 manually. Please contact milestone owner if you have questions. Owners: amineer@(Android), kariahda@(iOS), cindyb@(ChromeOS), govind@(Desktop) For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot
,
Aug 1
The following revision refers to this bug: https://skia.googlesource.com/skia/+/f4bd5f148c150c7ffe73bdf1aa6f9d565c0c2125 commit f4bd5f148c150c7ffe73bdf1aa6f9d565c0c2125 Author: skcms-skia-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com <skcms-skia-autoroll@skia-buildbots.google.com.iam.gserviceaccount.com> Date: Wed Aug 01 22:18:58 2018 backport skia-review.googlesource.com/143710 to M69 branch This is https://skia-review.googlesource.com/c/skcms/+/143710 exactly backported to the M69 branch. Bug: chromium:867813 Change-Id: Ia7ad00383e86429b3af7098137f2c1b7453eb6ef Reviewed-on: https://skia-review.googlesource.com/143713 Commit-Queue: Mike Klein <mtklein@google.com> Reviewed-by: Brian Osman <brianosman@google.com> [modify] https://crrev.com/f4bd5f148c150c7ffe73bdf1aa6f9d565c0c2125/third_party/skcms/skcms.cc
,
Aug 1
Seems like it is already merged to M69 at #13. Is there anything pending for M69? If not, pls remove "Merge-Approved-69" label. Thank you.
,
Aug 1
,
Aug 2
,
Aug 3
,
Aug 5
Windows 10 additional info: I am able to see this bug in one Windows account, on one screen only, but not in another Windows account. There may be some video driver/settings involved with this. I will test it more, then try the version from Download Chrome Beta, and let you know if anything changes.
,
Aug 5
The bug appeared in my other account as soon as I wrote the previous comment. Maybe that was just a cache issue. Found no improvement in 69.0.3497.23. The current beta is still bugged.
,
Aug 5
I figured out how to install the dev channel Chrome. Version 70.0.3510.2 is fixed for me in Windows 10.
,
Aug 6
,
Aug 6
Issue 870598 has been merged into this issue.
,
Aug 6
brianosman@ / mtklein@ - do you know how widespread is this issue? How many users or types of machines this is impacting?
,
Aug 6
We think it's not that widespread. It requires: - Software rasterization (so Linux, or a blacklisted GPU) - A color profile that doesn't fully enclose sRGB. Most profiles are sRGB, and most non-sRGB profiles are strictly larger than sRGB. Given those conditions, any image with partial transparency can potentially exhibit the bug. That said, we also think the fix is very safe if we want to cherry pick to 68.
,
Aug 6
Great thanks - which CLs will be required? Would it just be the CL in #13?
,
Aug 6
It'd be https://skia-review.googlesource.com/c/skia/+/143712, the same change as #13 taking into account some code moving around between M68 and M69.
,
Aug 6
Great thanks! Is this impacting Linux and Windows both?
,
Aug 6
YES
,
Aug 6
Yes, it's not at all platform specific.
,
Aug 6
How safe is this fix overall? How safe is the merge to M68? Seems like this has been in Canary/Beta now for a while.
,
Aug 6
It's very safe - I'd feel fine about merging to M68. And to clarify the platform issue: The code is not platform specific at all, but Windows and Mac have GPU rasterization enabled by default, unless the GPU/drivers are disabled for a specific reason (some bug that we can't work around, etc...). The bug is only in the CPU code, so users on Linux (where GPU is off by default) are more likely to see it, but there are plenty of Windows users that can be affected *if* they don't have GPU enabled.
,
Aug 6
,
Aug 6
Approving merge for M68. Branch:3440
,
Aug 6
The following revision refers to this bug: https://skia.googlesource.com/skia/+/1e419fead7ff3b5247102b29b982b6472a60cd4d commit 1e419fead7ff3b5247102b29b982b6472a60cd4d Author: Mike Klein <mtklein@google.com> Date: Mon Aug 06 23:38:59 2018 backport skia-review.googlesource.com/143710 to M68 branch This is https://skia-review.googlesource.com/c/skcms/+/143710 qualitatively backported to the M68 branch. Bug: chromium:867813 Change-Id: Id313cd9cd290c55dfe154f33b6031b21c38a370f Reviewed-on: https://skia-review.googlesource.com/143712 Reviewed-by: Brian Osman <brianosman@google.com> [modify] https://crrev.com/1e419fead7ff3b5247102b29b982b6472a60cd4d/third_party/skcms/src/Transform.c
,
Aug 7
Issue 871057 has been merged into this issue.
,
Aug 7
Issue 871621 has been merged into this issue.
,
Aug 7
,
Aug 7
M69 Stable promotion is coming VERY soon. Your bug is labelled as Stable ReleaseBlock, pls make sure to land the fix and request a merge into the release branch ASAP. Thank you.
,
Aug 8
This should be fixed in all (?) branches now. #5 was the ToT fix, which will be in m70. #13 merged the fix to m69, and #35 merged to m68. Do we just need to update labels?
,
Aug 8
Pls mark bug as fixed if nothing else is pending. Thank you.
,
Aug 8
,
Aug 15
,
Aug 17
Issue 874842 has been merged into this issue.
,
Oct 17
This issue, or a very similar one, is happening on Chrome 70, you can see it live here in the menu icons at the left of the form: https://www.trabber.us/ I include an attachment in which you can see the green border which is not in the original image. Chrome Version 70.0.3538.67 (Official Build) (64-bit) on Ubuntu 18.04.
,
Oct 17
ofrias@, Thank you for the information. I've tried to reproduce this using the color profile that another user submitted (and previously caused this issue), but without any luck. Can you check to see if you have a color profile installed for your monitor? There are instructions on locating the file under the section 'For Linux Users' in this document: https://docs.google.com/document/d/1jMokB_OBkZVELu22li8vnHxAUoL1eGnLedP-1Gttv40/view#heading=h.2to15xum98j3 You may be able to find the color profile in one of the following locations: /usr/share/color/icc/ /usr/local/share/color/icc ~/.local/share/icc/ One of those directories is likely to contain several files with a .icc extension. If you can find such a file (it may be named similar to your monitor's brand, or with something like 'edid-<long string of letters and numbers>.icc'. Getting the file used by your monitor will make tracking this down much easier. Regardless, with the profile enabled (so the bug repros), can you navigate to chrome://gpu, and copy/paste the section under 'Display(s) Information'. That has some of the color profile information, and may be helpful.
,
Oct 17
pt- Brasil Sim, bug na versão 70.0.3538.67. Eu uso o Linux 16.04, foto em anexo. Solucionei com https://support.google.com/chrome/forum/AAAAP1KN0B0yD9L8W9zF4g/?hl=pt-BR . Será que isso vai ser corrigido sem ter que mudar essa configuração? ------------------------ Inglês - Google a translator Yes, bug in version 70.0.3538.67. I use Linux 16.04, attached photo. I've solved it with https://support.google.com/chrome/forum/AAAAP1KN0B0yD9L8W9zF4g/?hl=en. Will this be fixed without having to change this setting?
,
Oct 18
Thanks for your help. As you suggested, I have found the profile that my display (Asus BE24A) is using, called "BE24A", under .local/share/icc/ I attach it. More info: If I directly open the PNG in Chrome, not inside the page, the bug is not present, although in this case the background is black instead of white, and maybe this is the reason: https://www.trabber.es/images/home/icoshome.png If I go to chrome://flags#force-color-profile, and set it to sRGB, the bug dissapears.
,
Oct 18
Oops, I forgot to copy chrome://gpu contents: Graphics Feature Status Canvas: Hardware accelerated Flash: Hardware accelerated Flash Stage3D: Hardware accelerated Flash Stage3D Baseline profile: Hardware accelerated Compositing: Hardware accelerated Multiple Raster Threads: Enabled Native GpuMemoryBuffers: Software only. Hardware acceleration disabled Out-of-process Rasterization: Disabled Hardware Protected Video Decode: Unavailable Rasterization: Software only. Hardware acceleration disabled Skia Deferred Display List: Disabled Skia Renderer: Disabled Surface Synchronization: Enabled Video Decode: Unavailable Viz Service Display Compositor: Disabled WebGL: Hardware accelerated WebGL2: Hardware accelerated Driver Bug Workarounds adjust_src_dst_region_for_blitframebuffer clear_uniforms_before_first_program_use count_all_in_varyings_packing decode_encode_srgb_for_generatemipmap disable_framebuffer_cmaa disable_post_sub_buffers_for_onscreen_surfaces msaa_is_slow scalarize_vec_and_mat_constructor_args disabled_extension_GL_ARB_timer_query disabled_extension_GL_EXT_timer_query disabled_extension_GL_KHR_blend_equation_advanced disabled_extension_GL_KHR_blend_equation_advanced_coherent Problems Detected Accelerated video decode is unavailable on Linux: 137247 Disabled Features: accelerated_video_decode Protected video decoding with swap chain is for Windows and Intel only Disabled Features: protected_video_decode Clear uniforms before first program use on all platforms: 124764, 349137 Applied Workarounds: clear_uniforms_before_first_program_use Mesa drivers in Linux handle varyings without static use incorrectly: 333885 Applied Workarounds: count_all_in_varyings_packing Disable partial swaps on Mesa drivers (detected with GL_RENDERER): 339493 Applied Workarounds: disable_post_sub_buffers_for_onscreen_surfaces Always rewrite vec/mat constructors to be consistent: 398694 Applied Workarounds: scalarize_vec_and_mat_constructor_args On Intel GPUs MSAA performance is not acceptable for GPU rasterization: 527565 Applied Workarounds: msaa_is_slow Timer queries crash on Intel GPUs on Linux: 540543, 576991 Applied Workarounds: disable(GL_ARB_timer_query), disable(GL_EXT_timer_query) Use GL_INTEL_framebuffer_CMAA on ChromeOS: 535198 Applied Workarounds: disable_framebuffer_cmaa Disable partial swaps on Mesa drivers (detected with GL_VERSION): 339493 Applied Workarounds: disable_post_sub_buffers_for_onscreen_surfaces Decode and encode before generateMipmap for srgb format textures on os except macosx: 634519 Applied Workarounds: decode_encode_srgb_for_generatemipmap adjust src/dst region if blitting pixels outside read framebuffer on Linux Intel: 664740 Applied Workarounds: adjust_src_dst_region_for_blitframebuffer Disable KHR_blend_equation_advanced until cc shaders are updated: 661715 Applied Workarounds: disable(GL_KHR_blend_equation_advanced), disable(GL_KHR_blend_equation_advanced_coherent) Native GpuMemoryBuffers have been disabled, either via about:flags or command line. Disabled Features: native_gpu_memory_buffers Viz service display compositor is not enabled by default. Disabled Features: viz_display_compositor Skia renderer is not used by default. Disabled Features: skia_renderer Skia deferred display list is not used by default. Disabled Features: skia_deferred_display_list Version Information Data exported 2018-10-18T09:40:09.359Z Chrome version Chrome/70.0.3538.67 Operating system Linux 4.15.0-36-generic Software rendering list URL https://chromium.googlesource.com/chromium/src/+/9ab0cfab84ded083718d3a4ff830726efd38869f/gpu/config/software_rendering_list.json Driver bug list URL https://chromium.googlesource.com/chromium/src/+/9ab0cfab84ded083718d3a4ff830726efd38869f/gpu/config/gpu_driver_bug_list.json ANGLE commit id 05c729f336ef 2D graphics backend Skia/70 193565807f7fd2a2c14e21b44c545cc03ebe134f- Command Line /usr/bin/google-chrome-stable --flag-switches-begin --flag-switches-end Driver Information Initialization time 24 In-process GPU false Passthrough Command Decoder false Sandboxed true GPU0 VENDOR = 0x8086 [Intel Open Source Technology Center], DEVICE= 0x3e92 [Mesa DRI Intel(R) HD Graphics (Coffeelake 3x8 GT2) ] *ACTIVE* Optimus false AMD switchable false Driver vendor Mesa Driver version 18.0.5 Driver date GPU CUDA compute capability major version 0 Pixel shader version 1.30 Vertex shader version 1.30 Max. MSAA samples 16 Machine model name Machine model version GL_VENDOR Intel Open Source Technology Center GL_RENDERER Mesa DRI Intel(R) HD Graphics (Coffeelake 3x8 GT2) GL_VERSION 3.0 Mesa 18.0.5 GL_EXTENSIONS GL_3DFX_texture_compression_FXT1 GL_AMD_conservative_depth GL_AMD_draw_buffers_blend GL_AMD_seamless_cubemap_per_texture GL_AMD_shader_stencil_export GL_AMD_shader_trinary_minmax GL_ANGLE_texture_compression_dxt3 GL_ANGLE_texture_compression_dxt5 GL_APPLE_object_purgeable GL_APPLE_packed_pixels GL_ARB_ES2_compatibility GL_ARB_ES3_compatibility GL_ARB_arrays_of_arrays GL_ARB_blend_func_extended GL_ARB_buffer_storage GL_ARB_clear_buffer_object GL_ARB_clear_texture GL_ARB_clip_control GL_ARB_color_buffer_float GL_ARB_compressed_texture_pixel_storage GL_ARB_compute_shader GL_ARB_conditional_render_inverted GL_ARB_conservative_depth GL_ARB_copy_buffer GL_ARB_copy_image GL_ARB_cull_distance GL_ARB_debug_output GL_ARB_depth_buffer_float GL_ARB_depth_clamp GL_ARB_depth_texture GL_ARB_derivative_control GL_ARB_draw_buffers GL_ARB_draw_buffers_blend GL_ARB_draw_elements_base_vertex GL_ARB_draw_instanced GL_ARB_explicit_attrib_location GL_ARB_explicit_uniform_location GL_ARB_fragment_coord_conventions GL_ARB_fragment_program GL_ARB_fragment_program_shadow GL_ARB_fragment_shader GL_ARB_framebuffer_no_attachments GL_ARB_framebuffer_object GL_ARB_framebuffer_sRGB GL_ARB_get_program_binary GL_ARB_get_texture_sub_image GL_ARB_half_float_pixel GL_ARB_half_float_vertex GL_ARB_instanced_arrays GL_ARB_internalformat_query GL_ARB_internalformat_query2 GL_ARB_invalidate_subdata GL_ARB_map_buffer_alignment GL_ARB_map_buffer_range GL_ARB_multi_bind GL_ARB_multisample GL_ARB_multitexture GL_ARB_occlusion_query GL_ARB_occlusion_query2 GL_ARB_pipeline_statistics_query GL_ARB_pixel_buffer_object GL_ARB_point_parameters GL_ARB_point_sprite GL_ARB_polygon_offset_clamp GL_ARB_program_interface_query GL_ARB_provoking_vertex GL_ARB_query_buffer_object GL_ARB_robust_buffer_access_behavior GL_ARB_robustness GL_ARB_sample_shading GL_ARB_sampler_objects GL_ARB_seamless_cube_map GL_ARB_seamless_cubemap_per_texture GL_ARB_separate_shader_objects GL_ARB_shader_atomic_counter_ops GL_ARB_shader_atomic_counters GL_ARB_shader_ballot GL_ARB_shader_bit_encoding GL_ARB_shader_clock GL_ARB_shader_draw_parameters GL_ARB_shader_group_vote GL_ARB_shader_image_load_store GL_ARB_shader_image_size GL_ARB_shader_objects GL_ARB_shader_precision GL_ARB_shader_stencil_export GL_ARB_shader_storage_buffer_object GL_ARB_shader_texture_image_samples GL_ARB_shader_texture_lod GL_ARB_shading_language_100 GL_ARB_shading_language_420pack GL_ARB_shading_language_packing GL_ARB_shadow GL_ARB_stencil_texturing GL_ARB_sync GL_ARB_texture_barrier GL_ARB_texture_border_clamp GL_ARB_texture_compression GL_ARB_texture_compression_bptc GL_ARB_texture_compression_rgtc GL_ARB_texture_cube_map GL_ARB_texture_cube_map_array GL_ARB_texture_env_add GL_ARB_texture_env_combine GL_ARB_texture_env_crossbar GL_ARB_texture_env_dot3 GL_ARB_texture_filter_anisotropic GL_ARB_texture_float GL_ARB_texture_gather GL_ARB_texture_mirror_clamp_to_edge GL_ARB_texture_mirrored_repeat GL_ARB_texture_multisample GL_ARB_texture_non_power_of_two GL_ARB_texture_query_levels GL_ARB_texture_query_lod GL_ARB_texture_rectangle GL_ARB_texture_rg GL_ARB_texture_rgb10_a2ui GL_ARB_texture_stencil8 GL_ARB_texture_storage GL_ARB_texture_storage_multisample GL_ARB_texture_swizzle GL_ARB_texture_view GL_ARB_timer_query GL_ARB_transform_feedback2 GL_ARB_transform_feedback3 GL_ARB_transform_feedback_instanced GL_ARB_transform_feedback_overflow_query GL_ARB_transpose_matrix GL_ARB_uniform_buffer_object GL_ARB_vertex_array_bgra GL_ARB_vertex_array_object GL_ARB_vertex_attrib_binding GL_ARB_vertex_buffer_object GL_ARB_vertex_program GL_ARB_vertex_shader GL_ARB_vertex_type_10f_11f_11f_rev GL_ARB_vertex_type_2_10_10_10_rev GL_ARB_window_pos GL_ATI_blend_equation_separate GL_ATI_draw_buffers GL_ATI_separate_stencil GL_ATI_texture_env_combine3 GL_ATI_texture_float GL_EXT_abgr GL_EXT_bgra GL_EXT_blend_color GL_EXT_blend_equation_separate GL_EXT_blend_func_separate GL_EXT_blend_minmax GL_EXT_blend_subtract GL_EXT_compiled_vertex_array GL_EXT_copy_texture GL_EXT_draw_buffers2 GL_EXT_draw_instanced GL_EXT_draw_range_elements GL_EXT_fog_coord GL_EXT_framebuffer_blit GL_EXT_framebuffer_multisample GL_EXT_framebuffer_multisample_blit_scaled GL_EXT_framebuffer_object GL_EXT_framebuffer_sRGB GL_EXT_gpu_program_parameters GL_EXT_multi_draw_arrays GL_EXT_packed_depth_stencil GL_EXT_packed_float GL_EXT_packed_pixels GL_EXT_pixel_buffer_object GL_EXT_point_parameters GL_EXT_polygon_offset GL_EXT_polygon_offset_clamp GL_EXT_provoking_vertex GL_EXT_rescale_normal GL_EXT_secondary_color GL_EXT_separate_specular_color GL_EXT_shader_integer_mix GL_EXT_shader_samples_identical GL_EXT_shadow_funcs GL_EXT_stencil_two_side GL_EXT_stencil_wrap GL_EXT_subtexture GL_EXT_texture GL_EXT_texture3D GL_EXT_texture_array GL_EXT_texture_compression_dxt1 GL_EXT_texture_compression_rgtc GL_EXT_texture_compression_s3tc GL_EXT_texture_cube_map GL_EXT_texture_edge_clamp GL_EXT_texture_env_add GL_EXT_texture_env_combine GL_EXT_texture_env_dot3 GL_EXT_texture_filter_anisotropic GL_EXT_texture_integer GL_EXT_texture_lod_bias GL_EXT_texture_object GL_EXT_texture_rectangle GL_EXT_texture_sRGB GL_EXT_texture_sRGB_decode GL_EXT_texture_shared_exponent GL_EXT_texture_snorm GL_EXT_texture_swizzle GL_EXT_timer_query GL_EXT_transform_feedback GL_EXT_vertex_array GL_EXT_vertex_array_bgra GL_IBM_multimode_draw_arrays GL_IBM_rasterpos_clip GL_IBM_texture_mirrored_repeat GL_INGR_blend_func_separate GL_INTEL_performance_query GL_KHR_blend_equation_advanced GL_KHR_blend_equation_advanced_coherent GL_KHR_context_flush_control GL_KHR_debug GL_KHR_no_error GL_KHR_robust_buffer_access_behavior GL_KHR_robustness GL_KHR_texture_compression_astc_ldr GL_KHR_texture_compression_astc_sliced_3d GL_MESA_pack_invert GL_MESA_shader_integer_functions GL_MESA_texture_signed_rgba GL_MESA_window_pos GL_NV_blend_square GL_NV_conditional_render GL_NV_depth_clamp GL_NV_light_max_exponent GL_NV_packed_depth_stencil GL_NV_primitive_restart GL_NV_texgen_reflection GL_NV_texture_barrier GL_NV_texture_env_combine4 GL_NV_texture_rectangle GL_OES_EGL_image GL_OES_read_format GL_S3_s3tc GL_SGIS_generate_mipmap GL_SGIS_texture_border_clamp GL_SGIS_texture_edge_clamp GL_SGIS_texture_lod GL_SUN_multi_draw_arrays Disabled Extensions GL_ARB_timer_query GL_EXT_timer_query GL_KHR_blend_equation_advanced GL_KHR_blend_equation_advanced_coherent Disabled WebGL Extensions Window system binding vendor SGI Window system binding version 1.4 Window system binding extensions GLX_ARB_create_context GLX_ARB_create_context_profile GLX_ARB_create_context_robustness GLX_ARB_fbconfig_float GLX_ARB_framebuffer_sRGB GLX_ARB_multisample GLX_EXT_create_context_es_profile GLX_EXT_create_context_es2_profile GLX_EXT_fbconfig_packed_float GLX_EXT_framebuffer_sRGB GLX_EXT_import_context GLX_EXT_libglvnd GLX_EXT_texture_from_pixmap GLX_EXT_visual_info GLX_EXT_visual_rating GLX_MESA_copy_sub_buffer GLX_OML_swap_method GLX_SGI_make_current_read GLX_SGI_swap_control GLX_SGIS_multisample GLX_SGIX_fbconfig GLX_SGIX_pbuffer GLX_SGIX_visual_select_group GLX_INTEL_swap_event Window manager GNOME Shell XDG_CURRENT_DESKTOP ubuntu:GNOME GDMSESSION ubuntu Compositing manager Yes Direct rendering Yes Reset notification strategy 0x8261 GPU process crash count 0 System visual ID 33 RGBA visual ID 308 Compositor Information Tile Update Mode One-copy Partial Raster Enabled GpuMemoryBuffers Status ATC Software only ATCIA Software only DXT1 Software only DXT5 Software only ETC1 Software only R_8 Software only R_16 Software only RG_88 Software only BGR_565 Software only RGBA_4444 Software only RGBX_8888 Software only RGBA_8888 Software only BGRX_8888 Software only BGRX_1010102 Software only RGBX_1010102 Software only BGRA_8888 Software only RGBA_F16 Software only YVU_420 Software only YUV_420_BIPLANAR Software only UYVY_422 Software only Display(s) Information Info Display[1241976709638723] bounds=[0,0 1920x1200], workarea=[0,0 1920x1152], scale=1, external. Color space information {primaries_d50_referred: [[0.6511, 0.3357], [0.3456, 0.6050], [0.1495, 0.0590]], transfer:0.0000*x + 0.0000 if x < 0.0000 else (1.0000*x + 0.0000)**2.2000 + 0.0000, matrix:RGB, range:FULL} Bits per color component 8 Bits per pixel 24 Video Acceleration Information
,
Oct 23
I can reproduce this on Mac (Version 70.0.3538.67) by forcing software rasterization, in slightly different ways with both the new BE24A and older LV1911 profiles. I'll see if I can figure out what's up.
,
Oct 23
The following revision refers to this bug: https://skia.googlesource.com/skia/+/fd731e46846c0af350f3cefbfd8c8fecf7ee29ae commit fd731e46846c0af350f3cefbfd8c8fecf7ee29ae Author: Mike Klein <mtklein@google.com> Date: Tue Oct 23 17:57:18 2018 add a few more ways to convert images to imgcvt Interestingly, the draw ends up a little different than the writePixels() and skcms paths. Maybe just premul/unpremul? Bug: chromium:867813 Change-Id: I8bf645c8354cbeb9a07512add258e8defc755366 Reviewed-on: https://skia-review.googlesource.com/c/164621 Commit-Queue: Mike Klein <mtklein@google.com> Commit-Queue: Brian Osman <brianosman@google.com> Auto-Submit: Mike Klein <mtklein@google.com> Reviewed-by: Brian Osman <brianosman@google.com> [modify] https://crrev.com/fd731e46846c0af350f3cefbfd8c8fecf7ee29ae/tools/imgcvt.cpp
,
Oct 23
Good news is, it looks like this is fixed... I'm not able to reproduce this with a build at head on m72. Going to start bisecting to see what fixed this and if that's able to be backported to 70 and/or 71.
,
Oct 23
M71 also looks okay.
,
Oct 23
A bisect found https://skia-review.googlesource.com/c/skia/+/152200, which is kind of obvious in retrospect. I've created a cherry-pick for M70 here: https://skia-review.googlesource.com/c/skia/+/164624. That CL isn't the only change we've done to add this clamping, but it does cover almost all the practical impact of it, and it should patch to M70 cleanly.
,
Oct 23
Might want to merge this to M70? The bug's cosmetic and fixed in M71, but the fix should be safe. It's a clean cherry-pick back of the same patch that happened to fix this bug for M71+.
,
Oct 23
This bug requires manual review: Request affecting a post-stable build Please contact the milestone owner if you have questions. Owners: benmason@(Android), kariahda@(iOS), geohsu@(ChromeOS), abdulsyed@(Desktop) For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot
,
Oct 23
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/96c1165edca01bacf878979d1e6824a435da3fd7 commit 96c1165edca01bacf878979d1e6824a435da3fd7 Author: chromium-autoroll <chromium-autoroll@skia-public.iam.gserviceaccount.com> Date: Tue Oct 23 21:11:29 2018 Roll src/third_party/skia c88cc779efdb..07055a3ee7ce (6 commits) https://skia.googlesource.com/skia.git/+log/c88cc779efdb..07055a3ee7ce git log c88cc779efdb..07055a3ee7ce --date=short --no-merges --format='%ad %ae %s' 2018-10-23 reed@google.com update dox for vertical-text demise 2018-10-23 jvanverth@google.com Check for valid initialization of ReflexHash 2018-10-23 mtklein@google.com add a few more ways to convert images to imgcvt 2018-10-23 brianosman@google.com Hoist color space xform creation out of GrTextureProducer 2018-10-23 jvanverth@google.com Check for finite segments when inserting into ActiveEdgeList 2018-10-23 pirama@google.com Add back -Wover-align Created with: gclient setdep -r src/third_party/skia@07055a3ee7ce The AutoRoll server is located here: https://autoroll.skia.org/r/skia-autoroll Documentation for the AutoRoller is here: https://skia.googlesource.com/buildbot/+/master/autoroll/README.md If the roll is causing failures, please contact the current sheriff, who should be CC'd on the roll, and stop the roller if necessary. CQ_INCLUDE_TRYBOTS=luci.chromium.try:android_optional_gpu_tests_rel;luci.chromium.try:linux-chromeos-compile-dbg;luci.chromium.try:linux_optional_gpu_tests_rel;luci.chromium.try:mac_optional_gpu_tests_rel;luci.chromium.try:win_optional_gpu_tests_rel;master.tryserver.blink:linux_trusty_blink_rel BUG= chromium:867813 TBR=brianosman@chromium.org Change-Id: Ic019cb3732c7d0131e372f8bb90dcead78146935 Reviewed-on: https://chromium-review.googlesource.com/c/1297014 Reviewed-by: chromium-autoroll <chromium-autoroll@skia-public.iam.gserviceaccount.com> Commit-Queue: Mike Klein <mtklein@chromium.org> Cr-Commit-Position: refs/heads/master@{#602095} [modify] https://crrev.com/96c1165edca01bacf878979d1e6824a435da3fd7/DEPS
,
Oct 23
,
Oct 29
M71 Stable promotion is coming VERY soon. Your bug is labelled as Stable ReleaseBlock, pls make sure to land the fix and request a merge into the release branch ASAP. Thank you.
,
Oct 29
M71 Stable promotion is coming VERY soon. Your bug is labelled as Stable ReleaseBlock, pls make sure to land the fix and request a merge into the release branch ASAP. Thank you.
,
Oct 30
Based on offline chat with Mike, the bug was fixed before we knew about this issue in M71. So there isn't any merges required.
,
Oct 31
,
Nov 1
,
Nov 1
,
Nov 1
Since M71 has the fix naturally and we've not targeted this for M70, I think there's nothing left to do here except wait.
,
Nov 5
Issue 901315 has been merged into this issue.
,
Nov 5
Issue 901755 has been merged into this issue.
,
Nov 5
mtklein@ - we are planning a respin for M70. How safe will this merge be? Is this only affecting Linux?
,
Nov 5
Ought to be quite safe. The CL that happened to fix this in M71 patches cleanly back on M70, and it's just a small logic change to run some already well tested code. Linux is the platform that's mostly affected, but really this affects any software rendering. That can happen with blacklisted GPU drivers, old machines, etc. https://skia-review.googlesource.com/c/skia/+/164624
,
Nov 5
Approving merge for M70. Branch:3538
,
Nov 5
The following revision refers to this bug: https://skia.googlesource.com/skia/+/8adc53ec0ee4bbafbd9925024df38437c0fcf88a commit 8adc53ec0ee4bbafbd9925024df38437c0fcf88a Author: Mike Klein <mtklein@google.com> Date: Mon Nov 05 17:00:54 2018 clamp after color pipeline when dst is normalized Most code working with unorm values doesn't expect to see out-of-gamut colors, but unorms _can_ store some out-of-gamut colors when premultiplied and alpha is <1. Clamping after the color (= shader + color filter) pipeline prevents us from creating those values. This CL does appear to fix the problem we were seeing with xfermodes2 in the narrow config. I've kept scalepixels_unpremul working by using our old trick, remembering that unpremul->unpremul involves the same steps as premul->opaque. We'll take the unpremul pixels, pun them to premul to avoid data loss, scale them up, and then avoid unpremul on the way out by punning the surface as opaque. It all kind of just barely hangs together. Bug: chromium:867813 Change-Id: I677d00ecfe4692ba7799ae963d8b8d1ccd4b00ac Reviewed-on: https://skia-review.googlesource.com/152200 Commit-Queue: Mike Klein <mtklein@google.com> Reviewed-by: Brian Osman <brianosman@google.com> Auto-Submit: Mike Klein <mtklein@google.com> Reviewed-on: https://skia-review.googlesource.com/c/164624 [modify] https://crrev.com/8adc53ec0ee4bbafbd9925024df38437c0fcf88a/src/core/SkRasterPipelineBlitter.cpp [modify] https://crrev.com/8adc53ec0ee4bbafbd9925024df38437c0fcf88a/src/core/SkPixmap.cpp
,
Nov 7
,
Nov 13
,
Nov 28
Issue 908816 has been merged into this issue. |
||||||||||||||||||||||||||||||||
►
Sign in to add a comment |
||||||||||||||||||||||||||||||||
Comment 1 by tudorv.p...@gmail.com
, Jul 26765 bytes
765 bytes View Download