Skip to content

A
angle
Switch branch/tag
  1. 20 Nov, 2019 9 commits
  3. 19 Nov, 2019 12 commits
  5. 18 Nov, 2019 6 commits
  7. 15 Nov, 2019 10 commits
  9. 14 Nov, 2019 3 commits
    • Manually copy dEQP data files to output directory · 11c487a8
      Tim Van Patten authored 
      dEQP has data files that live in several directories within dEQP's
source tree. For example, GLES3 has data files that live within:

data/gles3/data/
external/graphicsfuzz/data/gles3/

However, we can only tell dEQP about a single data directory during
initialization of dEQP. To get around this, we are manually copying all
of the necessary data files to the output generated files directory and
pointing dEQP to this single directory. This also helps us solve a
second problem related to the paths that dEQP uses when accessing
graphicsfuzz data files.

For the graphicsfuzz tests, dEQP will attempt to open the necessary
shaders by accessing them with the path:

data/gles3/graphicsfuzz/

However, those files would normally live at the path that matches their
location within the source tree:

external/graphicsfuzz/data/gles3/graphicsfuzz/

As part of the manual copy of these data files, we are also able to
strip the extra 'external/graphicsfuzz/' portion of the path.

Bug: angleproject:2322
Test: dEQP
Change-Id: Ibc96442c221485e2f246890fa8fe51f090c5e222
Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/1881759
Commit-Queue: Tim Van Patten <timvp@google.com>
Reviewed-by: 's avatarTim Van Patten <timvp@google.com>
    • Refactor end2end test macros · 1a01b4b3
      Tobin Ehlis authored 
      This is a foundational CL to enabling the end2end tests on swiftshader.

Refactored infrastructure with new ANGLE_INSTANTIATE_TEST_ES*
macros that will run tests over all various combinations of all
platforms for different ES versions.

Just skipping failing tests initially to get the refactor landed.

Bug: angleproject:4081
Bug: angleproject:4092
Change-Id: I017f6c3267179e49b6ae08cc7488096b423dcdb5
Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/1904635
Commit-Queue: Tobin Ehlis <tobine@google.com>
Reviewed-by: 's avatarCourtney Goeltzenleuchter <courtneygo@google.com>
    • Treat VK_TIMEOUT as an error · 968df09e
      Tim Van Patten authored 
      The fence wait time was increased to 120 seconds, so we don't expect
VK_TIMEOUT to be returned by a healty GPU. This change makes it so that
when VK_TIMEOUT is returned, we will treat this as an error and
propagate it to the GLES application as a GL_INVALID_OPERATION.

It is expected that the GL application will re-initialize ANGLE which
will also reinitialize the Vulkan device (and hopefully clean up the
error). This is not our final design for ANGLE's VK_TIMEOUT handling,
since we expect to expand our Vulkan device lost error handling when
implementing the GLES 3.2 robustness requirements. This will likely
improve ANGLE's VK_TIMEOUT handling as well.

Bug: angleproject:4043
Test: Manually force VK_TIMEOUT (lower kMaxFenceWaitTimeNs)
Change-Id: I2449ad214ada176014a1702a9b3417d4498d070f
Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/1910722
Commit-Queue: Tim Van Patten <timvp@google.com>
Reviewed-by: 's avatarShahbaz Youssefi <syoussefi@chromium.org>
Reviewed-by: 's avatarCody Northrop <cnorthrop@google.com>