-
Notifications
You must be signed in to change notification settings - Fork 5.3k
Comparing changes
Open a pull request
base repository: dotnet/runtime
base: cf20e76
head repository: dotnet/runtime
compare: 1e96637
- 6 commits
- 28 files changed
- 6 contributors
Commits on Aug 31, 2021
-
Add s390x support in a few places in eng/native build scripts (#58366)
* Make isMSBuildOnNETCoreSupported return true on s390x (fixes generating version headers) * Add s390x platform-specific library name in find_unwind_libs (enables linking against system-provided libunwind) Co-authored-by: Ulrich Weigand <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for 5a17bd8 - Browse repository at this point
Copy the full SHA 5a17bd8View commit details -
[release/6.0] [Libraries][Android] Fix alignment for padding System.D…
…ecimal (#58380) Fixes #49872 There is a difference in the current alignment of a decimal type on Android x86 (32bit) MONO_ABI_ALIGNOF (gpointer) which evaluates to 4 bytes and whats expected, which is 8 bytes. The behavior of MONO_ABI_ALIGNOF (gpointer) seems correct as gpointer is void* and 32-bit aligned pointers on a 32-bit platform makes sense. Instead, it's suspected that using MONO_ABI_ALIGNOF (gpointer) as the alignment is not correct for Android x86 if we are expecting an align of 8 bytes. Attempting to remove the entire special casing for System Decimal in marshal.c didn't seem to work as noted in #49872 (comment) This PR will adhere to the alignment value of 4 for Android x86, remove the special casing for System.Decimal because CoreCLR had removed their special casing, and modifies the expected values in the tests. https://github.com/dotnet/corefx/pull/30690/files?file-filters%5B%5D=.cs#r199212623 makes me suspect these tests have yet to be modified to fully accomodate Linux platforms.
Configuration menu - View commit details
-
Copy full SHA for 114aa9a - Browse repository at this point
Copy the full SHA 114aa9aView commit details -
Use OutputRid in test projects (#58340)
Co-authored-by: Adeel <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for 9f6ad57 - Browse repository at this point
Copy the full SHA 9f6ad57View commit details -
[release/6.0] Enable win registry install location for all architectu…
…res (#57851) * Multi-arch support for win registry install locations * Change get_dotnet_self_registered_config_location signature * Link advapi32.lib on win-arm * PR Feedback * Indent * Update get_dotnet_self_registered_config_location() Co-authored-by: Mateo Torres Ruiz <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for 08b249b - Browse repository at this point
Copy the full SHA 08b249bView commit details -
[release/6.0] [wasm] Require workloads if using `@(NativeFileReferenc…
…e)` (#58290) * [wasm] Require workloads, if a project is using native references Currently, if the `wasm-tools` workload is not installed, and a project uses AOT, then the build fails with an error saying that the workload is needed. But if the project is using native references, but not AOT, then the build does not fail. Instead, the `@(NativeFileReference)` just gets ignored. Even though the wasm workload is needed to relink dotnet.wasm with the native libraries. Implementation: - `$(RunAOTCompilation)` is a property, so it can be checked, and wasm workload imports can be enabled. - But `@(NativeFileReference)` is an item, and that gets evaluated in the second phase, so we can't use that to affect the imports. - Instead, we emit a warning from a target run before Build, if the project has any native references, but the workload isn't enabled. - Users can explicitly enable the workload by setting `$(WasmBuildNative)==true`. * Bump sdk for workload testing to 6.0.100-rc.2.21425.12 * Fix path to workload.stamp file Co-authored-by: Ankit Jain <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for e8aab6a - Browse repository at this point
Copy the full SHA e8aab6aView commit details -
[release/6.0] Fix fgValueNumberArrIndexVal for wide reads (#58427)
* fix fgValueNumberArrIndexVal for wide reads * Update test * Fix regressions * Update valuenum.cpp Co-authored-by: EgorBo <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for 1e96637 - Browse repository at this point
Copy the full SHA 1e96637View commit details
This comparison is taking too long to generate.
Unfortunately it looks like we can’t render this comparison for you right now. It might be too big, or there might be something weird with your repository.
You can try running this command locally to see the comparison on your machine:
git diff cf20e76...1e96637