Avoid collisions in cs files generated by Grpc.Tools by Falco20019 · Pull Request #22869 · grpc/grpc

Falco20019 · 2020-05-05T09:45:19Z

Fixes #17672
Based on the idea from #22852

Falco20019 · 2020-05-05T09:50:43Z

@jtattermusch Your idea mostly worked great. I only did the following modifications:

Ensure that generated paths will be created by MSBuild
Generate correct path if GrpcOutputDir is empty
Use same path for protodep files and unified calls
Added documentation and tests

I tested it with my complex use case (files on different drive, relative paths not possible, ...) and it works just great.

jtattermusch · 2020-05-05T11:01:24Z

Thanks!
Adhoc artifact build http://sponge/f30437b2-8244-4046-9350-dadee6a10f37 (will post the link to Grpc.Tools nuget shortly).

Update: the nugets are here:
https://storage.googleapis.com/grpc-testing-kokoro-prod/test_result_public/prod/grpc/core/master/windows/grpc_build_packages/11880/20200505-052419/github/grpc/artifacts/csharp_nugets_windows_dotnetcli.zip

jtattermusch · 2020-05-05T11:03:03Z

CC @kkm000

jtattermusch · 2020-05-05T11:03:23Z

CC @JunTaoLuo as you contributed to Grpc.Tools in the past.

src/csharp/Grpc.Tools/build/_protobuf/Google.Protobuf.Tools.targets

src/csharp/Grpc.Tools/DepFileUtil.cs

Falco20019 · 2020-05-05T12:47:38Z

I put some more thoughts into the dependency files. If Protobuf_DepFilesPath would not point to the OutputDirectory, the hash folders would break. Therefore, I reverted it back to {dirhash}_{filename}.protodep as before. Since we have no guessing (ProtoCompilerOutputs) there, we also have no good option to piggy-back for folder creation. And this would also only have been optional as there are no conflicts anyway.

I also made the MakeDir call conditional by creating _Protobuf_ExpectedGenerated explicitly. This allows us to only generate folder for modified files and reduce some overhead. Usually, after the clean build, the folders should already exist anyway and on first call, this will contain ALL folders.

src/csharp/Grpc.Tools/ProtoCompile.cs

Falco20019 · 2020-05-06T13:29:27Z

@jtattermusch I redid most of the stuff and adjusted the logic. I thought a bit more about the problem, especially after seeing how the guessing for C++ works. I think this is a lot cleaner and better to maintain.

The new approach that I would like to discuss:

Added PatchedProtobuf to ProtoCompilerOutputs as way to manipulate OutputDir and GrpcOutputDir
- Added PatchOutputDirectory to GeneratorService to do the patching
- Will be used by C# to add the relative dir into the OutputDir
- GetPossibleOutputs is easier for C# since it only works in the dir without doing the hashing stuff
- This allows us to keep ProtoCompile close to protoc call and don't have special logic only for C# generator (MaybeEnhanceOutputDir would have introduced a bug for Generator = cpp)
- _Protobuf_CoreCompile does not need the cryptic adjustment (_Protobuf_ExpectedGenerated) for adding the generated paths since %(_Protobuf_OutOfDateProto.[Grpc]OutputDir) works again
Added Protobuf_ProtoRoot to define a ProtoRoot default if outside of the project
- This allows the relative paths to be generated a lot better
- It keeps the old behavior to generate everything directly in [Grpc]OutputDir if no ProtoRoot was given
- This also means, that if no ProtoRoot or Protobuf_ProtoRoot is given, it could still create duplicates, but they are very easy and generic fixable: <Protobuf Include="$(LIProtoPath)/**/*.proto" ProtoRoot="$(LIProtoPath)" /> which is more natural than setting OutputDir = "$(Protobuf_OutputPath)%(RelativeDir)"
- This works for local, relative and absolute paths (including other drives)
I kept the changes to DepFileUtil (separation of GetDepFilenameForProto into GetOutputDirWithHash and GetDirectoryHash) as it was basically a refactoring with added documentation
- If you would like to keep it in one method for performance, I will just revert it back to master

I also added a fix for C++ as this was broken for protos outside of the project (ProtoRoot was not available in Protobuf_PrepareClean, therefore the relative dirs were calculated wrong, leading to nothing being cleared). This allowed us also to use this for C#:

Extracted target _Protobuf_SetProtoRoot from _Protobuf_SelectFiles
- _Protobuf_SetProtoRoot added as dependency to Protobuf_Clean to fix the bug in Protobuf_PrepareClean

If you are good with this idea, I would adjust the documentation for it. Didn't want to start until I get your feedback.

jtattermusch · 2020-05-07T06:01:01Z

Adhoc package build: https://sponge.corp.google.com/c5b95681-bf71-4ecc-943d-35d2fbf41ca8

https://storage.googleapis.com/grpc-testing-kokoro-prod/test_result_public/prod/grpc/core/master/windows/grpc_build_packages/11908/20200507-004410/github/grpc/artifacts/csharp_nugets_windows_dotnetcli.zip

jtattermusch · 2020-05-07T12:54:17Z

@jtattermusch I redid most of the stuff and adjusted the logic. I thought a bit more about the problem, especially after seeing how the guessing for C++ works. I think this is a lot cleaner and better to maintain.

The new approach that I would like to discuss:

Added PatchedProtobuf to ProtoCompilerOutputs as way to manipulate OutputDir and GrpcOutputDir

Added PatchOutputDirectory to GeneratorService to do the patching

Will be used by C# to add the relative dir into the OutputDir

GetPossibleOutputs is easier for C# since it only works in the dir without doing the hashing stuff

This allows us to keep ProtoCompile close to protoc call and don't have special logic only for C# generator (MaybeEnhanceOutputDir would have introduced a bug for Generator = cpp)

_Protobuf_CoreCompile does not need the cryptic adjustment (_Protobuf_ExpectedGenerated) for adding the generated paths since %(_Protobuf_OutOfDateProto.[Grpc]OutputDir) works again

Added Protobuf_ProtoRoot to define a ProtoRoot default if outside of the project

This allows the relative paths to be generated a lot better

It keeps the old behavior to generate everything directly in [Grpc]OutputDir if no ProtoRoot was given

This also means, that if no ProtoRoot or Protobuf_ProtoRoot is given, it could still create duplicates, but they are very easy and generic fixable: <Protobuf Include="$(LIProtoPath)/**/*.proto" ProtoRoot="$(LIProtoPath)" /> which is more natural than setting OutputDir = "$(Protobuf_OutputPath)%(RelativeDir)"

This works for local, relative and absolute paths (including other drives)

I kept the changes to DepFileUtil (separation of GetDepFilenameForProto into GetOutputDirWithHash and GetDirectoryHash) as it was basically a refactoring with added documentation

If you would like to keep it in one method for performance, I will just revert it back to master

I also added a fix for C++ as this was broken for protos outside of the project (ProtoRoot was not available in Protobuf_PrepareClean, therefore the relative dirs were calculated wrong, leading to nothing being cleared). This allowed us also to use this for C#:

Extracted target _Protobuf_SetProtoRoot from _Protobuf_SelectFiles

_Protobuf_SetProtoRoot added as dependency to Protobuf_Clean to fix the bug in Protobuf_PrepareClean

If you are good with this idea, I would adjust the documentation for it. Didn't want to start until I get your feedback.

To be honest I'm having trouble to understand how exactly the new approach works. I'm going to add some comments to pinpoint the places in the code where I'm unsure about what exactly it does (and therefore is hard for me to tell whether it's good/bad idea to do so).

jtattermusch

I'm a bit confused by the new changes and I'm not fully convinced that this new approach is better and simpler (but granted I don't understand the logic fully yet).
Let some comment that will hopefully help us clarify

src/csharp/Grpc.Tools.Tests/CSharpGeneratorTest.cs

jtattermusch · 2020-05-07T12:58:39Z

src/csharp/Grpc.Tools/DepFileUtil.cs

+        // Calculate part of proto path relative to root. Protoc is very picky
+        // about them matching exactly, so can be we. Expect root be exact prefix
+        // to proto, minus some slash normalization.
+        internal static string GetRelativeDir(string root, string proto, TaskLoggingHelper log)


It doesn't seem that there are any tests for this logic so it's hard for me to see what is this method supposed to do and how exactly it fits into the bigger picture.
What is the benefit of having this method over the previous state? It's fine to introduce more logic in case where it makes sense (in this case I don't quite understand what the goal is), but otherwise I'd like to keep things as simple as possible.

No new logic, only moved it without changes from GeneratorServices.cs to DepFileUtil.cs. Before, it was used only in CppGeneratorServices and now we use it with CsharpGeneratorServices too. We could also keep it in GeneratorServices but move it from CppGeneratorServices to GeneratorServices. Might be the better option. In an earlier version, I also needed this method in ProtoCompile, but that's gone.

I think moving it to GeneratorServices seems like the better choice.

jtattermusch · 2020-05-07T13:02:49Z

src/csharp/Grpc.Tools/GeneratorServices.cs

+            {
+                pathStem = Path.Combine(grpcdir, relative);
+            }
+            protoItem.SetMetadata(Metadata.GrpcOutputDir, pathStem);


I feel that updating the GrpcOutputDir and OutputDir makes the logic quite hard to follow?
Who will be the changes visible to? Does this actually modify the input item (visible to the entire rest of the build) or is this local to this task (in which case we should probably remember the values in a different way?)

This will be used to update Protobuf_Compile for the rest of the build and it's documented in the targets file to make it easier to follow. This allows us to streamline the build. We could also add a separate metadata instead of overwriting it. But then, the original OutputDir would not be used anywhere in the targets.

For C++, this is skipped as protoc is generating the sub-folders according to this logic already. For C#, we would adjust the OutputDir to point to the relative path as C++ and other languages would do it.

Identity: protos/sub/folder/foo.proto ProtoRoot: protos OutputDir (before patch): obj/Debug/netstandard2.1 OutputDir (after patch): obj/Debug/netstandard2.1/sub/folder

Therefore the plugin will generate the files to obj/Debug/netstandard2.1/sub/folder as if the user had manually set OutputDir = "$(Protobuf_OutputPath)%(RelativeDir)" but this also works for protos not on the drive since ProtoRoot is used and not the project location. This just replicates the C++ behavior.

jtattermusch · 2020-05-07T13:07:23Z

src/csharp/Grpc.Tools/ProtoCompilerOutputs.cs

+            var patched = new List<ITaskItem>();
            foreach (var proto in Protobuf)
            {
+                // This operates on a local copy and has no effect on the MSBuild instance!


Just saw this comment. It seems to contradict the doc comment of the method: // Update OutputDir and GrpcOutputDir for the item and all subsequent // targets using this item. This should only be done if the real // output directories for protoc should be modified.

If we're actually only modifying the output dir values in this instance, there should be a less confusing way of doing this (store in a different variable / metadata?)

I might rework this comment. The generator.PatchOutputDirectory will patch the instance of proto which is local-only to avoid having to create a copied instance. Protobuf is only a copy based on the input and any changes done to it won't affect MSBuild.

I could store it in a different variable, but then we would either have to decide to deep-copy proto which decreases performance and makes no sense since it's a local-only copy already, or if we assign it to another variable but still modify proto, we would need to document it properly so that everyone knows that proto is also patched, not just the return value.

This seemed like the most obvious without side-effects.

jtattermusch · 2020-05-07T13:11:46Z

src/csharp/Grpc.Tools/ProtoCompilerOutputs.cs

            {
+                // This operates on a local copy and has no effect on the MSBuild instance!
+                generator.PatchOutputDirectory(proto);
+                patched.Add(proto);


this seems odd. You're taking items from Protobuf and you're modifying them (without making a defensive copy), so once done, both Protobuf and PatchedProtobuf will contain the same instances of ITaskItem, which is very confusing (and the items in Protobuf will all be patched, which is counterintuitive given the naming).

Protobuf is a local copy of the MSBuild items (see above). I can create copies and assign them to a separate variable to make it more clear, but this will result in duplicating items twice (once from MSBuild into the task and again in PatchOutputDirectory).

Only PatchedProtobuf will affect the MSBuild flow. Talking about it, making the copy explicit seems like the better approach.

jtattermusch · 2020-05-07T13:14:29Z

src/csharp/Grpc.Tools/build/_protobuf/Google.Protobuf.Tools.targets

    </FindUnderPath>
+
+    <ItemGroup>
+      <Protobuf>


how do we know that this change won't break existing users?

It's the same logic but on Protobuf instead of Protobuf_Compile as we have to patch it for all items iterated over in Protobuf_Clean, not just the one that should be compiled. I only also added the logic for $(Protobuf_ProtoRoot) in additionally. So if you prefer, I can split it into more PRs. One for fixing the C++ bug, one for adding $(Protobuf_ProtoRoot) and keep this one only for the Avoid collisions part and have it blocked by the other 2 PRs.

jtattermusch · 2020-05-07T13:20:44Z

src/csharp/Grpc.Tools/build/_protobuf/Google.Protobuf.Tools.targets

          Condition=" '$(Language)' == 'C#' " />

-  <Target Name="_Protobuf_SelectFiles">
+  <Target Name="_Protobuf_SetProtoRoot">


Can you explain the logic around Protobuf_ProtoRoot a bit more? I'm not sure I understand the purpose of it.

_Protobuf_SetProtoRoot is an extract from _Protobuf_SelectFiles. Protobuf_Clean has an bug in C++ because it's missing the ProtoRoot. As this PR would also use ProtoRoot for calculating OutputDir in C#, this bug would also get copied over to C#. Therefore I fixed it in here instead of in a separate PR.

But we can also put it in a separate PR which blocks this one as it's also a bug by itself.

jtattermusch · 2020-05-07T13:25:35Z

Correct me if I'm mistaken, but it looks like the new approach doesn't use dir hashing for avoiding the collisions at all?
Given that this is a complete revamp of the original logic, it would have been better if there was a separate PR for it, rather than replacing the logic in this PR - can we restore the original logic in this PR and create a separate PR with the new alternative approach?
I think I like the approach with dir hashing a bit more because it seems simpler (that said I think some ideas from the new approach still apply to the dir hashing approach).

Falco20019 · 2020-05-07T15:07:13Z

I think you are right with having a separate PR and reverting this. I would even go a step further and split this up into 3 more PRs:

Fix C++ cleanup bug
Introduce $(Protobuf_ProtoRoot)
Avoid collisions in cs file generated by Grpc.Tools (new approach)

This approach does not need the hashing, as it's duplicating the logic of the C++ generator. It just put's the files in the ObjectDir in folders replicating the folder structure of the proto up to the ProtoRoot.

So if you have only one file K:\protos\my\sub\folder\foo.proto and you do <Protobuf Include="K:\**\*.proto" ProtoRoot="K:\protos" />, the files will be generated as obj\Debug\netstandard2.1\my\sub\folder\Foo.cs as it already was for C++. This is also the more natural approach to protoc as you replicate the folder structure as protoc is using it in the imports. It also generates folder names that you can rely on. This is also independent of where your csproj is located at, since the folders are relative to ProtoRoot and not the csproj.

So if you don't think too deep into it, this is a lot easier to understand for the user than random hashes. It just replicates the structure that he already had for his proto files.

jtattermusch · 2020-05-07T15:25:34Z

I think you are right with having a separate PR and reverting this. I would even go a step further and split this up into 3 more PRs:

Fix C++ cleanup bug

Introduce $(Protobuf_ProtoRoot)

Avoid collisions in cs file generated by Grpc.Tools (new approach)

Sg, let's split this into multiple PRs. The smaller PRs, the easier they are to integrate.

This approach does not need the hashing, as it's duplicating the logic of the C++ generator. It just put's the files in the ObjectDir in folders replicating the folder structure of the proto up to the ProtoRoot.

So if you have only one file K:\protos\my\sub\folder\foo.proto and you do <Protobuf Include="K:\**\*.proto" ProtoRoot="K:\protos" />, the files will be generated as obj\Debug\netstandard2.1\my\sub\folder\Foo.cs as it already was for C++. This is also the more natural approach to protoc as you replicate the folder structure as protoc is using it in the imports. It also generates folder names that you can rely on. This is also independent of where your csproj is located at, since the folders are relative to ProtoRoot and not the csproj.

So if you don't think too deep into it, this is a lot easier to understand for the user than random hashes. It just replicates the structure that he already had for his proto files.

Ok, thanks for the explanation, I'll take another look as soon as you split the PR.

Falco20019 · 2020-05-07T17:51:59Z

Extracted the first two PRs:

Introducing $(Protobuf_ProtoRoot) and avoiding collisions is build on top of #22895. I will create the remaining PRs once these two are fixed to not have to rebase a lot.

Needed #if NETCORE

Falco20019 · 2020-06-16T12:20:09Z

@jtattermusch @JunTaoLuo
I rebased and cleaned this up so that only the relevant parts remained and that the changes are easier to follow.

Added PatchedProtobuf to ProtoCompilerOutputs as way to manipulate OutputDir and GrpcOutputDir
- Added PatchOutputDirectory to GeneratorService to do the patching
- Will be used by C# to add the relative dir into the OutputDir
- GetPossibleOutputs is easier for C# since it only works in the dir without doing the hashing stuff
- This allows us to keep ProtoCompile close to protoc call and don't have special logic only for C# generator
Added Protobuf_ProtoRoot to define a ProtoRoot default if outside of the project
- This allows the relative paths to be generated a lot better
- It keeps the old behavior to generate everything directly in [Grpc]OutputDir if no ProtoRoot was given
- This also means, that if no ProtoRoot or Protobuf_ProtoRoot is given, it could still create duplicates, but they are very easy and generic fixable: <Protobuf Include="$(LIProtoPath)/**/*.proto" ProtoRoot="$(LIProtoPath)" /> which is more natural than setting OutputDir = "$(Protobuf_OutputPath)%(RelativeDir)"
- This works for local, relative and absolute paths (including other drives)

JunTaoLuo

Changes look reasonable to me

src/csharp/Grpc.Tools/build/_protobuf/Google.Protobuf.Tools.targets

jtattermusch · 2020-06-23T12:04:00Z

adhoc artifact build: https://sponge.corp.google.com/invocation?tab=Build+Log&id=3cb0cb47-36d1-4a9e-a67d-f488e3107f49&searchFor=

jtattermusch · 2020-06-23T15:41:51Z

Adhoc build of nugets is here: https://storage.googleapis.com/grpc-testing-kokoro-prod/test_result_public/prod/grpc/core/master/windows/grpc_build_packages/12602/20200623-064408/github/grpc/artifacts/csharp_nugets_windows_dotnetcli.zip

jtattermusch

The change is looking good, thanks for the contribution!

I also tested the resulting nugets with #23375 and things seem to be working as expected.

jtattermusch · 2020-07-01T11:40:47Z

@apolcyn please add another Googler's LGTM.

Fix plugin initializers Previous code not threadsafe nor idiomatic Fix server_builder_plugin_test Bump version to v1.30.0-pre1 Regenerate projects Merge pull request grpc#23155 from srini100/v1.30.x bump v1.30.x branch to v1.30.0-pre1 Changes needed to enable TLS 1.3. Cleanup |ssl_bytes_remaining| per Yang's comments. Check unused bytes size is smaller than received bytes size, and fix Python format in _server_ssl_cert_config_test.py Revert "Check unused bytes size is smaller than received bytes size, and fix Python format in _server_ssl_cert_config_test.py" This reverts commit 7d0ae9c. Add check that unused bytes size is at most received bytes size. Add back source-only ruby packages for 1.30.x Fix memory leak, per Yang's comment. Initial working implementation Avoid collisions in cs files generated by Grpc.Tools Create generated directories Updated unit tests Reverted GetDepFilenameForProto Otherwise it creates conflicts with non-standard $(Protobuf_DepFilesPath) Improved folder generation Use same logic in csharp as in cpp Removed unused usings Reverted whitespace changes Readded InteropServices Needed #if NETCORE Bring closer to master Add support for Protobuf_ProtoRoot Fix false positive in grpc_is_epollexclusive_available on ESXi Improve third_party readme Fix up aio interop client Allow the user to specify grpclb config Merge remote-tracking branch 'upstream/master' into matt-tls13 Merge pull request grpc#23206 from ericgribkoff/dynamic_server_health Add gentle failover test Merge pull request grpc#23248 from ericgribkoff/use_grpc_health_check Use grpc health checks for PHP and Ruby xDS tests Run clang-format on is_epollexclusive_available.cc Merge pull request grpc#23250 from ericgribkoff/backport_gentle_failover Backport gentle failover Add TLS 1.2 testing. Merge remote-tracking branch 'upstream/master' into matt-tls13 Add #ifdef around use of SSL_ctx_min/max_proto_version Add include of tls1.h for OpenSSL. I is a reserved identifier Update links to grpc.io guides in header files Fix the grpcpp.h formatting. Merge pull request grpc#23186 from apolcyn/upmerge_source_only_to_1_30x Add back source-only ruby packages for 1.30.x Trust google-auth library will handle their dependency properly Install rsa==4.0 to avoid the egg cache race issue Add tls1.h header regardless of what SSL library is used. Bump version to 1.30.0 Regenerate projects Add #if defined(TLS1_3_VERSION). Merge pull request grpc#23264 from lidizheng/v1.30.x-patch Backport: Trust google-auth library will handle their dependency properly Merge pull request grpc#23266 from srini100/v1.30.x Bump to v1.30.0 Fix whitespaces Release dispatch queue on CFStreamHandle destroy Address David's comments. Move ServerPosix from ::grpc_impl to ::grpc Revert https://github.com/grpc/grpc/pull/18372/files Avoid attribute error in __del__ from _ChannelCallState Guard _ChannelCallState.__del__ from cygrpc being deallocated early Merge pull request grpc#23291 from lidizheng/v1.30.x-patch-2 Backport: Avoid attribute error in __del__ of _ChannelCallState Revert "Revert "Move ServerBuilder back to ::grpc from ::grpc_impl"" fix from sanitize.sh Merge pull request grpc#23303 from karthikravis/revert-23294-revert-23182-server-builder fix from sanitize.sh Merge branch 'master' into server-posix Move create_channel_posix from ::grpc_impl to ::grpc Revert grpc#18373 Merge remote-tracking branch 'upstream/master' into matt-tls13 Move create_channel and credentials from ::grpc_impl to ::grpc Reverts: grpc#18374 and grpc#18444 Credentials and create_channel are very closely intertwined, so it is easier to migrate them together. Make build and formatting fixes Add tests for unused bytes. Formatting fixes Implement gce_channel_credentials in core Fix formatting and build files Regenerate projects Initial implementation of gce_channel_creds Format Address Jiangtao's comments. Change name to compute_engine_channel_credentials Let the wrapped language create the composite credential Remove use of TLSv1_2_method. Expose gce tenancy as a function Regenerate projects Yapf :( Remove inline Stop double caching Add test Remove weird, seemingly pointless variable Clang format Main Makefile should require protobuf 3.12+ now Remove platform-specific assertion Buildifier Fix use-after-free in ruby call creds Lower log level to INFO from ERROR from keepalive watchdog firing Mark core API as experimental Use a gerund for a function name Simplify reference count juggling Clang format Merge pull request grpc#19756 from apolcyn/fix_ruby_md Fix auth plugin context use-after-free in ruby Make sure call creds user callbacks can't kill the ruby call credentials thread Update license date; fix format and typo Added missing deps_linkage Regenerated projects Fix typo Merge pull request grpc#23221 from chalin/chalin-grpc-io-link-fixes-200616 Update links to grpc.io guides in header files Merge pull request grpc#23335 from veblush/deps-linkage Added missing `deps_linkage` property Added missing deps_linkage Generated projects Fix gRPC-C++.podspec file Merge branch 'master' into revert-23294-revert-23182-server-builder Merge branch 'revert-23294-revert-23182-server-builder' of https://github.com/grpc/grpc into revert-23294-revert-23182-server-builder Merge pull request grpc#23324 from stanley-cheung/makefile-min-protobuf-version Main Makefile should require protobuf 3.12+ now Merge pull request grpc#23337 from CorralPeltzer/python-doc-typo Fix typo in docstring Merge branch 'master' into create-channel Fix credentials test. Address David's comments. Formatting fixes Suppress exceptions from the __del__ of channel object Merge pull request grpc#23304 from karthikravis/server-posix Move server posix from ::grpc_impl to ::grpc Merge pull request grpc#23305 from karthikravis/channel-posix Move create_channel_posix from ::grpc_impl to ::grpc Add back ciphersuites in ssl_utils.cc Merge pull request grpc#23345 from karthikravis/revert-23294-revert-23182-server-builder Revert "Revert "Move ServerBuilder back to ::grpc from ::grpc_impl"" Fix fetch oauth2 test Fix formatting issue Replaced grpc::string with std::string Merge pull request grpc#23350 from veblush/grpc-string2 Replaced grpc::string with std::string Removed grpc_artifact_protoc/Dockerfile Refactored AR in makefile Remove things_queued_ever counter from callback CQ Refactored AR in makefile Add abseil compiler options unify Grpc.Tools projects with other csproj projects Add missing include. Merge pull request grpc#23331 from yashykt/keepalivewatchdoginfolog Lower log level to INFO from ERROR from keepalive watchdog firing Fix connect deadline issue in settings_timeout_test Fix possible deadlock in RemoveExternalConnectivityWatcher Objc support PB runtime import override Fix possible deadlock in RemoveExternalConnectivityWatcher Merge branch 'master' into create-channel Merge pull request grpc#23354 from grpc/vjpai-patch-1 Remove things_queued_ever counter from callback CQ Merge pull request grpc#23364 from yashykt/settingstimeouttest Fix connect deadline issue in settings_timeout_test Removed GRPC_USE_ABSL Merge pull request grpc#23237 from grpc/jtattermusch-patch-1 Improve third_party readme (checklist for adding dependencies) Merge pull request grpc#22869 from Falco20019/grpc-fix-collisions Avoid collisions in cs files generated by Grpc.Tools update submodule boringssl-with-bazel with origin/master-with-bazel update boringssl dependency to master-with-bazel commit SHA Merge pull request grpc#23360 from markdroth/llvm_build_fix Add missing include. Merge pull request grpc#23165 from matthewstevenson88/matt-tls13 Enable TLS 1.3 in the C-core and all wrapped languages. Fix message_allocator_e2e_test Merge pull request grpc#23356 from veblush/ar-makefile Refactored AR in makefile Merge pull request grpc#23357 from veblush/absl-cppflags Add abseil compiler options Merge pull request grpc#23352 from veblush/protoc-docker Removed grpc_artifact_protoc/Dockerfile Merge pull request grpc#23298 from yulin-liang/prefix_import_path Objc support PB runtime import override Merge pull request grpc#23371 from veblush/del-GRPC_USE_ABSL Removed GRPC_USE_ABSL Merge pull request grpc#23376 from yang-g/allocator_test Fix message_allocator_e2e_test Added call to grpc::testing::TestEnvironment in tests Merge pull request grpc#23367 from yashykt/130fixws Fix possible deadlock in RemoveExternalConnectivityWatcher Merge pull request grpc#23378 from veblush/harden-tests Added call to grpc::testing::TestEnvironment in tests Added call to grpc::testing::TestEnvironment in more tests Merge branch 'master' into create-channel Merge pull request grpc#23379 from veblush/harden-tests2 Added call to grpc::testing::TestEnvironment in more tests Merge pull request grpc#23276 from grpc/cfstream-queue-release Release dispatch queue on CFStreamHandle destroy Add comment Merge pull request grpc#23333 from apolcyn/bad_md_drops_call_creds_thread Make sure call creds user callbacks can't kill the ruby call credentials thread Merge pull request grpc#23365 from yashykt/deadlockws Fix possible deadlock in RemoveExternalConnectivityWatcher Passed the repo manager to veblush Merge pull request grpc#23392 from veblush/to-veblush Passed the repo manager to veblush Increase size of a test to match internal guidance Pin isort to fix sanity test breakage New Matchers Implementation - All except for Regex (which will be submitted with RE2 import PR) Merge pull request grpc#22813 from mehrdada/fix-static-initialization-crap Fix channelz plugin initializer Merge pull request grpc#23395 from gnossen/lint_fix Pin isort to fix sanity test breakage Merge remote-tracking branch 'upstream/master' into matcher_new Merge pull request grpc#22951 from donnadionne/matcher_new New Matchers Add missing string include Get new core API design building Fix use-after-free in ruby call creds Add Python wrapper Pin isort to fix sanity test breakage Added TCP_USER_TIMEOUT auto-detection Merge pull request grpc#23393 from grpc/vjpai-patch-1 Increase size of a test to match internal guidance Merge pull request grpc#23400 from apolcyn/backport_sanity_check_fix Backport pylint sanity check fix to 1.30.x Bump 1.30.x branch to 1.30.1 Fix inproc transport bugs on client WritesDone or Read after status Don't drop message on WriteAndFinish with non-ok status regenerate projects generate boringssl prefix headers fix nits in third_party/README.md Increment podspec version regenerate projects Update README.md Add TLS 1.2 and 1.3 specific tests to h2_tls fixture. Merge pull request grpc#23380 from vjpai/inproc_fix Fix inproc transport bug on client WritesDone and server send message after status Merge pull request grpc#23409 from grpc/jtattermusch-patch-2 fix nits in third_party/README.md Merge pull request grpc#23386 from Kernald/fix-string-includes Add missing string include Merge pull request grpc#23408 from grpc/vjpai-patch-1 Don't drop message on WriteAndFinish with non-ok status Merge pull request grpc#23401 from veblush/tcp_user_timeout Added TCP_USER_TIMEOUT auto-detection Add gRPC-specific docker images for Ruby build Make pluing embed zlib Allows poller to bound to ephemeral loops in multiple threads Address comments Add type annotation in comments Merge pull request grpc#23399 from apolcyn/backport_auth_plugin_ruby_fix Backport fix for use-after-free in ruby auth plugin Replace most uses of gpr_asprintf() with absl calls. Merge pull request grpc#23387 from veblush/protoc-docker Make plugins embed zlib Merge pull request grpc#23377 from lidizheng/aio-multi-loop [Aio] Allows poller to bound to ephemeral loops in multiple threads Fix FailHijackedRecvMessage for generic async APIs Update grpclb with optional field "name" Add test for the new config field Address comments Add support for xDS regex matchers. Merge pull request grpc#23351 from lidizheng/silence-del Suppress exceptions from the __del__ of channel object Make sure that some ops don't start concurrently with Finish Merge pull request grpc#23405 from apolcyn/bump_version Bump 1.30.x branch to 1.30.1 Merge pull request grpc#22987 from donnadionne/new Add support for xDS regex matchers. Merge pull request grpc#23157 from markdroth/gpr_asprintf Replace most uses of gpr_asprintf() with absl calls. Store ref to the ExternalConnectivityWatcher in external_watchers_ map. Add comment Fix ruby 2.7 keyword arguments deprecation Similar to grpc#22915 Properly count messages sent Merge pull request grpc#23413 from matthewstevenson88/tls13-h2-tls-fixture Add TLS 1.2 and 1.3 specific tests to h2_tls fixture. Stop trying to handle SIGSEGV Make request path more easily visible to LB policies. Make warning more dire Merge pull request grpc#23416 from vjpai/ccet Make sure that some test ops don't start concurrently with Finish Merge pull request grpc#23421 from gnossen/test_runner_segfault Stop trying to handle SIGSEGV in Test Runner Merge pull request grpc#23258 from tweej/I_is_reserved I is a reserved identifier Merge pull request grpc#23417 from markdroth/lb_policy_path Make request path more easily visible to LB policies. Reviewer comments WIP EM-agnostic callback completion queue Fix TCP_USER_TIMEOUT definition Address comments Another comment Merge pull request grpc#23415 from yashykt/genericfailsend Fix FailHijackedRecvMessage for generic async APIs Merge pull request grpc#23425 from veblush/fix-TCP_USER_TIMEOUT Fix TCP_USER_TIMEOUT definition Fix data race in RpcWithEarlyFreeRequest by using atomics Fix data race in RpcWithEarlyFreeRequest by using atomics Merge pull request grpc#23358 from jtattermusch/cleanup_grpc_tools unify Grpc.Tools projects with other csproj projects Merge pull request grpc#23410 from jtattermusch/upgrade_boringssl_submodule Upgrade boringssl submodule to latest master-with-bazel add Grpc.Tools test for generating for .proto with the same name Merge pull request grpc#23427 from vjpai/message_allocator_race Fix data race in RpcWithEarlyFreeRequest by using atomics Merge pull request grpc#23361 from vjpai/em_agnostic_core_callback_cq EM-agnostic core callback CQ Remove unnecessary RPC Remove references to skipping callback API tests Reviewer comments Merge pull request grpc#22345 from muxi/grpclb-name-config Update grpclb configuration with field "service_name" Add sunny day core API path Merge pull request grpc#23419 from miyucy/patch-1 Fix ruby 2.7 keyword arguments deprecation Update the hint for install Python header to depend on version Stop using mutex for non-polling engine marker Make sure >3 Python version triggers a failure Merge pull request grpc#23418 from yashykt/dataraceexternalcancel Store ref to the ExternalConnectivityWatcher in external_watchers_ map. Merge pull request grpc#23430 from vjpai/remove_do_not_test Remove references to skipping callback API tests Remove unused variable (fixes builder error) Merge pull request grpc#23435 from vjpai/fix_iomgr_non_polling Fix-forward: Stop using mutex for non-polling engine marker Add test Clean up Reuse mdelem Merge remote-tracking branch 'origin/master' into python_google_default_creds Merge pull request grpc#23440 from grpc/vjpai-patch-1 Remove unused variable (fixes builder error) Merge pull request grpc#23375 from jtattermusch/csharp_proto_collision_test Add Grpc.Tools test for generating for .proto with the same name Added change_repo_manager.sh Changed a repo manager to karthikrs add comments for constants Fix initialization order Complete synchronously Accomodate PHP examples/README rework - Fix link to "Intro to gRPC" (formerly "Overview") - Don't list wrapped languages twice (once in intro paragraph, then in subdirectory list); move subdirectory links to the start of the page. - Sort subdirectory list - Link to grpc.io "Supported Languages" page for the remaining languages Merge pull request grpc#23446 from yang-g/init_order Fix initialization order Merge pull request grpc#23445 from HannahShiSFB/add-comments4constants PHP: add comments for constants make port server python3 compatible fix livelock in concurrent_connectivity_test on windows update bigquery partition expiration Update grpc_release_schedule.md Remove xds-experimental URI scheme. Merge pull request grpc#23444 from veblush/to-karthikrs Changed a repo manager to karthikrs Merge pull request grpc#22201 from chrisse74/memoryLeakBuildAndStart fix memory leak of grpc_resource_user_quota Merge pull request grpc#23437 from lidizheng/update-support-message Update the hint for install Python header to depend on version Merge branch 'master' into create-channel Fix build error Further reduce the spam log by doing nothing if GC has error Reorganize header Change signature Merge pull request grpc#23463 from markdroth/xds_experimental Remove xds-experimental URI scheme. Clean up Format Typo Merge remote-tracking branch 'origin/master' into python_google_default_creds Regenerate projects Merge pull request grpc#23447 from chalin/chalin-ex-readme-200710 examples/README rework Add XdsConfigSelector and populate call attribute containing the deadline. Fix secure test Fix formatting Merge pull request grpc#23467 from lidizheng/dealloc-message Further reduce the spam log by doing nothing if GC has error Update by review Move AuthMetaDataProcessor from ::grpc_impl to ::grpc Reverts: https://github.com/grpc/grpc/pull/18436/files Move GenericStub from ::grpc_impl to ::grpc Reverts: grpc#18434 Move LoadReportingServiceServerBuilderOption from ::grpc_impl to ::grpc Reverts: https://github.com/grpc/grpc/pull/18419/files Move ProtoServerReflectionPlugin from grpc_impl to grpc namespace Reverts: grpc#18600 Merge pull request grpc#23460 from markdroth/xds_config_selector Add XdsConfigSelector and populate call attribute containing the deadline. Fix tests Fix formatting Defer allocation of error object Add to Python documentation Support asyncio interop client Call out unexpected behavior adjust comment Merge pull request grpc#23462 from grpc/jtattermusch-patch-3 Update grpc_release_schedule.md Merge pull request grpc#23457 from jtattermusch/livelock_concurrent_connectivity_test fix potential livelock in global_subchannel_pool.cc bytes(string, encoding) is not supported in python2 results in error "TypeError: str() takes at most 1 argument (2 given)" Merge pull request grpc#23458 from jtattermusch/bigquery_partition_expirations Update bigquery partition expiration (to 365 days) Merge pull request grpc#23455 from jtattermusch/port_server_python3 Make port server python3 compatible Change repo manager to donnadionne This will be merged on 7/20/2020. Merge pull request grpc#23343 from veblush/v1.30.x-ruby [v1.30.x] Backport grpc#23335 (Added missing `deps_linkage` property) Updating C++ interop client and server to do path matching and header matching tests. Merge pull request grpc#23464 from donnadionne/new Updating C++ interop client and server Makefile: stop supporting make install remove rpath trick regenerate projects move routeguide c++ distribtest under cmake_pkgconfig fix C++ interop test build fix PHP interop tests build update PHP readme Merge pull request grpc#23412 from jtattermusch/makefile_remove_installation Simplify makefile: Get rid of "install" rules with pure make, recommend cmake and bazel instead. Include the target name in c-ares and native DNS summary error messages Adding Fake headers for header matching. Construct/destruct ExecCtx between init/shutdown only Add --verbose to dart pub get command Merge pull request grpc#23494 from donnadionne/ga Adding Fake headers for header matching. Concatenate duplicate header keys for matching. Merge pull request grpc#23495 from vjpai/fix_tls Construct/destruct ExecCtx between init/shutdown only Fixing call to absl::StrSplit to deal with empty string case. Merge pull request grpc#23499 from donnadionne/new Fixing call to absl::StrSplit to deal with empty string case. Merge pull request grpc#23496 from stanley-cheung/dart-debug Add --verbose to dart pub get command fix concurrent file read flake Avoid unused param warning Merge pull request grpc#23506 from grpc/vjpai-patch-2 Avoid unused param warning Merge pull request grpc#23493 from apolcyn/error_messages Include the target name in top-level DNS error messages Merge pull request grpc#23468 from karthikravis/auth_metadata Move AuthMetaDataProcessor from ::grpc_impl to ::grpc Merge pull request grpc#23469 from karthikravis/generic-stub Move GenericStub from ::grpc_impl to ::grpc Make more fixes for moving GenericStub from ::grpc_impl to ::grpc Merge pull request grpc#23512 from karthikravis/generic-stub Make more fixes for moving GenericStub from ::grpc_impl to ::grpc Merge pull request grpc#23470 from karthikravis/load-reporting Move LoadReportingServiceServerBuilderOption from ::grpc_impl to ::grpc Fix ruby protoc plugin when message is in another package Ruby protoc plugin fix: fix test breakage Merge pull request grpc#23471 from karthikravis/proto-server Move ProtoServerReflectionPlugin from grpc_impl to grpc namespace Merge pull request grpc#22589 from veblush/ruby-docker Ruby docker for gRPC Manylinux2010-based Ruby images Remove proto_server_reflection_plugin_impl.h from build_autogenerated.yaml Fixing unsafe std::move. Merge pull request grpc#23501 from stanley-cheung/fix-ruby-plugin Fix ruby protoc plugin when message is in another package Fix ruby protoc plugin when message is in another package Merge pull request grpc#23516 from karthikravis/fix-1 Remove proto_server_reflection_plugin_impl.h from build_autogenerated.yaml Merge pull request grpc#23517 from donnadionne/new Fixing unsafe std::move. Merge pull request grpc#23519 from stanley-cheung/ruby-protoc-plugin-fix-backport Backport grpc#23501: Fix ruby protoc plugin when message is in another package Bump version to 1.30.2 Avoid unused param warnings in c_ares directory Merge pull request grpc#23504 from jtattermusch/unflake_verify_peer_options Do not trigger grpc_load_file() flake in verify_peer_options_test Merge pull request grpc#23520 from stanley-cheung/bump-version-1-30-2 Bump version to 1.30.2 Merge pull request grpc#23505 from grpc/vjpai-patch-1 Remove some unused parameter names Merge pull request grpc#23491 from markdroth/xds_header_combining Concatenate duplicate header keys for matching. Revert "Merge pull request grpc#21941 from markdroth/xds_logging" This reverts commit 37d5d93, reversing changes made to 0e8190a. Fix PHP build by adding re2 dependencies Merge pull request grpc#23344 from veblush/ruby-docker-ml Ruby docker image based on manylinux2010 Upgrade BoringSSL to latest on master-with-bazel branch Uses sources.json uses sources.json update rules remove sha256 update sha256 update one more place Merge pull request grpc#23524 from markdroth/xds_logging_revert Revert "Merge pull request grpc#21941 from markdroth/xds_logging" Merge pull request grpc#23526 from stanley-cheung/fix-php-build Fix PHP build by adding re2 dependencies Fix php build from source Revert "Adding Fake headers for header matching." This reverts commit 8c013bf. Fix concatenated_value Fix formatting issue Merge pull request grpc#23538 from karthikravis/fix-xds-breakage Revert "Adding Fake headers for header matching." Update PHP artifact test generate api reference documents by doxygen Added grpc::testing::TestEnvironment Merge pull request grpc#23531 from stanley-cheung/fix-php-build-from-source Fix php build from source Updated comments C++ify core server Merge remote-tracking branch 'upstream/v1.30.x' into HEAD Merge pull request grpc#23236 from tpetkov-VMW/fix-grpc_is_epollexclusive_available Fix false positive in grpc_is_epollexclusive_available on ESXi Remove env var protection of new xds routing code. increase timeout for PHP distribtests Merge pull request grpc#23540 from stanley-cheung/php-artifact-test Update PHP artifact test Merge pull request grpc#23545 from veblush/test-shutdown Added call to grpc::testing::TestEnvironment in tests Merge pull request grpc#23476 from karthikravis/change-owner Change repo manager to donnadionne revert gen_build_yaml.py generate boringssl prefix headers Merge pull request grpc#23548 from jtattermusch/upmerge_1_30_x_to_master Upmerge 1.30.x branch into master Second regeneration update grpc-core podspec template regenerate project Merge pull request grpc#23537 from markdroth/xds_routing_env Remove env var protection of new xds routing code. Protect access Fix docstring Implement fake and ignored headers. Take care to ensure we use fake headers in testing so we don't conflict with headers used by census. Decouple checking tenancy from checking for metadata server Provide an objc-generator option to import system files in a local way Merge pull request grpc#23550 from jtattermusch/php_distrib_15min increase timeout for PHP distribtests Merge pull request grpc#23539 from donnadionne/fix Adding Fake headers and ignored headers for header matching Change xDS ConfigSelector not to pass the timeout header value to the LB policy. Merge pull request grpc#23475 from yulin-liang/local_import_prefix Objc support grpc files import override Initialize the grpc_local_import variable to false. Include re2 in source wheel && solve the missing --all complain Stop checking g_is_on_gce in security connector Merge pull request grpc#23555 from yulin-liang/local_import_prefix Initialize the grpc_local_import variable to false. Log the peer address of grpc_cli CallMethod RPCs to stderr Explicitly fake status if cancelling stream Drop min thread count of threads doing Next to 1 Merge pull request grpc#23255 from HannahShiSFB/api-doxygen-step2 PHP: generate api reference documents by doxygen Merge pull request grpc#23534 from markdroth/xds_fake_header_revert Change xDS ConfigSelector not to pass the timeout header value to the LB policy. stop stripping .hpp files from python source packages Merge pull request grpc#23562 from yang-g/one_core Drop min thread count of threads doing Next to 1 Merge pull request grpc#23528 from emkornfield/upgrade_boring Upgrade BoringSSL to latest on master-with-bazel branch Merge pull request grpc#23552 from lidizheng/re2-source-wheel Include re2 in source wheel && solve the missing --all complain Merge pull request grpc#23565 from jtattermusch/python_dev_stop_excluding_hpp Stop stripping .hpp files from python source packages avoid destroy channel more than once Don't check tenancy if credentials specified Use RAII lock guard Support tuple and aio.Metadata interaction Make pytype happy Only >=3.7 allows annotate oneself as return type Metadata value has to be string or bytes Remove MAX_EPOLL_EVENTS_HANDLED_EACH_POLL_CALL to ensure timely processing of events Remove unused variable Add logs to a flakey ruby test that can help to diagnose issue Add comment Merge pull request grpc#23556 from lidizheng/fix-gapic-tests [Aio] Support tuple and aio.Metadata interaction Merge pull request grpc#23535 from apolcyn/repro_epollex_issue_minimal Remove MAX_EPOLL_EVENTS_HANDLED_EACH_POLL_CALL to ensure timely processing of events Output grpc-dotnet interop apps to specific directories Merge pull request grpc#23570 from apolcyn/fix_ruby_flake Add logs to a flakey ruby test that can help to diagnose issue Also fake status in grpc_chttp2_mark_stream_closed if already closed but there is an error Make ABSL header block consistent with other code. Merge pull request grpc#23558 from yashykt/cancelstreamdeadline Explicitly fake the status for when we cancel the streams Fix Mac and Windows Merge pull request grpc#22757 from vjpai/core_server_cxx C++ify grpc_server Fix a typo Convert multi-line comments to single line comments in chttp2_transport.cc Merge branch 'master' into chttp2transportcomment add g-stands-for for the next release update version to 1.32.0-dev Merge pull request grpc#23523 from JamesNK/jamesnk/interoptests-remove-version Output grpc-dotnet interop apps to specific directories regenerate projects Merge pull request grpc#23203 from gnossen/python_google_default_creds Implement compute_engine_channel_credentials in Python Clang format Plumb absl::Status through connectivity state notifiers Merge pull request grpc#23480 from yashykt/abslstatusconn Plumb absl::Status through connectivity state notifiers Merge branch 'master' into chttp2transportcomment Flag protect new logs Merge pull request grpc#23576 from jtattermusch/bump_master_1_32 update master version to 1.32.0-dev Merge branch 'master' into create-channel Consistency check core_version and version Fixes to match shellcheck Merge pull request grpc#23578 from vjpai/core_check Add a consistency check on version and core_version Merge pull request grpc#23571 from vjpai/typos Typo fix xDS v3 support Merge pull request grpc#23557 from apolcyn/display_peer_address Log the peer address of grpc_cli CallMethod RPCs to stderr fixed merge issues Merge pull request grpc#23281 from markdroth/xds_v3 xDS v3 support Make fixes to cronet_credentials.h Fixing Ruby 1.31.0.pre1 compilation failure with source-only package missing re2 from build_handwritten.yaml and generated grpc.gemspec Merge pull request grpc#23599 from donnadionne/ruby Fixing Ruby compilation failure with source-only package Merge pull request grpc#23308 from karthikravis/create-channel Move create_channel and credentials from ::grpc_impl to ::grpc Disable TLS 1.3 in the C-core. Revert "Move create_channel and credentials from ::grpc_impl to ::grpc" Disable TLS 1.3-specific unit tests in ssl_transport_security_test.cc. Fix number of TLS version tests to run. Merge pull request grpc#23609 from grpc/revert-23308-create-channel Revert "Move create_channel and credentials from ::grpc_impl to ::grpc" Merge pull request grpc#23567 from HannahShiSFB/23477 PHP: avoid destroy channel more than once Merge pull request grpc#23608 from matthewstevenson88/matt-disable-tls13 Disable TLS 1.3 in the C-core. Fix BUILD file for import. Member variable should not be a reference; causing crash when used in destructor. Merge pull request grpc#23575 from yashykt/chttp2transportcomment Comment formatting in chttp2_transport Merge pull request grpc#23610 from donnadionne/crash Member variable should not be a reference; causing crash when used in destructor. Merge pull request grpc#23612 from markdroth/xds_v3_build_fix Fix BUILD file for import. Passing repo manager to markdroth Merge pull request grpc#23624 from donnadionne/rm Passing repo manager to markdroth resolved merge conflicts with envoy v3 api

jtattermusch self-assigned this May 5, 2020

jtattermusch added lang/C# kokoro:run labels May 5, 2020

grpc-kokoro removed the kokoro:run label May 5, 2020

jtattermusch added the release notes: yes Indicates if PR needs to be in release notes label May 5, 2020

jtattermusch reviewed May 5, 2020

View reviewed changes

src/csharp/Grpc.Tools/build/_protobuf/Google.Protobuf.Tools.targets Outdated Show resolved Hide resolved

jtattermusch reviewed May 5, 2020

View reviewed changes

src/csharp/Grpc.Tools/DepFileUtil.cs Show resolved Hide resolved

Falco20019 requested a review from jtattermusch May 5, 2020 12:48

jtattermusch added the kokoro:run label May 5, 2020

grpc-kokoro removed the kokoro:run label May 5, 2020

jtattermusch reviewed May 5, 2020

View reviewed changes

src/csharp/Grpc.Tools/ProtoCompile.cs Outdated Show resolved Hide resolved

Falco20019 requested a review from jtattermusch May 6, 2020 13:19

Falco20019 mentioned this pull request May 6, 2020

Grpc.Tools: Accented characters in nuget path breaks c# grpc tool compiler #17995

Closed

jtattermusch added the kokoro:run label May 6, 2020

grpc-kokoro removed the kokoro:run label May 6, 2020

jtattermusch reviewed May 7, 2020

View reviewed changes

jtattermusch added the kokoro:run label May 8, 2020

Falco20019 added 2 commits June 16, 2020 12:10

Reverted whitespace changes

861cdf3

Readded InteropServices

d0c7eac

Needed #if NETCORE

Falco20019 force-pushed the grpc-fix-collisions branch from 1db3cff to d0c7eac Compare June 16, 2020 10:14

grpc-kokoro removed kokoro:run labels Jun 16, 2020

Falco20019 added 2 commits June 16, 2020 12:31

Bring closer to master

446f72d

Add support for Protobuf_ProtoRoot

e165a33

Falco20019 changed the title ~~[WIP] Avoid collisions in cs files generated by Grpc.Tools~~ Avoid collisions in cs files generated by Grpc.Tools Jun 16, 2020

Falco20019 requested a review from jtattermusch June 16, 2020 12:21

JunTaoLuo approved these changes Jun 16, 2020

View reviewed changes

src/csharp/Grpc.Tools/build/_protobuf/Google.Protobuf.Tools.targets Outdated Show resolved Hide resolved

jtattermusch added the kokoro:run label Jun 17, 2020

grpc-kokoro removed the kokoro:run label Jun 17, 2020

Fix whitespaces

91e44c2

jtattermusch added the kokoro:run label Jun 23, 2020

grpc-kokoro removed the kokoro:run label Jun 23, 2020

jtattermusch mentioned this pull request Jul 1, 2020

Add Grpc.Tools test for generating for .proto with the same name #23375

Merged

jtattermusch approved these changes Jul 1, 2020

View reviewed changes

jtattermusch merged commit 88cd5b0 into grpc:master Jul 1, 2020

Falco20019 deleted the grpc-fix-collisions branch July 1, 2020 15:25

jtattermusch mentioned this pull request Jul 9, 2020

Grpc.Tools readme update - duplicated .proto names #22628

Merged

jtattermusch mentioned this pull request Jul 20, 2020

Avoid collisions in cs files generated by Grpc.Tools #22852

Closed

YSZhuoyang mentioned this pull request Aug 15, 2020

Grpc.Tools: Incorrect warnings when multiple paths are specified in ProtoRoot attribute #23852

Closed

jtattermusch mentioned this pull request Nov 23, 2020

Add generation to multiple directories for C# plugin #24426

Closed

jtattermusch mentioned this pull request Dec 1, 2020

C# generator overwrites files with same *.proto in different folders/namespaces #24697

Closed

jtattermusch mentioned this pull request Mar 18, 2021

Grpc.Tools: Code Generation fails with VS2019 when proto files are added outside of project path #23583

Closed

Conversation

Falco20019 commented May 5, 2020

Uh oh!

Falco20019 commented May 5, 2020

Uh oh!

jtattermusch commented May 5, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

jtattermusch commented May 5, 2020

Uh oh!

jtattermusch commented May 5, 2020

Uh oh!

Uh oh!

Uh oh!

Falco20019 commented May 5, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Uh oh!

Falco20019 commented May 6, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

jtattermusch commented May 7, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

jtattermusch commented May 7, 2020

Uh oh!

jtattermusch left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Falco20019 May 7, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

jtattermusch commented May 7, 2020

Uh oh!

Falco20019 commented May 7, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

jtattermusch commented May 7, 2020

Uh oh!

Falco20019 commented May 7, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Falco20019 commented Jun 16, 2020

Uh oh!

JunTaoLuo left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

jtattermusch commented Jun 23, 2020

Uh oh!

jtattermusch commented Jun 23, 2020

Uh oh!

jtattermusch left a comment

Choose a reason for hiding this comment

jtattermusch commented May 5, 2020 •

edited

Loading

Falco20019 commented May 5, 2020 •

edited

Loading

Falco20019 commented May 6, 2020 •

edited

Loading

jtattermusch commented May 7, 2020 •

edited

Loading

Falco20019 May 7, 2020 •

edited

Loading

Falco20019 commented May 7, 2020 •

edited

Loading

Falco20019 commented May 7, 2020 •

edited

Loading