ping

@dschuff ptal

corresponding lld change is here: https://reviews.llvm.org/D38297

@dschuff PTAL

• [WebAssembly] Update test output to match llvm changes
• [WebAssembly] only run wasm tests when wasm is enabled
• [WebAssembly] Review feedback

Thank you for all the feedback @ruiu. It not surprise to me that a patch of this size takes a long time to review. I'm currently working on a fairly large change to allow data sections with the same name to be combined into a single section.. in order to allow things like init_array/fini_array. My aim is to get that done before the initial version of this patch lands.

Remove unrelated change

• [WebAssembly] Fix index for R_WEBASSEMBLY_MEMORY_ADDR_* relocations
• Merge remote-tracking branch 'origin/master' into wasm
• [WebAssembly] Don't re-export entry point as 'main'
• [WebAssembly] Allow mulitple wasm data segments
• [WebAssembly] Update tests to match changes in llvm
• [WebAssembly] Update LEB to match changes to LEB128.h
• [WebAssembly] don't test data-layout with -r
• [WebAssembly] Refactor code to create OutputSections.cpp

• add bss test case

• add test

Yes, there is the potential that we might need more metadata down the road (type, flags, etc) but right now I think name is enough for to get init_array and fini_array working.

• cleanup tests

• don't use const ArrayRef
• fix test

• rename test

Added support for explicit custom section names, and added new test

• support custom data sections

fix test

fix test

• fix

More section types

I created a more minimal version of this change https://reviews.llvm.org/D37876. Abandoning this one.

I guess I was just trying to match the other native toolchains. One can still pass -fdata-sections, but why make it the default?

Rebase

Yes, this change is a WIP that depends on D37834.

• review feedback
• Merge remote-tracking branch 'origin/master' into data_sections

Yes, I was refering to the waterfall tests.

Making the change here first means that I can test the new default and land it in llvm. If clang is continues to set its own default then anyone using clang (i.e. a lot of the tests) won't see the effect of the change.

• restore debugging

remove debugging

• .

remove debugging

• Merge branch 'master' into wasm
• [WebAssembly] Update expectations now that SymbolInfo not a required YAML field.
• [WebAssembly] Don't try to export main if its undefined
• [WebAssembly] Review feedback

@sunfish could you take a look?

• feedback

Could you take another look derek? This is currently causing test failures for big endian architectures. If this change is too big I can lang a more targeted one line change in the interim.

OK makes sense. We now only create symbols from imports/exports when the 'linking' section is present.

• .

• Don't treat imports/exports as symbols table unless linking

• handle 64bit global exports

My understanding is that with LEB one can pad to any number of bytes. We use 5 when we write 32-bit values in wasm because 5 is the number of bytes required to encode any 32-bit number as an LEB. We likewise use 10 bytes when writing padded 64-values. One might conceivably use 2 and 3 when writing 8-bit and 16-bit values as well.