Emscripten support take 2 (with bonus i686-unknown-linux-musl). #30629
Conversation
| llvm::LLVMWriteBitcodeToFile(llmod, out.as_ptr()); | ||
| // Change what we write and cleanup based on whether obj files are | ||
| // just llvm bitcode. In that case write bitcode, and possibly | ||
| // delete the bitcode if it wasn't requisted. Don't generate the |
There was a problem hiding this comment.
requested (sorry about the nit, my eyes just fell on it)
|
r=me with one minor nit, but otherwise this'll just need libc to land first as well |
|
☔ The latest upstream changes (presumably #30601) made this pull request unmergeable. Please resolve the merge conflicts. |
|
Rebasing this now. |
e1ced26
to
b9dd828
Compare
|
Well, I've rebased and addressed comments, and this builds, but testing against emscripten incoming it no longer produces working js. Still, I'm up for landing this for the wip. |
|
@bors r=alexcrichton |
|
📌 Commit b9dd828 has been approved by |
|
@bors r- I've got a commit that changes jemalloc in there. |
b9dd828
to
7afb56f
Compare
|
@bors r=alexcrichton |
|
📌 Commit 7afb56f has been approved by |
Backtraces, and the compilation of libbacktrace for asmjs, are disabled. This port doesn't use jemalloc so, like pnacl, it disables jemalloc *for all targets* in the configure file. It disables stack protection.
This tells trans::back::write not to LLVM codegen to create .o files but to put LLMV bitcode in .o files. Emscripten's emcc supports .o in this format, and this is, I think, slightly easier than making rlibs work without .o files.
|
⌛ Testing commit 7afb56f with merge 8c604dc... |
Here's another go at adding emscripten support. This needs to wait again on new [libc definitions](rust-lang/libc#122) landing. To get the libc definitions right I had to add support for i686-unknown-linux-musl, which are very similar to emscripten's, which are derived from arm/musl. This branch additionally removes the makefile dependency on the `EMSCRIPTEN` environment variable by not building the unused compiler-rt. Again, this is not sufficient for actually compiling to asmjs since it needs additional LLVM patches. r? @alexcrichton
|
@brson @alexcrichton a bit late to this party, but |
|
AFAIK that's just to tell the linker to emit a 32-bit executable in case it's default a 64-bit one |
|
Does that make it redundant with On Mon, May 2, 2016 at 1:10 AM, Alex Crichton [email protected]
|
|
In this case, probably yes |
|
@tamird Memory is hazy, but if I didn't write that it didn't work, and I don't know if it's redundant with |
Here's another go at adding emscripten support. This needs to wait again on new libc definitions landing. To get the libc definitions right I had to add support for i686-unknown-linux-musl, which are very similar to emscripten's, which are derived from arm/musl.
This branch additionally removes the makefile dependency on the
EMSCRIPTENenvironment variable by not building the unused compiler-rt.Again, this is not sufficient for actually compiling to asmjs since it needs additional LLVM patches.
r? @alexcrichton