similar to: Contributing LLD for Mach-O

On 2020-02-28, James Y Knight via llvm-dev wrote: >Nice! > >Your plan sounds great, and it'll be awesome to finally have a good MachO >LLD available. > >On Fri, Feb 28, 2020 at 4:32 PM Shoaib Meenai via llvm-dev < >llvm-dev at lists.llvm.org> wrote: > >> Hi all, >> >> We’re planning to contribute a new implementation of LLD for Mach-O, using

James, many thanks. Is there any linker available for Macs that has a freely available binary version? I thought maybe that GNU’s linker might fit the bill? I cannot use Apple’s linker from /usr/bin/ as it is not allowed to make external calls from a sandboxed app. Hence my interest in the LLVM lld. > On 7 May 2020, at 19:21, James Y Knight <jyknight at google.com> wrote: > > On

I looked into this further. ld64 has a macho_nlist abstraction over the various underlying nlist structures [1]. On x86-64, the P::getP referenced in n_value will resolve to [2], which in turn goes to [3], which calls OSReadLittleInt64. On a little endian machine, OSReadLittleEndian just calls _OSReadInt64 [4], which in turn does a pointer arithmetic and cast and then dereferences the pointer [5].

As part of our work on LLD for Mach-O, we’ve observed that the object files produced by LLVM don’t always have aligned nlist_64 entries. For context, nlist_64 is the 64-bit symbol table entry structure for Mach-O, and the symbol table is an array of nlist_64 entries. nlist_64 has an 8 byte member, so it should be 8-byte aligned, but we’ve seen object files where the symbol table only has a 4-byte

Adding Juergen to the thread. On Fri, Mar 15, 2019, 12:20 PM Shoaib Meenai via llvm-dev < llvm-dev at lists.llvm.org> wrote: > Ping. > > > > Would anyone be opposed to me trying to revive the in-tree Mach-O TAPI > code? > > > > Thanks, > > Shoaib > > > > *From: *Shoaib Meenai <smeenai at fb.com> > *Date: *Wednesday, March 6, 2019 at

Hi all, It looks like Juergen landed the TextAPI libraries for Mach-O in https://reviews.llvm.org/D53945, and also had several follow-ups, but then it was deleted entirely by r347874. I'm not sure if the entire deletion was intentional, since the commit message only mentions reverting a certain series of patches; CC Hans. In any case, I'm wondering if there are plans to revive it, since

I believe what's happening here is that clang translates the -fuse-ld=lld into calling the ld.lld executable, which is actually the ELF LLD linker, not the Mach-O one. On 6.0, the Mach-O linker symlink is called ld64.lld instead (and clang has been changed to call out to that name) to disambiguate the two. For 5.0, I'm not sure how best to force the Mach-O linker (I'm not familiar with

We're planning to contribute an LLVM version of Apple's libtool utility. To quote its man page, "Libtool with -static is intended to replace ar(5) and ranlib". In other words, libtool is the preferred tool to create archives for Apple platforms, and it also has good defaults for that platform (e.g. automatically creating a table of contents with the right format). It provides a

Hi. I'm using LLVM 5.0.1 on macOS 10.12. I have a very simple program (program.c): int main() {} When attempting to compile with LLD, I get this output: $ clang -fuse-ld=lld program.c /opt/llvm/5.0.1/bin/ld.lld: error: unknown argument: -no_deduplicate /opt/llvm/5.0.1/bin/ld.lld: error: unknown argument: -dynamic /opt/llvm/5.0.1/bin/ld.lld: error: unknown argument: -arch

I’m happy to run them, although I’d appreciate a pointer to the appropriate documentation. From: llvm-dev <llvm-dev-bounces at lists.llvm.org> on behalf of Tobias Hieta via llvm-dev <llvm-dev at lists.llvm.org> Reply-To: Tobias Hieta <tobias at plexapp.com> Date: Wednesday, September 30, 2020 at 11:15 AM To: Hans Wennborg <hans at chromium.org> Cc: llvm-dev <llvm-dev at

Actually, sorry, my fix was for the case where you had other arguments beginning with @ that weren't response files, whereas yours has actual response files, so my patch won't help there. CCing Reid and Hans, who did a bunch of the implementation in this area. From: llvm-dev <llvm-dev-bounces at lists.llvm.org> on behalf of Shoaib Meenai via llvm-dev <llvm-dev at

On Tue, Jun 6, 2017 at 11:14 PM, Michael Clark via llvm-dev < llvm-dev at lists.llvm.org> wrote: > OK. I see that the Mach-O linker is not even built when LLD is enabled in > Release_40, only the PE/COFF and ELF linkers are built. > > From looking at reviews it appears that Clang was able to be linked with > LLD on Darwin about 2 years ago, so Mach-O support seems to have

I'm an idiot and sent to llvm-commits instead of llvm-dev. Fixing. On 10/17/17, 5:09 PM, "llvm-commits on behalf of Shoaib Meenai via llvm-commits" <llvm-commits-bounces at lists.llvm.org on behalf of llvm-commits at lists.llvm.org> wrote: Hi all (CC beanz for cmake advice), I'm running into a cmake problem when I try to cross-compile a

I discovered that I can hack around my particular problem by placing set_property(TARGET clang PROPERTY INTERFACE_LINK_LIBRARIES) at the end of tools/driver/CMakeLists.txt. I'd prefer to fix this properly though, of course. On 12/1/17, 4:18 PM, "llvm-dev on behalf of Shoaib Meenai via llvm-dev" <llvm-dev-bounces at lists.llvm.org on behalf of llvm-dev at lists.llvm.org>

https://reviews.llvm.org/rL353906 introduced (presumably accidentally) the "b" top-level directory to the monorepo. The files should be moved to their proper location, but I'm also wondering if there's any way to prevent accidental top-level additions like this. Thanks, Shoaib -------------- next part -------------- An HTML attachment was scrubbed... URL:

Judging by this path: needed by 'tools/lld/Common/VCSVersion.inc' It looks to me like this is **not** a monorepo layout (if it were, lld would not appear in the tools directory). Therefore the LLVM_ENABLE_PROJECTS=lld is not even doing anything. I don't know how to build without a monorepo these days, and I also don't know what the most recent guidance setting up a monorepo is,

I have made some further investigation. My conclusion is that GNU ld does not do better than lld. Making the --defsym behavior ideal is difficult in the current framework. GNU ld has some unintended behaviors. ld.bfd a.o --defsym 'd=foo' --gc-sections -o a => GNU ld retains .text_foo ld.bfd a.o --defsym 'd=foo+3' --gc-sections -o a => GNU ld drops .text_foo ld.bfd a.o

Hi Sjoerd, On Fri, May 3, 2019 at 8:19 AM Sjoerd Meijer via llvm-dev < llvm-dev at lists.llvm.org> wrote: > > I read that out-of-order cores are supported. How about in-order cores? > Would it be easy/difficult to add support for that? > > Cheers, > Sjoerd. > > I don't think that it would be difficult to support in-order superscalar cores. However, it would

This is a silly question, but I am in a situation where I need to build x86 and arm assembly sources for some sources while the rest will be built with C. I do know that just adding `-fembed-bitcode` to a C sources would embed bitcode, but doing the same for the assembly files will not do that (at least, it will add the 1-byte `_LLVM,__asm` section, but not the `__LLVM,__bitcode` section).

Hi, I recently hit this on a project using a build system that relies heavily on nested response files. We found we could only have 21 response files total before getting errors related to the unexpanded response files. I tracked it down to this code in llvm/lib/Support/CommandLine.cpp // If we have too many response files, leave some unexpanded. This avoids // crashing on