NFC, just makes it easier to access from non templated code.

llvm-svn: 309152

Was discussed in D35797 thread.

llvm-svn: 309096

Previously we handled this option implicitly, only
for infering unresolved symbols handling policy.

ld man says: "--noinhibit-exec Retain the executable
output file whenever it is still usable",
and we may want to handle other cases too.

Differential revision: https://reviews.llvm.org/D35793

llvm-svn: 309091

This is PR30422,
previously LLD did not render all option aliases in --help.
With this patch it will.

Differential revision: https://reviews.llvm.org/D35477

llvm-svn: 309089

Since the flag is ignored anyway, it doesn't matter whether
it is an alias or not.

llvm-svn: 309055

llvm-svn: 309053

llvm-svn: 309051

These can be referenced with __start_/__stop_ symbols. I will try to make
this more precise in a followup patch.

llvm-svn: 309048

Differential Revision: https://reviews.llvm.org/D35768

llvm-svn: 309017

The test used /manifestinput: without /manifest:embed, which isn't actually
supported.  Just remove this part of the test for now; if it's important to
check this the llvm-readobj part should be extended to check this.

llvm-svn: 309002

Also emit an error if /manifestinput: is used without /manifest:embed.
Increases compatibility with link.exe

https://reviews.llvm.org/D35842

llvm-svn: 308998

Previously .init_array/.fini_array sections
were not unique and we had 3 .init_array sections +
3 .fini_array input sections passed to linker,
instead of 5 + 5.

Differential revision: https://reviews.llvm.org/D35552

llvm-svn: 308958

This is PR33714.

Previously for each input section offset of memory region
was incremented on a size of output section.

That resulted in a wrong error message saying about
overflow. Patch fixes that.

Differential revision: https://reviews.llvm.org/D35803

llvm-svn: 308955

The same adjustment is already done for the entry point in
Writer.cpp and for relocations that point to executable code
in Chunks.cpp.

Differential Revision: https://reviews.llvm.org/D35767

llvm-svn: 308953

This is a test for LLVM SVN r308951.

Differential Revision: https://reviews.llvm.org/D35814

llvm-svn: 308952

llvm-svn: 308939

Summary:
This fixes test/mach-o/objc-image-info-pass-output.yaml on
big-endian hosts.

Reviewers: lhames, kledzik, ruiu

Reviewed By: ruiu

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D35052

llvm-svn: 308935

If user requested section alignment is greater than MaxPageSize, propagate it to segment headers correctly.

Differential Revision: https://reviews.llvm.org/D35813

llvm-svn: 308930

Thanks to Rafael for finding it in a post-commit review.

llvm-svn: 308921

Follow-up to r308915 suggested by Rafael and Peter.

llvm-svn: 308920

llvm-svn: 308916

We do this emitting a section for every function when LTO is used.
Fixes PR33888.

Differential Revision:  https://reviews.llvm.org/D35809

llvm-svn: 308915

llvm-svn: 308812

llvm-svn: 308809

llvm-svn: 308775

Eq helper allows to define `XXX` and `XXX=` options forms easily.
Patch adds testcases for few aliases.

Differential revision: https://reviews.llvm.org/D35619

llvm-svn: 308752

Differential Revision: https://reviews.llvm.org/D35537

llvm-svn: 308728

This method is no longer being used anywhere.

Differential Revision: https://reviews.llvm.org/D35706

llvm-svn: 308706

Also add the test cases for the addition and subtraction both for
the relative and absolute case.

Differential Revision: https://reviews.llvm.org/D35346

llvm-svn: 308692

the dyld compact export entries in libObject.

llvm-svn: 308691

This dependency was introduced in r308681.

llvm-svn: 308686

We were not looking at Repl and so thinking there was no output
section associated with the merged symbol. Because of that it was
produced as absolute.

This was found by an internal round of testing.

llvm-svn: 308681

SHF_GROUP flag should have been removed when the control reaches here
because InputSectionBase turns the flag off. So this code should be
redundant.

llvm-svn: 308680

Summary:
If the linker is invoked with `--chroot /foo` and `/bar/baz.o`, it
tries to read the file from `/foo/bar/baz.o`. This feature is useful
when you are dealing with files created by the --reproduce option.

Reviewers: grimar

Subscribers: llvm-commits, emaste

Differential Revision: https://reviews.llvm.org/D35517

llvm-svn: 308646

It was fixed by r308544.

llvm-svn: 308634

Also extend the tests for IMAGE_REL_ARM64_PAGEOFFSET_12L to test
all 8/16/32/64 bit GPR and 8/16/32/64/128 SIMD/FP bit ldr/str variants,
and a ldr with an existing offset.

Differential revision: https://reviews.llvm.org/D35647

llvm-svn: 308631

Fix issues found in existing code, while reviewing other changes.

Change the data type of a variable to uint32_t, to avoid potential issues
with signedness in shifts.

Differential revision: https://reviews.llvm.org/D35646

llvm-svn: 308586

This fixes cases on ARM64 when importing from more than one DLL,
in case the imports from the first DLL ended up unaligned.

When fixing up a IMAGE_REL_ARM64_PAGEOFFSET_12L, which shifts the
offset by the load/store size, check that the shift doesn't discard
any bits. (This would also detect if the import address chunks were
unaligned.)

Differential revision: https://reviews.llvm.org/D35640

llvm-svn: 308585

With that in place we can use lld's own infrastructure for the low
level detail of dwarf parsing.

With this we don't decompress sections twice, we don't scan all
realocations and even with this simplistic implementation linking
clang with gdb index goes from 34.09 seconds to 20.80 seconds.

llvm-svn: 308544

llvm-svn: 308529