This reverts commit ef82063207.
- It conflicts with the existing llvm::size in STLExtras, which will now
never be called.
- Calling it without llvm:: breaks C++17 compat
To fix
../../chromeclang/bin/../include/c++/v1/__algorithm/min.h:39:1: note: candidate template ignored: deduced conflicting types for parameter '_Tp' ('unsigned long' vs. 'unsigned long long')
on macOS arm64.
In start-end.s there is a lit check line `# SEG: _main` to begin the
check at the start of the function main where `_main` is the Darwin name
mangling for C main. Because the text file that FileCheck is getting as
input has the path of the compiler build in it from llvm-mc and
llvm-objdump, and because of the lack of a trailing colon in this check
line we end up inadvertently matching against the line of text with the
compiler path in it in the case where said path contains "_main" some
place. This can be very likely if the compiler branch has "main" or
"_main" in it.
To fix this I include the training : since that will match on the
function label and not the path line.
When linking a Debug build clang (265MiB SHF_ALLOC sections, 920MiB uncompressed
debug info), in a --threads=1 link "Compress debug sections" takes 2/3 time and
in a --threads=8 link "Compress debug sections" takes ~70% time.
This patch splits a section into 1MiB shards and calls zlib `deflake` parallelly.
DEFLATE blocks are a bit sequence. We need to ensure every shard starts
at a byte boundary for concatenation. We use Z_SYNC_FLUSH for all shards
but the last to flush the output to a byte boundary. (Z_FULL_FLUSH can
be used as well, but Z_FULL_FLUSH clears the hash table which just
wastes time.)
The last block requires the BFINAL flag. We call deflate with Z_FINISH
to set the flag as well as flush the output to a byte boundary. Under
the hood, all of Z_SYNC_FLUSH, Z_FULL_FLUSH, and Z_FINISH emit a
non-compressed block (called stored block in zlib). RFC1951 says "Any
bits of input up to the next byte boundary are ignored."
In a --threads=8 link, "Compress debug sections" is 5.7x as fast and the total
speed is 2.54x. Because the hash table for one shard is not shared with the next
shard, the output is slightly larger. Better compression ratio can be achieved
by preloading the window size from the previous shard as dictionary
(`deflateSetDictionary`), but that is overkill.
```
# 1MiB shards
% bloaty clang.new -- clang.old
FILE SIZE VM SIZE
-------------- --------------
+0.3% +129Ki [ = ] 0 .debug_str
+0.1% +105Ki [ = ] 0 .debug_info
+0.3% +101Ki [ = ] 0 .debug_line
+0.2% +2.66Ki [ = ] 0 .debug_abbrev
+0.0% +1.19Ki [ = ] 0 .debug_ranges
+0.1% +341Ki [ = ] 0 TOTAL
# 2MiB shards
% bloaty clang.new -- clang.old
FILE SIZE VM SIZE
-------------- --------------
+0.2% +74.2Ki [ = ] 0 .debug_line
+0.1% +72.3Ki [ = ] 0 .debug_str
+0.0% +69.9Ki [ = ] 0 .debug_info
+0.1% +976 [ = ] 0 .debug_abbrev
+0.0% +882 [ = ] 0 .debug_ranges
+0.0% +218Ki [ = ] 0 TOTAL
```
Bonus in not using zlib::compress
* we can compress a debug section larger than 4GiB
* peak memory usage is lower because for most shards the output size is less
than 50% input size (all less than 55% for a large binary I tested, but
decreasing the initial output size does not decrease memory usage)
Reviewed By: ikudrin
Differential Revision: https://reviews.llvm.org/D117853
This is in preparation for moving the code that parses and processes
order files into this file.
See https://reviews.llvm.org/D117354 for context and discussion.
Notation: dst is `t->getThunkTargetSym()->getVA()`
On AArch64, when `src-0x8000000-r_addend <= dst < src-0x8000000`, the condition
`target->inBranchRange(rel.type, src, rel.sym->getVA(rel.addend))` may
incorrectly consider a thunk reusable.
`rel.addend = -getPCBias(rel.type)` resets the addend to 0 for AArch64/PPC
and the zero addend is used by `rel.sym->getVA(rel.addend)` to check
out-of-range relocations.
See the test for a case this computation is wrong:
`error: a.o:(.text_high+0x4): relocation R_AARCH64_JUMP26 out of range: -134217732 is not in [-134217728, 134217727]`
I have seen a real world case with r_addend=19960.
Reviewed By: peter.smith
Differential Revision: https://reviews.llvm.org/D117734
Only using that change in StringRef already decreases the number of
preoprocessed lines from 7837621 to 7776151 for LLVMSupport
Perhaps more interestingly, it shows that many files were relying on the
inclusion of StringRef.h to have the declaration from STLExtras.h. This
patch tries hard to patch relevant part of llvm-project impacted by this
hidden dependency removal.
Potential impact:
- "llvm/ADT/StringRef.h" no longer includes <memory>,
"llvm/ADT/Optional.h" nor "llvm/ADT/STLExtras.h"
Related Discourse thread:
https://llvm.discourse.group/t/include-what-you-use-include-cleanup/5831
D54759 introduced aarch64-combined-dynrel.s and
aarch64-combined-dynrel-ifunc.s . Unfortunately the requires line
at the top was AArch64 instead of aarch64 which means they were never
run. Update the tests to use aarch64 and fix to match current lld output.
Differential Revision: https://reviews.llvm.org/D117896
Unresolved symbols are not currently reported when building with
`-shared` or `-pie` so setting unresolvedSymbols doesn't have any
effect.
Differential Revision: https://reviews.llvm.org/D117737
Its defaulting logic must go after `project(..)` to work correctly, but `project(..)` is often in a standalone condition making this
awkward, since the rest of the condition code may also need GNUInstallDirs.
The good thing is there are the various standalone booleans, which I had missed before. This makes splitting the conditional blocks less awkward.
Reviewed By: arichardson, phosek, beanz, ldionne, #libunwind, #libc, #libc_abi
Differential Revision: https://reviews.llvm.org/D117639
StringRefZ does not improve performance. Non-local symbols always have eagerly
computed nameSize. Most local symbols's lengths will be updated in either:
* shouldKeepInSymtab
* SymbolTableBaseSection::addSymbol
Its benefit is offsetted by strlen in every call site (sums up to 5KiB code in a
release x86-64 build), so using StringRefZ may be slower.
In a -s link (uncommon) there is minor speedup, like ~0.3% for clang and chrome.
Reviewed By: alexander-shaposhnikov
Differential Revision: https://reviews.llvm.org/D117644
This patch writes the full -cc1 command into the resulting .OBJ, like MSVC does. This allows for external tools (Recode, Live++) to rebuild a source file without any external dependency but the .OBJ itself (other than the compiler) and without knowledge of the build system.
The LF_BUILDINFO record stores a full path to the compiler, the PWD (CWD at program startup), a relative or absolute path to the source, and the full CC1 command line. The stored command line is self-standing (does not depend on the environment). In the same way, MSVC doesn't exactly store the provided command-line, but an expanded version (a somehow equivalent of CC1) which is also self-standing.
For more information see PR36198 and D43002.
Differential Revision: https://reviews.llvm.org/D80833
If you're building this on macOS 12.x+ this produces a deprecation
warning. I'm not sure what this means for the bitcode format going
forward, but it seems safe to silence for now.
Do we need to worry about GCC for this?
Differential Revision: https://reviews.llvm.org/D117718
This flag is the default, so in ld64 it is not implemented, but it can
be useful to negate previous -all_load arguments. Specifically if your
build system has some global linker flags, that you may want to negate
for specific links. We use something like this today to make sure some
C++ symbols are automatically discovered for all links, which passing
-all_load hides.
Differential Revision: https://reviews.llvm.org/D117629
Fix a regression after aabe901d57 (`[ELF] Remove
one redundant computeBinding`): isLocal() does not indicate that the symbol is
originally local. For simplicity, just drop this optimization.
In ld.lld, when an ObjFile/BitcodeFile is read in --start-lib state, the file is
given archive semantics. --end-lib closes the previous --start-lib. A build
system can use this feature as an alternative to archives. This patch ports
the feature to lld-macho.
--start-lib and --end-lib are positional, unlike usual ld64 options.
I think the slight drawback does not matter as (a) reusing option names
make build systems convenient (b) `--start-lib a.o b.o --end-lib` conveys more
information than an alternative design: `-objlib a.o -objlib b.o` because
--start-lib makes it clear which objects are in the same conceptual archive.
This provides flexibility (c) `-objlib`/`-filelist` interaction may be weird.
Close https://github.com/llvm/llvm-project/issues/52931
Reviewed By: #lld-macho, Jez Ng, oontvoo
Differential Revision: https://reviews.llvm.org/D116913
When an archive with an empty index contains only bitcode files, it is
handled as a group of lazy (--start-lib) object files. If there is a
non-bitcode file, there will be a diagnostic a la GNU ld.
For some programs, the archive member extraction ratio is high (e.g. for chrome,
79% archive members are extracted according to --print-archive-stats=). Because
symbol interning is cached for ObjFile::parseLazy but not for ArchiveFile,
parsing an archive as a group of --start-lib object files may be faster.
If the linker speculatively creates section representations for archive members,
the archive index will not be used.
If we take the above view, the archive index is essentially useless. If a user
wants a fast build without using --start-lib, they may just build thin archives
without index (`ar rcS --thin`).
Therefore, I suggest that we no longer treat the code as a hack, instead as a
supported feature. I believe we will do this anyway if we add parallel symbol
interning (parallel symbol interning for lazy object files is simpler than that
for archives).
Ecosystem issues:
* parseLazy actually has nearly the same behavior as ArchiveFile::parse, but the symbol order may be different.
* users may get addicted to the behavior and build archives not working with GNU ld and gold. I think it is easy to rebuild archives to be compatible.
Reviewed By: ikudrin
Differential Revision: https://reviews.llvm.org/D117284
D46245 added support for this in llvm-libtool, but while lld-link can
also create .lib files from .def files it didn't support aliases.
I compared the Inputs/library.def test against the output from
llvm-libtool and it matches, except for the fact that lld-link reorders
functions for some reason.
I have also verified that this fixes a bug I was running into while
trying to compile .def files to .lib files in MinGW-w64 (using lld-link
instead of llvm-libtool).
Differential Revision: https://reviews.llvm.org/D113365
This simplifies the code a bit. While here,
* change the `multiple relocation sections` diagnostic from `fatal` to `error` and include the relocated section name.
* drop less useful name from `getRelocTarget`. Without -r/--emit-relocs we don't need to get SHT_REL/SHT_RELA names.
It's still uncertain but whether we want to have `deduplicate-literals` be the
default flag for LLD out of the box or not. If `deduplicate-literals` is the default
behavior, then we will need a way override it and not deduplicate. Luckily, we
have `no_deduplicate` to fill this gap. For now, I've set the default to be false
which aligns with the existing behavior. That can only always be changed after
discussions on D117250.
Reviewed By: #lld-macho, int3
Differential Revision: https://reviews.llvm.org/D117387
We already perform memory initialization and apply global relocations
during start. It makes sense to performs data relocations too. I think
the reason we were not doing this already is solely historical.
Differential Revision: https://reviews.llvm.org/D117412
After the D33630 fallout was properly fixed by a4c5db30be.
Tested by D37462/D44986 tests, the new --no-rosegment test in build-id.s, and a few --rosegment/--no-rosegment programs.
Move all variables at file-scope or function-static-scope into a hosting structure (lld::CommonLinkerContext) that lives at lldMain()-scope. Drivers will inherit from this structure and add their own global state, in the same way as for the existing COFFLinkerContext.
See discussion in https://lists.llvm.org/pipermail/llvm-dev/2021-June/151184.html
Differential Revision: https://reviews.llvm.org/D108850
Sorting dynamic relocations is a bottleneck. Simplifying the comparator improves
performance. Linking clang is 4~5% faster with --threads=8.
This change may shuffle R_MIPS_REL32 for Mips and is a NFC for non-Mips.
Tail merge is slow and of low value. With regular string deduplication, we can
just use the return value of StringTableBuilder::add.
There is no noticeable performance increase because without deduplication
`__cstring` is quite small (7.6MiB for chromium_framework).
Reviewed By: #lld-macho, Jez Ng
Differential Revision: https://reviews.llvm.org/D117273
The PlatformKind/PlatformType enums contain the same information, which requires
them to be kept in-sync. This commit changes over to PlatformType as the sole
source of truth, which allows the removal of the redundant PlatformKind.
The majority of the changes were in LLD and TextAPI.
Reviewed By: cishida
Differential Revision: https://reviews.llvm.org/D117163
This extends D81784. Sections can be discarded when linking a
relocatable output. Before the patch, LLD did not update the content
of debug sections and only replaced the corresponding relocations with
R_*_NONE, which could break the debug information.
Differential Revision: https://reviews.llvm.org/D116946
db08df0570 does not work because part.relrDyn is
a unique_ptr and `reset` destroys the object which may still be referenced.
This commit uses the D114180 approach. Also improve the test to check that there
is no R_X86_64_RELATIVE.
Fix https://github.com/llvm/llvm-project/issues/53073
In case of a relocation error, GNU ld's link map includes
the archive member extraction information but not output sections.
Our -Map and --why-extract= are currently no-op in case of an error.
This change makes the two options work.
Reviewed By: ikudrin, peter.smith
Differential Revision: https://reviews.llvm.org/D116838
to prepare for D116838, otherwise for linkerscript/discard-section-err.s,
there will be a null pointer dereference in `part.relrDyn->getParent()->size`
in `finalizeSynthetic(part.relrDyn.get())`.
Depends on D112160
This adds the new options `--call-graph-profile-sort` (default),
`--no-call-graph-profile-sort` and `--print-symbol-order=`. If call graph
profile sorting is enabled, reads `__LLVM,__cg_profile` sections from object
files and uses the resulting graph to put callees and callers close to each
other in the final binary via the C3 clustering heuristic.
Differential Revision: https://reviews.llvm.org/D112164
rG1bb0caf56168 changed the datalayout of f80 on Windows 32 bits. But it
missed the related use in the LLD tests. This patch will fix the
problem catched by buildbot.
After {D117069}, map-file.s seems flaky. It seems that the "Total Write
map file" section always exists, but the "Write map file" sub-section
may or may not be emitted. So we check for the former.
After {D115416}, the "Write map file" event no longer shows up
in the time trace. Each time trace profiler instance is thread-local,
but we had neglected to initialize a separate instance for the mapfile
worker thread.
Reviewed By: keith
Differential Revision: https://reviews.llvm.org/D117069
D116913 will add LazyObject. Rename LazySymbol to LazyArchive to avoid confusion
and mirror ELF.
Reviewed By: #lld-macho, Jez Ng
Differential Revision: https://reviews.llvm.org/D116914
Currently the way some relocation-related static functions pass around
states is clumsy. Add a Resolver class to store some states as member
variables.
Advantages:
* Avoid the parameter `InputSectionBase &sec` (this offsets the cost passing around `this` paramemter)
* Avoid the parameter `end` (Mips and PowerPC hacks)
* `config` and `target` can be cached as member variables to reduce global state accesses. (potential speedup because the compiler didn't know `config`/`target` were not changed across function calls)
* If we ever want to reduce if-else costs (e.g. `config->emachine==EM_MIPS` for non-Mips) or introduce parallel relocation scan not handling some tricky arches (PPC/Mips), we can templatize Resolver
`target` isn't used as much as `config`, so I change it to a const reference
during the migration.
There is a minor performance inprovement for elf::scanRelocations.
Reviewed By: ikudrin, peter.smith
Differential Revision: https://reviews.llvm.org/D116881
When `file->fetch(sym)` is replaced with a no-op, no test fails.
The new test catches the case.
Reviewed By: #lld-macho, oontvoo
Differential Revision: https://reviews.llvm.org/D116916
We only support both TLSDESC and TLS GD for x86 so this is an x86-specific
problem. If both are used, only one R_X86_64_TLSDESC is produced and TLS GD
accesses will incorrectly reference R_X86_64_TLSDESC. Fix this by introducing
SymbolAux::tlsDescIdx.
Reviewed By: ikudrin
Differential Revision: https://reviews.llvm.org/D116900
to decrease sizeof(SymbolUnion) by 8 on ELF64 platforms.
Symbols needing such information are typically 1% or fewer (5134 out of 560520
when linking clang, 19898 out of 5550705 when linking chrome). Storing them
elsewhere can decrease memory usage and symbol initialization time.
There is a ~0.8% saving on max RSS when linking a large program.
Future direction:
* Move some of dynsymIndex/verdefIndex/versionId to SymbolAux
* Support mixed TLSDESC and TLS GD without increasing sizeof(SymbolUnion)
Reviewed By: peter.smith
Differential Revision: https://reviews.llvm.org/D116281
This is continuing in the path of D51714, which did this for Clang.
I have rearranged the source code Clang so one can diff the top-level
CMakeLists.txt of Clang and LLD, ensuring we use the same strategy for
both.
Besides diffing the two files, `git diff --color-moved` on LLD also helps review.
Reviewed By: beanz
Differential Revision: https://reviews.llvm.org/D116492
See the docs in the new function for details.
I think I found every instance of this copy pasted code. Polly could
also use it, but currently does something different, so I will save the
behavior change for a future revision.
We get the shared, non-installed CMake modules following the pattern
established in D116472.
It might be good to have LLD and Flang also use this, but that would be
a functional change and so I leave it as future work.
Reviewed By: beanz, lebedev.ri
Differential Revision: https://reviews.llvm.org/D116521
Previously compounding was all-or-nothing. Now, the
compounding attempts will iterate and yield the most
compounds that still result in a valid packet.
One of our internal arm64 apps hit a thunk out of range error when building
with LLD. Per the comment, I'm arbitrarily increasing slop size to 256.
Reviewed By: #lld-macho, thakis
Differential Revision: https://reviews.llvm.org/D116705
Fangrui Song