This diff is motivated by my work to add proper DWARF unwind support. As
detailed in PR50956 functions that need DWARF unwind need to have
compact unwind entries synthesized for them. These CU entries encode an
offset within `__eh_frame` that points to the corresponding DWARF FDE.
In order to encode this offset during
`UnwindInfoSectionImpl::finalize()`, we need to first assign values to
`InputSection::outSecOff` for each `__eh_frame` subsection. But
`__eh_frame` is ordered after `__unwind_info` (according to ld64 at
least), which puts us in a bit of a bind: `outSecOff` gets assigned
during finalization, but `__eh_frame` is being finalized after
`__unwind_info`.
But it occurred to me that there's no real need for most
ConcatOutputSections to be finalized sequentially. It's only necessary
for text-containing ConcatOutputSections that may contain branch relocs
which may need thunks. ConcatOutputSections containing other types of
data can be finalized in any order.
This diff moves the finalization logic for non-text sections into a
separate `finalizeContents()` method. This method is called before
section address assignment & unwind info finalization takes place. In
theory we could call these `finalizeContents()` methods in parallel, but
in practice it seems to be faster to do it all on the main thread.
Reviewed By: #lld-macho, oontvoo
Differential Revision: https://reviews.llvm.org/D123279
`config->priorities` has been used to hold the intermediate state during the construction of the order in which sections should be laid out. This is not a good place to hold this state since the intermediate state is not a "configuration" for LLD. It should be encapsulated in a class for building a mapping from section to priority (which I created in this diff as the `PriorityBuilder` class).
The same thing is being done for `config->callGraphProfile`.
Reviewed By: #lld-macho, int3
Differential Revision: https://reviews.llvm.org/D122156
All references to interposable symbols can be redirected at runtime to
point to a different symbol definition (with the same name). For
example, if both dylib A and B define symbol _foo, and we load A before
B at runtime, then all references to _foo within dylib B will point to
the definition in dylib A.
ld64 makes all extern symbols interposable when linking with
`-flat_namespace`.
TODO 1: Support `-interposable` and `-interposable_list`, which should
just be a matter of parsing those CLI flags and setting the
`Defined::interposable` bit.
TODO 2: Set Reloc::FinalDefinitionInLinkageUnit correctly with this info
(we are currently not setting it at all, so we're erring on the
conservative side, but we should help the LTO backend generate more
optimal code.)
Reviewed By: modimo, MaskRay
Differential Revision: https://reviews.llvm.org/D119294
Earlier in LLD's evolution, I tried to create the illusion that
subsections were indistinguishable from "top-level" sections. Thus, even
though the subsections shared many common field values, I hid those
common values away in a private Shared struct (see D105305). More
recently, however, @gkm added a public `Section` struct in D113241 that
served as an explicit way to store values that are common to an entire
set of subsections (aka InputSections). Now that we have another "common
value" struct, `Shared` has been rendered redundant. All its fields can
be moved into `Section` instead, and the pointer to `Shared` can be replaced
with a pointer to `Section`.
This `Section` pointer also has the advantage of letting us inspect other
subsections easily, simplifying the implementation of {D118798}.
P.S. I do think that having both `Section` and `InputSection` makes for
a slightly confusing naming scheme. I considered renaming `InputSection`
to `Subsection`, but that would break the symmetry with `OutputSection`.
It would also make us deviate from LLD-ELF's naming scheme.
This change is perf-neutral on my 3.2 GHz 16-Core Intel Xeon W machine:
base diff difference (95% CI)
sys_time 1.258 ± 0.031 1.248 ± 0.023 [ -1.6% .. +0.1%]
user_time 3.659 ± 0.047 3.658 ± 0.041 [ -0.5% .. +0.4%]
wall_time 4.640 ± 0.085 4.625 ± 0.063 [ -1.0% .. +0.3%]
samples 49 61
There's also no stat sig change in RSS (as measured by `time -l`):
base diff difference (95% CI)
time 998038627.097 ± 13567305.958 1003327715.556 ± 15210451.236 [ -0.2% .. +1.2%]
samples 31 36
Reviewed By: #lld-macho, oontvoo
Differential Revision: https://reviews.llvm.org/D118797
Added some comments (particularly around finalize() and
finalizeContents()) as well as doing some rephrasing / grammar fixes for
existing comments.
Also did some minor style fixups, such as by putting methods together in
a class definition and having fields of similar types next to each
other.
Reviewed By: #lld-macho, oontvoo
Differential Revision: https://reviews.llvm.org/D118714
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.
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
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
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
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
References from thread-local variable sections are treated as offsets
relative to the start of the thread-local data memory area, which is
initialized via copying all the TLV data sections (which are all
contiguous). If later data sections require a greater alignment than
earlier ones, the offsets of data within those sections won't be
guaranteed to aligned unless we normalize alignments. We therefore use
the largest alignment for all TLV data sections.
Reviewed By: #lld-macho, int3
Differential Revision: https://reviews.llvm.org/D116263
For large applications that write to map files, writing map files can take quite
a bit of time. Sorting the biggest contributors to link times, writing map files
ranks in at 2nd place, with load input files being the biggest contributor of
link times. Avoiding writing map files on the critical path (and having its own
thread) saves ~2-3 seconds when linking chromium framework on a 16-Core
Intel Xeon W.
```
base diff difference (95% CI)
sys_time 1.617 ± 0.034 1.657 ± 0.026 [ +1.5% .. +3.5%]
user_time 28.536 ± 0.245 28.609 ± 0.180 [ -0.1% .. +0.7%]
wall_time 23.833 ± 0.271 21.684 ± 0.194 [ -9.5% .. -8.5%]
samples 31 24
```
Reviewed By: #lld-macho, oontvoo, int3
Differential Revision: https://reviews.llvm.org/D115416
This is an NFC diff that prepares for pruning & relocating `__eh_frame`.
Along the way, I made the following changes to ...
* clarify usage of `section` vs. `subsection`
* remove `map` & `vec` from type names
* disambiguate class `Section` from template parameter `SectionHeader`.
Differential Revision: https://reviews.llvm.org/D113241
We need to reset global state between runs, similar to the other ports.
There's some file-static state which needs to be reset as well and we
need to add some new helpers for that.
With this change, most LLD Mach-O tests pass with `LLD_IN_TEST=2` (which
runs the linker twice on each test). Some tests will be fixed by the
remainder of this stack, and the rest are fundamentally incompatible
with that mode (e.g. they intentionally throw fatal errors).
Fixes PR52070.
Reviewed By: #lld-macho, int3
Differential Revision: https://reviews.llvm.org/D112878
Having to remember to call `canonical()` all over the place is
error-prone; let's do it in a centralized location instead. It also
appears to improve performance slightly.
base diff difference (95% CI)
sys_time 0.984 ± 0.009 0.983 ± 0.014 [ -0.8% .. +0.6%]
user_time 6.508 ± 0.035 6.475 ± 0.036 [ -0.8% .. -0.2%]
wall_time 5.321 ± 0.034 5.300 ± 0.033 [ -0.7% .. -0.1%]
samples 36 23
Reviewed By: #lld-macho, thakis
Differential Revision: https://reviews.llvm.org/D112687
Otherwise tools like codesign_allocate will choke. We were already
handling this correctly for the other DYLD_INFO sections.
Doing this correctly is a bit subtle: we don't know if export_size will
be zero until we have run `ExportSection::finalizeContents()`. However,
we must still add the ExportSection to the `__LINKEDIT` segment in order
that it gets sorted during `sortSectionsAndSegments()`.
Reviewed By: #lld-macho, oontvoo
Differential Revision: https://reviews.llvm.org/D112589
This diff does away with `addEntriesForFunctionsWithoutUnwindInfo()`,
because `addSymbol()` can now determine which functions need those
entries.
While overhauling UnwindInfoSection, I also parallelized the relocation
of the contents of the CUEs. This somewhat offsets the time regression
from creating one InputSection per CUE (which was done in D109944).
Reviewed By: #lld-macho, oontvoo
Differential Revision: https://reviews.llvm.org/D109945
Compact unwind entries (CUEs) contain pointers to their respective
function symbols. However, during the link process, it's far more useful
to have pointers from the function symbol to the CUE than vice versa.
This diff adds that pointer in the form of `Defined::compactUnwind`.
In particular, when doing dead-stripping, we want to mark CUEs live when
their function symbol is live; and when doing ICF, we want to dedup
sections iff the symbols in that section have identical CUEs. In both
cases, we want to be able to locate the symbols within a given section,
as well as locate the CUEs belonging to those symbols. So this diff also
adds `InputSection::symbols`.
The ultimate goal of this refactor is to have ICF support dedup'ing
functions with unwind info, but that will be handled in subsequent
diffs. This diff focuses on simplifying `-dead_strip` --
`findFunctionsWithUnwindInfo` is no longer necessary, and
`Defined::isLive()` is now a lot simpler. Moreover, UnwindInfoSection no
longer has to check for dead CUEs -- we simply avoid adding them in the
first place.
Additionally, we now support stripping of dead LSDAs, which follows
quite naturally since `markLive()` can now reach them via the CUEs.
Reviewed By: #lld-macho, gkm
Differential Revision: https://reviews.llvm.org/D109944
We end up calling resolveBranchVA(), which asserts for Undefineds.
As fix, just return early in Writer::run() if there are any diagnostics
after processing relocations (which is where undefined symbol errors are
emitted). This matches what the ELF port does.
Differential Revision: https://reviews.llvm.org/D109079
These symbols are somewhat interesting in that they create non-existing
segments, which as far as I know is the only way to create segments
that don't contain any sections.
Final part of part of PR50760. Like D106629, but for segments instead
of sections. I'm not aware of anything that needs this in practice.
Differential Revision: https://reviews.llvm.org/D106767
With this, libclang_rt.profile_osx.a can be linked, that is coverage
and PGO-instrumented builds should now work with lld.
section$start and section$end symbols can create non-existing sections.
They're also undefined symbols that are only magic if there isn't a
regular symbol with their name, which means the need to be handled
in treatUndefined() instead of just looping over all existing
sections and adding start and end symbols like the ELF port does.
To represent the actual symbols, this uses absolute symbols that
get their value updated once an output section is layed out.
segment$start and segment$end are still missing for now, but they produce a
nicer error message after this patch.
Main part of PR50760.
Differential Revision: https://reviews.llvm.org/D106629
segment$start$/segment$end$ symbols allow creating segments without
sections, so getting the segment address off the first section
won't work there. Storing the address on the segment is arguably a
bit simpler too.
No behavior change, part of PR50760.
Differential Revision: https://reviews.llvm.org/D106665
Absolute symbols have a nullptr isec. buildInputSectionPriorities()
would defer isec, causing crashes. Ordering absolute symbols doesn't
make sense, so just ignore them. This seems to match ld64.
Differential Revision: https://reviews.llvm.org/D106628
We lost the `used` bit on the Undefined when we replaced it with a DylibSymbol
in treatUndefined().
Differential Revision: https://reviews.llvm.org/D106565
treatUndefinedSymbol() was previously called before gatherInputSections()
and markLive() for these special symbols, but after them for normal
undefineds.
For PR50760, treatUndefinedSymbol() will have to potentially create
sections, so it's good to move treatUndefinedSymbol() for special
undefineds later, so that it can assume that gatherInputSections()
and markLive() has already been called always.
No intended behavior change, but part of PR50760 (and covered in
tests in the patch for the full feature).
Differential Revision: https://reviews.llvm.org/D106552
ICF previously operated only within a given OutputSection. We would
merge all CFStrings first, then merge all regular code sections in a
second phase. This worked fine since CFStrings would never reference
regular `__text` sections. However, I would like to expand ICF to merge
functions that reference unwind info. Unwind info references the LSDA
section, which can in turn reference the `__text` section, so we cannot
perform ICF in phases.
In order to have ICF operate on InputSections spanning multiple
OutputSections, we need a way to distinguish InputSections that are
destined for different OutputSections, so that we don't fold across
section boundaries. We achieve this by creating OutputSections early,
and setting `InputSection::parent` to point to them. This is what
LLD-ELF does. (This change should also make it easier to implement the
`section$start$` symbols.)
This diff also folds InputSections w/o checking their flags, which I
think is the right behavior -- if they are destined for the same
OutputSection, they will have the same flags in the output (even if
their input flags differ). I.e. the `parent` pointer check subsumes the
`flags` check. In practice this has nearly no effect (ICF did not become
any more effective on chromium_framework).
I've also updated ICF.cpp's block comment to better reflect its current
status.
Reviewed By: #lld-macho, smeenai
Differential Revision: https://reviews.llvm.org/D105641
The mappings we were using had a small number of keys, so a vector is
probably better. This allows us to remove the last usage of std::map in
our codebase.
I also used `removeSimulator` to simplify the code a bit further.
Reviewed By: #lld-macho, thakis
Differential Revision: https://reviews.llvm.org/D105786
I think this is an old way for doing what is done with
-reexport_library these days, but it's e.g. still used in libunwind's
build (the opensource.apple.com one, not the llvm one).
Differential Revision: https://reviews.llvm.org/D105448
We have been creating many ConcatInputSections with identical values due
to .subsections_via_symbols. This diff factors out the identical values
into a Shared struct, to reduce memory consumption and make copying
cheaper.
I also changed `callSiteCount` from a uint32_t to a 31-bit field to save an
extra word.
All in all, this takes InputSection from 120 to 72 bytes (and
ConcatInputSection from 160 to 112 bytes), i.e. 30% size reduction in
ConcatInputSection.
Numbers for linking chromium_framework on my 3.2 GHz 16-Core Intel Xeon W:
N Min Max Median Avg Stddev
x 20 4.14 4.24 4.18 4.183 0.027548999
+ 20 4.04 4.11 4.075 4.0775 0.018027756
Difference at 95.0% confidence
-0.1055 +/- 0.0149005
-2.52211% +/- 0.356215%
(Student's t, pooled s = 0.0232803)
Reviewed By: #lld-macho, thakis
Differential Revision: https://reviews.llvm.org/D105305
This is a pretty big refactoring diff, so here are the motivations:
Previously, ICF ran after scanRelocations(), where we emitting
bind/rebase opcodes etc. So we had a bunch of redundant leftovers after
ICF. Having ICF run before Writer seems like a better design, and is
what LLD-ELF does, so this diff refactors it accordingly.
However, ICF had two dependencies on things occurring in Writer: 1) it
needs literals to be deduplicated beforehand and 2) it needs to know
which functions have unwind info, which was being handled by
`UnwindInfoSection::prepareRelocations()`.
In order to do literal deduplication earlier, we need to add literal
input sections to their corresponding output sections. So instead of
putting all input sections into the big `inputSections` vector, and then
filtering them by type later on, I've changed things so that literal
sections get added directly to their output sections during the 'gather'
phase. Likewise for compact unwind sections -- they get added directly
to the UnwindInfoSection now. This latter change is not strictly
necessary, but makes it easier for ICF to determine which functions have
unwind info.
Adding literal sections directly to their output sections means that we
can no longer determine `inputOrder` from iterating over
`inputSections`. Instead, we store that order explicitly on
InputSection. Bloating the size of InputSection for this purpose would
be unfortunate -- but LLD-ELF has already solved this problem: it reuses
`outSecOff` to store this order value.
One downside of this refactor is that we now make an additional pass
over the unwind info relocations to figure out which functions have
unwind info, since want to know that before `processRelocations()`. I've
made sure to run that extra loop only if ICF is enabled, so there should
be no overhead in non-optimizing runs of the linker.
The upside of all this is that the `inputSections` vector now contains
only ConcatInputSections that are destined for ConcatOutputSections, so
we can clean up a bunch of code that just existed to filter out other
elements from that vector.
I will test for the lack of redundant binds/rebases in the upcoming
cfstring deduplication diff. While binds/rebases can also happen in the
regular `.text` section, they're more common in `.data` sections, so it
seems more natural to test it that way.
This change is perf-neutral when linking chromium_framework.
Reviewed By: oontvoo
Differential Revision: https://reviews.llvm.org/D105044
Previously, we only applied the renames to
ConcatOutputSections.
Reviewed By: #lld-macho, thakis
Differential Revision: https://reviews.llvm.org/D105079
Fixes PR50637.
Downstream bug: https://crbug.com/1218958
Currently, we split __cstring along symbol boundaries with .subsections_via_symbols
when not deduplicating, and along null bytes when deduplicating. This change splits
along null bytes unconditionally, and preserves original alignment in the non-
deduplicated case.
Removing subsections-section-relocs.s because with this change, __cstring
is never reordered based on the order file.
Differential Revision: https://reviews.llvm.org/D104919
The two different thread_local_regular sections (__thread_data and
more_thread_data) had nondeterminstic ordering for two reasons:
1. https://reviews.llvm.org/D102972 changed concatOutputSections
from MapVector to DenseMap, so when we iterate it to make
output segments, we would add the two sections to the __DATA
output segment in nondeterministic order.
2. The same change also moved the two stable_sort()s for segments
and sections to sort(). Since sections with assigned priority
(such as TLV data) have the same priority for all sections,
this is incorrect -- we must use stable_sort() so that the
initial (input-order-based) order remains.
As a side effect, we now (deterministically) put the __common
section in front of __bss (while previously we happened to
put it after it). (__common and __bss are both zerofill so
both have order INT_MAX, but common symbols are added to
inputSections before normal sections are collected.)
Makes lld/test/MachO/tlv.s and lld/test/MachO/tlv-dylib.s pass with
LLVM_ENABLE_EXPENSIVE_CHECKS=ON.
Differential Revision: https://reviews.llvm.org/D105054
Literal sections can be deduplicated before running ICF. That makes it
easy to compare them during ICF: we can tell if two literals are
constant-equal by comparing their offsets in their OutputSection.
LLD-ELF takes a similar approach.
Reviewed By: #lld-macho, gkm
Differential Revision: https://reviews.llvm.org/D104671
`icfEqClass` only makes sense on ConcatInputSections since (in contrast
to literal sections) they are deduplicated as an atomic unit.
Similarly, `hasPersonality` and `replacement` don't make sense on
literal sections.
This mirrors LLD-ELF, which stores `icfEqClass` only on non-mergeable
sections.
Reviewed By: #lld-macho, gkm
Differential Revision: https://reviews.llvm.org/D104670
ICF = Identical C(ode|OMDAT) Folding
This is the LLD ELF/COFF algorithm, adapted for MachO. So far, only `-icf all` is supported. In order to support `-icf safe`, we will need to port address-significance tables (`.addrsig` directives) to MachO, which will come in later diffs.
`check-{llvm,clang,lld}` have 0 regressions for `lld -icf all` vs. baseline ld64.
We only run ICF on `__TEXT,__text` for reasons explained in the block comment in `ConcatOutputSection.cpp`.
Here is the perf impact for linking `chromium_framekwork` on a Mac Pro (16-core Xeon W) for the non-ICF case vs. pre-ICF:
```
N Min Max Median Avg Stddev
x 20 4.27 4.44 4.34 4.349 0.043029977
+ 20 4.37 4.46 4.405 4.4115 0.025188761
Difference at 95.0% confidence
0.0625 +/- 0.0225658
1.43711% +/- 0.518873%
(Student's t, pooled s = 0.0352566)
```
Reviewed By: #lld-macho, int3
Differential Revision: https://reviews.llvm.org/D103292
I *think* this is the fix, with the regression being introduced by
D104199. Not 100% sure since MSAN isn't supported on my Mac machine, and
it'll take some time to spin up a Linux box... will look at the
buildbots for answers
I wanted to see if we would get any perf wins out of this, but
it doesn't seem to be the case. But it still seems worth committing.
Reviewed By: MaskRay
Differential Revision: https://reviews.llvm.org/D104200
We don't need to define any special behavior for this section,
so creating a subclass for it is redundant.
Reviewed By: #lld-macho, thakis
Differential Revision: https://reviews.llvm.org/D104199
`outSecFileOff` and the associated `getFileOffset()` accessors were
unnecessary.
For all the cases we care about, `outSecFileOff` is the same as
`outSecOff`. The only time they deviate is if there are zerofill
sections within a given segment. But since zerofill sections are always
at the end of a segment, the only sections where the two values deviate
are zerofill sections themselves. And we never actually query the
outSecFileOff of zerofill sections.
As for `getFileOffset()`, the only place it was being used was to
calculate the offset of the entry symbol. However, we can compute that
value by just taking the difference between the address of the entry
symbol and the address of the Mach-O header. In fact, this appears to be
what ld64 itself does. This difference is the same as the file offset as
long as there are no intervening zerofill sections, but since `__text`
is the first section in `__TEXT`, this never happens, so our previous
use of `getFileOffset()` was not wrong -- just inefficient.
Reviewed By: #lld-macho, thakis
Differential Revision: https://reviews.llvm.org/D104177
Jez Ng