The R2 save stub will now support offsets up to 64 bits.
There are three cases that will be used.
1) The offset fits in 26 bits.
```
b <26 bit offset>
```
2) The offset does not fit in 26 bits but fits in 34 bits.
```
paddi r12, 0, <34 bit offset>, 1
mtctr r12
bctr
```
3) The offset does not fit in 34 bits. Since this is an R2 save stub we can use
the TOC in R2. We are not loading the offset but the actual address we want to
branch to.
```
addis r12, r2, <address in TOC lo>
ld r12 <address in TOC hi>(r12)
mtctr r12
bctr
```
In case 1) the stub is only 8 bytes while in cases 2) and 3) the stub will be
20 bytes.
Reviewed By: MaskRay, sfertile, NeHuang
Differential Revision: https://reviews.llvm.org/D87916
Library users should not need to call errorHandler().reset() explicitly.
google/iree calls lld:🧝:link and without the patch some global
variables are not cleaned up in the next invocation.
".text.split." holds symbols which are split out from functions in
other input sections. For example, with -fsplit-machine-functions,
placing the cold parts in .text.split instead of .text.unlikely mitigates
against poor profile inaccuracy. Techniques such as hugepage remapping can
make conservative decisions at the section granularity.
Differential Revision: https://reviews.llvm.org/D87840
In lit tests, we run each LLD invocation twice (LLD_IN_TEST=2), without shutting down the process in-between. This ensures a full cleanup is properly done between runs.
Only active for the COFF driver for now. Other drivers still use LLD_IN_TEST=1 which executes just one iteration with full cleanup, like before.
When the environment variable LLD_IN_TEST is unset, a shortcut is taken, only one iteration is executed, no cleanup for faster exit, like before.
A public API, lld::safeLldMain(), is also available when using LLD as a library.
Differential Revision: https://reviews.llvm.org/D70378
Update the thunk range error report for PPC64PCRelLongBranchThunk and add a range
error test case for PPC64R12SetupStub.
Differential Revision: https://reviews.llvm.org/D87381
Add Thread Local Storage Initial Exec support to LLD.
This patch adds the computation for the relocations as well as the relaxation from Initial Exec to Local Exec.
Initial Exec:
```
pld r9, x@got@tprel@pcrel
add r9, r9, x@tls@pcrel
```
or
```
pld r9, x@got@tprel@pcrel
lbzx r10, r9, x@tls@pcrel
```
Note that @tls@pcrel is actually encoded as R_PPC64_TLS with a one byte displacement.
For the above examples relaxing Intitial Exec to Local Exec:
```
paddi r9, r9, x@tprel
nop
```
or
```
paddi r9, r13, x@tprel
lbz r10, 0(r9)
```
Reviewed By: nemanjai, MaskRay, #powerpc
Differential Revision: https://reviews.llvm.org/D86893
Bitcode writer does not flush buffer until the end by default. This is
fine to small bitcode files. When -flto,--plugin-opt=emit-llvm,-gmlt are
used, the final bitcode file is large, for example, >8G. Keeping all
data in memory consumes a lot of memory.
This change allows bitcode writer flush data to disk early when buffered
data size is above some threshold. This is only enabled when lld emits
LLVM bitcode.
One issue to address is backpatching bitcode: subblock length, function
body indexes, meta data indexes need to backfill. If buffer can be
flushed partially, we introduced raw_fd_stream that supports
read/seek/write, and enables backpatching bitcode flushed in disk.
Reviewed-by: tejohnson, MaskRay
Differential Revision: https://reviews.llvm.org/D86905
I have noticed that a 374MiB powerpc64le 'ld.lld' requires 11 passes to link.
There is a ThunkSection (whose parent OutputSection is ".text" of 169MiB) with 12867 thunks.
Optimize the filename glob pattern matching in
LinkerScript::computeInputSections() and LinkerScript::shouldKeep().
Add InputFile::getNameForScript() which gets and if required caches the
Inputfile's name used for linker script matching. This avoids the
overhead of name creation that was in getFilename() in LinkerScript.cpp.
Add InputSectionDescription::matchesFile() and
SectionPattern::excludesFile() which perform the glob pattern matching
for an InputFile and make use of a cache of the previous result. As both
computeInputSections() and shouldKeep() process sections in order and
the sections of the same InputFile are contiguous, these single entry
caches can significantly speed up performance for more complex glob
patterns.
These changes have been seen to reduce link time with --gc-sections by
up to ~40% with linker scripts that contain KEEP filename glob patterns
such as "*crtbegin*.o".
Differential Revision: https://reviews.llvm.org/D87469
Add Thread Local Storage Local Exec support to LLD. This is to support PC Relative addressing of Local Exec.
The patch teaches LLD to handle:
```
paddi r9, r13, x1@tprel
```
The relocation is:
```
R_PPC_TPREL34
```
Reviewed By: NeHuang, MaskRay
Differential Revision: https://reviews.llvm.org/D86608
`ELFFile<ELFT>` has many methods that take pointers,
though they assume that arguments are never null and
hence could take references instead.
This patch performs such clean-up.
Differential revision: https://reviews.llvm.org/D87385
Prefer `errorOrWarn` to `fatal` for recoverable errors and graceful degradation
when --noinhibit-exec is specified.
Mention the destination symbol, otherwise the diagnostic is not really actionable.
Two errors are not tested but the patch does not intend to add the coverage.
Reviewed By: grimar
Differential Revision: https://reviews.llvm.org/D87486
MarkLive::scanEhFrameSection is used to retain personality/LSDA
functions when --gc-sections is enabled.
Improve its performance by only iterating over the .eh_frame relocations
that need to be resolved for an EhSectionPiece. This optimization makes
the same assumption as elsewhere in LLD that the .eh_frame relocations
are sorted by r_offset.
This appears to be a performance regression introduced in commit
e6c24299d2 (https://reviews.llvm.org/D59800).
This change has been seen to reduce link time by up to ~50%.
Differential Revision: https://reviews.llvm.org/D87245
Currently we treat SHT_RISCV_ATTRIBUTES like a normal section and
concatenate all such input sections, yielding invalid output unless only
a single attributes section is present in the input. Instead, pick the
first as with SHT_ARM_ATTRIBUTES. We do not currently need to condition
our behaviour on the contents, unlike Arm. In future, we should both do
stricter validation of the input and merge all sections together to
ensure we have, for example, the full arch string requirement, but this
rudimentary implementation is good enough for most common cases.
Reviewed By: MaskRay
Differential Revision: https://reviews.llvm.org/D86309
In this patch, a pc-rel based long branch thunk is added for the local
call protocol that caller and callee does not use TOC.
Reviewed By: sfertile, nemanjai
Differential Revision: https://reviews.llvm.org/D86706
The function `__tls_get_addr` is used to get the address of an object that is Thread Local Storage.
It needs to have two relocations on it.
One relocation is for the function call itself and it is either R_PPC64_REL24 or R_PPC64_REL24_NOTOC.
The other is R_PPC64_TLSGD or R_PPC64_TLSLD for the symbol that is having its address computed.
In the early days of the transition from the ELFv1 ABI that is used for big endian PowerPC Linux distributions to the ELFv2 ABI that is used for little endian PowerPC Linux distributions, there was some ambiguity in the specification of the relocations for TLS. The GNU linker has implemented support for correct handling of calls to __tls_get_addr with a missing relocation. Unfortunately, we didn't notice that the IBM XL compiler did not handle TLS according to the updated ABI until we tried linking XL compiled libraries with LLD. As a result, there is a lot of code out there in various libraries compiled with XL that have this problem.
This patch adds a new error check in LLD that makes sure calls to `__tls_get_addr` are not missing the TLSGD/TLSLD relocation.
Reviewed By: MaskRay
Differential Revision: https://reviews.llvm.org/D85994
PR46970: for `alias = aliasee`, the alias can be used in relocation processing
and on ARM st_type does affect Thumb interworking. It is thus desirable for the
alias to get the same st_type.
Note that the st_size field should not be inherited because some tools use
st_size=0 as a heuristic to detect aliases. Retaining st_size can thwart such
heuristics and cause aliases to be preferred over the original symbols.
Differential Revision: https://reviews.llvm.org/D86263
* GNU ld places non-SHF_ALLOC sections after SHF_ALLOC sections. This has the
advantage that the file offsets of a non-SHF_ALLOC cannot be contained in
a PT_LOAD. This patch matches the behavior.
* For non-SHF_ALLOC non-orphan sections, GNU ld may assign non-zero sh_addr and
treat them similar to SHT_NOBITS (not advance location counter). This
is an alternative approach to what we have done in D85100.
By placing non-SHF_ALLOC sections at the end, we can drop special
cases in createSection and findOrphanPos added by D85100.
Different from GNU ld, we set sh_addr to 0 for non-SHF_ALLOC sections. 0
arguably is better because non-SHF_ALLOC sections don't appear in the memory
image.
ELF spec says:
> sh_addr - If the section will appear in the memory image of a process, this
> member gives the address at which the section's first byte should
> reside. Otherwise, the member contains 0.
D85100 appeared to take a detour. If we take a combined view on D85100 and this
patch, the overall complexity slightly increases (one more 3-line loop) and
compatibility with GNU ld improves.
The behavior we don't want to match is the special treatment of .symtab
.shstrtab .strtab: they can be matched in LLD but not in GNU ld.
Reviewed By: jhenderson, psmith
Differential Revision: https://reviews.llvm.org/D85867
LLD currently does not allow non-contiguous SHF_LINK_ORDER components in an
output section. This makes it infeasible to add SHF_LINK_ORDER to an existing
metadata section if backward compatibility with older object files are
concerned.
We did not allow mixed components (like GNU ld) and D77007 relaxed to allow
non-contiguous SHF_LINK_ORDER components. This patch allows arbitrary mix, with
sorting performed within an InputSectionDescription. For example,
`.rodata : {*(.rodata.foo) *(.rodata.bar)}`, has two InputSectionDescription's.
If there is at least one SHF_LINK_ORDER and at least one non-SHF_LINK_ORDER in
.rodata.foo, they are ordered within `*(.rodata.foo)`: we arbitrarily place
SHF_LINK_ORDER components before non-SHF_LINK_ORDER components (like Solaris ld).
`*(.rodata.bar)` is ordered similarly, but the two InputSectionDescription's
don't interact. It can be argued that this is more reasonable than the previous
behavior where written order was not respected.
It would be nice if the two different semantics (ordering requirement & garbage
collection) were not overloaded on one section flag, however, it is probably
difficult to obtain a generic flag at this point
(https://groups.google.com/forum/#!topic/generic-abi/hgx_m1aXqUo
"SHF_LINK_ORDER's original semantics make upgrade difficult").
(Actually, without the GC semantics, SHF_LINK_ORDER would still have the
sh_link!=0 & sh_link=0 issue. It is just that people find the GC semantics more
useful and tend to use the feature more often.)
GNU ld feature request: https://sourceware.org/bugzilla/show_bug.cgi?id=16833
Differential Revision: https://reviews.llvm.org/D84001
This patch implements the handling for the R_PPC64_PCREL_OPT relocation as well
as the GOT relocation for the associated R_PPC64_GOT_PCREL34 relocation.
On Power10 targets with PC-Relative addressing, the linker can relax
GOT-relative accesses to PC-Relative under some conditions. Since the sequence
consists of a prefixed load, followed by a non-prefixed access (load or store),
the linker needs to replace the first instruction (as the replacement
instruction will be prefixed). The compiler communicates to the linker that
this optimization is safe by placing the two aforementioned relocations on the
GOT load (of the address).
The linker then does two things:
- Convert the load from the got into a PC-Relative add to compute the address
relative to the PC
- Find the instruction referred to by the second relocation (R_PPC64_PCREL_OPT)
and replace the first with the PC-Relative version of it
It is important to synchronize the mapping from legacy memory instructions to
their PC-Relative form. Hence, this patch adds a file to be included by both
the compiler and the linker so they're always in agreement.
Differential revision: https://reviews.llvm.org/D84360
Thunk alignment is added in thie patch when using pc-rel instructions
to avoid crossing the 64 byte boundary.
Patched by: nemanjai, NeHuang
Reviewed By: sfertile, MaskRay
Differential Revision: https://reviews.llvm.org/D85973
-gdwarf-5 -fdebug-types-section may produce multiple .debug_info sections. All
except one are type units (.debug_types before DWARF v5). When constructing
.gdb_index, we should ignore these type units. We use a simple heuristic: the
compile unit does not have the SHF_GROUP flag. (This needs to be revisited if
people place compile unit .debug_info in COMDAT groups.)
This issue manifests as a data race: because an object file may have multiple
.debug_info sections, we may concurrently construct `LLDDwarfObj` for the same
file in multiple threads. The threads may access `InputSectionBase::data()`
concurrently on the same input section. `InputSectionBase::data()` does a lazy
uncompress() and rewrites the member variable `rawData`. A thread running zlib
`inflate()` (transitively called by uncompress()) on a buffer with `rawData`
tampered by another thread may fail with `uncompress failed: zlib error: Z_DATA_ERROR`.
Even if no data race occurred in an optimistic run, if there are N .debug_info,
one CU entry and its address ranges will be replicated N times. The result
.gdb_index can be much larger than a correct one.
The new test gdb-index-dwarf5-type-unit.s actually has two compile units. This
cannot be produced with regular approaches (it can be produced with -r
--unique). This is used to demonstrate that the .gdb_index construction code
only considers the last non-SHF_GROUP .debug_info
Reviewed By: grimar
Differential Revision: https://reviews.llvm.org/D85579
* For .cfi_*, GCC/GNU as emits SHT_PROGBITS type .eh_frame sections.
* Since rL252300, clang emits SHT_X86_64_UNWIND type .eh_frame sections
(originated from Solaris, documented in the x86-64 psABI).
* Some assembly use `.section .eh_frame,"a",@unwind` to generate
SHT_X86_64_UNWIND .eh_frame sections.
In a non-relocatable link, input .eh_frame are combined and there is
only one SyntheticSection .eh_frame in the output section, so the
"section type mismatch" diagnostic does not fire.
In a relocatable link, there is no SyntheticSection .eh_frame. .eh_frame of
mixed types can trigger the diagnostic. This patch fixes it by adding another
special case 0x70000001 (= SHT_X86_64_UNWIND) to canMergeToProgbits().
ld.lld -r gcc.o clang.o => error: section type mismatch for .eh_frame
There was a discussion "RFC: Usefulness of SHT_X86_64_UNWIND" on the x86-64-abi
mailing list. Folks are not wild about making the psABI value 0x70000001 into
gABI, but a few think defining 0x70000001 for .eh_frame may be a good idea for a
new architecture.
Reviewed By: grimar
Differential Revision: https://reviews.llvm.org/D85785
For an InputSection, the `buf` argument of `InputSectionBase::relocate` points
to the content of the containing OutputSection, instead of the content of the
InputSection itself, so `outSecOff` needs to be added in its callees. This is
counter-intuitive and leads to many `- outSecOff` and `+ outSecOff`.
This patch makes `InputSection::writeTo` call `InputSectionBase::relocate` with
`outSecOff` added. relocAlloc/relocNonAlloc/relocateNonAllocForRelocatable can
thus be simplified now.
Updated test:
* non-abs-reloc.s: A minor offset bug is fixed for a diagnostic in `relocateNonAlloc`
Reviewed By: grimar
Differential Revision: https://reviews.llvm.org/D85618
glibc/sysdeps/unix/sysv/linux/x86_64/sigaction.c libc.a(sigaction.o) has a CIE
with the augmentation string "zRS". Support 'S' to allow --icf={safe,all}.
Compatibility checks for PPC64PltCallStub and PPC64PCRelPLTStub are
added in this patch to prevent the usage of incompatible thunk/stub.
Reviewed By: sfertile, nemanjai, stefanp
Differential Revision: https://reviews.llvm.org/D85459
tl;dr See D81784 for the 'tombstone value' concept. This patch changes our behavior to be almost the same as GNU ld (except that we also use 1 for .debug_loc):
* .debug_ranges & .debug_loc: 1 (LLD<11: 0+addend; GNU ld uses 1 for .debug_ranges)
* .debug_*: 0 (LLD<11: 0+addend; GNU ld uses 0; future LLD: -1)
We make the tweaks because:
1) The new tombstone is novel and needs more time to be adopted by consumers before it's the default.
2) The old (gold) strategy had problems with zero-length functions - so rather than going back that, we're going to the GNU ld strategy which doesn't have that problem.
3) One slight tweak to (2) is to apply the .debug_ranges workaround to .debug_loc for the same reasons it applies to debug_ranges - to avoid terminating lists early.
-----
http://lists.llvm.org/pipermail/llvm-dev/2020-July/143482.html
The tombstone value -1 in .debug_line caused problems to lldb (fixed by D83957;
will be included in 11.0.0) and breakpad (fixed by
https://crrev.com/c/2321300). It may potentially affects other DWARF consumers.
For .debug_ranges & .debug_loc: 1, an argument preferring 1 (GNU ld for .debug_ranges) over -2 is that:
```
{-1, -2} <<< base address selection entry
{0, length} <<< address range
```
may create a situation where low_pc is greater than high_pc. So we use
1, the GNU ld behavior for .debug_ranges
For other .debug_* sections, there haven't been many reports. One issue is that
bloaty (src/dwarf.cc) can incorrectly count address ranges in .debug_ranges . To
reduce similar disruption, this patch changes the tombstone values to be similar to GNU ld.
This does mean another behavior change to the default trunk behavior. Sorry
about it. The default trunk behavior will be similar to release/11.x while we work on a transition plan for LLD users.
Reviewed By: dblaikie, echristo
Differential Revision: https://reviews.llvm.org/D84825
GNU ld allows sections after a non-SHF_ALLOC section to be covered by PT_LOAD
(PR37607) and assigns addresses to non-SHF_ALLOC output sections (similar to
SHF_ALLOC NOBITS sections. The location counter is not advanced).
This patch tries to fix PR37607 (remove a special case in
`Writer<ELFT>::createPhdrs`). To make the created PT_LOAD meaningful, we cannot
reset dot to 0 for a middle non-SHF_ALLOC output section. This results in
removal of two special cases in LinkerScript::assignOffsets. Non-SHF_ALLOC
non-orphan sections can have non-zero addresses like in GNU ld.
The zero address rule for non-SHF_ALLOC sections is weakened to apply to orphan
only. This results in a special case in createSection and findOrphanPos, respectively.
Reviewed By: jhenderson
Differential Revision: https://reviews.llvm.org/D85100
Part of https://bugs.llvm.org/show_bug.cgi?id=41734
The semantics of SHF_LINK_ORDER have been extended to represent metadata
sections associated with some other sections (usually text).
The associated text section may be discarded (e.g. LTO) and we want the
metadata section to have sh_link=0 (D72899, D76802).
Normally the metadata section is only referenced by the associated text
section. sh_link=0 means the associated text section is discarded, and
the metadata section will be garbage collected. If there is another
section (.gc_root) referencing the metadata section, the metadata
section will be retained. It's the .gc_root consumer's job to validate
the metadata sections.
# This creates a SHF_LINK_ORDER .meta with sh_link=0
.section .meta,"awo",@progbits,0
1:
.section .meta,"awo",@progbits,foo
2:
.section .gc_root,"a",@progbits
.quad 1b
.quad 2b
Reviewed By: pcc, jhenderson
Differential Revision: https://reviews.llvm.org/D72904
GNU ld allows sections after a non-SHF_ALLOC section to be covered by PT_LOAD
(PR37607) and assigns addresses to non-SHF_ALLOC output sections (similar to
SHF_ALLOC NOBITS sections. The location counter is not advanced).
This patch tries to fix PR37607 (remove a special case in
`Writer<ELFT>::createPhdrs`). To make the created PT_LOAD meaningful, we cannot
reset dot to 0 for a middle non-SHF_ALLOC output section. This results in
removal of two special cases in LinkerScript::assignOffsets. Non-SHF_ALLOC
non-orphan sections can have non-zero addresses like in GNU ld.
The zero address rule for non-SHF_ALLOC sections is weakened to apply to orphan
only. This results in a special case in createSection and findOrphanPos, respectively.
Reviewed By: jhenderson
Differential Revision: https://reviews.llvm.org/D85100
Fix PR36272 and PR46835
A .eh_frame FDE references a text section and (optionally) a LSDA (in
.gcc_except_table). Even if two text sections have identical content and
relocations (e.g. a() and b()), we cannot fold them if their LSDA are different.
```
void foo();
void a() {
try { foo(); } catch (int) { }
}
void b() {
try { foo(); } catch (float) { }
}
```
Scan .eh_frame pieces with LSDA and disallow referenced text sections to be
folded. If two .gcc_except_table have identical semantics (usually identical
content with PC-relative encoding), we will lose folding opportunity.
For ClickHouse (an exception-heavy application), this can reduce --icf=all efficiency
from 9% to 5%. There may be some percentage we can reclaim without affecting
correctness, if we analyze .eh_frame and .gcc_except_table sections.
gold 2.24 implemented a more complex fix (resolution to
https://sourceware.org/bugzilla/show_bug.cgi?id=21066) which combines the
checksum of .eh_frame CIE/FDE pieces.
Reviewed By: grimar
Differential Revision: https://reviews.llvm.org/D84610
--oformat=binary is rare (used in a few places in FreeBSD, see `stand/i386/mbr/Makefile` `LDFLAGS_BIN`)
The result should be identical to a normal output transformed by `objcopy -O binary`.
The current implementation ignores addresses and lays out sections by
respecting output section alignments. It can fail when an output section
address is specified, e.g. `.rodata ALIGN(16) :` (PR33651).
Fix PR33651 by respecting LMA. The code is similar to
`tools/llvm-objcop/ELF/Object.cpp` BinaryWriter::finalize after D71035 and D79229.
Unforunately for an output section without PT_LOAD, we assume its LMA is equal
to its VMA. So the result is still incorrect when an output section LMA
(`AT(...)`) is specified
Also drop `alignTo(off, config->wordsize)`. GNU ld does not round up the file size.
Differential Revision: https://reviews.llvm.org/D85086
Clang and GCC have a feature (-MD flag) to create a dependency file
in a format that build systems such as Make or Ninja can read, which
specifies all the additional inputs such .h files.
This change introduces the same functionality to lld bringing it to
feature parity with ld and gold which gained this feature recently.
See https://sourceware.org/bugzilla/show_bug.cgi?id=22843 for more
details and discussion.
The implementation corresponds to -MD -MP compiler flag where the
generated dependency file also includes phony targets which works
around the errors where the dependency is removed. This matches the
format used by ld and gold.
Fixes PR42806
Differential Revision: https://reviews.llvm.org/D82437
D68049 created options for basic block sections: -fbasic-block-sections=,
-funique-basic-block-section-names. Rename options in llc and lld (--lto-)
to be consistent. Specifically,
+ Rename basicblock-sections to basic-block-sections
+ Rename unique-bb-section-names to unique-basic-block-section-names
Differential Revision: https://reviews.llvm.org/D84462