It's a common pattern that we have a machine type, but we don't care
which ARM64* platform we're dealing with. We already have
isAnyArm64 for that, but it does not fit cases where we use a switch
statement. With this helper, it's easy to simplify such cases by using
Triple::ArchType instead of machine type.
The content of a GOFF record is always dumped if NDEBUG is not defined,
which produces rather confusing output. This changes wrap the dumping
code in LLVM_DEBUG, so the dump is only done when debug output of this
module is requested.
This change introduces a version 3 of the PSV data that includes support
for the name of the entry function as an offset into StringTable data to
a null-terminated utf-8 string.
Additional tests were added to ensure that the new value was properly
serialized/deserialized from object data.
Fixes#80175
---------
Co-authored-by: Cooper Partin <coopp@ntdev.microsoft.com>
This change introduces a version 3 of the PSV data that includes support
for the name of the entry function as an offset into StringTable data to
a null-terminated utf-8 string.
Additional tests were added to ensure that the new value was properly
serialized/deserialized from object data.
Fixes#80175
---------
Co-authored-by: Cooper Partin <coopp@ntdev.microsoft.com>
This is compatible with MSVC, `-machine:arm64x` is essentially an alias
to `-machine:arm64ec`. To make a type library that exposes both native
and EC symbols, an additional `-defArm64Native` argument is needed in
both cases.
Defines a subset of attributes and emits them to a section called
.hexagon.attributes.
The current attributes recorded are the attributes needed by
llvm-objdump to automatically determine target features and eliminate
the need to manually pass features.
This is refactoring preparing to move UseECMap computation to the
archive writer. We currently require writeArchive caller to pass that.
This is not practical for llvm-ar, which currently interprets at most
one passed object. For a reliable UseECMap, we need to interpret all
symbolic objects: we may have, for example, a list of x86_64 files
followed by aarch64 file, which indicates that we should use EC map for
x86_64 objects.
This commit interprets symbolic files in a separated pass, which will be
a convenient place to implement UseECMap computation in the follow up.
It also makes accessing the next member for AIX big archive offset
computation a bit easier.
Detect COFF files by default and allow specifying it with --format
argument.
This is important for ARM64EC, which uses a separated symbol map for EC
symbols. Since K_COFF is mostly compatible with K_GNU, this shouldn't
really make a difference for other targets.
This originally landed as #82642, but was reverted due to test failures
in tests using no symbol table. Since COFF symbol can't express it,
fallback to GNU format in that case.
As noted in <https://github.com/llvm/llvm-project/pull/78537>, MSVC
places import descriptors in both the EC and regular map - that PR moved
the descriptors to ONLY the regular map, however this causes linking
errors when linking as Arm64EC:
```
bcryptprimitives.lib(bcryptprimitives.dll) : error LNK2001: unresolved external symbol __IMPORT_DESCRIPTOR_bcryptprimitives (EC Symbol)
```
This change copies import descriptors from the regular map to the EC
map, which fixes this linking error.
LLVM is inconsistent about how it converts `errno` to `std::error_code`.
This can cause problems because values outside of `std::errc` compare
differently if one is system and one is generic on POSIX systems.
This is even more of a problem on Windows where use of the system
category is just wrong, as that is for Windows errors, which have a
completely different mapping than POSIX/generic errors. This patch fixes
one instance of this mistake in `JSONTransport.cpp`.
This patch adds `errnoAsErrorCode()` which makes it so people do not
need to think about this issue in the future. It also cleans up a lot of
usage of `errno` in LLVM and Clang.
Detect COFF files by default and allow specifying it with --format
argument.
This is important for ARM64EC, which uses a separated symbol map for EC
symbols. Since K_COFF is mostly compatible with K_GNU, this shouldn't
really make a difference for other targets.
getSymbolSize was recently added to WasmObjectFile and has correct sizes
for most symbol types. This makes llvm-symbolizer correctly symbolize
addresses in the middle of the symbol.
When reworking the test I also noticed that the DWARF info seems to be
wrong for the first instruction in each function. I noted that in the test
comments but didn't attempt to fix here.
The section headers for XCOFF files have a subtype flag for Dwarf
sections. This PR updates obj2yaml, yaml2obj, and llvm-readobj so that
they recognize the subtype.
Currently the address reported by binutils for a global is its index;
but its offset (in the file or section) is more useful for binary size
attribution.
This PR treats globals similarly to functions, and tracks their offset
and size. It also centralizes the logic differentiating linked from object
and dylib files (where section addresses are 0).
Add new target and a new -n option allowing to specify native module
definition file, similar to how -defArm64Native works in llvm-lib. This
also changes archive format to use K_COFF like non-mingw targets. It's
required on ARM64EC, but it should be fine for other targets too.
The dot is too confusing for tools. Output temporaries would have
'10.3-generic' so tools could parse it as an extension, device libs &
the associated clang driver logic are also confused by the dot.
After discussions, we decided it's better to just remove the '.' from
the target name than fix each issue one by one.
This can be used to create import libraries that contain both ARM64EC
and native exports. The implementation follows observed MSVC lib.exe
behaviour. It's ignored on targets other than ARM64EC.
These generic targets include multiple GPUs and will, in the future,
provide a way to build once and run on multiple GPU, at the cost of less
optimization opportunities.
Note that this is just doing the compiler side of things, device libs an
runtimes/loader/etc. don't know about these targets yet, so none of them
actually work in practice right now. This is just the initial commit to
make LLVM aware of them.
This contains the documentation changes for both this change and #76954
as well.
ARM64EC import libraries expose two additional symbols: mangled thunk
symbol (like `#func`) and auxiliary import symbol (like`__imp_aux_func`).
The main functional change with this patch is that those symbols are
properly added to static library ECSYMBOLS.
nm already prints sizes for data symbols. Do that for function symbols
too, and update objdump to also print size information.
Implements item 3 from https://github.com/llvm/llvm-project/issues/76107
Summary:
Currently, the linker wrapper sorts input files into different link
jobs according to their architectures. Here we assume each architecture
is a unique and incompatible link job unless they are specifically
marked compatible. This patch simply adds an `all` target to represent
an architecture that should be linked against every single other
architecture.
This will be useful for modelling generic IR such as the ROCm device
libraries or the NVPTX libdevice.
Currently symbol info is generated from a linking section or from export
names. This PR generates symbols in a WasmObjectFile from the name
section as well, which allows tools like objdump and nm to show useful
information for more linked binaries. There are some limitations:
most notably that we don't assume any particular ABI, so we don't get
detailed information about data symbols if the segments are merged
(which is the default).
Covers most of the desired functionality from #76107
Wasm has no unified virtual memory space as other object formats and
architectures do, so previously WasmObjectFile reported 0 for all
section addresses, and until 428cf71ff used section offsets for function
symbols. Now we use file offsets for function symbols, and this change
switches section addresses to do the same (in linked files). The main
result of this is that objdump now reports VMAs in section listings, and
also uses file offets rather than section offsets when disassembling
linked binaries (matching the behavior of other disassemblers and stack
traces produced by browwsers). To make this work, this PR also updates
objdump's generation of synthetics fallback symbols to match lib/Object
and also correctly plumbs symbol types for regular and dummy symbols
through to the backend to avoid needing special knowledge of address 0.
This also paves the way for generating symbols from name sections rather
than symbol tables or imports (see #76107) by allowing the
disassembler's synthetic fallback symbols match the name-section
generated symbols (in a followup PR).
This code appears to be a hack to set the features to include compressed
instructions if the ELF EFLAGS flags bit is present, but the ELF
attribute for the ISA string is no present or not accurate.
We can't remove the hack because llvm-mc doesn't create ELF attributes
by default so a lot of tests fail to disassembler properly. Using clang
as the assembler does set the attributes.
This patch changes the hack to only set Zca since that is the minimum
implied by the flag. Setting anything else potentially conflicts with
the ISA string containing Zcmp or Zcmt.
JITLink also needs to be updated to recognize Zca in addition to C.
Summary:
The standard GPU compilation process embeds each intermediate object
file into the host file at the `.llvm.offloading` section so it can be
linked later. We also use a special section called something like
`omp_offloading_entries` to store all the globals that need to be
registered by the runtime. The linker-wrapper's job is to link the
embedded device code stored at this section and then emit code to
register the linked image and the kernels and globals in the offloading
entry section.
One downside to RDC linking is that it can become quite big for very
large projects that wish to make use of static linking. This patch
changes the support for relocatable linking via `-r` to support a kind
of "partial" RDC compilation for offloading languages.
This primarily requires manually editing the embedded data in the
output object file for the relocatable link. We need to rename the
output section to make it distinct from the input sections that will be
merged. We then delete the old embedded object code so it won't be
linked further. We then need to rename the old offloading section so
that it is private to the module. A runtime solution could also be done
to defer entries that don't belong to the given GPU executable, but this
is easier. Note that this does not work with COFF linking, only the ELF
method for handling offloading entries, that could be made to work
similarly.
Given this support, the following compilation path should produce two
distinct images for OpenMP offloading.
```
$ clang foo.c -fopenmp --offload-arch=native -c
$ clang foo.c -lomptarget.devicertl --offload-link -r -o merged.o
$ clang main.c merged.o -fopenmp --offload-arch=native
$ ./a.out
```
Or similarly for HIP to effectively perform non-RDC mode compilation for
a subset of files.
```
$ clang -x hip foo.c --offload-arch=native --offload-new-driver -fgpu-rdc -c
$ clang -x hip foo.c -lomptarget.devicertl --offload-link -r -o merged.o
$ clang -x hip main.c merged.o --offload-arch=native --offload-new-driver -fgpu-rdc
$ ./a.out
```
One question is whether or not this should be the default behavior of
`-r` when run through the linker-wrapper or a special option. Standard
`-r` behavior is still possible if used without invoking the
linker-wrapper and it guaranteed to be correct.
getExportName implementation is based on lld-link. In its current form,
it's mostly about convenience, but it will be more useful for EXPORTAS
support, for which export name is not possible to deduce from other
printed properties.
Move the WasmSymbolInfos from their own vector on the WasmLinkingData
directly into the WasmSymbol object. Removing the const-ref to an
external object allows the vector of WasmSymbols to be safely
expanded/reallocated; generating symbol info from the name section will
require this, as the numbers of function and data segment names are
stored separately.
This is a step toward generating symbol information from name sections
for #76107
Today `-split-machine-functions` and `-fbasic-block-sections={all,list}`
cannot be combined with `-basic-block-sections=labels` (the labels
option will be ignored).
The inconsistency comes from the way basic block address map -- the
underlying mechanism for basic block labels -- encodes basic block
addresses
(https://lists.llvm.org/pipermail/llvm-dev/2020-July/143512.html).
Specifically, basic block offsets are computed relative to the function
begin symbol. This relies on functions being contiguous which is not the
case for MFS and basic block section binaries. This means Propeller
cannot use binary profiles collected from these binaries, which limits
the applicability of Propeller for iterative optimization.
To make the `SHT_LLVM_BB_ADDR_MAP` feature work with basic block section
binaries, we propose modifying the encoding of this section as follows.
First let us review the current encoding which emits the address of each
function and its number of basic blocks, followed by basic block entries
for each basic block.
| | |
|--|--|
| Address of the function | Function Address |
| Number of basic blocks in this function | NumBlocks |
| BB entry 1
| BB entry 2
| ...
| BB entry #NumBlocks
To make this work for basic block sections, we treat each basic block
section similar to a function, except that basic block sections of the
same function must be encapsulated in the same structure so we can map
all of them to their single function.
We modify the encoding to first emit the number of basic block sections
(BB ranges) in the function. Then we emit the address map of each basic
block section section as before: the base address of the section, its
number of blocks, and BB entries for its basic block. The first section
in the BB address map is always the function entry section.
| | |
|--|--|
| Number of sections for this function | NumBBRanges |
| Section 1 begin address | BaseAddress[1] |
| Number of basic blocks in section 1 | NumBlocks[1] |
| BB entries for Section 1
|..................|
| Section #NumBBRanges begin address | BaseAddress[NumBBRanges] |
| Number of basic blocks in section #NumBBRanges |
NumBlocks[NumBBRanges] |
| BB entries for Section #NumBBRanges
The encoding of basic block entries remains as before with the minor
change that each basic block offset is now computed relative to the
begin symbol of its containing BB section.
This patch adds a new boolean codegen option `-basic-block-address-map`.
Correspondingly, the front-end flag `-fbasic-block-address-map` and LLD
flag `--lto-basic-block-address-map` are introduced.
Analogously, we add a new TargetOption field `BBAddrMap`. This means BB
address maps are either generated for all functions in the compiling
unit, or for none (depending on `TargetOptions::BBAddrMap`).
This patch keeps the functionality of the old
`-fbasic-block-sections=labels` option but does not remove it. A
subsequent patch will remove the obsolete option.
We refactor the `BasicBlockSections` pass by separating the BB address
map and BB sections handing to their own functions (named
`handleBBAddrMap` and `handleBBSections`). `handleBBSections` renumbers
basic blocks and places them in their assigned sections.
`handleBBAddrMap` is invoked after `handleBBSections` (if requested) and
only renumbers the blocks.
- New tests added:
- Two tests basic-block-address-map-with-basic-block-sections.ll and
basic-block-address-map-with-mfs.ll to exercise the combination of
`-basic-block-address-map` with `-basic-block-sections=list` and
'-split-machine-functions`.
- A driver sanity test for the `-fbasic-block-address-map` option
(basic-block-address-map.c).
- An LLD test for testing the `--lto-basic-block-address-map` option.
This reuses the LLVM IR from `lld/test/ELF/lto/basic-block-sections.ll`.
- Renamed and modified the two existing codegen tests for basic block
address map (`basic-block-sections-labels-functions-sections.ll` and
`basic-block-sections-labels.ll`)
- Removed `SHT_LLVM_BB_ADDR_MAP_V0` tests. Full deprecation of
`SHT_LLVM_BB_ADDR_MAP_V0` and `SHT_LLVM_BB_ADDR_MAP` version less than 2
will happen in a separate PR in a few months.
A distinction that doesn't _usually_ matter is that the
MachO::SymbolKind is really a mapping of entries in TBD files not
symbols. To better understand this, rename the enum so it represents an
encoding mapped to TBDs as opposed to symbols alone.
For example, it can be a bit confusing that "GlobalSymbol" is a enum
value when all of those values can represent a GlobalSymbol.
This change allows a WasmObjectFile to be created from a wasm file even
if it uses typed funcrefs and GC types. It does not significantly change how
lib/Object models its various internal types (e.g. WasmSignature,
WasmElemSegment), so LLVM does not really "support" or understand such
files, but it is sufficient to parse the type, global and element sections, discarding
types that are not understood. This is useful for low-level binary tools such as
nm and objcopy, which use only limited aspects of the binary (such as function
definitions) or deal with sections as opaque blobs.
This is done by allowing `WasmValType` to have a value of `OTHERREF`
(representing any unmodeled reference type), and adding a field to
`WasmSignature` indicating it's a placeholder for an unmodeled reference
type (since there is a 1:1 correspondence between WasmSignature objects
and types in the type section).
Then the object file parsers for the type and element sections are expanded
to parse encoded reference types and discard any unmodeled fields.
