It is a fairly direct extension of the gc algorithm. For merge sections
instead of remembering just a live bit, we remember which offsets were
used.
This reduces the .rodata sections in chromium from 9648861 to 9477472
bytes.
llvm-svn: 267164
You can instruct the linker to not discard sections even if they
are unused and --gc-sections option is given. The linker script
command for doing that is KEEP. The syntax is KEEP(foo) where foo
is a section name. KEEP commands are written in SECTIONS command,
so you can specify the order of sections *and* which sections
will be kept.
Each sub-command in SECTIONS command are translated into SectionRule
object. Previously, each SectionRule has `Keep` bit. However,
if you think about it, this hid information in too deep in elements
of a list. Semantically, KEEP commands aren't really related to
SECTIONS subcommands. We can keep the section list for KEEP in a
separate list. This patch does that.
llvm-svn: 267065
Since there is a copy in every translation unit that uses them, they can
be omitted from the symbol table if the address is not significant.
This still doesn't catch as many cases as the gold plugin. The
difference is that we check canBeOmittedFromSymbolTable in each file and
use lazy loading which limits what it can do. Gold checks it in the merged file.
I think the correct way of getting the same results as gold is just to
cache in the IR the result of canBeOmittedFromSymbolTable.
llvm-svn: 267063
Rafael reported on the mailing list that this reduces peak memory
usage while linking llvm-as by 15%. It makes sense to make it
the default, and introduce an inverse knob -lto-no-discard-value-names
for those who want to restore the old behavior.
llvm-svn: 267020
The gold plugin logic for common symbols is a little bit convoluted
as the plugin API has not an explicit way to update the alignment.
In gold, then, we need to keep the bitcode symbol @a around because
that's the only way to get the alignment right in the final object.
In lld, this is not true. We already have all the informations we
need about common symbols (size/alignment) so we don't have to
keep the existing symbol and pass it to the mover.
llvm-svn: 267007
SectionOrder vector was a part of LinkerScript class.
It can be removed because Commands vector contains the
same information and SectiorOrder is just a subset.
Differential revision: http://reviews.llvm.org/D19171
llvm-svn: 266974
These relocations are calculated as S + A - DTPREL or S + A - TPREL,
where DTPREL = TlsVA - 0x8000, TPREL = TlsVA - 0x7000. So the result
is relative to the TLS output section and is not an absolut value
The fix allows to escape creation of unnecessary dynamic relocations
in case of DSO linking.
llvm-svn: 266923
MIPS ABI turns using of GOT and dynamic relocations inside out. While
regular ABI uses dynamic relocations to fill up GOT entries MIPS ABI
requires dynamic linker to fills up GOT entries using specially sorted
dynamic symbol table. This affects even dynamic relocations against
symbols which do not require GOT entries creation explicitly, i.e. do
not have any GOT-relocations. So if a preemptible symbol has a dynamic
relocation we anyway have to create a GOT entry for it.
If a non-preemptible symbol has a dynamic relocation against it, dynamic
linker takes it st_value, adds offset and writes down result of the
dynamic relocation. In case of preemptible symbol dynamic linker
performs symbol resolution, writes the symbol value to the GOT entry and
reads the GOT entry when it needs to perform a dynamic relocation.
Differential Revision: http://reviews.llvm.org/D18948
llvm-svn: 266921
This patch is based heavily on George Rimor's patch
http://reviews.llvm.org/D19221.
In the linker script, you can write expressions to compute addresses.
Currently we only support "+" operator. This adds a few more operators.
Since this patch adds * and /, we need to handle operator precedences
properly. I implemented that using the operator-precedence grammar.
Differential Revision: http://reviews.llvm.org/D19237
llvm-svn: 266798
Manifest file is a separate or embedded XML file having metadata
of an executable. As it is XML, it can contain various types of
information. Probably the most popular one is to request escalated
priviledges.
Usually the linker creates an XML file and embed that file into
an executable. However, there's a way to supply an XML file from
command line. /manifestniput is it.
Apparently it is over-designed here, but if you supply two or more
manifest files, then the linker needs to merge the files into a
single XML file. A good news is that we don't need to do that ourselves.
MT.exe command can do that, so we call the command from the linker
in this patch.
llvm-svn: 266704
* Do script driven layout only if SECTIONS section exist.
Initial commit message:
[ELF] - Implemented basic location counter support.
This patch implements location counter support.
It also separates assign addresses for sections to assignAddressesScript() if it scipt exists.
Main testcase is test/ELF/linkerscript-locationcounter.s, It contains some work with location counter. It is basic now.
Implemented location counter assignment and '+' operations.
Patch by myself with LOTS of comments and design suggestions from Rui Ueyama.
Differential revision: http://reviews.llvm.org/D18499
llvm-svn: 266526
At the moment almost every lit.site.cfg.in contains two lines comment:
## Autogenerated by LLVM/Clang configuration.
# Do not edit!
The patch adds variable LIT_SITE_CFG_IN_HEADER, that is replaced from
configure_lit_site_cfg with the note and some useful information.
llvm-svn: 266521
Parallelism level can be chosen using the new --lto-jobs=K option
where K is the number of threads used for CodeGen. It currently
defaults to 1.
llvm-svn: 266484
This patch implements location counter support.
It also separates assign addresses for sections to assignAddressesScript() if it scipt exists.
Main testcase is test/ELF/linkerscript-locationcounter.s, It contains some work with location counter. It is basic now.
Implemented location counter assignment and '+' operations.
Patch by myself with LOTS of comments and design suggestions from Rui Ueyama.
Differential revision: http://reviews.llvm.org/D18499
llvm-svn: 266457
The _gp_disp symbol designates offset between start of function and 'gp'
pointer into GOT. The following code is a typical MIPS function preamble
used to setup $gp register:
lui $gp, %hi(_gp_disp)
addi $gp, $gp, %lo(_gp_disp)
To calculate R_MIPS_HI16 / R_MIPS_LO16 relocations results we use
the following formulas:
%hi(_gp - P + A)
%lo(_gp - P + A + 4),
where _gp is a value of _gp symbol, A is addend, and P current address.
The R_MIPS_LO16 relocation references _gp_disp symbol is always the second
instruction. That is why we need four byte adjustments. The patch assigns
R_PC type for R_MIPS_LO16 relocation and adjusts its addend by 4. That fix
R_MIPS_LO16 calculation.
For details see p. 4-19 at ftp://www.linux-mips.org/pub/linux/mips/doc/ABI/mipsabi.pdf
Differential Revision: http://reviews.llvm.org/D19115
llvm-svn: 266368
They are unnecessary, as the dynamic loader can apply the original relocations
directly. This was also resulting in the creation of copy relocations in PIEs.
Differential Revision: http://reviews.llvm.org/D19089
llvm-svn: 266273
This patch implements the --dynamic-list option, which adds a list of
global symbol that either should not be bounded by default definition
when creating shared libraries, or add in dynamic symbol table in the
case of creating executables.
The patch modifies the ScriptParserBase class to use a list of Token
instead of StringRef, which contains information if the token is a
quoted or unquoted strings. It is used to use a faster search for
exact match symbol name.
The input file follow a similar format of linker script with some
simplifications (it does not have scope or node names). It leads
to a simplified parser define in DynamicList.{cpp,h}.
Different from ld/gold neither glob pattern nor mangled names
(extern 'C++') are currently supported.
llvm-svn: 266227
Previously each archive file was reported no matter were it's member used or not,
like:
lib/libLLVMSupport.a
Now lld prints line for each used internal file, like:
lib/libLLVMSupport.a(lib/Support/CMakeFiles/LLVMSupport.dir/StringSaver.cpp.o)
lib/libLLVMSupport.a(lib/Support/CMakeFiles/LLVMSupport.dir/Host.cpp.o)
lib/libLLVMSupport.a(lib/Support/CMakeFiles/LLVMSupport.dir/ConvertUTF.c.o)
That should be consistent with what gold do.
This fixes PR27243.
Differential revision: http://reviews.llvm.org/D19011
llvm-svn: 266220
This simplifies the code by allowing us to remove the visibility argument
to functions that create synthetic symbols.
The only functional change is that the visibility of the MIPS "_gp" symbol
is now hidden. Because this symbol is defined in every executable or DSO, it
would be difficult to observe a visibility change here.
Differential Revision: http://reviews.llvm.org/D19033
llvm-svn: 266208
We need to ensure that the address of an undefined weak symbol evaluates to
zero. We were getting this right for non-PIC executables (where the symbol
can be evaluated directly) and for DSOs (where we emit a symbolic relocation
for these symbols, as they are preemptible). But we weren't getting it right
for PIEs. Probably the simplest way to ensure that these symbols evaluate
to zero is by not creating a relocation in .got for them.
Differential Revision: http://reviews.llvm.org/D19044
llvm-svn: 266161
With this patch we use the first scan over the relocations to remember
the information we found about them: will them be relaxed, will a plt be
used, etc.
With that the actual relocation application becomes much simpler. That
is particularly true for the interfaces in Target.h.
This unfortunately means that we now do two passes over relocations for
non SHF_ALLOC sections. I think this can be solved by factoring out the
code that scans a single relocation. It can then be used both as a scan
that record info and for a dedicated direct relocation of non SHF_ALLOC
sections.
I also think it is possible to reduce the number of enum values by
representing a target with just an OutputSection and an offset (which
can be from the start or end).
This should unblock adding features like relocation optimizations.
llvm-svn: 266158
The _gp* family of symbols is defined as an offset in .got, and it is
not at all clear what should happen when .got is not defined.
This will allow some simplifications on how these symbols are handled.
llvm-svn: 266063
It is possible to have FDEs with duplicate PCs if ICF was able to merge
functions with FDEs, or if the input files for some reason contained duplicate
FDEs. We previously weren't handling this correctly when producing the
contents of the .eh_frame_hdr section; we were dropping entries and leaving
null entries at the end of the section, which confused consumers of unwind
data, such as the backtrace() function.
Fix the bug by setting the FDE count to the number of FDEs actually emitted
into .eh_frame_hdr, rather than the number of FDEs in .eh_frame.
Differential Revision: http://reviews.llvm.org/D18911
llvm-svn: 265957
It is possible that the same symbol referenced by two kinds of
relocations at the same time. The first type requires say GOT entry
creation, the second type requires dynamic copy relocation. For MIPS
targets they might be R_MIPS_GOT16 and R_MIPS_HI16 relocations. For X86
target they might be R_386_GOT32 and R_386_32 respectively.
Now LLD never creates GOT entry for a symbol if this symbol already has
related copy relocation. This patch solves this problem.
Differential Revision: http://reviews.llvm.org/D18862
llvm-svn: 265910
Previously, Lazy symbols were created for undefined symbols even though
such symbols cannot be resolved by loading object files. This patch
fixes that bug.
llvm-svn: 265847
Now MustBeInDynSym is only true if the symbol really must be in the
dynamic symbol table.
IsUsedInRegularObj is only true if the symbol is used in a .o or -u. Not
a .so or a .bc.
A benefit is that this is now done almost entirilly during symbol
resolution. The only exception is copy relocations because of aliases.
This includes a small fix in that protected symbols in .so don't force
executable symbols to be exported.
This also opens the way for implementing internalize for -shared.
llvm-svn: 265826
The spec says:
If a symbol definition with STV_PROTECTED visibility from a shared
object is taken as resolving a reference from an executable or another
shared object, the SHN_UNDEF symbol table entry created has STV_DEFAULT
visibility.
llvm-svn: 265792
This patch fixes dynamic relocation creation from GOT access in dynamic
objects on aarch64. Current code creates a plt relative one
(R_AARCH64_JUMP_SLOT) instead of a got relative (R_AARCH64_GLOB_DAT).
It leads the programs fails with:
$ cat t.cc
std::string test = "hello...\n";
int main ()
{
printf ("%s\n", test.c_str());
return 0;
}
$ clang++ t.cc -fpic -o t
$ ./t
hello...
Segmentation fault (core dumped)
Due the fact it will try to access the plt instead of the got for
__cxa_atexit registration for the std::string destruction. It will
lead in a bogus function address in atexit.
llvm-svn: 265784
Previously, we supported only one hash function, FNV-1, so
BuildIdSection directly handled hash computation. In this patch,
I made BuildIdSection an abstract class and defined two subclasses,
BuildIdFnv1 and BuildIdMd5.
llvm-svn: 265737
start-lib and end-lib are options to link object files in the same
semantics as archive files. If an object is in start-lib and end-lib,
the object is linked only when the file is needed to resolve
undefined symbols. That means, if an object is in start-lib and end-lib,
it behaves as if it were in an archive file.
In this patch, I introduced a new notion, LazyObjectFile. That is
analogous to Archive file type, but that works for a single object
file instead of for an archive file.
http://reviews.llvm.org/D18814
llvm-svn: 265710
This requires knowing input section offsets in output sections before
scanRelocs. This is generally a good thing and should allow further
simplifications in the creation of dynamic relocations.
llvm-svn: 265673
Similar to r265462, TLS related relocations aren't marked as relative,
meaning that we end up generating R_AARCH64_RELATIVE relocations for
them. This change adds TLS relocations that I've seen on my system. With
this patch applied CloudABI's unit testing binary now passes on aarch64.
Approved by: ruiu
Differential Revision: http://reviews.llvm.org/D18816
llvm-svn: 265575
We have to differentiate undefined symbols from bitcode and undefined
symbols from other sources.
Undefined symbols from bitcode should not inhibit the symbol being
internalized. Undefined symbols from other sources should.
llvm-svn: 265536
When error, this adds the text line of script to the output
and a marks exact incorrect token under it:
line 1: <error text here>
UNKNOWN_TAG {
^
Differential revision: http://reviews.llvm.org/D18699
llvm-svn: 265523
While trying to get PIE work on CloudABI for x86-64, I noticed that even
though GNU ld would generate functional binaries, LLD would not. It
turns out that we generate relocations for referencing TLS objects
inside of the text segment, which shouldn't happen.
This change extends the isRelRelative() function to list some additional
relocation types that should be treated as relative. This makes my C
library unit testing binary work on x86-64.
Approved by: ruiu
Differential Revision: http://reviews.llvm.org/D18688
Fixes bug: https://llvm.org/bugs/show_bug.cgi?id=27174
llvm-svn: 265462
Where Clang's AArch64 backend seems to differ from the X86 backend is
that it tends to use the GOT more aggressively.
After getting CloudABI PIEs working on x86-64, I noticed that accessing
global variables would still crash on aarch64. Tracing it down, it turns
out that the GOT was filled with entries assuming the base address was
zero.
It turns out that we skip generating relocations for GOT entries in case
the relocation pointing towards the GOT is relative. Whether the thing
pointing to the GOT is absolute or relative shouldn't make any
difference; the GOT entry itself should contain the absolute address,
thus needs a relocation regardless.
Approved by: rafael
Differential Revision: http://reviews.llvm.org/D18739
llvm-svn: 265453
Make sure to copy the MustBeInDynSym field when replacing shared symbols with
bitcode symbols, and when replacing bitcode symbols with regular symbols
in addCombinedLtoObject. Fixes interposition of DSO symbols with bitcode
symbols in the main executable.
Differential Revision: http://reviews.llvm.org/D18780
llvm-svn: 265371
For each copy relocation that we create, look through the DSO's symbol table
for aliases and create a dynamic symbol for each one. This causes the copy
relocation to correctly interpose any aliases.
Copy relocations are relatively uncommon (on my machine, 56% of binaries in
/usr/bin have no copy relocations probably due to being PIEs, 97% of them
have <10, and the binary with the largest number of them has 97) so it's
probably fine to do this in a relatively inefficient way.
Differential Revision: http://reviews.llvm.org/D18731
llvm-svn: 265354
This test didn't actually test the functionality. The new test
verifies that "-verify" is passed if and only if -disable-verify
is not given.
llvm-svn: 265316
The only way to get an object file with symbols marked by the STO_MIPS_PIC
flag is to link PIC and non-PIC object files and generate a relocatable
output using '-r' command line option. Now LLD is able to generate a relocatable
output but does not mark PIC symbols by the STO_MIPS_PIC flag. So I have
to use binary input mips-sto-pic.o generated by GNU BFD linker.
llvm-svn: 265310
Our symbol representation was redundant, and some times would get out of
sync. It had an Elf_Sym, but some fields were copied to SymbolBody.
Different parts of the code were checking the bits in SymbolBody and
others were checking Elf_Sym.
There are two general approaches to fix this:
* Copy the required information and don't store and Elf_Sym.
* Don't copy the information and always use the Elf_Smy.
The second way sounds tempting, but has a big problem: we would have to
template SymbolBody. I started doing it, but it requires templeting
*everything* and creates a bit chicken and egg problem at the driver
where we have to find ELFT before we can create an ArchiveFile for
example.
As much as possible I compared the test differences with what gold and
bfd produce to make sure they are still valid. In most cases we are just
adding hidden visibility to a local symbol, which is harmless.
In most tests this is a small speedup. The only slowdown was scylla
(1.006X). The largest speedup was clang with no --build-id, -O3 or
--gc-sections (i.e.: focus on the relocations): 1.019X.
llvm-svn: 265293
So, there are some cases when the IR Linker produces a broken
module (which doesn't pass the verifier) and we end up asserting
inside the verifier. I think it's always a bug producing a module
which does not pass the verifier but there are some cases in which
people can live with the broken module (e.g. if only DebugInfo
metadata are broken). The gold plugin has something similar.
This commit is motivated by a situation I found in the
wild. It seems that somebody else discovered it independently
and reported in PR24923.
llvm-svn: 265258
We already got this right, but it never hurts adding another
test, in case we'll change the handling in the future, to ensure
we don't break it.
llvm-svn: 265256
Currently we create a file called .lto.bc. In UNIX,
ls(1) by default doesn't show up files starting with
a dot, as they're (only by convention) hidden.
This makes the output of -save-temps a little bit
hard to find. Use "a.out.lto.bc" instead if the
output file is not specified.
While here, change an existing -save-temps test to
exercise this more interesting behaviour.
llvm-svn: 265254
If a symbol is defined in an archive, when we replace its body
the isUsedInRegularObj wasn't set correctly. Internalize makes
its decision based on that bit so we ended up internalizing
symbols that we shouldn't (because they're referenced).
This should fix. Thanks to Peter and Rafael for discussion
and help diagnosing the issue!
Found during LTO of unittests.
llvm-svn: 265208
This fixes bootstrap of llvm-tblgen (with LTO) and PR27150.
Slightly longer explanation follows.
Emission of .init_array instead of .ctors is supported only on a
subset of the Target LLVM supports. Codegen needs to be conservative
and always emit .ctors unless instructed otherwise (based on target).
If the dynamic linker sees .init_array it completely ignores
what's inside .ctors and therefore some constructors are not called
(and this causes llvm-tblgen to crash on startup).
Teach LLD/LTO about the Codegen options so we end up always emitting
.init_array and avoid this issue.
In future, we might end up supporting mix of .ctors and .init_array
in different input files if this shows up as a real-world use case.
The way gold handles this case is mapping .ctors from input into
.init_array in output. There's also another caveat because
as far as I understand .ctors run in reverse order so when we do
the copy/mapping we need to reverse copy in the output if there's
more than one ctor. That's why I'd rather avoid this complicate logic
unless there's a real need.
An analogous reasoning holds for .dtors/.fini_array.
llvm-svn: 265085
Some targets might require creation of thunks. For example, MIPS targets
require stubs to call PIC code from non-PIC one. The patch implements
infrastructure for thunk code creation and provides support for MIPS
LA25 stubs. Any MIPS PIC code function is invoked with its address
in register $t9. So if we have a branch instruction from non-PIC code
to the PIC one we cannot make the jump directly and need to create a small
stub to save the target function address.
See page 3-38 ftp://www.linux-mips.org/pub/linux/mips/doc/ABI/mipsabi.pdf
- In relocation scanning phase we ask target about thunk creation necessity
by calling `TagetInfo::needsThunk` method. The `InputSection` class
maintains list of Symbols requires thunk creation.
- Reassigning offsets performed for each input sections after relocation
scanning complete because position of each section might change due
thunk creation.
- The patch introduces new dedicated value for DefinedSynthetic symbols
DefinedSynthetic::SectionEnd. Synthetic symbol with that value always
points to the end of the corresponding output section. That allows to
escape updating synthetic symbols if output sections sizes changes after
relocation scanning due thunk creation.
- In the `InputSection::writeTo` method we write thunks after corresponding
input section. Each thunk is written by calling `TargetInfo::writeThunk` method.
- The patch supports the only type of thunk code for each target. For now,
it is enough.
Differential Revision: http://reviews.llvm.org/D17934
llvm-svn: 265059
We have to check the final value that is written.
I don't think this has any real word implications (unless something
supports unaligned instructions), but unblocks simplifying the handling
of PC relative relocations.
llvm-svn: 265009
If we make R_MIPS_LO16 a relative relocation, linker:
- never creates R_MIPS_COPY relocation for it
- attempts to create R_MIPS_REL32 dynamic relocation if R_MIPS_LO16's
target is a preemptible symbol
Differential Revision: http://reviews.llvm.org/D18607
llvm-svn: 264956
Windows seems to complain that the file cannot be removed because
it is still in use. We don't have to remove the file but instead
just overwrite it, so do that.
llvm-svn: 264915
This patch add a TLS relax optimization test when transforming
Initial-Exec to Local-Exec for local symbols (which can not be preempted).
llvm-svn: 264903
gold and bfd do not include the undefined locals in symtab.
We have no reasons to support that either.
That fixes PR27016
Differential revision: http://reviews.llvm.org/D18554
llvm-svn: 264843
Local symbol which requires GOT entry initialized by "page" address.
This address is high 16 bits of sum of the symbol value and the relocation
addend. In the relocation scanning phase final values of symbols are unknown
so to reduce number of allocated GOT entries do the following trick. Save
all output sections referenced by GOT relocations during the relocation
scanning phase. Then later in the `GotSection::finalize` method calculate
number of "pages" required to cover all saved output sections and allocate
appropriate number of GOT entries. We assume the worst case - each 64kb
page of the output section has at least one GOT relocation against it.
Differential Revision: http://reviews.llvm.org/D18349
llvm-svn: 264730
When R_X86_64_PC32/R_X86_64_32 relocations are
used against preemptible symbol and output is position independent,
error should be generated.
Differential revision: http://reviews.llvm.org/D18190
llvm-svn: 264707
Some optimizations, e.g. SimplifyLibCalls, can replace functions with
others as part of the lowering, e.g. printf => puts.
The new symbols don't have the isUsedInRegularObj flag set so they
don't get included in the final symbol table (and dynamic symbol
table), and the dynamic linker gets confused. Include them as a fix.
Differential Revision: http://reviews.llvm.org/D18357
llvm-svn: 264688
IPO doesn't work very well across symbols referenced
by others TUs. The linker here tries to evaluate
which symbols are safe to internalize and switches
their linkage.
Differential Revision: http://reviews.llvm.org/D18415
llvm-svn: 264585
This patch simplifies the isRelRelative for AArch64 and add the missing
ones for bootstrap and test-suite. It also adds more testing for
shared object creation.
llvm-svn: 264322
The .o path always makes sure to store a power of 2 value in the
Section alignment. However, the YAML code didn't verify this.
Added verification and updated all the tests which had a 3 but meant
to have 2^3.
llvm-svn: 264228
Ensure we keep the symbol we need to before it reaches
the Writer (and hit an assertion), changing its linkage
from linkonce_odr to weak. For a more detailed description
of the problem, see PR19901 where a similar problem was
fixed for the gold plugin. Thanks to Rafael for providing
a testcase.
llvm-svn: 264111
When a tls access is optimized, a group of relocations is converted at a
time.
We were already skipping relocations that were optimized out in
relocate, but not in scanRelocs.
This is a small optimization. I got here while working on a patch that
will always keep scanRelocs and relocate in sync.
llvm-svn: 264048
Rafael Espindola