This patch implements a new design for the symbol table that stores
SymbolBodies within a memory region of the Symbol object. Symbols are mutated
by constructing SymbolBodies in place over existing SymbolBodies, rather
than by mutating pointers. As mentioned in the initial proposal [1], this
memory layout helps reduce the cache miss rate by improving memory locality.
Performance numbers:
old(s) new(s)
Without debug info:
chrome 7.178 6.432 (-11.5%)
LLVMgold.so 0.505 0.502 (-0.5%)
clang 0.954 0.827 (-15.4%)
llvm-as 0.052 0.045 (-15.5%)
With debug info:
scylla 5.695 5.613 (-1.5%)
clang 14.396 14.143 (-1.8%)
Performance counter results show that the fewer required indirections is
indeed the cause of the improved performance. For example, when linking
chrome, stalled cycles decreases from 14,556,444,002 to 12,959,238,310, and
instructions per cycle increases from 0.78 to 0.83. We are also executing
many fewer instructions (15,516,401,933 down to 15,002,434,310), probably
because we spend less time allocating SymbolBodies.
The new mechanism by which symbols are added to the symbol table is by calling
add* functions on the SymbolTable.
In this patch, I handle local symbols by storing them inside "unparented"
SymbolBodies. This is suboptimal, but if we do want to try to avoid allocating
these SymbolBodies, we can probably do that separately.
I also removed a few members from the SymbolBody class that were only being
used to pass information from the input file to the symbol table.
This patch implements the new design for the ELF linker only. I intend to
prepare a similar patch for the COFF linker.
[1] http://lists.llvm.org/pipermail/llvm-dev/2016-April/098832.html
Differential Revision: http://reviews.llvm.org/D19752
llvm-svn: 268178
There seems to be no reason to keep st_size of undefined symbols.
This patch removes the member for it. This patch will change outputs
in cases that undefined symbols are copied to output, but I think
this is unimportant.
Differential Revision: http://reviews.llvm.org/D19574
llvm-svn: 267826
The semantics of the -u flag are to load the lazy symbol named by the flag. We
were previously relying on this behavior falling out of symbol resolution
against a synthetic undefined symbol, but that didn't quite give us the
correct behavior, so we needed a flag to mark symbols created with -u so
we could treat them specially in the writer. However, it's simpler and less
error prone to implement the required behavior directly and remove the flag.
This fixes an issue where symbols loaded with -u would receive hidden
visibility even when the definition in an object file had wider visibility.
Differential Revision: http://reviews.llvm.org/D19560
llvm-svn: 267639
Fixes check-llvm when bootstrapping.
Also remove mostly dead and most likely incorrect logic regarding preemption
of weak undefined symbols.
llvm-svn: 267314
This patch only implements support for version scripts of the form:
{ [ global: symbol1; symbol2; [...]; symbolN; ] local: *; };
No wildcards are supported, other than for the local entry. Symbol versioning
is also not supported.
It works by introducing a new Symbol flag which tracks whether a symbol
appears in the global section of a version script.
This patch also simplifies the logic in SymbolBody::isPreemptible(), and
teaches it to handle the case where symbols with default visibility in DSOs
do not appear in the dynamic symbol table because of a version script.
Fixes PR27482.
Differential Revision: http://reviews.llvm.org/D19430
llvm-svn: 267208
These are properties of a symbol name, rather than a particular instance
of a symbol in an object file. We can simplify the code by collecting these
properties in Symbol.
The MustBeInDynSym flag has been renamed ExportDynamic, as its semantics
have been changed to be the same as those of --dynamic-list and
--export-dynamic-symbol, which do not cause hidden symbols to be exported.
Differential Revision: http://reviews.llvm.org/D19400
llvm-svn: 267183
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
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
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
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
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
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
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
DefinedElf was a superclass of DefinedRegular and SharedSymbol classes
and represented the notion of defined symbols created for ELF symbols.
It turned out that we didn't use that class often. We had only two
occurrences of dyn_cast'ing to DefinedElf, and both were easily
rewritten without it.
The class was also a bit confusing. The concept of "created for ELF
symbol" is orthogonal to defined/undefined types. However, we had
two distinct classes, DefinedElf and UndefinedElf.
This patch simply removes the class. Now the class hierarchy is one
level shallower.
llvm-svn: 265234
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
-warn-common
Warn when a common symbol is combined with another common symbol
or with a symbol definition. Unix linkers allow this somewhat
sloppy practice, but linkers on some other operating systems do
not. This option allows you to find potential problems from
combining global symbols.
Differential revision: http://reviews.llvm.org/D17998
llvm-svn: 263413
We want to make SymbolBody the central place to query symbol information.
This patch also renames canBePreempted to isPreemptible because I feel that
the latter is slightly better (the former is three words and the latter
is two words.)
llvm-svn: 263386
which was reverted because included
unrelative changes by mistake.
Original commit message:
[ELF] - Change all messages to lowercase to be consistent.
That is directly opposite to http://reviews.llvm.org/D18045,
which was reverted.
This patch changes all messages to start from lowercase letter if
they were not before.
That is done to be consistent with clang.
Differential revision: http://reviews.llvm.org/D18085
llvm-svn: 263337
That is directly opposite to http://reviews.llvm.org/D18045,
which was reverted.
This patch changes all messages to start from lowercase letter if
they were not before.
That is done to be consistent with clang.
Differential revision: http://reviews.llvm.org/D18085
llvm-svn: 263252
We can argue about a maximum alignment of a group of symbols,
but for each symbol, there is only one alignment.
So it is a bit weird that each symbol has a "maximum alignment".
llvm-svn: 263151
It looks a bit wierd that we have to initialize symbols for all ELFT
types when we use only one ELFT for link. We can only init those
that we need. Patch fixes it.
Differential revision: http://reviews.llvm.org/D18047
llvm-svn: 263133
Get rid of few accessors in that class, and replace
them with direct fields access.
Differential revision: http://reviews.llvm.org/D17879
llvm-svn: 262796
SymbolBody constructor and friends take isFunc and isTLS boolean arguments.
ELF symbols have already a type so than be easily passed as argument.
If we want to support another type, this scheme is not good enough, that is,
the current code logic would require passing another `bool isObject` around.
Up to two argument, this stretching exercise was a little bit goofy but
still acceptable, but with more types to support, is just too much, IMHO.
Change the code so that the type is passed instead.
Differential Revision: http://reviews.llvm.org/D17871
llvm-svn: 262684
https://docs.oracle.com/cd/E53394_01/html/E54766/u-etext-3c.html
It is said that:
_etext - The address of _etext is the first
location after the last read-only loadable segment.
_edata - The address of _edata is the first
location after the last read-write loadable segment.
_end - If the address of _edata is greater than the address
of _etext, the address of _end is same as the address of _edata.
In real life _end and _edata has different values for that case.
Both gold/bfd set _edata to the end of the last non SHT_NOBITS section.
This patch do the same for consistency.
It should fix the https://llvm.org/bugs/show_bug.cgi?id=26729.
Differential revision: http://reviews.llvm.org/D17601
llvm-svn: 262019
This patch implements the same algorithm as LLD/COFF's ICF. I'm
not going to repeat the same description about how it works, so you
want to read the comment in ICF.cpp in this patch if you want to know
the details. This algorithm should be more powerful than the ICF
algorithm implemented in GNU gold. It can even merge mutually-recursive
functions (which is harder than one might think).
ICF is a fairly effective size optimization. Here are some examples.
LLD: 37.14 MB -> 35.80 MB (-3.6%)
Clang: 59.41 MB -> 57.80 MB (-2.7%)
The lacking feature is "safe" version of ICF. This merges all
identical sections. That is not compatible with a C/C++ language
requirement that two distinct functions must have distinct addresses.
But as long as your program do not rely on the pointer equality
(which is in many cases true), your program should work with the
feature. LLD works fine for example.
GNU gold implements so-called "safe ICF" that identifies functions
that are safe to merge by heuristics -- for example, gold thinks
that constructors are safe to merge because there is no way to
take an address of a constructor in C++. We have a different idea
which David Majnemer suggested that we add NOPs at beginning of
merged functions so that two or more pointers can have distinct
values. We can do whichever we want, but this patch does not
include neither.
http://reviews.llvm.org/D17529
llvm-svn: 261912
This patch moves BitcodeFile instantiation into createObjectFile.
Previously, we handle bitcode files outside that function and did
not handle for --whole-archive.
http://reviews.llvm.org/D17527
llvm-svn: 261663
This is the function equivalent of a copy relocation.
Since functions are expected to change sizes, we cannot use copy
relocations. In situations where one would be needed, what is done
instead is:
* Create a plt entry
* Output an undefined symbol whose addr is the plt entry.
The dynamic linker makes sure any shared library uses the plt entry as
the function address.
llvm-svn: 260224
The variable was marking various cases where a symbol must be included
in the dynamic symbol table. Being used by a dynamic relocation was only
one of them.
llvm-svn: 259889
-Bsymbolic-functions:
When creating a shared library, bind references to global
function symbols to the definition within the shared library, if any.
This patch also fixed behavior of already existent -Bsymbolic:
previously PLT entries were created even if -Bsymbolic was specified.
Differential revision: http://reviews.llvm.org/D16411
llvm-svn: 259481
If object files are drawn from archive files, the error message should
be something like "conflict symbols in foo.a(bar.o) and baz.o" instead
of "conflict symbols in bar.o and baz.o". This patch implements that.
llvm-svn: 259475
Previously, the methods to get symbol addresses were somewhat scattered
in many places. You can use getEntryAddr returns the address of the symbol,
but if you want to get the GOT address for the symbol, you needed to call
Out<ELFT>::Got->getEntryAddr(Sym). This change adds new functions, getVA,
getGotVA, getGotPltVA, and getPltVA to SymbolBody, so that you can use
SymbolBody as the central place to ask about symbols.
http://reviews.llvm.org/D16710
llvm-svn: 259404
Main executable did not export symbols that exist both in the main executable and in DSOs before this patch.
Symbols from object files that override symbols in DSO should be added to .dynsym table.
Differential revision: http://reviews.llvm.org/D16405
llvm-svn: 258672
It seems that __cxa_demangle function on the buildbot tried to demangle
a variable "tlsvar" as a C++ symbol. Do not call that function unless
it does not start with "_Z" which is the prefix of mangled names.
llvm-svn: 257661
On MIPS O32 ABI, _gp_disp is a magic symbol designates offset between
start of function and gp pointer into GOT. To make seal with such symbol
we add new method addIgnoredStrong(). It adds ignored symbol with global
binding to prevent the symbol substitution. The addIgnored call is not
enough here because this call adds a weak symbol which might be
substituted by symbol from shared library.
Differential Revision: http://reviews.llvm.org/D16084
llvm-svn: 257449
There are 3 symbol types that a .bc can provide during lto: defined,
undefined, common.
Defined and undefined symbols have already been refactored. I was
working on common and noticed that absolute symbols would become an
oddity: They would be the only symbol type present in a .o but not in
a.bc.
Looking a bit more, other than the special section number they were only
used for special rules for computing values. In that way they are
similar to TLS, and we don't have a DefinedTLS.
This patch deletes it. With it we have a reasonable rule of the thumb
for having a symbol kind: It exists if it has special resolution
semantics.
llvm-svn: 256383
I am working on adding LTO support to the new ELF lld.
In order to do that, it will be necessary to represent defined and
undefined symbols that are not from ELF files. One way to do it is to
change the symbol hierarchy to look like
Defined : SymbolBody
Undefined : SymbolBody
DefinedElf<ELFT> : Defined
UndefinedElf<ELFT> : Undefined
Another option would be to use bogus Elf_Sym, but I think that is
getting a bit too hackish.
This patch does the Undefined/UndefinedElf. Split. The next one
will do the Defined/DefinedElf split.
llvm-svn: 256289
The ELF spec says:
... if any reference to or definition of a name is a symbol with a
non-default visibility attribute, the visibility attribute must be
propagated to the resolving symbol in the linked object. If different
visibility attributes are specified for distinct references to or
definitions of a symbol, the most constraining visibility attribute
must be propagated to the resolving symbol in the linked object. The
attributes, ordered from least to most constraining, are:
STV_PROTECTED, STV_HIDDEN and STV_INTERNAL.
llvm-svn: 246603
Now that we print a symbol table and all symbol kinds are at least declared,
we can test all combinations that don't produce an error.
This also includes a few fixes to keep the test passing:
* Keep the strong symbol in a weak X strong pair
* Handle common symbols.
The common X common case will be finished in a followup patch.
llvm-svn: 246401
This is a direct port of the new PE/COFF linker to ELF.
It can take a single object file and generate a valid executable that executes at the first byte in the text section.
llvm-svn: 242088
Peter Collingbourne