In the resolver, we maintain a list of undefined symbols, and when we
visit an archive file, we check that file if undefined symbols can be
resolved using files in the archive. The archive file class provides
find() function to lookup a symbol.
Previously, we call find() for each undefined symbols. Archive files
may be visited multiple times if they are in a --start-group and
--end-group. If we visit a file M times and if we have N undefined
symbols, find() is called M*N times. I found that that is one of the
most significant bottlenecks in LLD when linking a large executable.
find() is not a very cheap operation because it looks up a hash table
for a given string. And a string, or a symbol name, can be pretty long
if you are dealing with C++ symbols.
We can eliminate the bottleneck.
Calling find() with the same symbol multiple times is a waste. If a
result of looking up a symbol is "not found", it stays "not found"
forever because the symbol simply doesn't exist in the archive.
Thus, we should call find() only for newly-added undefined symbols.
This optimization makes O(M*N) O(N).
In this patch, all undefined symbols are added to a vector. For each
archive/shared library file, we maintain a start position P. All
symbols [0, P) are already searched. [P, end of the vector) are not
searched yet. For each file, we scan the vector only once.
This patch changes the order in which undefined symbols are looked for.
Previously, we iterated over the result of _symbolTable.undefines().
Now we iterate over the new vector. This is a benign change but caused
differences in output if remaining undefines exist. This is why some
tests are updated.
The performance improvement of this patch seems sometimes significant.
Previously, linking chrome.dll on my workstation (Xeon 2.4GHz 8 cores)
took about 70 seconds. Now it takes (only?) 30 seconds!
http://reviews.llvm.org/D8091
llvm-svn: 231434
_reverseRef is a multimap from atoms to atoms. The map contains
reverse edges of "layout-before" and "group" edges for dead-stripping.
The type of the variable was DenseMap<Atom *, DenseSet<Atom *>>.
This patch changes that to std::unordered_multimap<Atom *, Atom *>.
A DenseMap with a value type of DenseSet was not fast. Inserting 900k
items to the map took about 1.6 seconds on my workstation.
unordered_multimap on the other hand took only 0.6 seconds.
Use of the map also got faster -- originally markLive took 1.3 seconds
in the same test case, and it now took 1.0 seconds. In total we shove
off 1.3 seconds out of 27 seconds in that test case.
llvm-svn: 231432
We maintain a map from symbols to archive files for the archive file
pre-loading. That map is created at the beginning of the resolve()
and is never updated. However, the input file list may be updated by
File::beforeLink(). This is a patch to update the map after beforeLink.
llvm-svn: 231395
Merge::mergeByLargestSection is half-baked since it's defined
in terms of section size, there's no way to get the section size
of an atom.
Currently we work around the issue by traversing the layout edges
to both directions and calculate the sum of all atoms reachable.
I wrote that code but I knew it's hacky. It's even not guaranteed
to work. If you add layout edges before the core linking, it
miscalculates a size.
Also it's of course slow. It's basically a linked list traversal.
In this patch I added DefinedAtom::sectionSize so that we can use
that for mergeByLargestSection. I'm not very happy to add a new
field to DefinedAtom base class, but I think it's legitimate since
mergeByLargestSection is defined for section size, and the section
size is currently just missing.
http://reviews.llvm.org/D7966
llvm-svn: 231290
File objects are not really const in the resolver. We set ordinals to
them and call beforeLink hooks. Also, File's member functions marked
as const are not really const. ArchiveFile never returns the same
member file twice, so it remembers files returned before. find() has
side effects.
In order to deal with the inconsistencies, we sprinkled const_casts
and marked member varaibles as mutable.
This patch removes const from there to reflect the reality.
llvm-svn: 231212
std::promise and std::future in old version of libstdc++ are buggy.
I think that's the reason why LLD tests were flaky on Ubuntu 13
buildbots until we disabled file preloading.
In this patch, I implemented very simple future and used that in
FileArchive. Compared to std::promise and std::future, it lacks
many features, but should serve our purpose.
http://reviews.llvm.org/D8025
llvm-svn: 231153
"virtual" was present at a wrong place. FileArchive is a subclass of
ArchiveLibraryFile, and a FileArchive can be deleted through a
pointer of ArchiveLibraryFile. We want to make the destructor of the
base class virtual.
llvm-svn: 231033
In doParse, we shouldn't do anything that has side effects. That function may be
called speculatively and possibly in parallel.
We called WinLinkDriver::parse from doParse to parse a command line in a .drectve
section. The parse function updates a linking context object, so it has many side
effects. It was not safe to call that function from doParse. beforeLink is a
function for a File object to do something that has side effects. Moving a call
of WinLinkDriver::parse to there.
llvm-svn: 230791
Nothing wrong with reinterpret_cast<llvm::support::ulittle32_t *>(loc),
but that's redundant and not great from readability point of view.
The new functions are wrappers for that kind of reinterpet_casts.
Surprisingly or unsurprisingly, there was no use of big endian read
and write. {read,write}{16,32,64}be have no user. But I think they
still worth to be there in the header for completeness.
http://reviews.llvm.org/D7927
llvm-svn: 230725
Previously we have a string -> string map to keep the weak alias
symbol mapping. Naturally we can't define more than one weak alias
with that data structure.
This patch is to allow multiple aliases for the same symbol by
changing the map type to string -> set of string map.
llvm-svn: 230702
This is mainly for back-compatibility with GNU ld.
Ideally --stats should be a general option in LinkingContext, providing
individual stats for every pass in the linking process.
In the GNU driver, a better wording could be used, but there's no need
to change it for now.
Differential Revision: D7657
Reviewed by: ruiu
llvm-svn: 230157
The round-trip passes were introduced in r193300. The intention of
the change was to make sure that LLD is capable of reading end
writing such file formats.
But that turned out to be yet another over-designed stuff that had
been slowing down everyday development.
The passes ran after the core linker and before the writer. If you
had an additional piece of information that needs to be passed from
front-end to the writer, you had to invent a way to save the data to
YAML/Native. These passes forced us to do that even if that data
was not needed to be represented neither in an object file nor in
an executable/DSO. It doesn't make sense. We don't need these passes.
http://reviews.llvm.org/D7480
llvm-svn: 230069
LinkerScript AST nodes are never destroyed which means that their
std::vector members will never be destroyed.
Instead, allocate the operand list itself in the Parser's
BumpPtrAllocator. This ensures that the storage will be destroyed along
with the nodes when the Parser is destroyed.
llvm-svn: 229967
Wrap functionality was using a std::set to record symbols that need to be
wrapped. This changes the implementation to use a StringSet instead.
No change in functionality.
llvm-svn: 229165
Use a wrapper function for symbol. Any undefined reference to symbol will be
resolved to "__wrap_symbol". Any undefined reference to "__real_symbol" will be
resolved to symbol.
This can be used to provide a wrapper for a system function. The wrapper
function should be called "__wrap_symbol". If it wishes to call the system
function, it should call "__real_symbol".
Here is a trivial example:
void * __wrap_malloc (size_t c)
{
printf ("malloc called with %zu\n", c);
return __real_malloc (c);
}
If you link other code with this file using --wrap malloc, then all calls
to "malloc" will call the function "__wrap_malloc" instead. The call to
"__real_malloc" in "__wrap_malloc" will call the real "malloc" function.
llvm-svn: 228906
This adds the LinkingContext parameter to the ELFReader. Previously the flags in
that were needed in the Context was passed to the ELFReader, this made it very
hard to access data structures in the LinkingContext when reading an ELF file.
This change makes the ELFReader more flexible so that required parameters can be
grabbed directly from the LinkingContext.
Future patches make use of the changes.
There is no change in functionality though.
llvm-svn: 228905
The values are already arranged in ascending order, and all tests still pass.
Removing the values as its confusing when new enumerations need to be added.
llvm-svn: 228381
Use the environment variable "LLD_RUN_ROUNDTRIP_TEST" in the test that you want
to disable, as
RUN: env LLD_RUN_ROUNDTRIP_TEST= <run>
This was a patch that I made, but I find this a better way to accomplish what we
want to do.
llvm-svn: 228376
Only search library directories explicitly specified
on the command line. Library directories specified in linker
scripts (including linker scripts specified on the command
line) are ignored.
llvm-svn: 228375
Previously we only have File::path() to get the path name of a file.
If a file was a member of an archive file, path() returns a concatenated
string of the file name in the archive and the archive file name.
If we wanted to get a file name or an archive file name, we had to
parse that string. That's of course not good.
This patch adds new member functions, archivePath and memberPath, to File.
http://reviews.llvm.org/D7447
llvm-svn: 228352
The real user of the LayoutPass is now only Mach-O, so move that
pass out of the common directory to Mach-O directory.
"Core" architecture were using the LayoutPass. I modified that
to use a simple OrderPass. I think no one actually have authority
what feature should be in Core and what's not, but I believe the
LayoutPass is not very suitable for Core. Before more code starts
depending on the complex pass, it's better to remove that from
Core.
I could have simplified that pass because Mach-O is the only user
of the LayoutPass. For example, the second parameter of the
LayoutPass constructor can be converted from optional to mandatory.
I didn't do that in this patch to keep it simple. I'll do in a
followup patch.
http://reviews.llvm.org/D7311
llvm-svn: 228341
INPUT directive is a variant of GROUP in the sense that that specifies
a list of input files. The only difference is whether the entire file
list is wrapped with a --start-group/--end-group or not.
http://reviews.llvm.org/D7390
llvm-svn: 228060
Currently, no one owns script::Parser buffers, but yet ELFLinkingContext gets
updated with StringRef pointers to data inside Parser buffers. Since this buffer
is locally owned inside GnuLdDriver::evalLinkerScript(), as soon as this
function finishes, all pointers in ELFLinkingContext that comes from linker
scripts get invalid. The problem is that we need someone to own linker scripts
data structures and, since ELFLinkingContext transports references to linker
scripts data, we can simply make it also own all linker scripts data.
Differential Revision: http://reviews.llvm.org/D7323
llvm-svn: 227975
This is needed, among others by the FreeBSD kernel linker script.
Patch by Davide Italiano!
Reviewers: ruiu, rafaelauler
Differential Revision: http://reviews.llvm.org/D7220
llvm-svn: 227694
These fields were made protected in r193585. The aim of that change is to
expose these fields to SimpleFileWrapper. Because SimpleFileWrapper class
was removed in r227549, we can make them private.
llvm-svn: 227672
Previously we applied the LayoutPass to order atoms and then
apply elf::ArrayOrderPass to sort them again. The first pass is
basically supposed to sort atoms in the normal fashion (which
is to sort symbols in the same order as the input files).
The second pass sorts atoms in {init,fini}_array.<priority> by
priority.
The problem is that the LayoutPass is overkill. It analyzes
references between atoms to make a decision how to sort them.
It's slow, hard to understand, and above all, it doesn't seem
that we need its feature for ELF in the first place.
This patch remove the LayoutPass from ELF pass list. Now all
reordering is done in elf::OrderPass. That pass sorts atoms by
{init,fini}_array, and if they are not in the special section,
they are ordered as the same order as they appear in the command
line. The new code is far easier to understand, faster, and
still able to create valid executables.
Unlike the previous layout pass, elf::OrderPass doesn't count
any attributes of an atom (e.g. permissions) except its
position. It's OK because the writer takes care of them if we
have to.
This patch changes the order of final output, although that's
benign. Tests are updated.
http://reviews.llvm.org/D7278
llvm-svn: 227666
SimpleFileWrapper was a class to wrap an existing (possibly non-mutable)
file as a mutable file. We used instances of the class in RoundTrip*
passes, because the passes convert mutable files to non-mutable files,
and we needed to convert them back to mutable.
That feature can be implemented without defining a new class. Generally
speaking, if we can implement a feature without defining a class and
using only public interface of exsiting classes, that's preferred way
to do that. And this is the case.
llvm-svn: 227549
