equalsConstants() is the heaviest function in ICF, and that consumes
more than half of total ICF execution time. Of which, section content
comparison accounts for roughly one third.
Previously, we compared section contents at the beginning of the
function after comparing their checksums. The comparison is very
likely to succeed because when the control reaches that comparison,
their checksums are always equal. And because checksums are 64-bit
CRC, they are unlikely to collide.
We compared relocations and associative sections after that.
If they are different, the time we spent on byte-by-byte comparison
of section contents were wasted.
This patch moves the comparison at the end of function. If the
comparison fails, the time we spent on relocation comparison are
wasted, but as I wrote it's very unlikely to happen.
LLD took 1198 ms to link itself to produce a 27.11 MB executable.
Of which, ICF accounted for 536 ms. This patch cuts it by 90 ms,
which is 17% speedup of ICF and 7.5% speedup overall. All numbers
are median of ten runs.
llvm-svn: 247961
Basically the concept of "liveness" is for sections (or chunks in LLD
terminology) and not for symbols. Symbols are always available or live,
or otherwise it indicates a link failure.
Previously, we had isLive() and markLive() methods for DefinedSymbol.
They are confusing methods. What they actually did is to act as a proxy
to backing section chunks. We can simplify eliminate these methods
and call section chunk's methods directly.
llvm-svn: 247869
Only live symbols are written to the symbol table. Because isLive()
returned false if dead-stripping was disabled entirely, only
non-COMDAT sections were written to the symbol table. This patch fixes
the issue.
llvm-svn: 247856
This patch defines ICF class and defines ICF-related functions as
members of the class. By doing this we can move code that are
related only to ICF from SectionChunk to the newly-defined class.
This also eliminates a global variable "NextID".
llvm-svn: 247802
This is a patch to make LLD to be on par with MSVC in terms of ICF
effectiveness. MSVC produces a 27.14MB executable when linking LLD.
LLD previously produced a 27.61MB when self-linking. Now the size
is reduced to 27.11MB. Note that without ICF the size is 29.63MB.
In r247387, I implemented an algorithm that handles section graphs
as cyclic graphs and merge them using SCC. The algorithm did not
always work as intended as I demonstrated in r247721. The new
algortihm implemented in this patch is different from the previous
one. If you are interested the details, you want to read the file
comment of ICF.cpp.
llvm-svn: 247770
Previously, LLD's ICF couldn't merge cyclic graphs. That was unfortunate
because, in COFF, cyclic graphs are not exceptional at all. That is
pretty common.
In this patch, sections are grouped by Tarjan's strongly connected
component algorithm to get acyclic graphs. And then we try to merge
SCCs whose outdegree is zero, and remove them from the graph. This
makes other SCCs to have outdegree zero, so we can repeat the
process until all SCCs are removed. When comparing two SCCs, we handle
cycles properly.
This algorithm works better than previous one. Previously, self-linking
produced a 29.0MB executable. It now produces a 27.7MB. There's still some
gap compared to MSVC linker which produces a 27.1MB executable for the
same input. So the gap is narrowed, but still LLD is not on par with MSVC.
I'll investigate that later.
llvm-svn: 247387
Identical COMDAT Folding is a feature to merge COMDAT sections
by contents. Two sections are considered the same if their contents,
relocations, attributes, etc, are all the same.
An interesting fact is that MSVC linker takes "iterations" parameter
for ICF because the algorithm they are using is iterative. Merging
two sections could make more sections to be mergeable because
different relocations could now point to the same section. ICF is
repeated until we get a convergence (until no section can be merged).
This algorithm is not fast. Usually it needs three iterations until a
convergence is obtained.
In the new algorithm implemented in this patch, we consider sections
and relocations as a directed acyclic graph, and we try to merge
sections whose outdegree is zero. Sections with outdegree zero are then
removed from the graph, which makes other sections to have outdegree
zero. We repeat that until all sections are processed. In this
algorithm, we don't iterate over the same sections many times.
There's an apparent issue in the algorithm -- the section graph is
not guaranteed to be acyclic. It's actually pretty often cyclic.
So this algorithm cannot eliminate all possible duplicates.
That's OK for now because the previous algorithm was not able to
eliminate cycles too. I'll address the issue in a follow-up patch.
llvm-svn: 246878
Previously, we calculated our own hash values for section contents.
Of coruse that's slow because we had to access all bytes in sections.
Fortunately, COFF objects usually contain hash values for COMDAT
sections. We can use that to speed up Identical COMDAT Folding.
llvm-svn: 246869
The option is added in MSVC 2015, and there's no documentation about
what the option is. This patch is to ignore the option for now, so that
at least LLD is usable with MSVC 2015.
llvm-svn: 246780
This patch fixes a subtle incompatibility with MSVC linker.
MSVC linker preserves the original spelling of a DLL in the
import descriptor table. LLD previously converted all
characters to lowercase. Usually this difference is benign,
but if a program explicitly checks for DLL file names, the
program could fail.
llvm-svn: 246620
In r246424, I made a change that disables non-DLL to export
symbols. It turned out that the change was not correct. Both
DLLs and executables are able to export symbols (although the
latter is relatively rare). This change restores the feature.
llvm-svn: 246537
I have totally no idea why, but MSVC linker is sensitive about
file names of archive members. If we do not make import library
file names to the same as the DLL name, MSVC link *crashes*
when it is processing the library file. This patch is to set
the same name.
llvm-svn: 246535
The rules for dllexported symbols are overly complicated due to
x86 name decoration, fuzzy symbol resolution, and the fact that
one symbol can be resolved by so many different names. The rules
are probably intended to be "intuitive", so that users don't have
to understand the name mangling schemes, but it seems that it can
lead to unintended symbol exports.
To make it clear what I'm trying to do with this patch, let me
write how the export rules are subtle and complicated.
- x86 name decoration: If machine type is i386 and export name
is given by a command line option, like /export:foo, the
real symbol name the linker has to search for is _foo because
all symbols are decorated with "_" prefixes. This doesn't happen
on non-x86 machines. This automatic name decoration happens only
when the name is not C++ mangled.
However, the symbol name exported from DLLs are ones without "_"
on all platforms.
Moreover, if the option is given via .drectve section, no
symbol decoration is done (the reason being that the .drectve
section is created by a compiler and the compiler should always
know the exact name of the symbol, I guess).
- Fuzzy symbol resolution: In addition to x86 name decoration,
the linker has to look for cdecl or C++ mangled symbols
for a given /export. For example, it searches for not only
_foo but also _foo@<number> or ??foo@... for /export:foo.
Previous implementation didn't get it right. I'm trying to make
it as compatible with MSVC linker as possible with this patch
however the rules are. The new code looks a bit messy to me, but
I don't think it can be simpler due to the ad-hoc-ness of the rules.
llvm-svn: 246424
This is exposed via a new flag /opt:lldltojobs=N, where N is the number of
code generation threads.
Differential Revision: http://reviews.llvm.org/D12309
llvm-svn: 246342
lib.exe has a feature to create import library files (which contain
short import files) from module-definition files. Previously, we were
using that feature, but it turned out that the feature is not complete
for us.
There seems no way to specify "Import Types" in module-definition file.
lib.exe always adds "_" to given symbols and specify IMPORT_NAME_UNDECORATE.
We need more fine-grainded control on that value.
This patch teaches LLD to create short import files itself.
We are still using lib.exe, but the use of the tool is limited to create
empty import library files. We then create short import files and add them
to the empty files as new members.
This patch does not intend to change the functionality. LLD produces
the same import libraries as before. I'll make another change to create
different import libraries in a follow-up patch.
llvm-svn: 246292
__NULL_IMPORT_DESCRIPTOR is a symbol used by MSVC liner to construct
the import descriptor table. We do not use the symbol. Previously,
we had code to skip that symbol. That code does not actually do
anything meaningful because no one is referencing the symbol, the
symbol would naturally be ignored. This patch stops recognizing
the symbol.
llvm-svn: 245280
Previously, weak external symbols could reference only symbols that
appeared before them. Although that covers almost all use cases
of weak externals, there are object files out there which contains
weak externals that have forward references.
This patch supports such weak externals.
llvm-svn: 245258
There are some DLLs whose initializers depends on other DLLs'
initializers. The initialization order matters for them.
MSVC linker uses the order of the libraries from the command line.
LLD used ASCII-betical order. So they were incompatible.
This patch makes LLD compatible with MSVC.
llvm-svn: 245201
This is more convenient than the offset from the start of the file as we
don't have to worry about it changing when we move the output section.
This is a port of r245008 from ELF.
llvm-svn: 245018
Sections must start at page boundaries in memory, but they
can be aligned to sector boundaries (512-bytes) on disk.
We aligned them to 4096-byte boundaries even on disk, so we
wasted disk space a bit.
llvm-svn: 244691
MSVC 2015's load configuration object (__load_config_used) contains
references to these symbols. I don't fully understand how it works,
but looks like these symbols are linker-defined ones. So I define them
here in the Driver. With this patch, LLD can self-host with MSVC 2015.
This patch is to link MSVC 2015-produced object files. It does not
implement Control Flow Protection. If I understand correctly, the
linker has to create a bitmap of function entry point addresses for
the CFG runtime. We don't do that yet. Produced executables will not
be protected by CFG.
llvm-svn: 244425
SymbolTable::find(mangle(X)) is equivalent to SymbolTable::findUnderscore(X)
except that the latter is slightly efficient as that doesn't allocate a new
string.
llvm-svn: 244377
This has a few advantages
* Less C++ code (about 300 lines less).
* Less machine code (about 14 KB of text on a linux x86_64 build).
* It is more debugger friendly. Just set a breakpoint on the exit function and
you get the complete lld stack trace of when the error was found.
* It is a more robust API. The errors are handled early and we don't get a
std::error_code hot potato being passed around.
* In most cases the error function in a better position to print diagnostics
(it has more context).
llvm-svn: 244215
Various parameters are passed implicitly using Config global variable
already. Output file path is no different from others, so there was no
special reason to handle that differnetly.
This patch changes the signature of writeResult(SymbolTable *, StringRef)
to writeResult(SymbolTable *).
llvm-svn: 244180
We were printing an error but exiting with 0.
Not sure how to test this. We could add a no-winlib feature,
but that is probably not worth it.
llvm-svn: 244109
I don't remember why I thought that only functions are subject
of garbage collection, but the comment here said so, which is
not correct. Moreover, the code just below the comment does not
do what the comment says -- it handles non-COMDAT, non-function
sections as GC root. As a result, it just handles non-COMDAT
sections as GC root.
This patch cleans that up by removing SectionChunk::isRoot and
use isCOMDAT instead.
llvm-svn: 243700
We want to convince the NT loader not to map these sections into memory.
A good first step is to move them to the end of the executable.
Differential Revision: http://reviews.llvm.org/D11655
llvm-svn: 243680
We create a module-definition file and give that to lib.exe to
create an import library file. A module-definition has to be
syntactically and semantically correct, of course.
There was a case that we created a module-definition file that
lib.exe would complain for duplicate entries. If a user gives
an unmangled and mangled name for the same symbol, we would end
up having two duplicate lines for the mangled name in a module-
definition file.
This patch fixes that issue by uniquefying entries by mangled
symbol name.
llvm-svn: 243587
Windows ARM is the thumb ARM environment, and pointers to thumb code
needs to have its LSB set. When we apply relocations, we need to
adjust the LSB if it points to an executable section.
llvm-svn: 243560
SECREL should sets the 32-bit offset of the target from the beginning
of *target's* output section. Previously, the offset from the beginning
of source's output section was used instead.
SECTION means the target section's index, and not the source section's
index. This patch fixes that issue too.
llvm-svn: 243535
I don't fully understand the rationale behind the name mangling
scheme used for the DLL export table and the import library.
Why only leading "_" is dropped for the import library while
both "_" and "@" are dropped from DLL symbol table? But this seems
to be what MSVC linker does.
llvm-svn: 243490
The linker is now able to link not only LLVM/Clang/LLD for x86 but
even larger programs. I confirmed that it successsfully linked Chrome
for x86. Because the browser is a pretty large program, I think I can
say that the linker is now mostly feature complete. (I'm pretty sure
that there are hidden bugs somewhere, but they shouldn't be significant.)
llvm-svn: 243377
Previously, we ignore /merge option if /debug is specified
because I thought that was MSVC linker did. This was wrong.
/merge shouldn't be ignored even in debug mode.
llvm-svn: 243375
Leaving them in an executable is basically harmless but wastes disk space.
Because no one is using non-DWARF debug info linked by LLD, we can just
remove them.
llvm-svn: 243364
On x64 and x86, we use only one base relocation type, so we handled
base relocations just as a list of RVAs. That doesn't work well for
ARM becuase we have to handle two types of base relocations on ARM.
This patch changes the type of base relocation from uint32_t to
{reltype, uint32_t} to make it easy to port this code to ARM.
llvm-svn: 243197
In many places we assumed that is64() means AMD64 and i386 otherwise.
This assumption is not sound because Windows also supports ARM.
The linker doesn't support ARM yet, but this is a first step.
llvm-svn: 243188
An object file compatible with Safe SEH contains a .sxdata section.
The section contains a list of symbol table indices, each of which
is an exception handler function. A safe SEH-enabled executable
contains a list of exception handler RVAs. So, what the linker has
to do to support Safe SEH is basically to read the .sxdata section,
interpret the contents as a list of symbol indices, unique-fy and
sort their RVAs, and then emit that list to .rdata. This patch
implements that feature.
llvm-svn: 243182
__ImageBase is a special symbol whose value is the image base address.
Previously, we handled __ImageBase symbol as an absolute symbol.
Absolute symbols point to specific locations in memory and the locations
never change even if an image is base-relocated. That means that we
don't have base relocation entries for absolute symbols.
This is not a case for __ImageBase. If an image is base-relocated, its
base address changes, and __ImageBase needs to be shifted as well.
So we have to have base relocations for __ImageBase. That means that
__ImageBase is not really an absolute symbol but a different kind of
symbol.
In this patch, I introduced a new type of symbol -- DefinedRelative.
DefinedRelative is similar to DefinedAbsolute, but it has not a VA but RVA
and is a subject of base relocation. Currently only __ImageBase is of
the new symbol type.
llvm-svn: 243176
Load Configuration field points to a structure containing information
for SEH. That data strucutre is not created by the linker but provided
by an external file. What we have to do is just to set __load_config_used
address to the header.
llvm-svn: 242427
If a symbol is exported as /export:foo, and foo is resolved as a
mangled name (_foo@<number> or ?foo@@Y...), that mangled name should
be written to the export table. Previously, we wrote the original
name to the export table.
llvm-svn: 242342
Because thunks for dllimported symbols contain absolute addresses on x86,
they need to be relocated at load-time. This bug was a cause of crashes
in DLL initialization routines.
llvm-svn: 242259
I am adding support for thin archives. On those, getting the buffer
involves reading another file.
Since we only need an id in here, use the member offset in the archive.
llvm-svn: 242205
Entry name selection rule is already complicated on x64, but it's more
complicated on x86 because of the underscore name mangling scheme.
If one of _main, _main@<number> (a C function) or ?main@@... (a C++ function)
is defined, entry name is _mainCRTStartup. If _wmain, _wmain@<number or
?wmain@@... is defined, entry name is _wmainCRTStartup. And so on.
llvm-svn: 242110
If /delayload option is given, we have to resolve __delayLoadHelper2
since the function is the dynamic loader to delay-load DLLs.
The function name is mangled in x86 as ___delayLoadHelper2@8.
llvm-svn: 242078
clang-cl doesn't compile std::atomic_flag correctly (PR24101). Since the COFF
linker doesn't use threads yet, just revert r241420 and r241481 for now to
work around this clang-cl bug.
llvm-svn: 242006
Symbol foo is mangled as _foo in C and ?foo@@... in C++ on x86.
findMangle has to remove prefix underscore before mangle a given name
as a C++ symbol.
llvm-svn: 241874
Symbol names are usually mangled by appending "_" prefix on x86.
But the mangled name is not used in DLL export table. The export
table contains unmangled names.
llvm-svn: 241872
With this patch, LLD is now able to self-link an .exe file for x86
that runs correctly, although I don't think some headers (particularly
SEH) are not correct. DLL support is coming soon.
llvm-svn: 241857
Previously, we infer machine type at the very end of linking after
all symbols are resolved. That's actually too late because machine
type affects how we mangle symbols (whether or not we need to
add "_").
For example, /entry:foo adds "_foo" to the symbol table if x86 but
"foo" if x64.
This patch moves the code to infer machine type, so that machine
type is inferred based on input files given via the command line
(but not based on .directives files).
llvm-svn: 241843
Symbols exported by DLLs are listed in import library files.
Exported names may be mangled by "Import Name Type" field as
described in PE/COFF spec 7.3. This patch implements that
mangling scheme.
llvm-svn: 241719
Providing a symbol table in the executable is quite useful when
debugging a fully-linked executable without having to reconstruct one
from DWARF.
Differential Revision: http://reviews.llvm.org/D11023
llvm-svn: 241689
Previously we were unnecessarily loading lazy symbols if they appeared in an
archive multiple times, as can happen with comdat symbols. This change fixes
the bug by only loading symbols from archives at load time if the original
symbol was undefined.
Differential Revision: http://reviews.llvm.org/D10980
llvm-svn: 241538
The previous code was not even safe with MSVC 2013 because the compiler
doesn't guarantee that static variables (in this case, a mutex) are
initialized in a thread-safe manner.
llvm-svn: 241481
TLS table header field is supposed to have address and size of TLS table.
The linker doesn't have to understand what TLS table is. TLS table's name
is always "_tls_used", so if there's that symbol, the linker simply sets
that symbol's RVA to the header. The size of the TLS table is always 40 bytes.
llvm-svn: 241426
This function is called SymbolTable::readObjects, so in order to
parallelize that function, we have to make this function thread-safe.
llvm-svn: 241420
In the new design, mutation of Symbol pointers is the name resolution
operation. This patch makes them atomic pointers so that they can
be mutated by multiple threads safely. I'm going to use atomic
compare-exchange on these pointers.
dyn_cast<> doesn't recognize atomic pointers as pointers,
so we need to call load(). This is unfortunate, but in other places
automatic type conversion works fine.
llvm-svn: 241416
We were previously hitting assertion failures in the writer in cases where
a regular object file defined a weak external symbol that was defined by
a bitcode file. Because /export and /entry name mangling were implemented
using weak externals, the same problem affected mangled symbol names in
bitcode files.
The underlying cause of the problem was that weak external symbols were
being resolved before doing LTO, so the symbol table may have contained stale
references to bitcode symbols. The fix here is to defer weak external symbol
resolution until after LTO.
Also implement support for weak external symbols in bitcode files
by modelling them as replaceable DefinedBitcode symbols.
Differential Revision: http://reviews.llvm.org/D10940
llvm-svn: 241391
Looks like clang-cl sets a bogus value to the field, which makes
getSectionContents() to truncate section contents. This patch directly
uses SizeOfRawData field instead of VirtualSize to see if this can
make buildbot green.
llvm-svn: 241386
This worked before, but only by accident, and only with assertions disabled.
We ended up storing a DefinedRegular symbol in the WeakAlias field,
and never using it as an Undefined.
Differential Revision: http://reviews.llvm.org/D10934
llvm-svn: 241376
This change cut the link time of chrome.dll from 24 seconds
to 22 seconds (5% gain). When the control reaches end of link(),
all output files have already been written. All in-memory
objects can just vanish. There is no use to call their dtors.
llvm-svn: 241320
DLLs can export symbols only by ordinal, and DLLs are also able to be
delay-loaded. The combination of the two is valid. I didn't expect
that combination. This patch implements that feature.
With this patch, LLD is now able to link a working executable of Chrome
for 64-bit debug build. The browser seemed to be working fine. Chrome is
good for testing because of its variety and size. It contains various
open-source libraries written by various people. The largest file in
Chrome is chrome.dll whose size is 496MB. LLD can link it in 24 seconds.
MSVC linker takes 48 seconds. So it is exactly 2x faster. (I measured
that with debug info and ICF being turned off.)
With this achievement, I think I can say that the new COFF linker is
now mostly feature complete for x86-64 Windows. I believe there are
still many lingering bugs, though.
llvm-svn: 241318
Previously, __ImageBase symbol got a different value than the one
specified by /base:<number> because the symbol was created in the
SymbolTable's constructor. When the constructor is called,
no command line options are processed yet, so the symbol was
created always with the initial value. This caused wrong relocations
and thus caused mysterious crashes of some executables linked by LLD.
llvm-svn: 241313
Previously, pointers pointed by locally-imported symbols were broken.
It has only 4 bytes although the correct size is 8 byte. This patch
fixes that bug.
llvm-svn: 241295
Previously, SymbolBody::compare(A, B) didn't satisfy weak ordering.
There was a case that A < B and B < A could have been true.
This is because we just pick LHS if A and B are consisdered equivalent.
This patch is to make symbols being weakly ordered. If A and B are
not tie, one of A < B && B > A or A > B && B < A is true.
This is not an improtant property for a single-threaded environment
because everything is deterministic anyways. However, in a multi-
threaded environment, this property becomes important.
If a symbol is defined or lazy, ties are resolved by its file index.
For simple types that we don't really care about their identities,
symbols are compared by their addresses.
llvm-svn: 241294
On Windows, we have four different main functions, {w,}{main,WinMain}.
The linker has to choose a corresponding entry point function among
{w,}{main,WinMain}CRTStartup. These entry point functions are defined
in the standard library. The linker resolves one of them by looking at
which main function is defined and adding a corresponding undefined
symbol to the symbol table.
Object files containing entry point functions conflicts each other.
For example, we cannot resolve both mainCRTStartup and WinMainCRTStartup
because other symbols defined in the files conflict.
Previously, we inferred CRT function name at the very end of name
resolution. I found that that is sometimes too late. If the linker
already linked one of these four archive member objects, it's too late
to change the decision.
The right thing to do here is to infer entry point name after adding
all symbols from command line files and before adding any other files
(which are specified by directive sections). This patch does that.
llvm-svn: 241236
Previously, we use SymbolTable::rename to resolve AlternateName symbols.
This patch is to merge that mechanism with weak aliases, so that we
remove that function.
llvm-svn: 241230
I think Undefined symbols are a bit more convenient than StringRefs
since SymbolBodies are handles for symbols. You can get resolved
symbols for undefined symbols just by calling getReplacmenet without
looking up the symbol table.
llvm-svn: 241214
Occasionally we have to resolve an undefined symbol to its
mangled symbol. Previously, we did that on calling side of
findMangle by explicitly updating SymbolBody.
In this patch, mangled symbols are handled as weak aliases
for undefined symbols.
llvm-svn: 241213
Previously, the order of adding symbols to the symbol table was simple.
We have a list of all input files. We read each file from beginning of
the list and add all symbols in it to the symbol table.
This patch changes that order. Now all archive files are added to the
symbol table first, and then all the other object files are added.
This shouldn't change the behavior in single-threading, and make room
to parallelize in multi-threading.
In the first step, only lazy symbols are added to the symbol table
because archives contain only Lazy symbols. Member object files
found to be necessary are queued. In the second step, defined and
undefined symbols are added from object files. Adding an undefined
symbol to the symbol table may cause more member files to be added
to the queue. We simply continue reading all object files until the
queue is empty.
Finally, new archive or object files may be added to the queues by
object files' directive sections (which contain new command line
options).
The above process is repeated until we get no new files.
Symbols defined both in object files and in archives can make results
undeterministic. If an archive is read before an object, a new member
file gets linked, while in the other way, no new file would be added.
That is the most popular cause of an undeterministic result or linking
failure as I observed. Separating phases of adding lazy symbols and
undefined symbols makes that deterministic. Adding symbols in each
phase should be parallelizable.
llvm-svn: 241107
The size of this class actually matters because this is the
most popular class among all classes. We create a Defined symbol
for each defined symbol in a symbol table. That can be millions
for a large program. For example, linking LLD instantiates this
class millions times.
llvm-svn: 241025
its members into the base class.
First, to help motivate this kind of change, understand that in
a self-link, LLD creates 5.5 million defined regular symbol bodies (and
6 million symbol bodies total). A significant portion of its time is
spent allocating the memory for these symbols, and befor ethis patch
the defined regular symbol body objects alone consumed some 420mb of
memory during the self link.
As a consequence, I think it is worth expending considerable effort to
make these objects as memory efficient as possible. This is the first of
several components of that. This change starts with the goal of removing
the virtual functins from SymbolBody so that it can avoid having a vptr
embedded in it when it already contains a "kind" member, and that member
can be much more compact than a vptr.
The primary way of doing this is to sink as much of the logic that we
would have to dispatch for into data in the base class. As part of this,
I made the various flags bits that will pack into a bitfield with the
kind tag. I also sank the Name down to eliminate the dispatch for that,
and used LLVM's RTTI-style dispatch for everything else (most of which
is cold and so doesn't matter terribly if we get minutely worse lowering
than a vtable dispatch).
As I was doing this, I wanted to make the RTTI-dispatch (which would
become much hotter than before) as efficient as possible, so I've
re-organized the tags somewhat. Notably, the common case (regular
defined symbols) is now zero which we can test for faster.
I also needed to rewrite the comparison routine used during resolving
symbols. This proved to be quite complex as the semantics of the
existing one were very subtle due to the back-and-forth virtual dispatch
caused by re-dispatching with reversed operands. I've consolidated it to
a single function and tried to comment it quite a bit more to help
explain what is going on. However, this may need more comments or other
explanations. It at least passes all the regression tests. I'm not
working on Windows, so I can't fully test it.
With all of these changes, the size of a DefinedRegular symbol on
a 64-bit build goes from 80 bytes to 64 bytes, and we save approximately
84mb or 20% of the memory consumed by these symbol bodies during the
link.
The link time appears marginally faster as well, and the profile hotness
of the memory allocation subsystem got a bit better, but there is still
a lot of allocation traffic.
Differential Revision: http://reviews.llvm.org/D10792
llvm-svn: 241001
This uses a single cast and test to get the section for the symbol, and
uses the cast_or_null<> pattern throughout to handle the known type but
unknown non-null-ness.
No functionality changed.
Differential Revision: http://reviews.llvm.org/D10791
llvm-svn: 241000
This flattens the entire liveness walk from a recursive mark approach to
a worklist approach. It also sinks the worklist management completely
out of the SectionChunk and into the Writer by exposing the ability to
iterato over children of a chunk and over the symbol bodies of relocated
symbols. I'm not 100% happy with the API names, so suggestions welcome
there.
This allows us to use a single worklist for the entire recursive walk
and would also be a natural place to take advantage of parallelism at
some future point.
With this, we completely inline away the GC walk into the
Writer::markLive function and it makes it very easy to profile what is
slow. Currently, time is being wasted checking whether a Chunk isa
SectionChunk (it essentially always is), finding (or skipping)
a replacement for a symbol, and chasing pointers between symbols and
their chunks. There are a bunch of things we can do to fix this, and its
easier to do them after this change IMO.
This change alone saves 1-2% of the time for my self-link of lld.exe
(which I'm running and benchmarking on Linux ironically).
Perhaps more notably, we'll no longer blow out the stack for large
links. =]
Just as an FYI, at this point, I/O is starting to really dominate the
profile. Well over 10% of the time appears to be inside the kernel doing
page table silliness. I think a decent chunk of this can be nuked as
well, but it's a little odd as cross-linking in this way isn't really
the primary goal here.
Differential Revision: http://reviews.llvm.org/D10790
llvm-svn: 240995
method to get a SymbolBody and into the callers, and kill now dead
includes.
This removes the need to have the SymbolBody definition when we're
defining the inline method and makes it a better inline method. That was
the only reason for a lot of header includes here. Removing these and
using forward declarations actually uncovers a bunch of cross-header
dependencies that I've fixed while I'm here, and will allow me to
introduce some *important* inline code into Chunks.h that requires the
definition of ObjectFile.
No functionality changed at this point.
Differential Revision: http://reviews.llvm.org/D10789
llvm-svn: 240982
Compilers recognize "main" function and don't mangle its name.
But if you use a different function as a user-defined entry name,
and if you didn't define that function with extern C, your entry
point function name is mangled. And the linker has to be able to
find that. This is relatively rare but can happen.
llvm-svn: 240953
Most build system depends on existence or time stamp of a file.
This patch is to create an empty file for /pdb:<filename> option
just to satisfy some build rules.
llvm-svn: 240948
The previous logic to find default entry name or subsystem does not
seem correct (i.e. was not compatible with MSVC linker). Previously,
default entry name was inferred from CRT functions and user-defined
entry functions. Subsystem was inferred from CRT functions.
Default entry name and subsystem are now inferred based on the
following table. Note that we no longer use CRT functions to infer
them.
Entry name Subsystem
main mainCRTStartup console
wmain wmainCRTStartup console
WinMain WinMainCRTStartup windows
wWinMain wWinMainCRTStartup windows
llvm-svn: 240922
Usually dllexported symbols are defined with 'extern "C"',
so identifying them is easy. We can just do hash table lookup
to look up exported symbols.
However, C++ non-member functions are also allowed to be exported,
and they can be specified with unmangled name. So, if /export:foo
is given, we need to look up not only "foo" but also its all
mangled names. In MSVC mangling scheme, that means that we need to
look up any symbol which starts with "?foo@@Y".
In this patch, we scan the entire symbol table to search for
a mangled symbol. The symbol table is a DenseMap, and that doesn't
support table lookup by string prefix. This is of course very
inefficient. But that should be probably OK because the user
should always add 'extern "C"' to dllexported symbols.
llvm-svn: 240919
This option is to ignore remaining undefined symbols and force
the linker to create an output file anyways.
The existing code assumes that there's no undefined symbol after
reportRemainingUndefines(). That assumption is legitimate.
I also don't want to mess up the existing code for this minor feature.
In order to keep it as is, remaining undefined symbols are replaced
with dummy defined symbols.
llvm-svn: 240913
When comparing two COMDAT sections, we need to take section values
and associative sections into account. This patch fixes that bug.
It fixes a crash bug of llvm-tblgen when linked with /opt:lldicf.
One thing I don't understand yet is that this logic seems to be
too strict. MSVC linker is able to create more compact executables
(which of course work correctly). With this ICF algorithm, LLD is
able to make executable smaller, but the outputs are larger than
MSVC's. There must be something I'm missing here.
llvm-svn: 240897
This function is actually *very* hot. It is hard to see currently
because the call graph is very recursive, but I'm working to remove that
and when I do this function becomes significantly higher on the profile
(up to 5%!) and so worth avoiding the call overhead.
No specific perf gain I can measure yet (below the noise), but likely to
have more impact as we stop cluttering the call graph.
Differential Revision: http://reviews.llvm.org/D10788
llvm-svn: 240873
StringRefs. This uses the LLVM hashing rather than the standard library
and a closed addressed hash table rather than chaining.
This improves the Windows self-link of LLD by 4.4% (averaged over 10
runs, with well under 1% of variance on each).
There is still some room to improve here. Two things I clearly see in
the profile:
1) This is one of the biggest stress tests for the LLVM hashing code. It
actually consumes something like 3-4% of the link time after the
change.
2) The way that StringRef keys are handled in the DenseMap interface is
pretty suboptimal. We pay the price of checking for empty and
tombstone keys when we could only possibly be looking for a normal
key. But fixing this requires invasive API changes.
So there is still some headroom here.
Differential Revision: http://reviews.llvm.org/D10684
llvm-svn: 240871
There were a few issues with the previous delay-import tables.
- "Attribute" field should have been 1 instead of 0.
(I don't know the meaning of this field, though.)
- LEA and CALL operands had wrong addresses.
- Address tables are in .didat (which is read-only).
They should have been in .data.
llvm-svn: 240837
This flag can be used to produce a map file, which is essentially a list
of objects linked into the final output file together with the RVAs of
their symbols. Because our format differs from MSVC's we expose it as a
separate flag.
Differential Revision: http://reviews.llvm.org/D10773
llvm-svn: 240812
We were resolving entry symbols and /include'd symbols after all other
symbols are resolved. But looks like it's too late. I found that it
causes some program to fail to link.
Let's say we have an object file A which defines symbols X and Y in an
archive. We also have another file B after A which defines X, Y and
_DLLMainCRTStartup in another archive. They conflict each other, so
either A or B can be linked.
If we have _DLLMainCRTStartup as an undefined symbol, file B is always
chosen. If not, there's a chance that A is chosen. If the linker
find it needs _DllMainCRTStartup after that, it's too late.
This patch adds undefined symbols to the symbol table as soon as
possible to fix the issue.
llvm-svn: 240757
Absolute symbols were always handled as external symbols, so if two
or more object files define the same absolute symbol, they would
conflict even if the symbol is private to each file.
This patch fixes that bug.
llvm-svn: 240756
ICF implemented in LLD is so experimental that we don't want to
enable that even if /opt:icf option is passed. I'll rename it back
once the feature is complete.
llvm-svn: 240721
I split them in r240319 because I thought they are different enough
that we should treat them as different types. It turned out that
that was not a good idea. They are so similar that we ended up having
many duplicate code.
llvm-svn: 240706
Only SectionChunk can be dead-stripped. Previously,
all types of chunks implemented these functions,
but their functions were blank.
Likewise, only DefinedRegular and DefinedCOMDAT symbols
can be dead-stripped. markLive() function was implemented
for other symbol types, but they were blank.
I started thinking that the change I made in r240319 was
a mistake. I separated DefinedCOMDAT from DefinedRegular
because I thought that would make the code cleaner, but now
we want to handle them as the same type here. Maybe we
should roll it back.
This change should improve readability a bit as this removes
some dubious uses of reinterpret_cast. Previously, we
assumed that all COMDAT chunks are actually SectionChunks,
which was not very obvious.
llvm-svn: 240675
The change I made in r240620 was not correct. If a symbol foo is
defined, and if you use __imp_foo, __imp_foo symbol is automatically
defined as a pointer (not just an alias) to foo.
Now that we need to create a chunk for automatically-created symbols.
I defined LocalImportChunk class for them.
llvm-svn: 240622
MSVC linker is able to link an object file created from the following code.
Note that __imp_hello is not defined anywhere.
void hello() { printf("Hello\n"); }
extern void (*__imp_hello)();
int main() { __imp_hello(); }
Function symbols exported from DLLs are automatically mangled by appending
__imp_ prefix, so they have two names (original one and with the prefix).
This "feature" seems to simulate that behavior even for non-DLL symbols.
This is in my opnion very odd feature. Even MSVC linker warns if you use this.
I'm adding that anyway for the sake of compatibiltiy.
llvm-svn: 240620
Getting an iterator to the relocation table is very hot operation
in the linker. We do that not only to apply relocations but also
to mark live sections and to do ICF.
libObject's interface is slow. By caching pointers to the first
relocation table entries makes the linker 6% faster to self-link.
We probably need to fix libObject as well.
llvm-svn: 240603
Identical COMDAT Folding (ICF) is an optimization to reduce binary
size by merging COMDAT sections that contain the same metadata,
actual data and relocations. MSVC link.exe and many other linkers
have this feature. LLD achieves on per with MSVC in terms produced
binary size with this patch.
This technique is pretty effective. For example, LLD's size is
reduced from 64MB to 54MB by enaling this optimization.
The algorithm implemented in this patch is extremely inefficient.
It puts all COMDAT sections into a set to identify duplicates.
Time to self-link with/without ICF are 3.3 and 320 seconds,
respectively. So this option roughly makes LLD 100x slower.
But it's okay as I wanted to achieve correctness first.
LLD is still able to link itself with this optimization.
I'm going to make it more efficient in followup patches.
Note that this optimization is *not* entirely safe. C/C++ require
different functions have different addresses. If your program
relies on that property, your program wouldn't work with ICF.
However, it's not going to be an issue on Windows because MSVC
link.exe turns ICF on by default. As long as your program works
with default settings (or not passing /opt:noicf), your program
would work with LLD too.
llvm-svn: 240519
Chunks are basically unnamed chunks of bytes, and we don't like
to give them names. However, for logging or debugging, we want to
know symbols names of functions for COMDAT chunks. (For example,
we want to print out "we have removed unreferenced COMDAT section
which contains a function FOOBAR.")
This patch is to do that.
llvm-svn: 240484
Previously, we added files in directive sections to the symbol
table as we read the sections, so the link order was depth-first.
That's not compatible with MSVC link.exe nor the old LLD.
This patch is to queue files so that new files are added to the
end of the queue and processed last. Now addFile() doesn't parse
files nor resolve symbols. You need to call run() to process
queued files.
llvm-svn: 240483
Before this change, you got to cast a symbol to DefinedRegular and then
call isCOMDAT() to determine if a given symbol is a COMDAT symbol.
Now you can just use isa<DefinedCOMDAT>().
As to the class definition of DefinedCOMDAT, I could remove duplicate
code from DefinedRegular and DefinedCOMDAT by introducing another base
class for them, but I chose to not do that to keep the class hierarchy
shallow. This amount of code duplication doesn't worth to define a new
class.
llvm-svn: 240319
DLLs are usually resolved at process startup, but you can
delay-load them by passing /delayload option to the linker.
If a /delayload is specified, the linker has to create data
which is similar to regular import table.
One notable difference is that the pointers in a delay-load
import table are originally pointing to thunks that resolves
themselves. Each thunk loads a DLL, resolve its name, and then
overwrites the pointer with the result so that subsequent
function calls directly call a desired function. The linker
has to emit thunks.
llvm-svn: 240250
.pdata section contains a list of triplets of function start address,
function end address and its unwind information. Linkers have to
sort section contents by function start address and set the section
address to the file header (so that runtime is able to find it and
do binary search.)
This change seems to resolve all but one remaining test failures in
check{,-clang,-lld} when building the entire stuff with clang-cl and
lld-link.
llvm-svn: 240231
This is a case that one mistake caused a very mysterious bug.
I made a mistake to calculate addresses of common symbols, so
each common symbol pointed not to the beginning of its location
but to the end of its location. (Ouch!)
Common symbols are aligned on 16 byte boundaries. If a common
symbol is small enough to fit between the end of its real
location and whatever comes next, this bug didn't cause any harm.
However, if a common symbol is larger than that, its memory
naturally overlapped with other symbols. That means some
uninitialized variables accidentally shared memory. Because
totally unrelated memory writes mutated other varaibles, it was
hard to debug.
It's surprising that LLD was able to link itself and all LLD
tests except gunit tests passed with this nasty bug.
With this fix, the new COFF linker is able to pass all tests
for LLVM, Clang and LLD if I use MSVC cl.exe as a compiler.
Only three tests are failing when used with clang-cl.
llvm-svn: 240216
This avoids undefined behaviour caused by an out-of-range access if the
vector is empty, which can happen if an object file's directive section
contains only whitespace.
llvm-svn: 240183
getName() does strlen() on the symbol table, so it's not very fast.
It's not as bad as r239332 because the number of symbols exported
from archive files are fewer than object files, and they are usually
shorter, though.
llvm-svn: 240178
Rui Ueyama