Commit Graph

663 Commits

Author SHA1 Message Date
Rui Ueyama f54e4fd3c6 PECOFF: Move error check for invalid command line combination to validateImpl.
llvm-svn: 228461
2015-02-06 23:09:13 +00:00
Rui Ueyama 92634be399 Use make_unique.
llvm-svn: 228453
2015-02-06 22:44:16 +00:00
Shankar Easwaran d050c50522 [Revert] [ELF] Determine default search directories from Context
It looks like the Driver manages search path for each Target lld would support
on the Gnu flavor.

llvm-svn: 228440
2015-02-06 21:23:50 +00:00
Shankar Easwaran 2ba4f5d9e3 [Cleanup] Remove member functions added to support nostdlib
No change in functionality.

llvm-svn: 228379
2015-02-06 05:01:38 +00:00
Shankar Easwaran 8f1b2d0930 [Core] Remove roundTripPass() function.
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
2015-02-06 04:15:02 +00:00
Shankar Easwaran e315edd747 [ELF] Fix -nostdlib option.
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
2015-02-06 04:15:00 +00:00
Rui Ueyama 2a96704b33 ELF: Support INPUT linker script directive
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
2015-02-03 23:00:19 +00:00
Rafael Auler ee95c3b5c5 Make ELFLinkingContext own LinkerScript buffers
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
2015-02-03 16:08:57 +00:00
Rui Ueyama a6d38a326a ELF: Improve linker script unit tests.
This patch is to enable to write unit tests for linker script with
less boilerplate code.

llvm-svn: 227902
2015-02-03 00:42:36 +00:00
Davide Italiano 104949df73 [ELF] Implement action for OUTPUT linker script command
Reviewed by:	ruiu, shankarke

llvm-svn: 227787
2015-02-02 06:28:12 +00:00
Shankar Easwaran 148cd8e8e9 [ELF] Determine default search directories from Context.
Target specific LinkingContext's  determine the default search directory.

No change in functionality.

llvm-svn: 227784
2015-02-02 06:00:04 +00:00
Davide Italiano 1ecdf4ad30 [ELF] Implement semantic action for ENTRY linker script command.
Reviewed by:	shankarke, ruiu

llvm-svn: 227708
2015-02-01 03:27:25 +00:00
Rafael Auler 8251d741f4 Implement semantic action for SEARCH_DIR linker script command
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
2015-01-31 22:42:19 +00:00
Davide Italiano bf2625d42b GNU LD option export-dynamic can accept single dash flag.
Differential Revision:	D7044
Reviewed by:	ruiu

llvm-svn: 227631
2015-01-30 22:56:22 +00:00
Rui Ueyama a9a6f55840 ELF: Fix a bug in -l.
If a linker script is given with -l, LLD failed to read the file.
The bug was introduced in 227341.

llvm-svn: 227610
2015-01-30 21:03:19 +00:00
Rui Ueyama 7f3639fb19 ELF: s/diagnostics/diag/
We use a variable name "diag" in many other places and I think
it's preferred because of its length.

llvm-svn: 227369
2015-01-28 21:20:28 +00:00
Rui Ueyama 48eec8a90d ELF: add a unit test for --as-needed.
llvm-svn: 227367
2015-01-28 21:01:24 +00:00
Rui Ueyama 0b55151d3e Add a unit test for LinkerScript.
llvm-svn: 227341
2015-01-28 18:38:50 +00:00
Greg Fitzgerald df0f5cd474 Remove PPC ELF target
Differential Revision: http://reviews.llvm.org/D7225

llvm-svn: 227320
2015-01-28 16:37:43 +00:00
Rui Ueyama 9d7a15c80d Accept -allow-shlib-undefined in addition to --allow-shlib-undefined.
GNU LD accepts both. CMake seems to use the former than the latter.

llvm-svn: 227144
2015-01-26 22:13:53 +00:00
Greg Fitzgerald 0dc86722f5 Fix shared library build
* Removed cyclic dependency between lldPECOFF and lldDriver
* Added missing dependencies in unit tests

Differential Revision: http://reviews.llvm.org/D7185

llvm-svn: 227134
2015-01-26 20:46:47 +00:00
Simon Atanasyan 0712650742 [Mips] Teach LLD to recognize one more MIPS target triple - mips64el
This is initial patch to support MIPS64 object files linking.
The patch just makes some classes more generalized, and rejects
attempts to interlinking O32 and N64 ABI object files.

I try to reuse the current MIPS target related classes as much as
possible because O32 and N64 MIPS ABI are tightly related and share
almost the same set of relocations, GOT, flags etc.

llvm-svn: 227058
2015-01-25 19:15:35 +00:00
Greg Fitzgerald 5aed2f1ea2 Fix single-arch builds broken by r226976
TODO: Move ELF/PPC to ELF/PowerPC
TODO: Move ELF/X86_64 into ELF/X86
llvm-svn: 226989
2015-01-24 04:51:26 +00:00
Greg Fitzgerald ad363ad804 Fix the ELF shared library build targets - take 2
lldELF is used by each ELF backend.  lldELF's ELFLinkingContext
 also held a reference to each backend, creating a link-time
 cycle.  This patch moves the backend references to lldDriver.

 Differential Revision: http://reviews.llvm.org/D7119

llvm-svn: 226976
2015-01-24 01:06:07 +00:00
Rui Ueyama 009c1a931e Fix spelling.
llvm-svn: 226966
2015-01-23 23:39:33 +00:00
Greg Fitzgerald b4eb64ef6a [MachO] Remove dependency on lldDriver
Moved getMemoryBuffer from DarwnLdDriver to MachOLinkingContext.
lldMachO shared library target now builds.

Differential Review: http://reviews.llvm.org/D7155

llvm-svn: 226963
2015-01-23 23:26:13 +00:00
Greg Fitzgerald ba2bcb0da3 Revert " Fix the ELF shared library build targets"
This reverts commit 6a3f545b44cea46321e025d9ab773786af86cb51.

llvm-svn: 226928
2015-01-23 19:24:32 +00:00
Greg Fitzgerald 0823ea636e Fix the ELF shared library build targets
lldELF is used by each ELF backend.  lldELF's ELFLinkingContext
 also held a reference to each backend, creating a link-time
 cycle.  This patch moves the backend references to lldDriver.

 Differential Revision: http://reviews.llvm.org/D7119

llvm-svn: 226922
2015-01-23 18:52:44 +00:00
Rui Ueyama 0330697bfc Don't use iterator for the same reason as r226883.
llvm-svn: 226893
2015-01-23 01:44:51 +00:00
Rui Ueyama 92fb392786 Make access to LinkingContext::getNode safe.
llvm-svn: 226883
2015-01-23 00:09:05 +00:00
Greg Fitzgerald 4b6a7e355b Fix five of the shared library build targets
Before this patch there was a cyclic dependency between lldCore and
lldReaderWriter.  Only lldConfig could be built as a shared library.

* Moved Reader and Writer base classes into lldCore.
* The following shared libraries can now be built:
     lldCore
     lldYAML
     lldNative
     lldPasses
     lldReaderWriter

Differential Revision: http://reviews.llvm.org/D7105

From: Greg Fitzgerald <garious@gmail.com>
llvm-svn: 226732
2015-01-21 22:54:56 +00:00
Greg Fitzgerald 57c29ad223 add_lld_library -> add_llvm_library
* Works better for shared libraries (sets PRIVATE instead of INTERFACE)
* Fixes http://llvm.org/bugs/show_bug.cgi?id=22269
* Also, use build-target names instead of component names

Differential Revision: http://reviews.llvm.org/D7074

From: Greg Fitzgerald <garious@gmail.com>
llvm-svn: 226702
2015-01-21 21:26:27 +00:00
Greg Fitzgerald 8a1887f1f1 [ELF] Minimal implementation for ARM static linking
The code is able to statically link the simplest case of:

  int main() { return 0; }

* Only works with ARM code - no Thumb code, no interwork (-marm -mno-thumb-interwork)
* musl libc built with no interwork and no Thumb code

Differential Revision: http://reviews.llvm.org/D6716

From: Denis Protivensky <dprotivensky@accesssoftek.com>
llvm-svn: 226643
2015-01-21 07:35:48 +00:00
Rafael Espindola 246b4bbfee Fix use after free.
This fixes running a few tests (including test/elf/Mips/e-flags-merge-1.test)
under asan.

llvm-svn: 226580
2015-01-20 17:08:27 +00:00
Rui Ueyama d1b00ea848 Update comments.
llvm-svn: 226380
2015-01-17 22:38:09 +00:00
Greg Fitzgerald c7a815bb3a Add LLVM_LINK_COMPONENTS for the shared object build
Differential Revision: http://reviews.llvm.org/D7023

From: Greg Fitzgerald <garious@gmail.com>
llvm-svn: 226346
2015-01-16 23:34:41 +00:00
Rui Ueyama 6d176f88ea [PECOFF] Remove ResolvableSymbols to simplify.
We had such class there because of InputGraph abstraction.
Previously, no one except InputGraph itself has complete picture of
input file list. In order to create a set of all defined symbols,
we had to use some indirections there to workaround InputGraph.

It can now be rewritten as simple code. No change in functionality.

llvm-svn: 226319
2015-01-16 20:48:46 +00:00
Rui Ueyama 949dc41502 Split a function for readability.
llvm-svn: 226292
2015-01-16 17:53:55 +00:00
Rui Ueyama d50d2e7af0 Simplify. No functionality change.
llvm-svn: 226287
2015-01-16 16:58:58 +00:00
Rui Ueyama cb9b1086ff [PECOFF] Improve parallelism.
llvm-svn: 226284
2015-01-16 15:58:36 +00:00
Rui Ueyama d4730ea555 Run the resolver in parallel with the reader.
This patch makes File::parse() multi-thread safe. If one thread is running
File::parse(), other threads will block if they try to call the same method.
File::parse() is idempotent, so you can safely call  multiple times.

With this change, we don't have to wait for all worker threads to finish
in Driver::link(). Previously, Driver::link() calls TaskGroup::sync() to
wait for all threads running File::parse(). This was not ideal because
we couldn't start the resolver until we parse all files.

This patch increase parallelism by making Driver::link() to not wait for
worker threads. The resolver calls parse() to make sure that the file
being read has been parsed, and then uses the file. In this approach,
the resolver can run with the parser threads in parallel.

http://reviews.llvm.org/D6994

llvm-svn: 226281
2015-01-16 15:54:13 +00:00
Rui Ueyama fa7e8a663f [ELF] Add --as-needed.
The previous default behavior of LLD is --as-needed. LLD linked
against a DSO only if the DSO file was actually used to link an
executable (i.e. at least one symbol was resolved using the shared
library file.)

In this patch I added a boolean flag to FileNode for --as-needed.
I also added an accessor to DSO name to shared library file class.

llvm-svn: 226274
2015-01-16 14:27:01 +00:00
Rui Ueyama c2836079df Simplify.
llvm-svn: 226225
2015-01-15 23:15:09 +00:00
Rui Ueyama cf14b63116 [ELF] Do not error on non-existent file in the driver.
This change makes it easy to write unit tests for the GNU driver.
No more "empty group" hack is needed. No change in functionality.

llvm-svn: 226150
2015-01-15 08:49:19 +00:00
Rui Ueyama 883afba228 Remove InputGraph and use std::vector<Node> instead.
In total we have removed more than 1000 lines!

llvm-svn: 226149
2015-01-15 08:46:36 +00:00
Rui Ueyama 61635440a9 Rename InputElement Node.
InputElement was named that because it's an element of an InputGraph.
It's losing the origin because the InputGraph is now being removed.
InputElement's subclass is FileNode, that naming inconsistency needed
to be fixed.

llvm-svn: 226147
2015-01-15 08:31:46 +00:00
Rui Ueyama 861c2d6c3a Remove InputGraph::addInputElement{,Front}.
They were the last member functions of InputGraph (besides members()).
Now InputGraph is just a container of a vector. We are ready to replace
InputGraph with plain File vector.

llvm-svn: 226146
2015-01-15 08:18:14 +00:00
Rui Ueyama 423177a54b Remove WrapperNode.
WrapperNode was a useless subclass of FileNode. We should just use
FileNode instead.

llvm-svn: 226145
2015-01-15 08:10:10 +00:00
Rui Ueyama df3fca2520 Remove FileNode::parse.
FileNode::parse was just a forwarder to File::parse so we could remove that.
Also removed dead code.

llvm-svn: 226144
2015-01-15 07:56:14 +00:00
Rui Ueyama 7c4ba12769 Remove FileNode::getPath().
Previously both FileNode and File keep filename. This patch removed
that redundancy.

llvm-svn: 226143
2015-01-15 07:38:32 +00:00
Rui Ueyama cdb1071be5 Remove InputGraph::size().
llvm-svn: 226140
2015-01-15 07:20:39 +00:00
Rui Ueyama 56206368f5 Simplify FileNode.
The member function was defined to allow subclasses to customize
error message. But since we only have one FileNode type, there's
no actual need for that.

llvm-svn: 226139
2015-01-15 07:15:36 +00:00
Rui Ueyama 8054342f9e Merge SimpleFileNode with WrapperNode.
llvm-svn: 226137
2015-01-15 07:05:46 +00:00
Rui Ueyama 5a831ee5fd [PECOFF] Remove an InputElement placeholder for the entry name.
llvm-svn: 226133
2015-01-15 06:56:26 +00:00
Rui Ueyama 80c04431ca Re-commit r225766, r225767, r225769, r225814, r225816, r225829, and r225832.
These changes depended on r225674 and had been rolled back in r225859.
Because r225674 has been re-submitted, it's safe to re-submit them.

llvm-svn: 226132
2015-01-15 06:49:21 +00:00
Rui Ueyama 2b5a4fbc30 Attempt to fix a Linux buildbot.
llvm-svn: 226127
2015-01-15 05:41:12 +00:00
Rui Ueyama df230b21e3 Re-commit r225674: Convert other drivers to use WrapperNode.
The original commit had an issue with Mac OS dylib files. It didn't
handle fat binary dylib files correctly. This patch includes a fix.
A test for that case has already been committed in r225764.

llvm-svn: 226123
2015-01-15 04:34:31 +00:00
Rui Ueyama cfb2534ef8 Revert "Convert other drivers to use WrapperNode" and subsequent commits.
r225764 broke a basic functionality on Mac OS. This change reverts
r225764, r225766, r225767, r225769, r225814, r225816, r225829, and r225832.

llvm-svn: 225859
2015-01-14 00:21:34 +00:00
Rui Ueyama d5bb5c2bfe Remove dead code.
Now every InputElement has exactly one File in it, so "expand"
method is now no-op.

llvm-svn: 225769
2015-01-13 05:59:17 +00:00
Rui Ueyama 64e3dedf3b Remove InputGraph::registerObserver.
PECOFF was the only user of the API, and the reason why we created
the API is because, although the driver creates a list of input files,
it has no knowledge on what files are being created. It was because
everything was hidden behind the InputGraph abstraction.

Now the driver knows what that's doing. We no longer need this
indirection to get the file list being processed.

llvm-svn: 225767
2015-01-13 05:24:53 +00:00
Rui Ueyama d2838c4958 Remove InputGraph::dump().
This is dead code.

llvm-svn: 225766
2015-01-13 05:11:05 +00:00
Rui Ueyama e8ecb2b144 Convert other drivers to use WrapperNode.
llvm-svn: 225764
2015-01-13 04:33:07 +00:00
Rui Ueyama 1e51e238ed Update comment.
llvm-svn: 225645
2015-01-12 18:32:24 +00:00
Rui Ueyama 5a25fd5962 [ELF] Remove {ELF,}GNULinkerScript.
Instead of representing a linker script file as an "InputElement",
parse and evaluate scripts in the driver as we see them.

Linker scripts are not regular input files (regular file is one of
object, archive, or shared library file). They are more like
extended command line options. Linker script handling was needlessly
complicated because of that inappropriate abstraction (besides
excessive class hierarchy -- there is no such thing like ELF linker
script but we had two classes there for some reason.)

LinkerScript was one of a few remaining InputElement subclasses
that can be expanded to multiple files. With this patch, we are one
step closer to retire InputElement.

http://reviews.llvm.org/D6648

llvm-svn: 225330
2015-01-07 01:13:08 +00:00
Rui Ueyama a5f2816b1c Convert CoreInputGraph.
This is a part of InputGraph cleanup to represent input files as a flat
list of Files (and some meta-nodes for group etc.)

We cannot achieve that goal in one gigantic patch, so I split the task
into small steps as shown below.

(Recap the progress so far: Currently InputGraph contains a list of
InputElements. Each InputElement contain one File (that used to have
multiple Files, but I eliminated that use case in r223867). Files are
currently instantiated in Driver::link(), but I already made a change
to separate file parsing from object instantiation (r224102), so we
can safely instantiate Files when we need them, instead of wrapping
a file with the wrapper class (FileNode class). InputGraph used to
act like a generator class by interpreting groups by itself, but it's
now just a container of a list of InputElements (r223867).)

1. Instantiate Files in the driver and wrap them with WrapperNode.
WrapperNode is a temporary class that allows us to instantiate Files
in the driver while keep using the current InputGraph data structure.

This patch demonstrates how this step 1 looks like, using Core driver
as an example.

2. Do the same thing for the other drivers.

When step 2 is done, an InputGraph consists of GroupEnd objects or
WrapperNodes each of which contains one File. Other types of
FileNode subclasses are removed.

3. Replace InputGraph with std::vector<std::unique_ptr<InputElement>>.
InputGraph is already just a container of list of InputElements,
so this step removes that needless class.

4. Remove WrapperNode.

We need some code cleanup between each step, because many classes
do a bit odd things (e.g. InputGraph::getGroupSize()). I'll straight
things up as I need to.

llvm-svn: 225313
2015-01-06 23:06:49 +00:00
Jean-Daniel Dupas 23dd15e26d [macho] -rpath support
Summary:
Work on adding -rpath support to the mach-o linker.
This patch is based on the ld64 behavior for the command line option validation.

It includes a basic test to check that the LC_RPATH load commands are properly generated when that option is used.

It also add LC_RPATH support to the binary reader, but I don't know how to test it though.


Reviewers: kledzik

Subscribers: llvm-commits

Projects: #lld

Differential Revision: http://reviews.llvm.org/D6724

llvm-svn: 224544
2014-12-18 21:33:38 +00:00
Rui Ueyama 7da4037c81 Replace ReaderError with DynamicError.
ReaderErrorCategory was used only at one place. We now have a
DynamicErrorCategory for this kind of one-time error, so use it.
The calling function doesn't really care the type of an error, so
ReaderErrorCategory was actually dead code.

llvm-svn: 224245
2014-12-15 12:20:13 +00:00
Rui Ueyama 0215c88f00 Remove PECOFFLibraryNode.
This class is empty, provides no additional feature to the base class.

llvm-svn: 224212
2014-12-14 05:08:53 +00:00
Rui Ueyama c39120e101 More WinLinkInputGraph cleanup.
This function is called only from WinLinkDriver.cpp.
Move the function to that file and mark as static.

llvm-svn: 224211
2014-12-14 05:04:22 +00:00
Rui Ueyama 11c4874d83 [PECOFF] Resolve file name in the driver, not in InputElement.
llvm-svn: 224209
2014-12-14 02:50:29 +00:00
Rui Ueyama 4726967a51 Remove code duplication.
llvm-svn: 224206
2014-12-14 02:03:47 +00:00
Rui Ueyama 92cdfc7cb8 [ELF] Clean up OPT_INPUT handler. NFC.
llvm-svn: 224192
2014-12-13 09:36:44 +00:00
Rui Ueyama ecf28e25de Move definitions to the correct file.
llvm-svn: 224190
2014-12-13 08:59:46 +00:00
Rui Ueyama 961f43fb70 Make File always take the ownership of a MemoryBuffer.
The documentation of parseFile() said that "the resulting File
object may take ownership of the MemoryBuffer." So, whether or not
the ownership of a MemoryBuffer would be taken was not clear.
A FileNode (a subclass of InputElement, which is being deprecated)
keeps the ownership if a File doesn't take it.

This patch makes File always take the ownership of a buffer.
Buffers lifespan is not always the same as File instances.
Files are able to deallocate buffers after parsing the contents.

llvm-svn: 224113
2014-12-12 10:27:33 +00:00
Simon Atanasyan 04da06ccee [ELF] Make -init/-fini options compatible with the gnu linker
The LLD linker searches initializer and finalizer function names
and emits DT_INIT/DT_FINI dynamic table tags to point to these symbols.
The -init/-fini command line options override initializer ("_init") and
finalizer ("_fini") function names used by default.

Now the -init/-fini options do not affect .init_array/.fini_array
sections. The corresponding code has been removed.

Differential Revision: http://reviews.llvm.org/D6578

llvm-svn: 223917
2014-12-10 05:38:46 +00:00
Rui Ueyama 00eb257f2e Re-commit r223330: Rewrite InputGraph's Group
llvm-svn: 223867
2014-12-10 00:33:00 +00:00
Rui Ueyama 5ae2050420 Revert "Rewrite InputGraph's Group"
This reverts commit r223330 because it broke Darwin and ELF
linkers in a way that we couldn't have caught with the existing
test cases.

llvm-svn: 223373
2014-12-04 18:29:03 +00:00
Rui Ueyama 60df72ff61 Rewrite InputGraph's Group
The aim of this patch is to reduce the excessive abstraction from
the InputGraph. We found that even a simple thing, such as sorting
input files (Mach-O) or adding a new file to the input file list
(PE/COFF), is nearly impossible with the InputGraph abstraction,
because it hides too much information behind it. As a result,
we invented complex interactions between components (e.g.
notifyProgress() mechanism) and tricky code to work around that
limitation. There were many occasions that we needed to write
awkward code.

This patch is a first step to make it cleaner. As a first step,
this removes Group class from the InputGraph. The grouping feature
is now directly handled by the Resolver. notifyProgress is removed
since we no longer need that. I could have cleaned it up even more,
but in order to keep the patch minimum, I focused on Group.

SimpleFileNode class, a container of File objects, is now limited
to have only one File. We shold have done this earlier.
We used to allow putting multiple File objects to FileNode.
Although SimpleFileNode usually has only one file, the Driver class
actually used that capability. I modified the Driver class a bit,
so that one FileNode is created for each input File.

We should now probably remove SimpleFileNode and directly store
File objects to the InputGraph in some way, because a container
that can contain only one object is useless. This is a TODO.

Mach-O input files are now sorted before they are passe to the
Resolver. DarwinInputGraph class is no longer needed, so removed.

PECOFF still has hacky code to add a new file to the input file list.
This will be cleaned up in another patch.

llvm-svn: 223330
2014-12-04 01:09:06 +00:00
Shankar Easwaran 2895527ec2 [Core] Add flag to check if RoundTripPasses need to be run.
This would allow other flavor specific contexts to override the default value,
if they want to optionally run the round trip passes.

There is some information lost like the original file owner of the atom with
RoundTripPasses. The Gnu flavor needs this information inorder to implement
LinkerScript matching and for other diagnostic outputs such as Map files.

The flag also can be used to record information in the Atom if the information
to the Writer needs to be conveyed through References too.

llvm-svn: 222983
2014-12-01 01:04:11 +00:00
Rui Ueyama 24a7f7eed8 [PECOFF] Ignore /maxilkfile option.
.ilk file is a file for incremental linking. We don't create nor use
that file.

/MAXILKFILE is an undocumented option to specify the maximum size
of the .ilk file, IIUC. We should just ignore the option.

llvm-svn: 222777
2014-11-25 19:51:28 +00:00
Rui Ueyama 0d9a181d9d [PECOFF] Create an empty PDB file if debug option is enabled.
There are many build files in the wild that depend on the fact that
link.exe produces a PDB file if /DEBUG option is given. They fail
if the file is not created.

This patch is to make LLD create an empty (dummy) file to satisfy
such build targets. This doesn't do anything other than "touching"
the file.

If a target depends on the content of the PDB file, this workaround
is no help, of course. Otherwise this patch should help build some
stuff.

llvm-svn: 222773
2014-11-25 19:01:01 +00:00
Rui Ueyama 4a259cf4a5 [PECOFF] Enable dead-stripping even if /debug option is given.
/debug makes MSVC link.exe to not remove unused sections from
the resulting executable. We did the same thing before. However,
I realized that the removal of associative section depends on
the dead-stripping pass in LLD, so we cannot disable that. Or
LLD may produce slightly broken executables that have too much
data in it (which could result in nasty subtle bugs).

This patch is a temporary measure to create correct executable.
Currently /debug does not have any real effect for LLD anyway.

I'll improve associative section handling in another patch, so that
they are removed from output without depending on the dead-stripping
pass.

llvm-svn: 222483
2014-11-20 23:59:40 +00:00
Shankar Easwaran 2e20049260 [ELF] Fix conditions for max-page-size.
Fix comments from Rui, also adds a test.

llvm-svn: 221860
2014-11-13 04:03:46 +00:00
Shankar Easwaran b43804b6f4 [ELF] Fix max-page-size option.
The user can use the max-page-size option and set the maximum page size. Dont
check for maximum allowed values for page size, as its what the kernel is
configured with.

Fix the test as well.

llvm-svn: 221858
2014-11-13 03:25:38 +00:00
Shankar Easwaran 1987862287 [Gnu][Driver] Use StringRef conversion functions
llvm-svn: 221648
2014-11-11 00:40:32 +00:00
Shankar Easwaran 7757f1ace6 [Gnu] Support --image-base option
The value for --image-base is used as the base address of the program.

llvm-svn: 221589
2014-11-10 14:55:21 +00:00
Shankar Easwaran 78aa81d6ed [Gnu] Add options that are ignored
Add options that are ignored and exists just for compatibility reasons.

llvm-svn: 221585
2014-11-10 14:54:49 +00:00
Shankar Easwaran ba154afa06 [ELF] Support -z max-page-size option
The GNU linker allows the user to change the page size by using the option -z
max-page-size.

llvm-svn: 221584
2014-11-10 14:54:43 +00:00
Shankar Easwaran 22c76a5d79 [ELF] Support --no-align-segments.
lld generates an ELF by adhering to the ELF spec by aligning vma/fileoffset to a
page boundary, but this becomes an issue when dealing with large pages. This
adds support so that lld generated executables adheres to the ELF spec with the
rule vma % p_align = offset % p_align.

This is supported by the flag --no-align-segments.

This could be the default in few targets like X86_64 to save space on disk.

llvm-svn: 221571
2014-11-08 03:44:49 +00:00
Nick Kledzik f04ab0a244 [mach-o] Fix MachOFileNode to own archives same as ELFFileNode
My previous fix to have FileArchive own the member MemoryBuffers was not a
complete solution for darwin because nothing owned the FileArchive object.
Fixed MachOFileNode to be like ELFFileNode and have the graph node own the
archive object.

llvm-svn: 221552
2014-11-07 22:00:23 +00:00
Nick Kledzik 82d24bc932 [mach-o] Add support for -order_file option
The darwin linker lets you rearrange functions and data for better locality
(less paging).  You do this with the -order_file option which supplies a text
file containing one symbol per line.

Implementing this required a small change to LayoutPass to add a custom sorter
hook.

llvm-svn: 221545
2014-11-07 21:01:21 +00:00
Nick Kledzik babee83257 remove unneeded { }
llvm-svn: 221478
2014-11-06 19:33:59 +00:00
Nick Kledzik 24f504001d [mach-o] Add support for -force_load option
The darwin linker has two ways to force all members of an archive to be loaded.
The -all_load option applies to all static libraries.  The -force_load takes
a path to a library and just that library's members are force loaded.

llvm-svn: 221477
2014-11-06 19:33:57 +00:00
Nick Kledzik 8f75da0db3 [mach-o] Add support for -S option
The darwin linker does not process dwarf debug info.  Instead it produces a
"debug map" in the output file which points back to the original .o files for
anything that wants debug info (e.g. debugger).

The -S option means "don't add a debug map".  lld for mach-o currently does
not generate the debug map, so there is nothing to do when this option is used.
But we need to process the option to get existing projects building.

llvm-svn: 221432
2014-11-06 03:03:42 +00:00
Nick Kledzik 80118548d2 Fix shadowed variable warning
llvm-svn: 221414
2014-11-05 23:40:04 +00:00
Rafael Espindola 5a899e332b Use llvm::sys::findProgramByName. NFC.
llvm-svn: 221257
2014-11-04 12:34:32 +00:00
Nick Kledzik cb955ae1a5 [mach-o] Turn -single_module error into a warning
To be more compatible with existing darwin linker.

llvm-svn: 220822
2014-10-28 21:11:02 +00:00
Rui Ueyama 8bae8189b2 [PECOFF] Fix exported symbol in the import library
There are two ways to specify a symbol to be exported in the module
definition file.

1) EXPORT <external name> = <symbol>
2) EXPORT <symbol>

In (1), you give both external name and internal name. In that case,
the linker tries to find a symbol using the internal name, and write
that address to the export table with the external name. Thus, from
the outer world, the symbol seems to be exported as the external name.

In (2), internal name is basically the same as the external name
with an exception: if you give an undecorated symbol to the EXPORT
directive, and if the linker finds a decorated symbol, the external
name for the symbol will become the decorated symbol.

LLD didn't implement that exception correctly. This patch fixes that.

llvm-svn: 220333
2014-10-21 21:41:28 +00:00
Nick Kledzik a721db24c0 Subclass InputGraph to get darwin linker library semantics
The darwin linker operates differently than the gnu linker with respect to
libraries. The darwin linker first links in all object files from the command
line, then to resolve any remaining undefines, it repeatedly iterates over
libraries on the command line until either all undefines are resolved or no
undefines were resolved in the last pass.

When Shankar made the InputGraph model, the plan for darwin was for the darwin
driver to place all libraries in a group at the end of the InputGraph. Thus
making the darwin model a subset of the gnu model. But it turns out that does
not work because the driver cannot tell if a file is an object or library until
it has been loaded, which happens later.

This solution is to subclass InputGraph for darwin and just iterate the graph
the way darwin linker needs.

llvm-svn: 220330
2014-10-21 21:14:11 +00:00
Shankar Easwaran 1b44519a6c [Gnu] Move certain options to a different group.
The -mllvm option is not specific to LTO.

llvm-svn: 220263
2014-10-21 04:02:32 +00:00
Shankar Easwaran 3dadd20fe5 [ELF] Implement demangle.
This adds functionality in the GNU flavor to demangle symbols when
undefined symbols are displayed to the user.

llvm-svn: 220184
2014-10-20 05:04:53 +00:00
Shankar Easwaran 2b67fca033 Sort include files according to convention.
llvm-svn: 220131
2014-10-18 05:33:55 +00:00
NAKAMURA Takumi 8c38df4e9e [CMake] lld: Introduce ${cmake_2_8_12_INTERFACE} onto each target_link_libraries. [PR20254]
FIXME: Dependencies should be reorganized.
llvm-svn: 220000
2014-10-17 00:37:49 +00:00
Nick Kledzik 3a118aa344 [mach-o] Fix crash when -all_load used with dylibs
-all_load tells the darwin linker to immediately load all members of all
archives.  The code do that used reinterpret_cast<> instead of dyn_cast<>.
If the file was a dylib, the reinterpret_cast<> turned a pointer to a dylib
into a pointer to an archive...boom.

Added test case to reproduce the crash, simplified the code and used dyn_cast<>.

llvm-svn: 219990
2014-10-16 23:42:42 +00:00
Nick Kledzik 5172067391 [mach-o] Add support for upward linking
To deal with cycles in shared library dependencies, the darwin linker supports
marking specific link dependencies as "upward".  An upward link is when a
lower level library links against a higher level library.

llvm-svn: 219949
2014-10-16 19:31:28 +00:00
Rui Ueyama b38b96ab4c [PECOFF] Support delay-load import table for x86
This patch creates the import address table and sets its
address to the delay-load import table. This also creates
wrapper functions for __delayLoadHelper2.

x86 only for now.

llvm-svn: 219948
2014-10-16 19:30:44 +00:00
Rui Ueyama 7be03d09cd [PECOFF] Emit the delay-import table
This is a partial patch to emit the delay-import table. With this,
LLD is now able to emit the table that llvm-readobj can read and
dump.

The table lacks a few fields, such as the address of HMODULE, the
import address table, etc. They'll be added in subsequent patches.

llvm-svn: 219384
2014-10-09 02:48:14 +00:00
Rui Ueyama ecbb309e5e [PECOFF] Remember DLL names given with /delayload option.
This is a step toward full support of /delayload.

llvm-svn: 219344
2014-10-08 21:38:10 +00:00
Rafael Auler ce1af1a201 [ELF] Implement --export-dynamic/-E
When creating a dynamic executable and receiving the -E flag, the linker should
export all globally visible symbols in its dynamic symbol table.

This commit also moves the logic that exports symbols in the dynamic symbol
table from OutputELFWriter to the ExecutableWriter class. It is not correct to
leave this at OutputELFWriter because DynamicLibraryWriter, another subclass of
OutputELFWriter, already exports all symbols, meaning we can potentially end up
with duplicated symbols in the dynamic symbol table when creating shared libs.

Reviewers: shankarke

http://reviews.llvm.org/D5585

llvm-svn: 219334
2014-10-08 18:54:26 +00:00
Rui Ueyama 0bc6d5f97d Fix typo in comment
llvm-svn: 219326
2014-10-08 17:50:15 +00:00
Shankar Easwaran c3550f9231 Add support to print version.
Summary: Add support in the universal driver to print the lld version and the
repository version.

Test Plan: A driver test is added

Reviewers: kledzik, ruiu

Reviewed By: ruiu

Subscribers: llvm-commits

Projects: #lld

Differential Revision: http://reviews.llvm.org/D5641

llvm-svn: 219277
2014-10-08 03:47:51 +00:00
Nick Kledzik 14b5d208cb [mach-o] Support fat archives
mach-o supports "fat" files which are a header/table-of-contents followed by a
concatenation of mach-o files (or archives of mach-o files) built for
different architectures.  Previously, the support for fat files was in the
MachOReader, but that only supported fat .o files and dylibs (not archives).

The fix is to put the fat handing into MachOFileNode.  That way any input file
kind (including archives) can be fat.  MachOFileNode selects the sub-range
of the fat file that matches the arch being linked and creates a MemoryBuffer
for just that subrange.

llvm-svn: 219268
2014-10-08 01:48:10 +00:00
Rui Ueyama 8f0c803c26 [PECOFF] Infer machine type from input object file
If /machine option is omitted, the linker needs to infer that from
input object files. This patch implements that.

llvm-svn: 219180
2014-10-07 01:38:46 +00:00
Nick Kledzik 15eba696f6 [mach-o] add -iphoneos_version_min as alias for -ios_version_min
llvm-svn: 219041
2014-10-04 00:19:56 +00:00
Nick Kledzik 09d00bb4d7 [mach-o] Add support for -dependency_info command line option
This option is added by Xcode when it runs the linker.  It produces a binary
file which contains the file the linker used.  Xcode uses the info to
dynamically update it dependency tracking.

To check the content of the binary file, the test case uses a python script
to dump the binary file as text which FileCheck can check.

llvm-svn: 219039
2014-10-04 00:16:13 +00:00
Rui Ueyama 711471ea1e [PECOFF] Resolve __delayLoadHelper2 if /delayload is given
DLL delay importing is a feature to load a DLL lazily, instead of
at program start-up time.

If the feature is turned on with the /delayload flag, the linker
resolves the delay-load helper function. All function pointer table
entries for the DLL are initially pointing to the helper function.
When called, the function loads and resolves the DLL symbols using
dlopen-ish Windows system calls and then write the reuslts to the
function pointer table. The helper function is in "delayimp.lib".

Note that this feature is not completely implemented yet. LLD
also needs to emit the table that's consumed by the delay-load
helper function. That'll be done in another patch.

llvm-svn: 218943
2014-10-03 01:14:50 +00:00
Nick Kledzik 22c9073ada Add MachOLinkingContext parameter to MachOFileNode constructor.
No functionality change.  This removes a down-cast from LinkingContext to
MachOLinkingContext.

Also, remove const from LinkingContext::createImplicitFiles() to remove
the need for another const cast.  Seems reasonable for createImplicitFiles()
to need to modify the context (MachOLinkingContext does).

llvm-svn: 218796
2014-10-01 20:24:30 +00:00
Nick Kledzik be43d7ef29 [mach-o] Implement -demangle.
The darwin linker has the -demangle option which directs it to demangle C++
(and soon Swift) mangled symbol names. Long term we need some Diagnostics object
for formatting errors and warnings. But for now we have the Core linker just
writing messages to llvm::errs(). So, to enable demangling, I changed the
Resolver to call a LinkingContext method on the symbol name.

To make this more interesting, the demangling code is done via __cxa_demangle()
which is part of the C++ ABI, which is only supported on some platforms, so I
had to conditionalize the code with the config generated HAVE_CXXABI_H.

llvm-svn: 218718
2014-09-30 23:15:39 +00:00
Rui Ueyama fa67adc28d [PECOFF] Allow /export:<symbol>,PRTVATE.
PRIVATE option is also an undocumented feature.

llvm-svn: 218696
2014-09-30 20:09:31 +00:00
Rui Ueyama 3837e10002 [PECOFF] Fix /export option.
MSDN doesn't say about /export:foo=bar style option, but
it turned out MSVC link.exe actually accepts that. So we need that
too.

It also means that the export directive in the module definition
file and /export command line option are functionally equivalent.

llvm-svn: 218695
2014-09-30 20:03:11 +00:00
Nick Kledzik b166d45730 [mach-o] Move implementation of MachOFileNode::parse
Move method implementation from header file to .cpp file.  No functionality
change.

llvm-svn: 218639
2014-09-29 23:04:24 +00:00
Rui Ueyama de18f9657b [PECOFF] Keep unmangled name in the export table descriptor
The export table descriptor is a data structure to keep information
about the export table. It contains a symbol name, and the name may
or may not be mangled.

We need unmangled names for the export table, so we demangle them
before writing them to the export table.

Obviously this is not a correct round-trip conversion. That could
drop a leading underscore from a symbol because that's
indistinguishable from a mangled name.

What we need to do is to keep unmangled names. This patch does that.

llvm-svn: 218345
2014-09-24 00:55:15 +00:00
Rui Ueyama 251d9a34e2 [PECOFF] Simplify /machine option handling
/machine:ebc was previously recognized but rejected. Unknown architecture
names were handled differently but eventually rejected too. We don't need
to distinguish them.

llvm-svn: 218344
2014-09-24 00:21:45 +00:00
Rui Ueyama a3eff3afd9 Make -core/-flavor options have higher priority than linker name
Also allows -core/flavor to appear at any position in the command line.
Patch from Oleg Ranevskyy!

http://reviews.llvm.org/D5384

llvm-svn: 218321
2014-09-23 18:09:38 +00:00
Rui Ueyama 117ef70c98 [PECOFF] Handle PRIVATE keyword in the module definition file
A symbol in a module definition file may be annotated with the
PRIVATE keyword like this.

  EXPORTS
    func PRIVATE

The PRIVATE keyword does not affect the resulting .dll file.
But it prevents the symbol to be listed in the .lib (import
library) file.

llvm-svn: 218273
2014-09-22 20:50:46 +00:00
Saleem Abdulrasool c2c2937d55 Driver: accept /machine:arm for Windows linker
Accept /machine:arm as an argument.  This is changed to support ARM NT.
Although there is no way to differentiate between ARM (Windows CE) and ARM NT
(Windows on ARM), since LLVM currently only supports Windows on ARM, simply take
/machine:arm to mean Windows on ARM.

llvm-svn: 218105
2014-09-19 06:09:30 +00:00
Rui Ueyama f47c7fab71 Make anonymous namespace as small as possible.
LLVM coding style says that "static" is preferred for file-scope
functions.

Differential Revision: http://reviews.llvm.org/D5323

llvm-svn: 217692
2014-09-12 17:30:13 +00:00
Nick Kledzik ad0184056f [mach-o] support "0x" or "0X" as prefix on hex numbers on command line
This matches the strtoull() behavior in ld64.

llvm-svn: 217650
2014-09-12 00:16:29 +00:00
Tim Northover 1684a614b3 [mach-o]: support optional "0x" prefix for -image_base
llvm-svn: 217578
2014-09-11 10:31:46 +00:00
Tim Northover 5d95bd7037 [mach-o]: tighten up diagnostics for -image_base option
The provided base must also be a multiple of the system's page size, which is a
reasonable enough demand.

Also check the other diagnostics more thoroughly.

llvm-svn: 217577
2014-09-11 10:31:42 +00:00
Tim Northover 7b33f21f3d [mach-o]: Support deprecated -seg1addr alias for -image_base
Because NO LINKER MAY CHANGE. EVER. Even if it's a complete rewrite
from scratch.

llvm-svn: 217572
2014-09-11 07:56:20 +00:00
Rui Ueyama a726ef12a4 Make getFlavor function.
The dangling "else" at the end of #if looked a bit error-prone.
Make it a separate function. No functionality change.

llvm-svn: 217568
2014-09-11 03:13:20 +00:00
Nick Kledzik 50bda292c8 If lld is renamed (or symlinked) to "ld" automatically pick the right flavor.
The existing system linkers on Darwin and Linux are called "ld".  We'd like to
eventually drop in lld as "ld" and have it just work.  But lld is a universal
linker that requires the first option to be -flavor to know which command line
mode to emulate (gnu or darwin).

This change tests if argv[0] is "ld" and if so, if the tool was built on MacOSX
then assume the darwin flavor otherwise the gnu flavor.  There are two test
cases which copy lld to "ld" and then run it. One for darwin and one for linux.

llvm-svn: 217566
2014-09-11 00:52:05 +00:00
Tim Northover 936e053da0 [MSVC]: use StringRef::getAsInteger instead of strtoull
This keeps non-conformant MSVC implementations happy.

llvm-svn: 217491
2014-09-10 11:15:36 +00:00
Tim Northover af3075b93f [mach-o]: implement -image_base option on Darwin.
As suggested by Nick, this will make __unwind_info implementation more natural,
and it'd have to be done at some point anyway.

llvm-svn: 217486
2014-09-10 10:39:57 +00:00
Nick Kledzik 1944bb6cf9 [mach-o] Fix missing { }
llvm-svn: 217411
2014-09-09 02:20:32 +00:00
Nick Kledzik 4ce63c2a54 [mach-o] Fix so no options causes usage message to be printed
llvm-svn: 217409
2014-09-09 00:45:35 +00:00
Nick Kledzik b7035ae367 [mach-o] Add support for -pie and -no_pie
There is a bit (MH_PIE) in the flags field of the mach_header which tells
the kernel is a program was built position independent (for ASLR).  The linker
automatically attempts to build programs PIE if they are built for a recent
OS version.  But the -pie and -no_pie options override that default behavior.

llvm-svn: 217408
2014-09-09 00:17:52 +00:00
Shankar Easwaran c7701e9a8c [ELF][Driver] Produce a proper error when file is not found
When a file is not found, produce a proper error message. The previous error
message produced a file format error, which made me wonder for a while why
there is a file format error, but essentially the file was not found.

This fixes the problem by producing a proper error message.

llvm-svn: 217359
2014-09-08 04:18:11 +00:00
Shankar Easwaran 595e056dce [ELF] Implement --rosegment
By default linker would not create a separate segment to hold read only data.

This option overrides that behavior by creating the a separate read only segment
for read only data.

llvm-svn: 217358
2014-09-08 04:05:52 +00:00
Shankar Easwaran 24511fc426 [ELF] Undefined symbols should be allowed when building dynamic libraries
When dynamic libraries are built, undefined symbols should always be allowed and
the linker should not exit with an error.

llvm-svn: 217356
2014-09-08 03:11:50 +00:00
Rui Ueyama 30a6bc286b [PECOFF] Run cvtres.exe with /machine:x64.
If we are creating a PE+ executable, we need to run cvtres with
/machine:x64 instead of /machine:x86. Otherwise the resulting executable
would be invalid.

llvm-svn: 217214
2014-09-05 00:26:28 +00:00
Nick Kledzik 635f9c7158 [mach-o] Let darwin driver infer arch from .o files if -arch not used.
Mach-O has a "fat" (or "universal") variant where the same contents built for
different architectures are concatenated into one file with a table-of-contents
header at the start.  But this leaves a dilemma for the linker - which
architecture to use.

Normally, the linker command line -arch is used to force which slice of any fat
files are used.  The clang compiler always passes -arch to the linker when
invoking it.  But some Makefiles invoke the linker directly and don’t specify
the -arch option.  For those cases, the linker scans all input files in command
line order and finds the first non-fat object file.  Whatever architecture it
is becomes the architecture for the link.

llvm-svn: 217189
2014-09-04 20:08:30 +00:00
Nick Kledzik cec33b84f2 If lld binary is named 'ld' on darwin, use darwin driver mode
llvm-svn: 217112
2014-09-04 00:03:36 +00:00
Rui Ueyama d718c73f0b [PECOFF] Add /HighEntropyVA.
This is yet another command line flag to set a bit in
DLLCharacteristics. Default on Win64 is "on".

llvm-svn: 216414
2014-08-25 22:23:34 +00:00
Rui Ueyama 768cc6d31d Use LLD naming style.
llvm-svn: 216402
2014-08-25 19:38:57 +00:00
Rafael Espindola 0941b568ce Update for LLVM api change
llvm-svn: 216395
2014-08-25 18:16:56 +00:00
Rui Ueyama 5711df44b8 [PECOFF] Fix PE+ relocations
The implementation of AMD64 relocations was imcomplete
and wrong. On AMD64, we of course have to use AMD64
relocations instead of i386 ones. This patch fixes the
issue.

LLD is now able to link hello64.obj (created from
hello64.asm) against user32.lib and kernel32.lib to
create a Win64 binary.

llvm-svn: 216253
2014-08-22 01:15:43 +00:00
Nick Kledzik e484075d82 [mach-o] Add support for -single_module and -multi_module
These are both obsolete options that controled how dylibs were built.

llvm-svn: 216205
2014-08-21 18:57:42 +00:00
Nick Kledzik 8c0bf75ef5 [mach-o] Add support for -exported_symbols_list and -keep_private_externs
Both options control the final scope of atoms.

When -exported_symbols_list <file> is used, the file is parsed into one
symbol per line in the file.  Only those symbols will be exported (global)
in the final linked image.

The -keep_private_externs option is only used with -r mode. Normally, -r
mode reduces private extern (scopeLinkageUnit) symbols to non-external. But
add the -keep_private_externs option keeps them private external.

llvm-svn: 216146
2014-08-21 01:59:11 +00:00