RFC: ELF Autolinking

At Sony we offer autolinking as a feature in our ELF toolchain. We would like to see full support for this feature upstream as there is anecdotal evidence that it would find use beyond Sony.

In general autolinking (https://en.wikipedia.org/wiki/Auto-linking) allows developers to specify inputs to the linker in their source code. LLVM and Clang already have support for autolinking on ELF via embedding strings, which specify linker behavior, into a .linker-options section in relocatable object files, see:

RFC - http://lists.llvm.org/pipermail/llvm-dev/2018-January/120101.html
LLVM - https://llvm.org/docs/Extensions.html#linker-options-section-linker-options, https://reviews.llvm.org/D40849
Clang - https://clang.llvm.org/docs/LanguageExtensions.html#specifying-linker-options-on-elf-targets, https://reviews.llvm.org/D42758

However, although support was added to Clang and LLVM, no support has been implemented in LLD; and, I get the sense, from reading the reviews, that there wasn’t agreement on the implementation when the changes landed. The original motivation seems to have been to remove the “autolink-extract” mechanism used by Swift to workaround the lack of autolinking support for ELF. However, looking at the Swift source code, Swift still seems to be using the “autolink-extract” method.

So my first question: Are there any users of the current implementation for ELF?

Assuming that no one is using the current code, I would like to suggest a different mechanism for autolinking.

For ELF we need limited autolinking support. Specifically, we only need support for “comment lib” pragmas (https://docs.microsoft.com/en-us/cpp/preprocessor/comment-c-cpp?view=vs-2017) in C/C++ e.g. #pragma comment(lib, “foo”). My suggestion that we keep the implementation as lean as possible.

Principles to guide the implementation:

  • Developers should be able to easily understand autolinking behavior.
  • Developers should be able to override autolinking from the linker command line.
  • Inputs specified via pragmas should be handled in a general way to allow the same source code to work in different environments.

I would like to propose that we focus on autolinking exclusively and that we divorce the implementation from the idea of “linker options” which, by nature, would tie source code to the vagaries of particular linkers. I don’t see much value in supporting other linker operations so I suggest that the binary representation be a mergable string section (SHF_MERGE, SHF_STRINGS), called .autolink, with custom type SHT_LLVM_AUTOLINK (0x6fff4c04), and SHF_EXCLUDE set (to avoid the contents appearing in the output). The compiler can form this section by concatenating the arguments of the “comment lib” pragmas in the order they are encountered. Partial (-r, -Ur) links can be handled by concatenating .autolink sections with the normal mergeable string section rules. The current .linker-options can remain (or be removed); but, “comment lib” pragmas for ELF should be lowered to .autolink not to .linker-options. This makes sense as there is no linker option that “comment lib” pragmas map directly to. As an example, #pragma comment(lib, “foo”) would result in:

.section “.autolink”,“eMS”,@llvm_autolink,1
.asciz “foo”

For LTO, equivalent information to the contents of a the .autolink section will be written to the IRSymtab so that it is available to the linker for symbol resolution.

The linker will process the .autolink strings in the following way:

  1. Inputs from the .autolink sections of a relocatable object file are added when the linker decides to include that file (which could itself be in a library) in the link. Autolinked inputs behave as if they were appended to the command line as a group after all other options. As a consequence the set of autolinked libraries are searched last to resolve symbols.
  2. It is an error if a file cannot be found for a given string.
  3. Any command line options in effect at the end of the command line parsing apply to autolinked inputs, e.g. --whole-archive.
  4. Duplicate autolinked inputs are ignored.
  5. The linker tries to add a library or relocatable object file from each of the strings in a .autolink section by; first, handling the string as if it was specified on the commandline; second, by looking for the string in each of the library search paths in turn; third, by looking for a lib.a or lib.so (depending on the current mode of the linker) in each of the library search paths.
  6. A new command line option --no-llvm-autolink will tell LLD to ignore the .autolink sections.

Rationale for the above points:

  1. Adding the autolinked inputs last makes the process simple to understand from a developers perspective. All linkers are able to implement this scheme.
  2. Error-ing for libraries that are not found seems like better behavior than failing the link during symbol resolution.
  3. It seems useful for the user to be able to apply command line options which will affect all of the autolinked input files. There is a potential problem of surprise for developers, who might not realize that these options would apply to the “invisible” autolinked input files; however, despite the potential for surprise, this is easy for developers to reason about and gives developers the control that they may require.
  4. Unlike on the command line it is probably easy to include the same input file twice via pragmas and might be a pain to fix; think of Third-party libraries supplied as binaries.
  5. This algorithm takes into account all of the different ways that ELF linkers find input files. The different search methods are tried by the linker in most obvious to least obvious order.
  6. I considered adding finer grained control over which .autolink inputs were ignored (e.g. MSVC has /nodefaultlib:); however, I concluded that this is not necessary: if finer control is required developers can recreate the same effect autolinking would have had using command line options.

Thoughts?

Hello,

I've put some comments on the proposal inline. Having to had to debug
library selection problems where all the libraries are visible on the
linker command line, I would prefer if people didn't embed difficult
to find directives in object files, but I'm guessing in some languages
this is the natural way of adding libraries.

At Sony we offer autolinking as a feature in our ELF toolchain. We would like to see full support for this feature upstream as there is anecdotal evidence that it would find use beyond Sony.

I've not got any use of the existing code. Personally I've not come
across anyone wanting this type of feature, but that is also anecdotal
on my part.

For ELF we need limited autolinking support. Specifically, we only need support for "comment lib" pragmas (https://docs.microsoft.com/en-us/cpp/preprocessor/comment-c-cpp?view=vs-2017) in C/C++ e.g. #pragma comment(lib, "foo"). My suggestion that we keep the implementation as lean as possible.

Principles to guide the implementation:
- Developers should be able to easily understand autolinking behavior.
- Developers should be able to override autolinking from the linker command line.
- Inputs specified via pragmas should be handled in a general way to allow the same source code to work in different environments.

I would like to propose that we focus on autolinking exclusively and that we divorce the implementation from the idea of "linker options" which, by nature, would tie source code to the vagaries of particular linkers. I don't see much value in supporting other linker operations so I suggest that the binary representation be a mergable string section (SHF_MERGE, SHF_STRINGS), called .autolink, with custom type SHT_LLVM_AUTOLINK (0x6fff4c04), and SHF_EXCLUDE set (to avoid the contents appearing in the output). The compiler can form this section by concatenating the arguments of the "comment lib" pragmas in the order they are encountered. Partial (-r, -Ur) links can be handled by concatenating .autolink sections with the normal mergeable string section rules. The current .linker-options can remain (or be removed); but, "comment lib" pragmas for ELF should be lowered to .autolink not to .linker-options. This makes sense as there is no linker option that "comment lib" pragmas map directly to. As an example, #pragma comment(lib, "foo") would result in:

.section ".autolink","eMS",@llvm_autolink,1
        .asciz "foo"

For LTO, equivalent information to the contents of a the .autolink section will be written to the IRSymtab so that it is available to the linker for symbol resolution.

I'm not sure I understand the bit about "for symbol resolution". I
think that what you mean is that you will encode the autolink section
using symbols instead of as a section, and the linker is expected to
extract this when it reads the symbol table?

The linker will process the .autolink strings in the following way:

1. Inputs from the .autolink sections of a relocatable object file are added when the linker decides to include that file (which could itself be in a library) in the link. Autolinked inputs behave as if they were appended to the command line as a group after all other options. As a consequence the set of autolinked libraries are searched last to resolve symbols.
2. It is an error if a file cannot be found for a given string.
3. Any command line options in effect at the end of the command line parsing apply to autolinked inputs, e.g. --whole-archive.

I've not got any experience of autolinking as a user, so I'm
struggling a bit with this one. I'm guessing that autolinking is
useful because someone can do the equivalent of #include <library.h>
and #pragma comment lib "library.so" in the same place without having
to fight the build system. I'm less convinced about --whole-archive as
I think this tends to be a way of structuring the build and would be
best made explicit in the build system. Moreover, what if someone
wants to not use --whole-archive, for their autolink, but one already
exists. This could be quite difficult to check with a large project.
Personally I'd have the user be explicit in the .autolink whether they
were intending it to be whole-archive or not.

4. Duplicate autolinked inputs are ignored.

If we take the issue of --whole-archive off the table does it matter
that there are duplicate libraries? Unresolved symbols will match
against the first library. I guess it might make a difference if this
feature is implemented in ld.lld and ld.gold, where you'd have to wrap
the libraries in a start-group, end-group, but is this likely to
happen?

5. The linker tries to add a library or relocatable object file from each of the strings in a .autolink section by; first, handling the string as if it was specified on the commandline; second, by looking for the string in each of the library search paths in turn; third, by looking for a lib<string>.a or lib<string>.so (depending on the current mode of the linker) in each of the library search paths.

There is some precedent for including files and libraries from
linkerscripts https://sourceware.org/binutils/docs/ld/File-Commands.html#File-Commands
, these distinguish between "-lfile" and "file". Would this be a
better fit for a ld.bfd interface compatible linker?

6. A new command line option --no-llvm-autolink will tell LLD to ignore the .autolink sections.

Personally I would have thought --no-llvm-autolink would error if it
found a .autolink section, on the grounds that I wanted all the
libraries to be defined on the command-line or linker script rather
than hidden in object files. I would have thought ignoring the
autolink sections would in most cases result in undefined symbols. If
there is a use case for it, perhaps --ignore-llvm-autolink.

Hi,

I guess I agree it would be best to remove the objfile linker option support and replace it with just auto-linking. We already have a mechanism for adding new features to object files: .note sections. Linkers already know to ignore ones that they don’t understand. If, in the future, we want to add a new feature that could be handled by embedding linker flags, we can instead implement it with a new .note section that other linkers and old versions of LLD will know to ignore.

On top of that, the generic ABI group has previously rejected proposals to embed linker options in object files (https://groups.google.com/forum/#!topic/generic-abi/iS_-m-X5ZwQ).

Given how ELF has done things in the past, maybe the section name should be “.note.autolink”. We could also be like GCC and namespace our extensions as “.note.LLVM.autolink”, but maybe that’s a step too far.

Reid

This proposal seems much better than the generic .linker-options scheme that potentially allows arbitrary linker options to be embedded to an object file. The proposed scheme is basically the same mechanism as the “comment lib” feature implemented on Microsoft linker, which I found mildly useful and at least not harmful.

As a use case, what I heard of was that in the game industry where many developers are using Visual Studio as an IDE and familiar with Windows’ semantics of linking, people find it annoying that to build the same program on Unix, they had to add bunch of -lfoo to the linker command line while they are automatically handled on Windows. I can understand that – if you have to add -lm 99.9% of the time when #include <math.h> for example, that’s not too odd to think why this is not processed automatically.

But the above story was from the game industry. Just like Ben, I’d like to hear from other people if they really want this feature.

Details inline:

At Sony we offer autolinking as a feature in our ELF toolchain. We would like to see full support for this feature upstream as there is anecdotal evidence that it would find use beyond Sony.

In general autolinking (https://en.wikipedia.org/wiki/Auto-linking) allows developers to specify inputs to the linker in their source code. LLVM and Clang already have support for autolinking on ELF via embedding strings, which specify linker behavior, into a .linker-options section in relocatable object files, see:

RFC - http://lists.llvm.org/pipermail/llvm-dev/2018-January/120101.html
LLVM - https://llvm.org/docs/Extensions.html#linker-options-section-linker-options, https://reviews.llvm.org/D40849
Clang - https://clang.llvm.org/docs/LanguageExtensions.html#specifying-linker-options-on-elf-targets, https://reviews.llvm.org/D42758

However, although support was added to Clang and LLVM, no support has been implemented in LLD; and, I get the sense, from reading the reviews, that there wasn’t agreement on the implementation when the changes landed. The original motivation seems to have been to remove the “autolink-extract” mechanism used by Swift to workaround the lack of autolinking support for ELF. However, looking at the Swift source code, Swift still seems to be using the “autolink-extract” method.

So my first question: Are there any users of the current implementation for ELF?

Assuming that no one is using the current code, I would like to suggest a different mechanism for autolinking.

For ELF we need limited autolinking support. Specifically, we only need support for “comment lib” pragmas (https://docs.microsoft.com/en-us/cpp/preprocessor/comment-c-cpp?view=vs-2017) in C/C++ e.g. #pragma comment(lib, “foo”). My suggestion that we keep the implementation as lean as possible.

Principles to guide the implementation:

  • Developers should be able to easily understand autolinking behavior.
  • Developers should be able to override autolinking from the linker command line.
  • Inputs specified via pragmas should be handled in a general way to allow the same source code to work in different environments.

I would like to propose that we focus on autolinking exclusively and that we divorce the implementation from the idea of “linker options” which, by nature, would tie source code to the vagaries of particular linkers. I don’t see much value in supporting other linker operations so I suggest that the binary representation be a mergable string section (SHF_MERGE, SHF_STRINGS), called .autolink, with custom type SHT_LLVM_AUTOLINK (0x6fff4c04), and SHF_EXCLUDE set (to avoid the contents appearing in the output). The compiler can form this section by concatenating the arguments of the “comment lib” pragmas in the order they are encountered. Partial (-r, -Ur) links can be handled by concatenating .autolink sections with the normal mergeable string section rules. The current .linker-options can remain (or be removed); but, “comment lib” pragmas for ELF should be lowered to .autolink not to .linker-options. This makes sense as there is no linker option that “comment lib” pragmas map directly to. As an example, #pragma comment(lib, “foo”) would result in:

.section “.autolink”,“eMS”,@llvm_autolink,1
.asciz “foo”

For LTO, equivalent information to the contents of a the .autolink section will be written to the IRSymtab so that it is available to the linker for symbol resolution.

The linker will process the .autolink strings in the following way:

  1. Inputs from the .autolink sections of a relocatable object file are added when the linker decides to include that file (which could itself be in a library) in the link. Autolinked inputs behave as if they were appended to the command line as a group after all other options. As a consequence the set of autolinked libraries are searched last to resolve symbols.
  2. It is an error if a file cannot be found for a given string.
  3. Any command line options in effect at the end of the command line parsing apply to autolinked inputs, e.g. --whole-archive.

I thought that the scope of this mechanism is essentially to add -lfoo automatically to the command line if you include a header that requires libfoo. From that perspective, the item 3 seems odd. Why do you need that?

  1. Duplicate autolinked inputs are ignored.

I’d say duplicate autolinked inputs are processed normally, but because of the same reason why the second parameter in -lfoo -lfoo is basically no-op, duplicated autolinked inputs are naturally ignored.

  1. The linker tries to add a library or relocatable object file from each of the strings in a .autolink section by; first, handling the string as if it was specified on the commandline; second, by looking for the string in each of the library search paths in turn; third, by looking for a lib.a or lib.so (depending on the current mode of the linker) in each of the library search paths.

Again, this seems like a little beyond the scope of what I expect (and it looks like you want to allow an .o file using this scheme).

Hello,

I’ve put some comments on the proposal inline. Having to had to debug
library selection problems where all the libraries are visible on the
linker command line, I would prefer if people didn’t embed difficult
to find directives in object files, but I’m guessing in some languages
this is the natural way of adding libraries.

At Sony we offer autolinking as a feature in our ELF toolchain. We would like to see full support for this feature upstream as there is anecdotal evidence that it would find use beyond Sony.

I’ve not got any use of the existing code. Personally I’ve not come
across anyone wanting this type of feature, but that is also anecdotal
on my part.

For ELF we need limited autolinking support. Specifically, we only need support for “comment lib” pragmas (https://docs.microsoft.com/en-us/cpp/preprocessor/comment-c-cpp?view=vs-2017) in C/C++ e.g. #pragma comment(lib, “foo”). My suggestion that we keep the implementation as lean as possible.

Principles to guide the implementation:

  • Developers should be able to easily understand autolinking behavior.
  • Developers should be able to override autolinking from the linker command line.
  • Inputs specified via pragmas should be handled in a general way to allow the same source code to work in different environments.

I would like to propose that we focus on autolinking exclusively and that we divorce the implementation from the idea of “linker options” which, by nature, would tie source code to the vagaries of particular linkers. I don’t see much value in supporting other linker operations so I suggest that the binary representation be a mergable string section (SHF_MERGE, SHF_STRINGS), called .autolink, with custom type SHT_LLVM_AUTOLINK (0x6fff4c04), and SHF_EXCLUDE set (to avoid the contents appearing in the output). The compiler can form this section by concatenating the arguments of the “comment lib” pragmas in the order they are encountered. Partial (-r, -Ur) links can be handled by concatenating .autolink sections with the normal mergeable string section rules. The current .linker-options can remain (or be removed); but, “comment lib” pragmas for ELF should be lowered to .autolink not to .linker-options. This makes sense as there is no linker option that “comment lib” pragmas map directly to. As an example, #pragma comment(lib, “foo”) would result in:

.section “.autolink”,“eMS”,@llvm_autolink,1
.asciz “foo”

For LTO, equivalent information to the contents of a the .autolink section will be written to the IRSymtab so that it is available to the linker for symbol resolution.

I’m not sure I understand the bit about “for symbol resolution”. I
think that what you mean is that you will encode the autolink section
using symbols instead of as a section, and the linker is expected to
extract this when it reads the symbol table?

Whoops… might have used a bit of a colloquialism there; sorry. All I mean is that there will be a method on the IRSymtab that LLD can use to retrieve the same set of strings that would be written into the the .autolink section of the relocatable object files by the backend.

The linker will process the .autolink strings in the following way:

  1. Inputs from the .autolink sections of a relocatable object file are added when the linker decides to include that file (which could itself be in a library) in the link. Autolinked inputs behave as if they were appended to the command line as a group after all other options. As a consequence the set of autolinked libraries are searched last to resolve symbols.
  2. It is an error if a file cannot be found for a given string.
  3. Any command line options in effect at the end of the command line parsing apply to autolinked inputs, e.g. --whole-archive.

I’ve not got any experience of autolinking as a user, so I’m
struggling a bit with this one. I’m guessing that autolinking is
useful because someone can do the equivalent of #include <library.h>
and #pragma comment lib “library.so” in the same place without having
to fight the build system.

Right. Consider that many codebases have multiple build configurations and the linker needs to be given the correct version of a library to use for the particular build configuration. This is often easier to do using the preprocessor than in the build system. Also, if a program is dependent on an external library, autolinking allows the library writer to reorganize how that library is structured transparently to the users of the library. There are notes about utility in https://stackoverflow.com/questions/1685206/pragma-commentlib-xxx-lib-equivalent-under-linux and https://stackoverflow.com/questions/3851956/whats-pragma-comment-lib-lib-glut32-lib?noredirect=1&lq=1.

I’m less convinced about --whole-archive as
I think this tends to be a way of structuring the build and would be
best made explicit in the build system. Moreover, what if someone
wants to not use --whole-archive, for their autolink, but one already
exists.

Then they can specify --no-whole-archive on the end of the command line, no?

This could be quite difficult to check with a large project.
Personally I’d have the user be explicit in the .autolink whether they
were intending it to be whole-archive or not.

I was hoping to avoid this as I want to avoid getting into how to specify linker specific options in the frontend. If we dislike the idea that the state of the command line parser at the end of the linker command line affects the autolinked libraries then I would rather go for a scheme in which the default state of the command line parser applies when linking the autolinked libraries; however, that seems harder to implement in LLD and gives the user less control over autolinking.

  1. Duplicate autolinked inputs are ignored.

If we take the issue of --whole-archive off the table does it matter
that there are duplicate libraries? Unresolved symbols will match
against the first library.

It doesn’t matter for libraries in LLD; but, it is important for object files. I think that this mechanism should be usable for object files an libraries. This is common in ELF linkers - for example the --library command line option can be used to link object files.

I guess it might make a difference if this
feature is implemented in ld.lld and ld.gold, where you’d have to wrap
the libraries in a start-group, end-group, but is this likely to
happen?

I would like the design to be such that it could be implemented by GNU.

  1. The linker tries to add a library or relocatable object file from each of the strings in a .autolink section by; first, handling the string as if it was specified on the commandline; second, by looking for the string in each of the library search paths in turn; third, by looking for a lib.a or lib.so (depending on the current mode of the linker) in each of the library search paths.

There is some precedent for including files and libraries from
linkerscripts https://sourceware.org/binutils/docs/ld/File-Commands.html#File-Commands
, these distinguish between “-lfile” and “file”. Would this be a
better fit for a ld.bfd interface compatible linker?

I was hoping to avoid GNUism’s and use a “general” mechanism. MSVC source code compatibility is a usecase.

  1. A new command line option --no-llvm-autolink will tell LLD to ignore the .autolink sections.

Personally I would have thought --no-llvm-autolink would error if it
found a .autolink section, on the grounds that I wanted all the
libraries to be defined on the command-line or linker script rather
than hidden in object files. I would have thought ignoring the
autolink sections would in most cases result in undefined symbols. If
there is a use case for it, perhaps --ignore-llvm-autolink.

The usecase that I had in mind is that you need to override autolinking. To do so you tell the linker to ignore the embedded autolinking information and construct an equivalent command line. I think your proposed --ignore-llvm-autolink is a better name for this option given the intended semantics.

Another situation where it is useful is when a 3rd party library supports multiple different ABI-incompatible build configurations (typically selected between via pre-processor settings). This way, the header file can choose the right version of the library to link against.

This comes up often when linking against python, for example, where if you have #defined _DEBUG then you need to link against python35_d.[dll|so] versus python35.[dll|so].

You can certainly represent this kind of logic in a build system, but it leads to even more maintenance burden in my experience, and in general it’s nice if things “just work”.

Hi,

I guess I agree it would be best to remove the objfile linker option support and replace it with just auto-linking. We already have a mechanism for adding new features to object files: .note sections. Linkers already know to ignore ones that they don’t understand. If, in the future, we want to add a new feature that could be handled by embedding linker flags, we can instead implement it with a new .note section that other linkers and old versions of LLD will know to ignore.

On top of that, the generic ABI group has previously rejected proposals to embed linker options in object files (https://groups.google.com/forum/#!topic/generic-abi/iS_-m-X5ZwQ).

Given how ELF has done things in the past, maybe the section name should be “.note.autolink”. We could also be like GCC and namespace our extensions as “.note.LLVM.autolink”, but maybe that’s a step too far.

A .note section consists of a series of type-length-value records. My understanding is that the static linker aggregates them to a single location and put it into PT_NOTE segment, and the records can still be read by the loader even after the section table is stripped from an executable. For the proposed purpose, the note section header would not be useful or meaningful, so a plain section that just contains an ASCII string would be simpler.

Hi,

I guess I agree it would be best to remove the objfile linker option support and replace it with just auto-linking. We already have a mechanism for adding new features to object files: .note sections. Linkers already know to ignore ones that they don’t understand. If, in the future, we want to add a new feature that could be handled by embedding linker flags, we can instead implement it with a new .note section that other linkers and old versions of LLD will know to ignore.

On top of that, the generic ABI group has previously rejected proposals to embed linker options in object files (https://groups.google.com/forum/#!topic/generic-abi/iS_-m-X5ZwQ).

Given how ELF has done things in the past, maybe the section name should be “.note.autolink”. We could also be like GCC and namespace our extensions as “.note.LLVM.autolink”, but maybe that’s a step too far.

A .note section consists of a series of type-length-value records. My understanding is that the static linker aggregates them to a single location and put it into PT_NOTE segment, and the records can still be read by the loader even after the section table is stripped from an executable. For the proposed purpose, the note section header would not be useful or meaningful, so a plain section that just contains an ASCII string would be simpler.

I agree that this should not be a note section because the section is meaningless in a final linked executable.

Note that the proposal is limited to adding -l flags only. I would not expect this particular section to be expanded in the future to cover other use cases; rather, we would likely invent more section types.

Peter

At Sony we offer autolinking as a feature in our ELF toolchain. We would like to see full support for this feature upstream as there is anecdotal evidence that it would find use beyond Sony.

In general autolinking (https://en.wikipedia.org/wiki/Auto-linking) allows developers to specify inputs to the linker in their source code. LLVM and Clang already have support for autolinking on ELF via embedding strings, which specify linker behavior, into a .linker-options section in relocatable object files, see:

RFC - http://lists.llvm.org/pipermail/llvm-dev/2018-January/120101.html
LLVM - https://llvm.org/docs/Extensions.html#linker-options-section-linker-options, https://reviews.llvm.org/D40849
Clang - https://clang.llvm.org/docs/LanguageExtensions.html#specifying-linker-options-on-elf-targets, https://reviews.llvm.org/D42758

However, although support was added to Clang and LLVM, no support has been implemented in LLD; and, I get the sense, from reading the reviews, that there wasn’t agreement on the implementation when the changes landed. The original motivation seems to have been to remove the “autolink-extract” mechanism used by Swift to workaround the lack of autolinking support for ELF. However, looking at the Swift source code, Swift still seems to be using the “autolink-extract” method.

So my first question: Are there any users of the current implementation for ELF?

Assuming that no one is using the current code, I would like to suggest a different mechanism for autolinking.

For ELF we need limited autolinking support. Specifically, we only need support for “comment lib” pragmas (https://docs.microsoft.com/en-us/cpp/preprocessor/comment-c-cpp?view=vs-2017) in C/C++ e.g. #pragma comment(lib, “foo”). My suggestion that we keep the implementation as lean as possible.

Principles to guide the implementation:

  • Developers should be able to easily understand autolinking behavior.
  • Developers should be able to override autolinking from the linker command line.
  • Inputs specified via pragmas should be handled in a general way to allow the same source code to work in different environments.

I would like to propose that we focus on autolinking exclusively and that we divorce the implementation from the idea of “linker options” which, by nature, would tie source code to the vagaries of particular linkers. I don’t see much value in supporting other linker operations so I suggest that the binary representation be a mergable string section (SHF_MERGE, SHF_STRINGS), called .autolink, with custom type SHT_LLVM_AUTOLINK (0x6fff4c04), and SHF_EXCLUDE set (to avoid the contents appearing in the output).

Should we set SHF_EXCLUDE on the section? I think it may be better not to set it. The semantics of the bit are that the section will be excluded from the final output file, but there’s no requirement that other sections are not excluded. For example SHT_REL(A) sections are interpreted by the linker and excluded from the output, but they do not have SHF_EXCLUDE set. The SHT_LLVM_AUTOLINK section could be treated similarly in compatible linkers. SHF_EXCLUDE is appropriate for sections that can be freely dropped without changing the semantics, such as the address significance table, but the autolink section is different because dropping it changes the semantics. By not setting the bit, we would cause incompatible linkers to leave the autolink section in the output file, which would allow downstream tools to be written that would detect (and possibly diagnose) such files.

Peter

I agree that this should not be a note section because the section is meaningless in a final linked executable.

Sounds good. I’m clearly super familiar with ELF. =P

Note that the proposal is limited to adding -l flags only. I would not expect this particular section to be expanded in the future to cover other use cases; rather, we would likely invent more section types.

Sounds like we agree on this, though.

The linker will process the .autolink strings in the following way:

  1. Inputs from the .autolink sections of a relocatable object file are added when the linker decides to include that file (which could itself be in a library) in the link. Autolinked inputs behave as if they were appended to the command line as a group after all other options. As a consequence the set of autolinked libraries are searched last to resolve symbols.
  2. It is an error if a file cannot be found for a given string.
  3. Any command line options in effect at the end of the command line parsing apply to autolinked inputs, e.g. --whole-archive.

I thought that the scope of this mechanism is essentially to add -lfoo automatically to the command line if you include a header that requires libfoo. >From that perspective, the item 3 seems odd. Why do you need that?

I replied to Peter already on this; but, what I’m basically saying is that if you had a command line like…

bob.o --whole-archive -lbar

… and bob.o was built with #pragma comment(lib, “foo”), then this effectively is transformed into…

bob.o --whole-archive -lbar -lfoo

…and the --whole-archive applies to the -lfoo. The alternative that you use some defaults when handling the autolinked libraries.

  1. Duplicate autolinked inputs are ignored.

I’d say duplicate autolinked inputs are processed normally, but because of the same reason why the second parameter in -lfoo -lfoo is basically no-op, duplicated autolinked inputs are naturally ignored.

Again, replied to Peter - .o files and I also want to leave the door open for GNU to be able to implement this.

  1. The linker tries to add a library or relocatable object file from each of the strings in a .autolink section by; first, handling the string as if it was specified on the commandline; second, by looking for the string in each of the library search paths in turn; third, by looking for a lib.a or lib.so (depending on the current mode of the linker) in each of the library search paths.

Again, this seems like a little beyond the scope of what I expect (and it looks like you want to allow an .o file using this scheme).

Again, replied to Peter. I think that .o files should be included, --library allows this.

Hello,

I’ve put some comments on the proposal inline. Having to had to debug
library selection problems where all the libraries are visible on the
linker command line, I would prefer if people didn’t embed difficult
to find directives in object files, but I’m guessing in some languages
this is the natural way of adding libraries.

At Sony we offer autolinking as a feature in our ELF toolchain. We would like to see full support for this feature upstream as there is anecdotal evidence that it would find use beyond Sony.

I’ve not got any use of the existing code. Personally I’ve not come
across anyone wanting this type of feature, but that is also anecdotal
on my part.

For ELF we need limited autolinking support. Specifically, we only need support for “comment lib” pragmas (https://docs.microsoft.com/en-us/cpp/preprocessor/comment-c-cpp?view=vs-2017) in C/C++ e.g. #pragma comment(lib, “foo”). My suggestion that we keep the implementation as lean as possible.

Principles to guide the implementation:

  • Developers should be able to easily understand autolinking behavior.
  • Developers should be able to override autolinking from the linker command line.
  • Inputs specified via pragmas should be handled in a general way to allow the same source code to work in different environments.

I would like to propose that we focus on autolinking exclusively and that we divorce the implementation from the idea of “linker options” which, by nature, would tie source code to the vagaries of particular linkers. I don’t see much value in supporting other linker operations so I suggest that the binary representation be a mergable string section (SHF_MERGE, SHF_STRINGS), called .autolink, with custom type SHT_LLVM_AUTOLINK (0x6fff4c04), and SHF_EXCLUDE set (to avoid the contents appearing in the output). The compiler can form this section by concatenating the arguments of the “comment lib” pragmas in the order they are encountered. Partial (-r, -Ur) links can be handled by concatenating .autolink sections with the normal mergeable string section rules. The current .linker-options can remain (or be removed); but, “comment lib” pragmas for ELF should be lowered to .autolink not to .linker-options. This makes sense as there is no linker option that “comment lib” pragmas map directly to. As an example, #pragma comment(lib, “foo”) would result in:

.section “.autolink”,“eMS”,@llvm_autolink,1
.asciz “foo”

For LTO, equivalent information to the contents of a the .autolink section will be written to the IRSymtab so that it is available to the linker for symbol resolution.

I’m not sure I understand the bit about “for symbol resolution”. I
think that what you mean is that you will encode the autolink section
using symbols instead of as a section, and the linker is expected to
extract this when it reads the symbol table?

Whoops… might have used a bit of a colloquialism there; sorry. All I mean is that there will be a method on the IRSymtab that LLD can use to retrieve the same set of strings that would be written into the the .autolink section of the relocatable object files by the backend.

The linker will process the .autolink strings in the following way:

  1. Inputs from the .autolink sections of a relocatable object file are added when the linker decides to include that file (which could itself be in a library) in the link. Autolinked inputs behave as if they were appended to the command line as a group after all other options. As a consequence the set of autolinked libraries are searched last to resolve symbols.
  2. It is an error if a file cannot be found for a given string.
  3. Any command line options in effect at the end of the command line parsing apply to autolinked inputs, e.g. --whole-archive.

I’ve not got any experience of autolinking as a user, so I’m
struggling a bit with this one. I’m guessing that autolinking is
useful because someone can do the equivalent of #include <library.h>
and #pragma comment lib “library.so” in the same place without having
to fight the build system.

Right. Consider that many codebases have multiple build configurations and the linker needs to be given the correct version of a library to use for the particular build configuration. This is often easier to do using the preprocessor than in the build system. Also, if a program is dependent on an external library, autolinking allows the library writer to reorganize how that library is structured transparently to the users of the library. There are notes about utility in https://stackoverflow.com/questions/1685206/pragma-commentlib-xxx-lib-equivalent-under-linux and https://stackoverflow.com/questions/3851956/whats-pragma-comment-lib-lib-glut32-lib?noredirect=1&lq=1.

I’m less convinced about --whole-archive as
I think this tends to be a way of structuring the build and would be
best made explicit in the build system. Moreover, what if someone
wants to not use --whole-archive, for their autolink, but one already
exists.

Then they can specify --no-whole-archive on the end of the command line, no?

This could be quite difficult to check with a large project.
Personally I’d have the user be explicit in the .autolink whether they
were intending it to be whole-archive or not.

I was hoping to avoid this as I want to avoid getting into how to specify linker specific options in the frontend. If we dislike the idea that the state of the command line parser at the end of the linker command line affects the autolinked libraries then I would rather go for a scheme in which the default state of the command line parser applies when linking the autolinked libraries; however, that seems harder to implement in LLD and gives the user less control over autolinking.

I think that handling .autolink’ed files in the default state is simpler, and it doesn’t seem too hard to implement.

The other option is to handle autolinked libraries as soon as we find them, so that if foo.o autolinks libbar, the linker would act as if foo.o in the command line is followed by -lbar. I’d think that’s not too bad or arguably more straightforward semantics than autolinking everything all at once at the end.

  1. Duplicate autolinked inputs are ignored.

If we take the issue of --whole-archive off the table does it matter
that there are duplicate libraries? Unresolved symbols will match
against the first library.

It doesn’t matter for libraries in LLD; but, it is important for object files. I think that this mechanism should be usable for object files an libraries. This is common in ELF linkers - for example the --library command line option can be used to link object files.

Do you actually often link .o file using -l? It seems a bit weird use of the option. To me, it seems better to limit the ability of autolinking to link against .so or .a.

Great point Peter, accepted. I’m rather afraid to say that I just copied the SHF_EXCLUDE from the existing .linker-options without thinking it through. When I prototyped this up SHF_EXCLUDE doesn’t actually buy you much because you still need to special case SHT_LLVM_AUTOLINK for -r links to undo the effect of SHF_EXCLUDE. Probably ELF needs some sort of EXCLUDED but not for -r links flag.

Exactly. To be clear I am not opposing a mechanism to embed linker command line options in input files, maybe someone does have a valid usecase for this. I am basing my arguments in the RFC on the idea that the syntax does matter, in the sense that #pragma comment(lib, “foo”) seems to imply a relatively “general” method of specifying a library to link, as opposed to some imagined syntax like #pragma linker-commandline-option("–library", “foo”) which is specific to the command line options of a particular linker. Also, given that it seems that autolinking is the only usecase currently, the proposed .autolink section is a simpler (e.g. binary tools can dump the section without needing special rules), and more efficient (bytes needed to convey the information) format than something like .linker-options.

Hello,

I've put some comments on the proposal inline. Having to had to debug
library selection problems where all the libraries are visible on the
linker command line, I would prefer if people didn't embed difficult
to find directives in object files, but I'm guessing in some languages
this is the natural way of adding libraries.

>
> At Sony we offer autolinking as a feature in our ELF toolchain. We would like to see full support for this feature upstream as there is anecdotal evidence that it would find use beyond Sony.
>

I've not got any use of the existing code. Personally I've not come
across anyone wanting this type of feature, but that is also anecdotal
on my part.

>
> For ELF we need limited autolinking support. Specifically, we only need support for "comment lib" pragmas (https://docs.microsoft.com/en-us/cpp/preprocessor/comment-c-cpp?view=vs-2017) in C/C++ e.g. #pragma comment(lib, "foo"). My suggestion that we keep the implementation as lean as possible.
>
> Principles to guide the implementation:
> - Developers should be able to easily understand autolinking behavior.
> - Developers should be able to override autolinking from the linker command line.
> - Inputs specified via pragmas should be handled in a general way to allow the same source code to work in different environments.
>
> I would like to propose that we focus on autolinking exclusively and that we divorce the implementation from the idea of "linker options" which, by nature, would tie source code to the vagaries of particular linkers. I don't see much value in supporting other linker operations so I suggest that the binary representation be a mergable string section (SHF_MERGE, SHF_STRINGS), called .autolink, with custom type SHT_LLVM_AUTOLINK (0x6fff4c04), and SHF_EXCLUDE set (to avoid the contents appearing in the output). The compiler can form this section by concatenating the arguments of the "comment lib" pragmas in the order they are encountered. Partial (-r, -Ur) links can be handled by concatenating .autolink sections with the normal mergeable string section rules. The current .linker-options can remain (or be removed); but, "comment lib" pragmas for ELF should be lowered to .autolink not to .linker-options. This makes sense as there is no linker option that "comment lib" pragmas map directly to. As an example, #pragma comment(lib, "foo") would result in:
>
> .section ".autolink","eMS",@llvm_autolink,1
> .asciz "foo"
>
> For LTO, equivalent information to the contents of a the .autolink section will be written to the IRSymtab so that it is available to the linker for symbol resolution.
>

I'm not sure I understand the bit about "for symbol resolution". I
think that what you mean is that you will encode the autolink section
using symbols instead of as a section, and the linker is expected to
extract this when it reads the symbol table?

Whoops... might have used a bit of a colloquialism there; sorry. All I mean is that there will be a method on the IRSymtab that LLD can use to retrieve the same set of strings that would be written into the the .autolink section of the relocatable object files by the backend.

> The linker will process the .autolink strings in the following way:
>
> 1. Inputs from the .autolink sections of a relocatable object file are added when the linker decides to include that file (which could itself be in a library) in the link. Autolinked inputs behave as if they were appended to the command line as a group after all other options. As a consequence the set of autolinked libraries are searched last to resolve symbols.
> 2. It is an error if a file cannot be found for a given string.
> 3. Any command line options in effect at the end of the command line parsing apply to autolinked inputs, e.g. --whole-archive.

I've not got any experience of autolinking as a user, so I'm
struggling a bit with this one. I'm guessing that autolinking is
useful because someone can do the equivalent of #include <library.h>
and #pragma comment lib "library.so" in the same place without having
to fight the build system.

Right. Consider that many codebases have multiple build configurations and the linker needs to be given the correct version of a library to use for the particular build configuration. This is often easier to do using the preprocessor than in the build system. Also, if a program is dependent on an external library, autolinking allows the library writer to reorganize how that library is structured transparently to the users of the library. There are notes about utility in https://stackoverflow.com/questions/1685206/pragma-commentlib-xxx-lib-equivalent-under-linux and https://stackoverflow.com/questions/3851956/whats-pragma-comment-lib-lib-glut32-lib?noredirect=1&lq=1.

I'm less convinced about --whole-archive as
I think this tends to be a way of structuring the build and would be
best made explicit in the build system. Moreover, what if someone
wants to not use --whole-archive, for their autolink, but one already
exists.

Then they can specify --no-whole-archive on the end of the command line, no?

This could be quite difficult to check with a large project.
Personally I'd have the user be explicit in the .autolink whether they
were intending it to be whole-archive or not.

I was hoping to avoid this as I want to avoid getting into how to specify linker specific options in the frontend. If we dislike the idea that the state of the command line parser at the end of the linker command line affects the autolinked libraries then I would rather go for a scheme in which the default state of the command line parser applies when linking the autolinked libraries; however, that seems harder to implement in LLD and gives the user less control over autolinking.

I think that handling .autolink'ed files in the default state is simpler, and it doesn't seem too hard to implement.

The other option is to handle autolinked libraries as soon as we find them, so that if foo.o autolinks libbar, the linker would act as if foo.o in the command line is followed by -lbar. I'd think that's not too bad or arguably more straightforward semantics than autolinking everything all at once at the end.

One of the difficulties of having --whole-archive applying to all the
libraries at the end is the case where some libraries need to be
--whole-archive and some do not. If I've understood correctly,
retaining the state at the end of the link line is all or nothing. I'm
torn over whether processing the library as soon as it is found is the
right one or not. Yes it does solve part of the --whole-archive
problem, but actually using it in practice would be difficult. It also
tends to move the autolink libraries ahead of the explicit ones and I
think user's would prefer the other way around.

> 4. Duplicate autolinked inputs are ignored.

If we take the issue of --whole-archive off the table does it matter
that there are duplicate libraries? Unresolved symbols will match
against the first library.

It doesn't matter for libraries in LLD; but, it is important for object files. I think that this mechanism should be usable for object files an libraries. This is common in ELF linkers - for example the --library command line option can be used to link object files.

Do you actually often link .o file using -l? It seems a bit weird use of the option. To me, it seems better to limit the ability of autolinking to link against .so or .a.

I think you need the explicit : in the namespec to load an object. I
can't say I've seen it used in anger before in an ELF context, but it
may be more useful in an autolink context to effectively (ab)use the
pre-processor into doing conditional inclusion of object files.

At Sony we offer autolinking as a feature in our ELF toolchain. We would like to see full support for this feature upstream as there is anecdotal evidence that it would find use beyond Sony.

In general autolinking (https://en.wikipedia.org/wiki/Auto-linking) allows developers to specify inputs to the linker in their source code. LLVM and Clang already have support for autolinking on ELF via embedding strings, which specify linker behavior, into a .linker-options section in relocatable object files, see:

RFC - http://lists.llvm.org/pipermail/llvm-dev/2018-January/120101.html
LLVM - https://llvm.org/docs/Extensions.html#linker-options-section-linker-options, https://reviews.llvm.org/D40849
Clang - https://clang.llvm.org/docs/LanguageExtensions.html#specifying-linker-options-on-elf-targets, https://reviews.llvm.org/D42758

However, although support was added to Clang and LLVM, no support has been implemented in LLD; and, I get the sense, from reading the reviews, that there wasn’t agreement on the implementation when the changes landed. The original motivation seems to have been to remove the “autolink-extract” mechanism used by Swift to workaround the lack of autolinking support for ELF. However, looking at the Swift source code, Swift still seems to be using the “autolink-extract” method.

So my first question: Are there any users of the current implementation for ELF?

Assuming that no one is using the current code, I would like to suggest a different mechanism for autolinking.

For ELF we need limited autolinking support. Specifically, we only need support for “comment lib” pragmas (https://docs.microsoft.com/en-us/cpp/preprocessor/comment-c-cpp?view=vs-2017) in C/C++ e.g. #pragma comment(lib, “foo”). My suggestion that we keep the implementation as lean as possible.

Principles to guide the implementation:

  • Developers should be able to easily understand autolinking behavior.
  • Developers should be able to override autolinking from the linker command line.
  • Inputs specified via pragmas should be handled in a general way to allow the same source code to work in different environments.

I would like to propose that we focus on autolinking exclusively and that we divorce the implementation from the idea of “linker options” which, by nature, would tie source code to the vagaries of particular linkers. I don’t see much value in supporting other linker operations so I suggest that the binary representation be a mergable string section (SHF_MERGE, SHF_STRINGS), called .autolink, with custom type SHT_LLVM_AUTOLINK (0x6fff4c04), and SHF_EXCLUDE set (to avoid the contents appearing in the output). The compiler can form this section by concatenating the arguments of the “comment lib” pragmas in the order they are encountered. Partial (-r, -Ur) links can be handled by concatenating .autolink sections with the normal mergeable string section rules. The current .linker-options can remain (or be removed); but, “comment lib” pragmas for ELF should be lowered to .autolink not to .linker-options. This makes sense as there is no linker option that “comment lib” pragmas map directly to. As an example, #pragma comment(lib, “foo”) would result in:

.section “.autolink”,“eMS”,@llvm_autolink,1
.asciz “foo”

For LTO, equivalent information to the contents of a the .autolink section will be written to the IRSymtab so that it is available to the linker for symbol resolution.

The linker will process the .autolink strings in the following way:

  1. Inputs from the .autolink sections of a relocatable object file are added when the linker decides to include that file (which could itself be in a library) in the link. Autolinked inputs behave as if they were appended to the command line as a group after all other options. As a consequence the set of autolinked libraries are searched last to resolve symbols.

If we want this to be compatible with GNU linkers, doesn’t the autolinked input need to appear at the point immediately after the object file appears in the link? I’m imagining the case where you have a statically linked libc as well as a libbar.a autolinked from a foo.o. The link command line would look like this:

ld foo.o -lc

Now foo.o autolinks against bar. The command line becomes:

ld foo.o -lc -lbar

If libbar.a requires an additional object file from libc.a, it will not be added to the link.

  1. It is an error if a file cannot be found for a given string.
  2. Any command line options in effect at the end of the command line parsing apply to autolinked inputs, e.g. --whole-archive.
  3. Duplicate autolinked inputs are ignored.

This seems like it would work in GNU linkers, as long as the autolinked file is added to the link immediately after the last mention, rather than the first. Otherwise a command line like:

ld foo1.o foo2.o

(where foo1.o and foo2.o both autolink bar) could end up looking like:

ld foo1.o -lbar foo2.o

and you will not link anything from libbar.a that only foo2.o requires. It may end up being simpler to not ignore duplicates.

  1. The linker tries to add a library or relocatable object file from each of the strings in a .autolink section by; first, handling the string as if it was specified on the commandline; second, by looking for the string in each of the library search paths in turn; third, by looking for a lib.a or lib.so (depending on the current mode of the linker) in each of the library search paths.

Is the second part necessary? “-l:foo” causes the linker to search for a file named “foo” in the library search path, so it seems that allowing the autolink string to look like “:foo” would satisfy this use case.

Hello,

I’ve put some comments on the proposal inline. Having to had to debug
library selection problems where all the libraries are visible on the
linker command line, I would prefer if people didn’t embed difficult
to find directives in object files, but I’m guessing in some languages
this is the natural way of adding libraries.

At Sony we offer autolinking as a feature in our ELF toolchain. We would like to see full support for this feature upstream as there is anecdotal evidence that it would find use beyond Sony.

I’ve not got any use of the existing code. Personally I’ve not come
across anyone wanting this type of feature, but that is also anecdotal
on my part.

For ELF we need limited autolinking support. Specifically, we only need support for “comment lib” pragmas (https://docs.microsoft.com/en-us/cpp/preprocessor/comment-c-cpp?view=vs-2017) in C/C++ e.g. #pragma comment(lib, “foo”). My suggestion that we keep the implementation as lean as possible.

Principles to guide the implementation:

  • Developers should be able to easily understand autolinking behavior.
  • Developers should be able to override autolinking from the linker command line.
  • Inputs specified via pragmas should be handled in a general way to allow the same source code to work in different environments.

I would like to propose that we focus on autolinking exclusively and that we divorce the implementation from the idea of “linker options” which, by nature, would tie source code to the vagaries of particular linkers. I don’t see much value in supporting other linker operations so I suggest that the binary representation be a mergable string section (SHF_MERGE, SHF_STRINGS), called .autolink, with custom type SHT_LLVM_AUTOLINK (0x6fff4c04), and SHF_EXCLUDE set (to avoid the contents appearing in the output). The compiler can form this section by concatenating the arguments of the “comment lib” pragmas in the order they are encountered. Partial (-r, -Ur) links can be handled by concatenating .autolink sections with the normal mergeable string section rules. The current .linker-options can remain (or be removed); but, “comment lib” pragmas for ELF should be lowered to .autolink not to .linker-options. This makes sense as there is no linker option that “comment lib” pragmas map directly to. As an example, #pragma comment(lib, “foo”) would result in:

.section “.autolink”,“eMS”,@llvm_autolink,1
.asciz “foo”

For LTO, equivalent information to the contents of a the .autolink section will be written to the IRSymtab so that it is available to the linker for symbol resolution.

I’m not sure I understand the bit about “for symbol resolution”. I
think that what you mean is that you will encode the autolink section
using symbols instead of as a section, and the linker is expected to
extract this when it reads the symbol table?

Whoops… might have used a bit of a colloquialism there; sorry. All I mean is that there will be a method on the IRSymtab that LLD can use to retrieve the same set of strings that would be written into the the .autolink section of the relocatable object files by the backend.

The linker will process the .autolink strings in the following way:

  1. Inputs from the .autolink sections of a relocatable object file are added when the linker decides to include that file (which could itself be in a library) in the link. Autolinked inputs behave as if they were appended to the command line as a group after all other options. As a consequence the set of autolinked libraries are searched last to resolve symbols.
  2. It is an error if a file cannot be found for a given string.
  3. Any command line options in effect at the end of the command line parsing apply to autolinked inputs, e.g. --whole-archive.

I’ve not got any experience of autolinking as a user, so I’m
struggling a bit with this one. I’m guessing that autolinking is
useful because someone can do the equivalent of #include <library.h>
and #pragma comment lib “library.so” in the same place without having
to fight the build system.

Right. Consider that many codebases have multiple build configurations and the linker needs to be given the correct version of a library to use for the particular build configuration. This is often easier to do using the preprocessor than in the build system. Also, if a program is dependent on an external library, autolinking allows the library writer to reorganize how that library is structured transparently to the users of the library. There are notes about utility in https://stackoverflow.com/questions/1685206/pragma-commentlib-xxx-lib-equivalent-under-linux and https://stackoverflow.com/questions/3851956/whats-pragma-comment-lib-lib-glut32-lib?noredirect=1&lq=1.

I’m less convinced about --whole-archive as
I think this tends to be a way of structuring the build and would be
best made explicit in the build system. Moreover, what if someone
wants to not use --whole-archive, for their autolink, but one already
exists.

Then they can specify --no-whole-archive on the end of the command line, no?

This could be quite difficult to check with a large project.
Personally I’d have the user be explicit in the .autolink whether they
were intending it to be whole-archive or not.

I was hoping to avoid this as I want to avoid getting into how to specify linker specific options in the frontend. If we dislike the idea that the state of the command line parser at the end of the linker command line affects the autolinked libraries then I would rather go for a scheme in which the default state of the command line parser applies when linking the autolinked libraries; however, that seems harder to implement in LLD and gives the user less control over autolinking.

I think that handling .autolink’ed files in the default state is simpler, and it doesn’t seem too hard to implement.

Right… definitely possible to implement. So the trade offs are that it is possibly confusing if options like --whole-archive start applying to the “invisible” autolinked inputs. OTOH why not allow command line options to affect the autolinked inputs? It gives developers some more control at no cost (apart form the possible confusion).

The other option is to handle autolinked libraries as soon as we find them, so that if foo.o autolinks libbar, the linker would act as if foo.o in the command line is followed by -lbar. I’d think that’s not too bad or arguably more straightforward semantics than autolinking everything all at once at the end.

So I played around with this idea a bit. Some background info:

MSVC searches libraries added via “comment lib” pragmas last, after searching all of the libraries specified on the command line; however, symbols that are unresolved when bringing in an object file from a library are searched for in that library first (https://docs.microsoft.com/en-us/cpp/build/reference/link-input-files?view=vs-2017).

In the upstream discussion for autolinking, Cary Coutant offered the following as a good compromise for traditional ELF linkers (http://lists.llvm.org/pipermail/llvm-dev/2018-January/120382.html.):

“”“I think what would work is to insert each requested object or shared
library into the link order immediately after the object that requests
it, but only if the object hasn’t already been inserted and isn’t
already listed on the command line (i.e., we won’t try to load the
same file twice); and to search each requested archive library
immediately after each object that requests it (of course, because of
how library searching works, we would load a given archive member once
at most). With this method, libm would be searched after both a.o and
b.o, so we’d load any members needed by a.o before b.o, and any
remaining members needed by b.o before c.o.”""

The problem with what your suggesting is that with the GNU linkers it is always possible to define “where” in the command line parsing you are. However for MSVC or LLD it is not always possible… think of a object file in a library that autolinks foo.a that gets pulled into the link (by a undefined symbol) much later on in the link order. My RFC is careful to try to set out a scheme that all linkers can implement (as much as is possible).

  1. Duplicate autolinked inputs are ignored.

If we take the issue of --whole-archive off the table does it matter
that there are duplicate libraries? Unresolved symbols will match
against the first library.

It doesn’t matter for libraries in LLD; but, it is important for object files. I think that this mechanism should be usable for object files an libraries. This is common in ELF linkers - for example the --library command line option can be used to link object files.

Do you actually often link .o file using -l? It seems a bit weird use of the option. To me, it seems better to limit the ability of autolinking to link against .so or .a.

I don’t personally but it does seem useful to be able to find .o files on the library search paths.

At Sony we offer autolinking as a feature in our ELF toolchain. We would like to see full support for this feature upstream as there is anecdotal evidence that it would find use beyond Sony.

In general autolinking (https://en.wikipedia.org/wiki/Auto-linking) allows developers to specify inputs to the linker in their source code. LLVM and Clang already have support for autolinking on ELF via embedding strings, which specify linker behavior, into a .linker-options section in relocatable object files, see:

RFC - http://lists.llvm.org/pipermail/llvm-dev/2018-January/120101.html
LLVM - https://llvm.org/docs/Extensions.html#linker-options-section-linker-options, https://reviews.llvm.org/D40849
Clang - https://clang.llvm.org/docs/LanguageExtensions.html#specifying-linker-options-on-elf-targets, https://reviews.llvm.org/D42758

However, although support was added to Clang and LLVM, no support has been implemented in LLD; and, I get the sense, from reading the reviews, that there wasn’t agreement on the implementation when the changes landed. The original motivation seems to have been to remove the “autolink-extract” mechanism used by Swift to workaround the lack of autolinking support for ELF. However, looking at the Swift source code, Swift still seems to be using the “autolink-extract” method.

So my first question: Are there any users of the current implementation for ELF?

Assuming that no one is using the current code, I would like to suggest a different mechanism for autolinking.

For ELF we need limited autolinking support. Specifically, we only need support for “comment lib” pragmas (https://docs.microsoft.com/en-us/cpp/preprocessor/comment-c-cpp?view=vs-2017) in C/C++ e.g. #pragma comment(lib, “foo”). My suggestion that we keep the implementation as lean as possible.

Principles to guide the implementation:

  • Developers should be able to easily understand autolinking behavior.
  • Developers should be able to override autolinking from the linker command line.
  • Inputs specified via pragmas should be handled in a general way to allow the same source code to work in different environments.

I would like to propose that we focus on autolinking exclusively and that we divorce the implementation from the idea of “linker options” which, by nature, would tie source code to the vagaries of particular linkers. I don’t see much value in supporting other linker operations so I suggest that the binary representation be a mergable string section (SHF_MERGE, SHF_STRINGS), called .autolink, with custom type SHT_LLVM_AUTOLINK (0x6fff4c04), and SHF_EXCLUDE set (to avoid the contents appearing in the output). The compiler can form this section by concatenating the arguments of the “comment lib” pragmas in the order they are encountered. Partial (-r, -Ur) links can be handled by concatenating .autolink sections with the normal mergeable string section rules. The current .linker-options can remain (or be removed); but, “comment lib” pragmas for ELF should be lowered to .autolink not to .linker-options. This makes sense as there is no linker option that “comment lib” pragmas map directly to. As an example, #pragma comment(lib, “foo”) would result in:

.section “.autolink”,“eMS”,@llvm_autolink,1
.asciz “foo”

For LTO, equivalent information to the contents of a the .autolink section will be written to the IRSymtab so that it is available to the linker for symbol resolution.

The linker will process the .autolink strings in the following way:

  1. Inputs from the .autolink sections of a relocatable object file are added when the linker decides to include that file (which could itself be in a library) in the link. Autolinked inputs behave as if they were appended to the command line as a group after all other options. As a consequence the set of autolinked libraries are searched last to resolve symbols.

If we want this to be compatible with GNU linkers, doesn’t the autolinked input need to appear at the point immediately after the object file appears in the link? I’m imagining the case where you have a statically linked libc as well as a libbar.a autolinked from a foo.o. The link command line would look like this:

ld foo.o -lc

Now foo.o autolinks against bar. The command line becomes:

ld foo.o -lc -lbar

Actually, I was thinking that on a GNU linker the command line would become “ld foo.o -lc -( -lbar )-”; but, this doesn’t affect your point.

If libbar.a requires an additional object file from libc.a, it will not be added to the link.

As it stands all the dependencies of an autolinked library must themselves be autolinked. I had imagined that this is a reasonable limitation. If not we need another scheme. I try to think about some motivating examples for this.

  1. It is an error if a file cannot be found for a given string.
  2. Any command line options in effect at the end of the command line parsing apply to autolinked inputs, e.g. --whole-archive.
  3. Duplicate autolinked inputs are ignored.

This seems like it would work in GNU linkers, as long as the autolinked file is added to the link immediately after the last mention, rather than the first. Otherwise a command line like:

ld foo1.o foo2.o

(where foo1.o and foo2.o both autolink bar) could end up looking like:

ld foo1.o -lbar foo2.o

and you will not link anything from libbar.a that only foo2.o requires. It may end up being simpler to not ignore duplicates.

Correct; but, given that the proposal was to handle the libraries as if they are appended to the link line after everything on the command line then I think this will work. With deduplication (and the use of SHF_MERGE) developers get no ordering guarantees. I claim that this is a feature! My rationale is that the order in which libraries are linked affects different linkers in different ways (e.g. LLD does not resolve symbols from archives in a compatible manner with either the Microsoft linker or the GNU linkers.), by not allowing the user to control the order I am essentially saying that autolinking is not suitable for libraries that offer competing copies of the same symbol. This ties into my argument that “comment lib” pragmas should be handled in as “general” a way as possible.

  1. The linker tries to add a library or relocatable object file from each of the strings in a .autolink section by; first, handling the string as if it was specified on the commandline; second, by looking for the string in each of the library search paths in turn; third, by looking for a lib.a or lib.so (depending on the current mode of the linker) in each of the library search paths.

Is the second part necessary? “-l:foo” causes the linker to search for a file named “foo” in the library search path, so it seems that allowing the autolink string to look like “:foo” would satisfy this use case.

I worded the proposal to avoid mapping “comment lib” pragmas to --library command line options. My reasons:

  1. I find the requirement that the user put ‘:’ in their lib strings slightly awkward. It means that the source code is now coupled to a GNU-style linker. So then this isn’t merely an ELF linking proposal, it’s a proposal for ELF toolchains with GNU-like linkers (e.g. the arm linker doesn’t support the colon prefix http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.dui0474c/Cjahbdei.html).

  2. The syntax is #pragma comment(lib, …) not #pragma linker-option(library, …) i.e. the only thing this (frankly rather bizarre) syntax definitely implies is that the argument is related to libraries (and comments ¯_(ツ)_/¯); it is a bit of a stretch to interpret “comment lib” pragmas as mapping directly to “specifying an additional --library command line option”.

AFAIK all linkers support two ways of specifying inputs; firstly, directly on the command line; secondly, with an option with very similar semantics to GNU’s --library option. I choose a method of finding a input files that encompasses both methods of specifying a library on the command line. I think that this method is actually more intuitive than either the method used by the linker script INPUT command or by --library. FWIW, I looked into the history of the colon prefix. It was added in https://www.sourceware.org/ml/binutils/2007-03/msg00421.html. Unfortunately, the rationale given is that it was merely a port of a vxworks linker extension. I couldn’t trace the history any further than that to find the actual design discussion. The linker script command INPUT uses a different scheme and the command already had this search order 20 years ago, which is the earliest version of the GNU linker I have history for; again, the rationale is not available.

Hello,

I’ve put some comments on the proposal inline. Having to had to debug
library selection problems where all the libraries are visible on the
linker command line, I would prefer if people didn’t embed difficult
to find directives in object files, but I’m guessing in some languages
this is the natural way of adding libraries.

At Sony we offer autolinking as a feature in our ELF toolchain. We would like to see full support for this feature upstream as there is anecdotal evidence that it would find use beyond Sony.

I’ve not got any use of the existing code. Personally I’ve not come
across anyone wanting this type of feature, but that is also anecdotal
on my part.

For ELF we need limited autolinking support. Specifically, we only need support for “comment lib” pragmas (https://docs.microsoft.com/en-us/cpp/preprocessor/comment-c-cpp?view=vs-2017) in C/C++ e.g. #pragma comment(lib, “foo”). My suggestion that we keep the implementation as lean as possible.

Principles to guide the implementation:

  • Developers should be able to easily understand autolinking behavior.
  • Developers should be able to override autolinking from the linker command line.
  • Inputs specified via pragmas should be handled in a general way to allow the same source code to work in different environments.

I would like to propose that we focus on autolinking exclusively and that we divorce the implementation from the idea of “linker options” which, by nature, would tie source code to the vagaries of particular linkers. I don’t see much value in supporting other linker operations so I suggest that the binary representation be a mergable string section (SHF_MERGE, SHF_STRINGS), called .autolink, with custom type SHT_LLVM_AUTOLINK (0x6fff4c04), and SHF_EXCLUDE set (to avoid the contents appearing in the output). The compiler can form this section by concatenating the arguments of the “comment lib” pragmas in the order they are encountered. Partial (-r, -Ur) links can be handled by concatenating .autolink sections with the normal mergeable string section rules. The current .linker-options can remain (or be removed); but, “comment lib” pragmas for ELF should be lowered to .autolink not to .linker-options. This makes sense as there is no linker option that “comment lib” pragmas map directly to. As an example, #pragma comment(lib, “foo”) would result in:

.section “.autolink”,“eMS”,@llvm_autolink,1
.asciz “foo”

For LTO, equivalent information to the contents of a the .autolink section will be written to the IRSymtab so that it is available to the linker for symbol resolution.

I’m not sure I understand the bit about “for symbol resolution”. I
think that what you mean is that you will encode the autolink section
using symbols instead of as a section, and the linker is expected to
extract this when it reads the symbol table?

Whoops… might have used a bit of a colloquialism there; sorry. All I mean is that there will be a method on the IRSymtab that LLD can use to retrieve the same set of strings that would be written into the the .autolink section of the relocatable object files by the backend.

The linker will process the .autolink strings in the following way:

  1. Inputs from the .autolink sections of a relocatable object file are added when the linker decides to include that file (which could itself be in a library) in the link. Autolinked inputs behave as if they were appended to the command line as a group after all other options. As a consequence the set of autolinked libraries are searched last to resolve symbols.
  2. It is an error if a file cannot be found for a given string.
  3. Any command line options in effect at the end of the command line parsing apply to autolinked inputs, e.g. --whole-archive.

I’ve not got any experience of autolinking as a user, so I’m
struggling a bit with this one. I’m guessing that autolinking is
useful because someone can do the equivalent of #include <library.h>
and #pragma comment lib “library.so” in the same place without having
to fight the build system.

Right. Consider that many codebases have multiple build configurations and the linker needs to be given the correct version of a library to use for the particular build configuration. This is often easier to do using the preprocessor than in the build system. Also, if a program is dependent on an external library, autolinking allows the library writer to reorganize how that library is structured transparently to the users of the library. There are notes about utility in https://stackoverflow.com/questions/1685206/pragma-commentlib-xxx-lib-equivalent-under-linux and https://stackoverflow.com/questions/3851956/whats-pragma-comment-lib-lib-glut32-lib?noredirect=1&lq=1.

I’m less convinced about --whole-archive as
I think this tends to be a way of structuring the build and would be
best made explicit in the build system. Moreover, what if someone
wants to not use --whole-archive, for their autolink, but one already
exists.

Then they can specify --no-whole-archive on the end of the command line, no?

This could be quite difficult to check with a large project.
Personally I’d have the user be explicit in the .autolink whether they
were intending it to be whole-archive or not.

I was hoping to avoid this as I want to avoid getting into how to specify linker specific options in the frontend. If we dislike the idea that the state of the command line parser at the end of the linker command line affects the autolinked libraries then I would rather go for a scheme in which the default state of the command line parser applies when linking the autolinked libraries; however, that seems harder to implement in LLD and gives the user less control over autolinking.

I think that handling .autolink’ed files in the default state is simpler, and it doesn’t seem too hard to implement.

Right… definitely possible to implement. So the trade offs are that it is possibly confusing if options like --whole-archive start applying to the “invisible” autolinked inputs. OTOH why not allow command line options to affect the autolinked inputs? It gives developers some more control at no cost (apart form the possible confusion).

The other option is to handle autolinked libraries as soon as we find them, so that if foo.o autolinks libbar, the linker would act as if foo.o in the command line is followed by -lbar. I’d think that’s not too bad or arguably more straightforward semantics than autolinking everything all at once at the end.

So I played around with this idea a bit. Some background info:

MSVC searches libraries added via “comment lib” pragmas last, after searching all of the libraries specified on the command line; however, symbols that are unresolved when bringing in an object file from a library are searched for in that library first (https://docs.microsoft.com/en-us/cpp/build/reference/link-input-files?view=vs-2017).

In the upstream discussion for autolinking, Cary Coutant offered the following as a good compromise for traditional ELF linkers (http://lists.llvm.org/pipermail/llvm-dev/2018-January/120382.html.):

“”“I think what would work is to insert each requested object or shared
library into the link order immediately after the object that requests
it, but only if the object hasn’t already been inserted and isn’t
already listed on the command line (i.e., we won’t try to load the
same file twice); and to search each requested archive library
immediately after each object that requests it (of course, because of
how library searching works, we would load a given archive member once
at most). With this method, libm would be searched after both a.o and
b.o, so we’d load any members needed by a.o before b.o, and any
remaining members needed by b.o before c.o.”""

The problem with what your suggesting is that with the GNU linkers it is always possible to define “where” in the command line parsing you are. However for MSVC or LLD it is not always possible… think of a object file in a library that autolinks foo.a that gets pulled into the link (by a undefined symbol) much later on in the link order. My RFC is careful to try to set out a scheme that all linkers can implement (as much as is possible).

  1. Duplicate autolinked inputs are ignored.

If we take the issue of --whole-archive off the table does it matter
that there are duplicate libraries? Unresolved symbols will match
against the first library.

It doesn’t matter for libraries in LLD; but, it is important for object files. I think that this mechanism should be usable for object files an libraries. This is common in ELF linkers - for example the --library command line option can be used to link object files.

Do you actually often link .o file using -l? It seems a bit weird use of the option. To me, it seems better to limit the ability of autolinking to link against .so or .a.

I don’t personally but it does seem useful to be able to find .o files on the library search paths.

Rui - I’m sure you know everything about MSVC linking already! For others benefit though, MSVC only allows loading of libraries via “comment lib” pragmas. It rejects .obj files.

C:\temp\library_semantics>more msvc_foo.c
int foo() {return 10;}

C:\temp\library_semantics>cl msvc_foo.c /c
msvc_foo.c

C:\temp\library_semantics>more msvc.c
#pragma comment(lib, "msvc_foo.obj")
int foo ();
int main () {return foo();}

C:\temp\library_semantics>cl msvc.c
Microsoft (R) C/C++ Optimizing Compiler Version 19.00.24213.1 for x64
Copyright (C) Microsoft Corporation.  All rights reserved.

msvc.c
Microsoft (R) Incremental Linker Version 14.00.24213.1
Copyright (C) Microsoft Corporation.  All rights reserved.

/out:msvc.exe
msvc.obj
msvc_foo.obj : warning LNK4003: invalid library format; library ignored
msvc.obj : error LNK2019: unresolved external symbol foo referenced in function main
msvc.exe : fatal error LNK1120: 1 unresolved externals

C:\temp\library_semantics>lib /out:msvc_foo.lib msvc_foo.obj
Microsoft (R) Library Manager Version 14.00.24213.1
Copyright (C) Microsoft Corporation.  All rights reserved.

C:\temp\library_semantics>more msvc.c
#pragma comment(lib, "msvc_foo.lib")
int foo ();
int main (){return foo();}

C:\temp\library_semantics>cl msvc.c /link /verbose | grep msvc
msvc.c
/out:msvc.exe
msvc.obj
Processed /DEFAULTLIB:msvc_foo.lib
    Searching msvc_foo.lib:
        Referenced in msvc.obj
        Loaded msvc_foo.lib(msvc_foo.obj)
 Processed /DISALLOWLIB:msvcrt.lib
 Processed /DISALLOWLIB:msvcrtd.lib
    Searching msvc_foo.lib:
    Searching msvc_foo.lib:
    Searching msvc_foo.lib:
     msvc.obj
     msvc_foo.lib(msvc_foo.obj)

Other interesting MSVC behaviour:

MSVC forms the library name to search for based on the file extension. An interesting difference is that on windows import libraries and static archives both have the same naming convention of .lib. Whereas on Unix dynamic libraries are conventionally named .so and static archives are lib.a.

#pragma comment(lib, “winmm”) → Searches for “winmm.lib” (doesn’t search for “winmm”)
#pragma comment(lib, “winmm.lib”) → Searches for “winmm.lib” (doesn’t search for “winmm.lib.lib”)
#pragma comment(lib, “winmm.lix”) → Searches for “winmm.lix” (doesn’t search for “winmm.lix.lib”)

MSVC allows specifying libraries on the command line as just file names or by using the /DEFAULTLIB option. In both cases the rules for locating the library are the same. If a path is specified with the library name, LINK searches for the library in that directory. If no path is specified, LINK looks first in the directory that LINK is running from, and then in any directories specified in the LIB environment variable, see : https://docs.microsoft.com/en-us/cpp/build/reference/dot-lib-files-as-linker-input?view=vs-2017. Additionally, LINK will search for any /LIBPATH paths before those specified in the LIB environment variable, see: https://docs.microsoft.com/en-us/cpp/build/reference/libpath-additional-libpath?view=vs-2017. LINK handles libraries specified via “comment lib” pragmas just as if you had named them at on the command line, see: https://docs.microsoft.com/en-us/cpp/preprocessor/comment-c-cpp?view=vs-2017.

MSVC rules for resolving symbols from libraries: A library specified with /DEFAULTLIB is searched after libraries specified explicitly on the command line and before default libraries named in .obj files (see https://docs.microsoft.com/en-us/cpp/build/reference/nodefaultlib-ignore-libraries?view=vs-2017).

MSVC allows passing not only libraries to the linker via pragams but also a subset of the linkers command line options (https://docs.microsoft.com/en-us/cpp/preprocessor/comment-c-cpp). In addition to the documented options MSVC also accepts some undocumented options. One of these is the /DISALLOWLIB which allows an object file to state that it is incompatible with a given library, see: https://stackoverflow.com/questions/761394/what-does-the-disallowlib-message-mean-in-vc-linker-output and https://stackoverflow.com/questions/3007312/resolving-lnk4098-defaultlib-msvcrt-conflicts-with.

One of the options supported is /DEFAULTLIB. This means you can specify libraries via pragmas with either #pragma comment(lib, ) or #pragma comment(linker, “/DEFAULTLIB:”).

MSVC has the “/NODEFAULTLIB” option which ignores any /DEFAULTLIB options from object files or the command-line. You can also ignore specific libraries, with “/NODEFAULTLIB:name.lib”.

Both Gold and GNU-ld allow loading of non-library files via -l/–library options; but, MSVC only allows adding libraries via its equivalent of the -l command:

C:\temp\library_semantics>more msvc_foo.c
int foo() {return 10;}

C:\temp\library_semantics>cl msvc_foo.c /c

C:\temp\library_semantics>lib /out:msvc_foo.lib msvc_foo.obj

C:\temp\library_semantics>type msvc_main.c
void main(){}

C:\temp\library_semantics>cl msvc_main.c /link /DEFAULTLIB:foo.obj

/out:msvc_main.exe
/DEFAULTLIB:foo.obj
msvc_main.obj
foo.obj : warning LNK4003: invalid library format; library ignored

MSVC also ignores duplicate .objs on the command line:

c:\temp\library_semantics>cl msvc.obj
/out:msvc.exe
msvc.obj

c:\temp\library_semantics>cl msvc.obj msvc.obj
/out:msvc.exe
msvc.obj
msvc.obj
msvc.obj : warning LNK4042: object specified more than once; extras ignored

At Sony we offer autolinking as a feature in our ELF toolchain. We would like to see full support for this feature upstream as there is anecdotal evidence that it would find use beyond Sony.

In general autolinking (https://en.wikipedia.org/wiki/Auto-linking) allows developers to specify inputs to the linker in their source code. LLVM and Clang already have support for autolinking on ELF via embedding strings, which specify linker behavior, into a .linker-options section in relocatable object files, see:

RFC - http://lists.llvm.org/pipermail/llvm-dev/2018-January/120101.html
LLVM - https://llvm.org/docs/Extensions.html#linker-options-section-linker-options, https://reviews.llvm.org/D40849
Clang - https://clang.llvm.org/docs/LanguageExtensions.html#specifying-linker-options-on-elf-targets, https://reviews.llvm.org/D42758

However, although support was added to Clang and LLVM, no support has been implemented in LLD; and, I get the sense, from reading the reviews, that there wasn’t agreement on the implementation when the changes landed. The original motivation seems to have been to remove the “autolink-extract” mechanism used by Swift to workaround the lack of autolinking support for ELF. However, looking at the Swift source code, Swift still seems to be using the “autolink-extract” method.

So my first question: Are there any users of the current implementation for ELF?

Assuming that no one is using the current code, I would like to suggest a different mechanism for autolinking.

For ELF we need limited autolinking support. Specifically, we only need support for “comment lib” pragmas (https://docs.microsoft.com/en-us/cpp/preprocessor/comment-c-cpp?view=vs-2017) in C/C++ e.g. #pragma comment(lib, “foo”). My suggestion that we keep the implementation as lean as possible.

Principles to guide the implementation:

  • Developers should be able to easily understand autolinking behavior.
  • Developers should be able to override autolinking from the linker command line.
  • Inputs specified via pragmas should be handled in a general way to allow the same source code to work in different environments.

I would like to propose that we focus on autolinking exclusively and that we divorce the implementation from the idea of “linker options” which, by nature, would tie source code to the vagaries of particular linkers. I don’t see much value in supporting other linker operations so I suggest that the binary representation be a mergable string section (SHF_MERGE, SHF_STRINGS), called .autolink, with custom type SHT_LLVM_AUTOLINK (0x6fff4c04), and SHF_EXCLUDE set (to avoid the contents appearing in the output). The compiler can form this section by concatenating the arguments of the “comment lib” pragmas in the order they are encountered. Partial (-r, -Ur) links can be handled by concatenating .autolink sections with the normal mergeable string section rules. The current .linker-options can remain (or be removed); but, “comment lib” pragmas for ELF should be lowered to .autolink not to .linker-options. This makes sense as there is no linker option that “comment lib” pragmas map directly to. As an example, #pragma comment(lib, “foo”) would result in:

.section “.autolink”,“eMS”,@llvm_autolink,1
.asciz “foo”

For LTO, equivalent information to the contents of a the .autolink section will be written to the IRSymtab so that it is available to the linker for symbol resolution.

The linker will process the .autolink strings in the following way:

  1. Inputs from the .autolink sections of a relocatable object file are added when the linker decides to include that file (which could itself be in a library) in the link. Autolinked inputs behave as if they were appended to the command line as a group after all other options. As a consequence the set of autolinked libraries are searched last to resolve symbols.

If we want this to be compatible with GNU linkers, doesn’t the autolinked input need to appear at the point immediately after the object file appears in the link? I’m imagining the case where you have a statically linked libc as well as a libbar.a autolinked from a foo.o. The link command line would look like this:

ld foo.o -lc

Now foo.o autolinks against bar. The command line becomes:

ld foo.o -lc -lbar

Actually, I was thinking that on a GNU linker the command line would become “ld foo.o -lc -( -lbar )-”; but, this doesn’t affect your point.

If libbar.a requires an additional object file from libc.a, it will not be added to the link.

As it stands all the dependencies of an autolinked library must themselves be autolinked. I had imagined that this is a reasonable limitation. If not we need another scheme. I try to think about some motivating examples for this.

  1. It is an error if a file cannot be found for a given string.
  2. Any command line options in effect at the end of the command line parsing apply to autolinked inputs, e.g. --whole-archive.
  3. Duplicate autolinked inputs are ignored.

This seems like it would work in GNU linkers, as long as the autolinked file is added to the link immediately after the last mention, rather than the first. Otherwise a command line like:

ld foo1.o foo2.o

(where foo1.o and foo2.o both autolink bar) could end up looking like:

ld foo1.o -lbar foo2.o

and you will not link anything from libbar.a that only foo2.o requires. It may end up being simpler to not ignore duplicates.

Correct; but, given that the proposal was to handle the libraries as if they are appended to the link line after everything on the command line then I think this will work. With deduplication (and the use of SHF_MERGE) developers get no ordering guarantees. I claim that this is a feature! My rationale is that the order in which libraries are linked affects different linkers in different ways (e.g. LLD does not resolve symbols from archives in a compatible manner with either the Microsoft linker or the GNU linkers.), by not allowing the user to control the order I am essentially saying that autolinking is not suitable for libraries that offer competing copies of the same symbol. This ties into my argument that “comment lib” pragmas should be handled in as “general” a way as possible.

Right. I think if you need a fine control over the link order, autolinking is not a feature you want to use. Or, in general, if your program is sensitive to a link order because its source object files have competing symbols of the same name, it’s perhaps unnecessarily fragile.

That being said, I think you need to address the issue that pcc pointed out. If you statically link a program foo with the following command line

ld -o foo foo.o -lc

, foo.o auto-imports libbar.a, and libbar.a depends on libc.a, can your proposed feature pull out object files needed for libbar.a?