bug#44462: Problem with get_multilibs on macOS

[Fred Wright]

On Sun, 8 Nov 2020, Jacob Bachmeyer wrote:
> Fred Wright wrote:
> > On Thu, 5 Nov 2020, Jacob Bachmeyer wrote:
> > > Fred Wright wrote:
> > >
> > > > Even when it's using gcc, it bloats the logfile with the verbose 
> > > > -dumpspecs output, in order to determine something of highly questionable 
> > > > value.
> > >
> > > That value may be less questionable for some targets, or for tests that 
> > > need to be run across each multilib target a GCC instance supports.
> >
> > Even in the useful multilib cases, making that behavior unconditional seems 
> > like a bad idea, since the client code may have its own iteration across 
> > architectures, and doing both would either not work or explode the 
> > multi-architecture handling from O(N) to O(N^2).
>
> The get_multilibs procedure does not perform iterations itself.  I am not yet 
> certain what exactly it is supposed to do, since it seems to do a large 
> amount of work to set the "multitop" board_info key.  Whatever it does, it is 
> clearly specific to GCC, but does not even check if the selected compiler 
> even looks like GCC.
>
> > > > I traced the results of get_multilibs (as used by the libffi tests) in 
> > > > many macOS versions, and even the "successful" cases seem to have 
> > > > questionably useful results:
> > > >
> > > > [...]
> > > >
> > > > The "/usr/." result returned on 10.4, 10.5, and 32-bit 10.6 is the same 
> > > > as what I see on Ubuntu 14.04, CentOS 7, and Fedora 25, though it's not 
> > > > clear to me what it's supposed to represent.
> > >
> > > I have a suspicion that this feature is designed to support testing with 
> > > an "experimental" compiler build that is not installed on the system and 
> > > may be useless with system compilers generally, or with Apple's compilers 
> > > specifically, if Apple does not use multilib.
> >
> > Apple supports the concept of multi-architecture binaries, but not in the 
> > same way that multilib does it (AFAIK).  Macs can have "universal" 
> > binaries, which are archives combining multiple per-architecture slices. 
> > This is applicable to object files, shared libraries, and executables.  If
> > the build setup allows it, it can be as simple as including multiple 
> > architecture options in the compile command.  E.g.:
> >
> >     cc -arch x86_64 -arch i386 -o hello hello.c
> >
> > Under the hood, the compiler driver runs a separate compile/assemble for 
> > each architecture, and then combines the object files.  The linker supports 
> > universal binaries directly.
> >
> > With this arrangement, architecture-related conditionals in the source code 
> > work just fine, but what *doesn't* work is having architecture-related 
> > parameters in a configure script, which is unfortunately not as uncommon as 
> > it ought to be.
>
> No, Autoconf pushes programs to respond to features instead of using known 
> architectures.  This approach has been very successful:  as I understand, 
> most programs using Autoconf needed no changes at all to port to RISC-V even 
> if they were written long before RISC-V existed.  Many programs, if they had 
> previously been ported to any 64-bit architecture, needed no changes at all 
> for x86_64.  Autoconf has achieved something architecture #ifdefs cannot do: 
> provide automatic portability to a new architecture that did not exist when 
> the program was written.  I consider Autoconf's approach here completely 
> vindicated.

I'm not talking about checking for specific architectures; this is about 
checking architecture-related *properties*.  For example, a configure 
check for sizeof(long) is incompatible with multi-architecture builds, 
while using LONG_BIT works just fine.  But there are build procedures that 
do things like the former, for no good reason.

> You could submit a patch to Autoconf to add support for multi-arch config.h 
> files, where configure would run tests for each of a list of architectures 
> and use arch #ifdefs in config.h to select the configure results for each 
> compile.

I don't plan to mess with Autoconf if I can possibly avoid it. :-)

> > > > Since get_multilibs already has code to return an empty string in the 
> > > > "remote" case (where it assumes this function won't work), I just added 
> > > > code to unix.exp to set multitop to "" for all "darwin" targets, thereby 
> > > > short-circuiting almost all og get_multilibs.  That certainly fixes the 
> > > > problem with the libffi tests, and doesn't change any non-Mac behavior, 
> > > > though I don't know if that's the ideal fix.  The whole get_mutilibs 
> > > > function looks pretty ugly anyway, and it's generally recognized that 
> > > > relying on -dumpspecs is a bad idea.
> > >
> > > It is most certainly not ideal.  A better solution is probably to add a 
> > > test to get_multilibs to return an empty string if the compiler is not 
> > > GCC.  Of course, if another compiler pretends to be GCC enough to pass 
> > > that check, but does not actually implement -dumpspecs, that is not our 
> > > bug.
> >
> > Limiting it to gcc would avoid actual failures, but wouldn't avoid bloating 
> > the logfile with the humongous -dumpspecs output in the many cases where 
> > the multilibs action isn't even wanted.
> >
> > The meaning of "pretends to be gcc" isn't well-defined.  It's not uncommon 
> > to have a compiler named "gcc" which is really clang, largely because there 
> > are so many projects that think that all compilers of interest are named 
> > "gcc".  And of course, clang tries to be highly gcc-compatible, to 
> > facilitate switching to it, but not to the extent of implementing 
> > -dumpspecs, which is is derived purely from gcc's internal implementation 
> > details, and was never intended to be used in this fashion.
>
> Autoconf has always allowed setting CC to select a compiler.  Apple *could* 
> have shipped Clang as "llcc" or similar or even simply the traditional "cc" 
> for a system compiler, but they chose to ship it as "gcc" instead.  Not that 
> the current version of get_multilibs even bothers to check that the compiler 
> has "gcc" in its name...
>
> > The libffi test suite comes up with a "compiler_vendor" variable which 
> > seems to be able to distinguish clang from gcc, though I haven't looked at 
> > the details.
> >
> > Fixing get_multilibs properly would probably mean making it both highly 
> > platform-specific and optional.
>
> The get_multilibs procedure is *not* platform-specific; it is GCC-specific. 
> I am still unsure how exactly it is used.

Since it seems to involve file paths, it may be specific to combinations 
of GCC and platforms.

But since neither of us seems to know very much about what get_multilibs 
is trying to do, it's hard to discuss it intelligently. :-)

> > > This issue on Mac OS X will probably be a known bug in 1.6.3 and fixed in 
> > > 1.6.4.
> > >
> > > > I primarily tested my patch against the 1.6.2 release, since the current 
> > > > master won't install from a non-git directory, and also has multiple 
> > > > failures in its own tests (even on Linux).  The patch is nearly identical 
> > > > between the 1.6.2 and master cases, anyway.
> > >
> > > Are we looking at the same current master?  I have commit 
> > > 3d62df24deedfb3c7c3e396a31b8ce431138eb49 here and all of the tests pass.
> > >
> > > ****These other problems are potential release blockers for 1.6.3.****
> > > Can you file another bug report with the test failures and more 
> > > information about these issues?
> >
> > I looked into this more closely and it's probably related to the non-git 
> > issue.  When running from a non-git directory, the configure script reports 
> > a "fatal" error, but then goes on to complete with a zero exit status and a 
> > more or less buildable setup, so you have to be paying close attention to 
> > the output to notice.

Actually it now looks like the two things are unrelated; I filed two 
separate bugs.

> > If this is a typical hack to provide git-based extra information in 
> > between-release version strings, it should have a fallback for the non-git 
> > case.  Consider the case of pushing all the git-tracked files to a test 
> > system with git ls-files and rsync.
>
> Please file another bug report for this issue.  This is separate from 
> get_multilibs mishandling non-GCC compilers.
>
> > > > I can send the current patch, either as a bare email or as an attachment. 
> > > > AFAIK, Savannah doesn't have the pull request / merge request concept.
> > >
> > > This will need to be fixed in libgloss.exp, not unix.exp.  I am putting my 
> > > foot down on fixing bugs in DejaGnu's own tree directly instead of hacking 
> > > around them like that.
> >
> > Well, OK, but there seem to be other similar hacks in unix.exp, and if the 
> > idea is that get_multilibs is completely useless on the Mac (which appears 
> > to be the case with the current implemenation, anyway), then disabling it 
> > in the target-related code doesn't seem unreasonable.
>
> It is not completely useless even on MacOS X -- a user could install GCC and 
> expect a testsuite to use it, particularly in the case of a cross-compiler 
> for embedded development.

If it's a cross compiler, then by definition it can't run the resulting 
code on the host platform.  But since get_multilibs already excludes all 
remote cases, it wouldn't be able to run it on a separate target platform, 
either.

Besides, aren't files like baseboards/unix.exp based on the *target* 
platform, not the host?  If so, then it seems like disabling get_multilibs 
for the Mac there is exactly the right thing, at least until such time as 
get_multilibs can behave usefully for a Mac target.

Fred Wright