Page MenuHomeFreeBSD

Export various 128 bit long double functions from libgcc_s.so.1
ClosedPublic

Authored by dim on Feb 15 2021, 7:56 PM.
Tags
None
Referenced Files
F82869712: D28690.id90890.diff
Fri, May 3, 9:40 AM
F82862775: D28690.diff
Fri, May 3, 8:00 AM
Unknown Object (File)
Mon, Apr 29, 9:15 AM
Unknown Object (File)
Thu, Apr 25, 8:04 AM
Unknown Object (File)
Tue, Apr 16, 11:55 PM
Unknown Object (File)
Fri, Apr 5, 7:27 AM
Unknown Object (File)
Mar 22 2024, 1:06 PM
Unknown Object (File)
Mar 22 2024, 1:06 PM

Details

Summary

These were already compiled for some time on aarch64 and riscv, by
including lib/libcompiler_rt/Makefile.inc, but never exported in the
shared library. Since gcc exports these under vesrion GCC_4.6.0, we do
the same.

This review should replace D11482 for now. For e.g. amd64 more work is
still to be done, as compiler-rt does not seem to support 128 bit long
double math for that architecture. (And as far as I know it doesn't have
it at all? But I'm probably missing something.)

Diff Detail

Repository
rS FreeBSD src repository - subversion
Lint
Lint Passed
Unit
No Test Coverage
Build Status
Buildable 37033
Build 33922: arc lint + arc unit

Event Timeline

dim requested review of this revision.Feb 15 2021, 7:56 PM
This revision is now accepted and ready to land.Feb 15 2021, 10:14 PM

Don't the symbol map entries just get ignored if the symbols aren't present, so it would be ok to have these in the MI Symbol.map?

In D28690#642826, @jhb wrote:

Don't the symbol map entries just get ignored if the symbols aren't present, so it would be ok to have these in the MI Symbol.map?

I'm unsure if that is the case. We also have a lib/libgcc_s/arm/Symbol.map, so that's why I copied the approach. But it would simplify things basically back to the old D11482 review, maybe plus or minus one line...

I do think they are ignored if not present, but I don't care strongly either way.

Can someone commit this please?
Thanks in advance!