llvm dev - Nov 2018 - RFC: Adding debug information to LLVM to support Fortran

We use DW_TAG_string_type on OpenVMS I64 for Fortran CHARACTER*n,
Pascal's PACKED ARRAY OF CHAR, COBOL, etc.  We'll want it as well.  I
had sorta assumed it was already there?

For arrays with run-time bounds, we have them too in Pascal (they are
called "Schema Types" in the Extended Pascal standard).  I'd
prefer that
you pick a more generic name than "DIFortranSubrange"   Again on
OpenVMS
I64, we support bounds as subranges with literals, simple locations, or
thunk calls (as Pascal's computation of array bounds or size can get
complicated with uplevel referencing in the mix - we didn't want to have
to describe how to walk the static link in a location expression)

And Fortran's assumed shape arrays is exactly mapped to an OpenVMS
descriptor format (we have several) so we'll use these as well for
non-Fortran too.

Other than suggesting to make the names less Fortran specific (the folks
looking at adding PL/1 support would probably want to leverage all of
this, Ada too if somebody tackles that), I'm OK with the approach.

John

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: OpenPGP digital signature
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20181102/8487c5ce/attachment.sig>

Hi John,

Thanks for the feedback.  Glad that this might have broader usefulness.

We've obviously been focused on Fortran implementations, but we're not
against generalizing.

--
Eric

-----Original Message-----
From: llvm-dev <llvm-dev-bounces at lists.llvm.org> On Behalf Of John
Reagan via llvm-dev
Sent: Friday, November 02, 2018 2:34 PM
To: llvm-dev at lists.llvm.org
Subject: Re: [llvm-dev] RFC: Adding debug information to LLVM to support Fortran

We use DW_TAG_string_type on OpenVMS I64 for Fortran CHARACTER*n, Pascal's
PACKED ARRAY OF CHAR, COBOL, etc.  We'll want it as well.  I had sorta
assumed it was already there?

For arrays with run-time bounds, we have them too in Pascal (they are called
"Schema Types" in the Extended Pascal standard).  I'd prefer that
you pick a more generic name than "DIFortranSubrange"   Again on
OpenVMS
I64, we support bounds as subranges with literals, simple locations, or thunk
calls (as Pascal's computation of array bounds or size can get complicated
with uplevel referencing in the mix - we didn't want to have to describe how
to walk the static link in a location expression)

And Fortran's assumed shape arrays is exactly mapped to an OpenVMS
descriptor format (we have several) so we'll use these as well for
non-Fortran too.

Other than suggesting to make the names less Fortran specific (the folks looking
at adding PL/1 support would probably want to leverage all of this, Ada too if
somebody tackles that), I'm OK with the approach.

John


-----------------------------------------------------------------------------------
This email message is for the sole use of the intended recipient(s) and may
contain confidential information.  Any unauthorized review, use, disclosure or
distribution is prohibited.  If you are not the intended recipient, please
contact the sender by reply email and destroy all copies of the original
message.
-----------------------------------------------------------------------------------