public inbox for gcc@gcc.gnu.org
 help / color / mirror / Atom feed
From: "Joseph S. Myers" <jsm@polyomino.org.uk>
To: Tim O'Callaghan <timo@dspsrv.com>
Cc: gcc@gcc.gnu.org, m.hayes@elec.canterbury.ac.nz
Subject: Re: GCC Target for TI C54x series of DSP.
Date: Fri, 01 Oct 2004 14:18:00 -0000	[thread overview]
Message-ID: <Pine.LNX.4.61.0410011253490.16447@digraph.polyomino.org.uk> (raw)
In-Reply-To: <415D4FE4.3090507@dspsrv.com>

On Fri, 1 Oct 2004, Tim O'Callaghan wrote:

> I'm interested in getting GCC to cross compile for the C54x architecture. I
> realize that GCC already supports the C4x series of DSP, and that is what i
> shall start with as a base. I have a few questions though, before i get stuck
> into it.

If you search GCC Bugzilla for bugs with c4x target, you'll find three 
bugs open (12018, 14436, 16621) of which the latter two suggest this 
target is completely broken at present, so you may need to start by 
restoring it to working order.  (Unless you find the bugs are fixed but 
haven't been closed in Bugzilla, in which case you should report this to 
those bugs.)

> 2. There is a source forge project created,
> http://tic54x-gcc.sourceforge.net/, and its waiting for a baseline. Should i
> start with 3.4.2? or would an earlier/later version be a better choice?

Start with GCC CVS mainline (which will become 4.0).  For a minor target 
discretion might be granted to include non-regression bugfixes on a 
release branch which would not be considered safe for more widely used 
targets.

> 3. Any pitfalls, gotchas, trips and traps i should look out for?

Ensure all copyright assignments are in place at an early stage.

I presume that like C4x, C54x uses 32-bit bytes.  Parts of GCC are 
confused about the distinction between host and target bytes.  It may be a 
fair amount of work to sort this out properly.  As some examples, 
libcpp/charset.c:emit_numeric_escape doesn't handle target bytes wider 
than host bytes (this might not be important for typical DSP uses); what 
sort of bytes ASM_OUTPUT_ASCII takes is not defined; code extracting 
characters from string constants tends to presume host and target strings 
use bytes of the same width (and the internal format isn't well-defined in 
other cases).

-- 
Joseph S. Myers               http://www.srcf.ucam.org/~jsm28/gcc/
  http://www.srcf.ucam.org/~jsm28/gcc/#c90status - status of C90 for GCC 4.0
    jsm@polyomino.org.uk (personal mail)
    jsm28@gcc.gnu.org (Bugzilla assignments and CCs)

      reply	other threads:[~2004-10-01 13:04 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2004-10-01 14:07 Tim O'Callaghan
2004-10-01 14:18 ` Joseph S. Myers [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=Pine.LNX.4.61.0410011253490.16447@digraph.polyomino.org.uk \
    --to=jsm@polyomino.org.uk \
    --cc=gcc@gcc.gnu.org \
    --cc=m.hayes@elec.canterbury.ac.nz \
    --cc=timo@dspsrv.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).