This is the mail archive of the
mailing list for the GCC project.
Re: [tree-ssa] Illegible dumps with dump_cfg_function_to_file
- From: Diego Novillo <dnovillo at redhat dot com>
- To: Andrew Macleod <amacleod at redhat dot com>
- Cc: Zdenek Dvorak <rakdver at atrey dot karlin dot mff dot cuni dot cz>, gcc-patches <gcc-patches at gcc dot gnu dot org>
- Date: Tue, 11 Nov 2003 14:22:33 -0500
- Subject: Re: [tree-ssa] Illegible dumps with dump_cfg_function_to_file
- Organization: Red Hat Canada
- References: <email@example.com> <20031107083114.GA16857@atrey.karlin.mff.cuni.cz> <firstname.lastname@example.org> <20031107143835.GA22295@atrey.karlin.mff.cuni.cz> <email@example.com> <20031108165602.GA23453@atrey.karlin.mff.cuni.cz> <firstname.lastname@example.org> <1068578453.18840.2.camel@p4>
On Tue, 2003-11-11 at 14:20, Andrew MacLeod wrote:
> On Tue, 2003-11-11 at 13:12, Diego Novillo wrote:
> > We should also show PHI nodes even when the user hasn't selected -vops.
> > In those cases, we should only use PHI nodes for renameable variables.
> > But we still can't tell whether a var needs a virtual operand (we will
> > soon).
> How do you intend to tell the difference between a PHI of virtual
> operands and a PHI of regular operands?
By calling var_needs_vops_p (SSA_NAME (PHI_RESULT (phi))).
var_needs_vops_p is implemented out of bits and pieces from tree-dfa.c.
(it's in my local tree)