This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
optimization/1521: Misleading documentation
- To: gcc-gnats at gcc dot gnu dot org
- Subject: optimization/1521: Misleading documentation
- From: lucier at math dot purdue dot edu
- Date: 28 Dec 2000 05:04:36 -0000
- Reply-To: lucier at math dot purdue dot edu
>Number: 1521
>Category: optimization
>Synopsis: Misleading documentation
>Confidential: no
>Severity: serious
>Priority: medium
>Responsible: unassigned
>State: open
>Class: doc-bug
>Submitter-Id: net
>Arrival-Date: Wed Dec 27 21:06:00 PST 2000
>Closed-Date:
>Last-Modified:
>Originator: Brad Lucier
>Release: development
>Organization:
>Environment:
all
>Description:
In toplev.c you find
{"schedule-insns", &flag_schedule_insns, 1,
"Reschedule instructions to avoid pipeline stalls"},
{"schedule-insns2", &flag_schedule_insns_after_reload, 1,
"Run two passes of the instruction scheduler"},
which I took to mean that if I set -fschedule-insns2
then two passes of the instruction scheduler would be
run, but this is not the case, only the second pass is
run.
Perhaps this is clearer from the info page:
`-fschedule-insns'
If supported for the target machine, attempt to reorder
instructions to eliminate execution stalls due to required data
being unavailable. This helps machines that have slow floating
point or memory load instructions by allowing other instructions
to be issued until the result of the load or floating point
instruction is required.
`-fschedule-insns2'
Similar to `-fschedule-insns', but requests an additional pass of
instruction scheduling after register allocation has been done.
This is especially useful on machines with a relatively small
number of registers and where memory load instructions take more
than one cycle.
but I think it should be fixed in the short documentation
in toplev.c
Brad
>How-To-Repeat:
>Fix:
>Release-Note:
>Audit-Trail:
>Unformatted: