This is the mail archive of the fortran@gcc.gnu.org mailing list for the GNU Fortran project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: optimization problems, how to obtain "processed source"?


Daniel Franke wrote:
Hi all.

This is a somewhat stupid question, but the manpages are no help here
(or I can't identify what is needed).

I try to hunt down a problem with optimization. From my tests, it is
to assume that a variable gets optimized away, that shouldn't. The
main problem is, that adding WRITEs changes the behaviour of the
compiled code, therefore I can't use them to debug. My ability to
decipher assembler is limited, therefore I would like to obtain
"processed source" as it sometimes added to PRs, e.g. PR29101, comment
#1 [1]. How can I obtain such "processed sources"? (Is there a proper
name for this stuff?)

Thanks in advance!
    Daniel

[1] http://gcc.gnu.org/bugzilla/show_bug.cgi?id=29101#c1

Try -fdump-tree-original for starters. This is the output before going to the next stages. Then I think there is a -fdump-tree-all which gives a lot of files to look at.

There are others here more versed than I in these features who may chime in.

Jerry


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]