avr-gcc-list
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [avr-gcc-list] May avr-gcc emit EIJMP/EICALL?


From: Georg-Johann Lay
Subject: Re: [avr-gcc-list] May avr-gcc emit EIJMP/EICALL?
Date: Sat, 05 Nov 2011 22:59:59 +0100
User-agent: Mozilla Thunderbird 1.0.7 (Windows/20050923)

Jan Waclawek schrieb:
There are two independent bugs.

One is, that the trampolines without relax are broken. http://www.avrfreaks.net/index.php?name=PNphpBB2&file=viewtopic&p=707285#707285

Second is a genuine -relax bug. http://www.avrfreaks.net/index.php?name=PNphpBB2&file=viewtopic&p=764532#764532

You have the relevant bug numbers from binutils or avr-libc?
You mean from their respective bug trackers?
Yes. If there is interest to have an issue fixed, the first step is
to do a qualifier problem report.

That's the problem - I don't feel qualified enough to report it in
the style and manner usual in the bug tracker. And of course I don't
have patches.

A bug report is a bug report.

And a patch is a patch.

These are different things. In a bug report you have to make the issue clear and enable others to reproduce the issue. No one expects you to come up with an explanation of what's wrong inside, what needs to be changed or even with a patch.

Here is an example of a really bad style report:

http://gcc.gnu.org/PR49764

But back on-topic: I think it's a good exercise for you to file these bug reports. Just think one minit what someone else needs to reproduce the artifact.

Some posts in avtfreaks contain C source that triggers the bug. Do binutils deal with C source code? NO! So get rid of the C-scrap by piping the source through the compiler by means of

$ foo-gcc -S file.c [your options]

et voilĂ : you have kicked of the C scrap away: one barrier less for the binutils guys.

- What version did you use? You know how to get it.
- What command line options did you use? Who will know it if not you?
- Waht's the error(message)? You know it.

No, I never did care to submit them.

Okay, understand. As is was already observed in a release from 2007
and still persists, it must be really unimportant artifact for
pathological corner-case code if no one took 5 minutes to file a
report till today...

They both are corner cases, of course, otherwise they would be
discovered earlier. And it was *not* me who discovered them, as you
can read in the respective threads - in both cases I was simply
trying to help others, as so many do on avrfreaks.

In fact, users might have come across any of these bugs, and then
having no clue, jumped through hoops until they got rid of the error
somehow. It would be nice if they would be fixed, so that others are
not caught by surprise. So I would like to ask you, as one of those
who are undoubtedly much more qualified to submit a bug report than
me, to spend those 5 minutes to file the reports.

As I already said: Just do that "no-excuses" reports.
You are skilled enough for that.

Without a report the prospects are really thin that a bug gets fixed.

But it's much more fun to chat in avrfreaks an post hacks around the bugs. And every some months some other guy is caught by the bug and the topic bubbles up in avrfreaks again and again (free software is scrap from hobbyists, we all know it).

All the poor developers would never come to avrfreaks if there were no bugs...

Johann



reply via email to

[Prev in Thread] Current Thread [Next in Thread]