llvm.org bug trend

Over most of the past year, I have been keeping an eye on the overall LLVM.org open-bug count.

Sampling the count (almost) every Monday morning, it is a consistently non-decreasing number.

I thought I’d post something about it to the Dev lists, as the count broke 4000 this past week.

For your entertainment here’s a chart that Excel produced from the data. (To make it more

dramatic, I carefully did not use a proper zero point on the X-axis.)

I do not have per-category breakdowns, sorry, just the raw total.

Makes me think more seriously about cruising the bug list for something that looks like

I could actually fix it…

–paulr

image001.png

More users and more bug reports - I wouldn't be surprised if the hidden number of "bugs" was over 10k. Is there a way to get bugs fixed per month or per week? Even if that number was decreasing I'd just attribute it to llvm/clang stabilizing and things becoming more complicated to fix. By comparison how many open issues does gcc currently have. I'm not trying to take away from your point, but there's probably not magic bullet.

Maybe a bug fix month? (Trying to sort out a bunch of low hanging fruit that new developers may be able to fix?)

A bug fix month is certainly not a bad idea. It gives people who want to contribute a good starting point if they don’t know where to begin. I fall under that category so I would benefit from that. A win-win situation as they say.

It would be really nice if our bug tracker had something like the “Issues: 30-day summary” that this jira instance has <https://issues.apache.org/jira/browse/LUCENE>.

Another interesting statistic that might be interesting to plot alongside the “open bugs” graph is “total lines of code” (or maybe just plot the ratio “bugs/LOC”).

image001.png

I’d say that bugzilla is in a very sad state. It doesn’t seem to be a part of the development flow (the way commit reviews or tests are). I did some cleanup a month or two ago, mainly trying to close non issues. Went through clang Driver and some C++ bugs and found a bunch of bugs that can’t be reproduced. Stuff from gcc stress tests that used to crash but no longer does. Some bugs are only partially true, like http://llvm.org/bugs/show_bug.cgi?id=14084, -print-stats is no longer exposed through the driver but -time is and it doesn’t work. It’s unclear if some bugs should be fixed, like http://llvm.org/bugs/show_bug.cgi?id=6488 (It’s unclear if clang community is interested in supporting this). Some bugs have attached patches that fix the issue in question, like this one http://llvm.org/bugs/show_bug.cgi?id=13992 (the comments also tell a story). And so on…

I would like to see bugzilla cleaned up and bug entry process somewhat formalized so that entries meet certain quality.

image001.png

You can get the list searching with the Bug Changes filter giving appropriate values in fields.

and shared it.

Other interested user can visit Preferences ->Saved Searches after logging into Bugzilla.

As we say, Just created and shared this bug group under the Last30DayIssues in bugzilla.

image001.png

This week, for the very first time since I started sampling (14 months ago),

the weekly open-bug count went DOWN.

Traffic on llvm-bugs indicates that Bill Wendling is doing a lot of this. Huzzah!

–paulr

image001.png

This week, for the very first time since I started sampling (14 months
ago),****

the weekly open-bug count went DOWN.****

** **

Traffic on llvm-bugs indicates that Bill Wendling is doing a lot of this.
Huzzah!

Bill is indeed on a bug-stomping rampage!

image001.png