Bug #60171 | use of SYS_gettid in stacktrace.c breaks legacy Linux | ||
---|---|---|---|
Submitted: | 18 Feb 2011 20:24 | Modified: | 19 Feb 2011 10:16 |
Reporter: | Todd Rinaldo | Email Updates: | |
Status: | Unsupported | Impact on me: | |
Category: | MySQL Server: Compiling | Severity: | S3 (Non-critical) |
Version: | 5.1.55 | OS: | Linux (pre-RH AS3) |
Assigned to: | Davi Arnaut | CPU Architecture: | Any |
Tags: | Contribution, regression |
[18 Feb 2011 20:24]
Todd Rinaldo
[18 Feb 2011 20:28]
Todd Rinaldo
Suggested patch should fix legacy systems
Attachment: SYS_gettid.patch (application/octet-stream, text), 830 bytes.
[18 Feb 2011 20:28]
Peter Laursen
Is it really relevant to support MySQL 5.5 on such antique Operating System? I think no! But I had a recent discussion wiht the documentation team and docs tell (they were updated) that 'Linux systems with a 2.6.x kernel are supported' (quoted from memory). Peter (not a MySQL person)
[18 Feb 2011 20:31]
Todd Rinaldo
5.1 not 5.5
[19 Feb 2011 1:27]
Davi Arnaut
It is pre- or on-AS3?
[19 Feb 2011 3:09]
Todd Rinaldo
Pre-AS3. We seem to be harping on what linux platforms don't work. I don't get it. The change I'm suggesting simply allows MySQL to support a wider user base. I'm not really clear why that's undesirable to everyone. Especially since the bit of code which doesn't support pre-AS3 is so small. I'd point out that this bit of code also isn't supported on Windows, BSD, etc. *shrug*
[19 Feb 2011 10:16]
Davi Arnaut
The maintainability cost is undesired. We would need to ensure that it works on such platform and on newer ones (plus patch review, testing, etc). We don't want to keep supporting hacks for old platforms forever, sometimes we have to draw a line on the sand -- no special cases for unsupported platforms.