| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | Tomas Vondra <tomas(dot)vondra(at)enterprisedb(dot)com> |
| Cc: | David Rowley <dgrowleyml(at)gmail(dot)com>, Amit Kapila <amit(dot)kapila16(at)gmail(dot)com>, Andres Freund <andres(at)anarazel(dot)de>, Robert Haas <robertmhaas(at)gmail(dot)com>, Yura Sokolov <y(dot)sokolov(at)postgrespro(dot)ru>, PostgreSQL Developers <pgsql-hackers(at)lists(dot)postgresql(dot)org> |
| Subject: | Re: Reducing the chunk header sizes on all memory context types |
| Date: | 2022-09-01 01:06:09 |
| Message-ID: | 225785.1661994369@sss.pgh.pa.us |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Tomas Vondra <tomas(dot)vondra(at)enterprisedb(dot)com> writes:
> You're probably right we'll notice the clobber cases due to corruption
> of the next chunk header. The annoying thing is having a corrupted
> header only tells you there's a corruption somewhere, but it may be hard
> to know which part of the code caused it.
Same's true of a sentinel, though.
> OTOH we have platforms where valgrind is either not supported or no one
> runs tests with (e.g. on rpi4 it'd take insane amounts of code).
According to
https://valgrind.org/info/platforms.html
valgrind supports a pretty respectable set of platforms. It might
be too slow to be useful on ancient hardware, of course.
I've had some success in identifying clobber perpetrators by putting
a hardware watchpoint on the clobbered word, which IIRC does work on
recent ARM hardware. It's tedious and far more manual than valgrind,
but it's possible.
regards, tom lane
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Kyotaro Horiguchi | 2022-09-01 01:12:37 | Re: Avoid overhead open-close indexes (catalog updates) |
| Previous Message | Michael Paquier | 2022-09-01 00:57:23 | Re: Support tls-exporter as channel binding for TLSv1.3 |