855081 - jprof should use abi::__cxa_demangle

Status: RESOLVED FIXED

(Core :: General, defect)

Description

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

I've had a patch in my tree for a while to make jprof compile (which it wasn't for me without the patch).  It switches from using the old cplus_demangle function to using abi::__cxa_demangle, which is sort of officially supported:
http://gcc.gnu.org/onlinedocs/libstdc++/libstdc++-html-USERS-4.3/a01696.html

cplus_demangle didn't link for me.

Comment 1

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

Attachment: Make jprof compile for me by switching from cplus_demangle (which doesn't link) to abi::__cxa_demangle.

Comment 2

[Randell Jesup [:jesup] (needinfo me)]

Comment on attachment 729808 [details] [diff] [review]
Make jprof compile for me by switching from cplus_demangle (which doesn't link) to abi::__cxa_demangle.

Review of attachment 729808 [details] [diff] [review]:
-----------------------------------------------------------------

::: tools/jprof/bfd.cpp
@@ +187,5 @@
>  	}
>  	sp->Init(dnm ? dnm : nm, syminfo.value + aBaseAddress);
> +        if (dnm) {
> +          free(dnm);
> +        }

It would be a lot less malloc traffic if you just reused the same buffer each time (it will be realloced larger as needed).  However, this only matters when processing, so it's not a big deal and this is provably safe.

Comment 3

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

Landed:
https://hg.mozilla.org/integration/mozilla-inbound/rev/261e3e928ce9

(I'll add a patch for the better way in my queue so I test it the next time I use jprof...)

Comment 4

[Ed Morley [:emorley]]

https://hg.mozilla.org/mozilla-central/rev/261e3e928ce9

Comment 5

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

Attachment: patch 2: Use a single buffer for the demangled names.

This has been working for me for a while.

Comment 6

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

https://hg.mozilla.org/integration/mozilla-inbound/rev/8a72d2620918

(also see try push at https://tbpl.mozilla.org/?tree=Try&rev=657c8987d64d)

Comment 7

[Ed Morley [:emorley]]

https://hg.mozilla.org/mozilla-central/rev/8a72d2620918