String to ascii char array converter for debug mode.

String to ascii char array converter for debug mode.

Committed: http://code.google.com/p/v8/source/detail?r=8763

[Yang, 2011-07-29 09:04:48]

PTAL.

[Rico, 2011-07-29 09:18:07]

nit drive by

http://codereview.chromium.org/7523052/diff/1/src/objects-printer.cc
File src/objects-printer.cc (right):

http://codereview.chromium.org/7523052/diff/1/src/objects-printer.cc#newcode565
src/objects-printer.cc:565: if (buffer != NULL)
{ } around if body unless moved to same line

[Yang, 2011-07-29 09:29:43]

I think that should be OK?

[Sven Panne, 2011-07-29 09:31:46]

http://codereview.chromium.org/7523052/diff/1/src/objects-printer.cc
File src/objects-printer.cc (right):

http://codereview.chromium.org/7523052/diff/1/src/objects-printer.cc#newcode564
src/objects-printer.cc:564: static char* buffer = NULL;
Drive-by commment: This looks scary... The real problem is that we mix 2 things
here (like in lots of other places): Memory management and the conversion
itself. This leads to obscure ownerships and lifetimes, like the "static" above.
IMHO it would be cleaner and simpler to transfer the ownership of the result to
the caller.

Long-term solution: Switch to C++-style << and >>, one rarely needs the whole
converted result/input all at once.

[Rico, 2011-07-29 09:48:07]

On 2011/07/29 09:29:43, Yang wrote:
> I think that should be OK?

Yes, on one line that is fine

[Yang, 2011-07-29 09:49:05]

I don't think there is a big issue with memory management as this method is
supposed to be called from gdb only, for debugging purposes. There are more
sophisticated alternatives for the runtime system.

On 2011/07/29 09:31:46, Sven wrote:
> http://codereview.chromium.org/7523052/diff/1/src/objects-printer.cc
> File src/objects-printer.cc (right):
> 
>
http://codereview.chromium.org/7523052/diff/1/src/objects-printer.cc#newcode564
> src/objects-printer.cc:564: static char* buffer = NULL;
> Drive-by commment: This looks scary... The real problem is that we mix 2
things
> here (like in lots of other places): Memory management and the conversion
> itself. This leads to obscure ownerships and lifetimes, like the "static"
above.
> IMHO it would be cleaner and simpler to transfer the ownership of the result
to
> the caller.
> 
> Long-term solution: Switch to C++-style << and >>, one rarely needs the whole
> converted result/input all at once.

[Jakob Kummerow, 2011-07-29 09:49:25]

LGTM, thanks.