J2SDK 1.4.0: SEGV

Richard rgollub at uninet.com.br
Sun Aug 11 08:01:09 PDT 2002

Greg Schafer wrote:
> Your first compile was already at -march=i686(pentiumpro) so I don't
> think you'll see any difference.

	Tks Greg for your interest and support. Indeed, I also do not expect
any, but, when it comes to gcc (not the only one, mind you) I am a bit
sceptical, and I am giving preference to hard evidence, as opposed to
taking at face value related docs. (Just remember the diff of options
enabled/passed when -march is given compared to the default... Who would
guarantee consistent sound behaviour on the part of the compiler once
submitted to a different set of (highly interrelated) options, taking 
into account all we read / hear / (and most importantly) experience? :))

-------------------------------------- (time lag between these

	Anyway, has just left j2dsk compiling overnight with explicit reference
to -march=i686 (pentiumpro :) and checked the outcome. Simply put: the
rabbit that hopped out of the hat is an exact clone of the previous one
:/ Exactly the same behavior and exception. (Well, at least, I am glad
with the compiler's "consistency"...)

> If it was me, I would explicitly pass -march=i386 somewhere in your
> CFLAGS and see if that makes the difference. Passing -march= on
> the command line (or in CFLAGS in this case) will always override
> whatever is the compiler's default.

	There is no problem with this package at it provides a reliable
external access via the OTHER_CFLAGS. Its makefiles are robust and well
thought out as to details like that one. Will keep you posted on the
outcome of compiling for i386 (or i486 more appropriately, judging from
info gathered from the makefiles.)

	Hence, in conclusion, my first message is correct all the way, and,
thus, await eagerly reports from others who might have any info to

Unsubscribe: send email to listar at linuxfromscratch.org
and put 'unsubscribe blfs-support' in the subject header of the message

More information about the blfs-support mailing list