This is the mail archive of the
mailing list for the glibc project.
Re: [PATCH] remove nested function hack_digit
- From: Roland McGrath <roland at hack dot frob dot com>
- To: Konstantin Serebryany <konstantin dot s dot serebryany at gmail dot com>
- Cc: Andreas Schwab <schwab at linux-m68k dot org>, GNU C Library <libc-alpha at sourceware dot org>
- Date: Mon, 22 Sep 2014 15:45:16 -0700 (PDT)
- Subject: Re: [PATCH] remove nested function hack_digit
- Authentication-results: sourceware.org; auth=none
- References: <CAGQ9bdxUJaUzz=ndu-qnhkPGAH7=m5mFKxpDag=H693TeA2ORw at mail dot gmail dot com> <87a960l9ze dot fsf at igel dot home> <CAGQ9bdyxCW-_3rLy6uLg4Vc2FPx+gUL7PChaXA4i6aKmnjGVZg at mail dot gmail dot com> <mvm38bsyppg dot fsf at hawking dot suse dot de> <CAGQ9bdya8w_OmD=1wKayhLN51H+Jqaio3RGqtATKWc6_hPgBxQ at mail dot gmail dot com> <20140922214338 dot 0D30A2C3971 at topped-with-meat dot com> <CAGQ9bdzKgTMEFM7-uL98nzqgJfOtm+U0AhzcnkgqTuqs3r_=UQ at mail dot gmail dot com>
> hack_digit becomes longer too due to longer function prologue epilogue:
Clearly that cannot be the only difference. Was the function's code itself
actually all nearly identical, modulo trivial differences like different
register allocation choices? Each access to one of the parent's locals
surely looks different, and how different that code looks is probably where
the most important differences are.
> I considered doing such a patch but it turned out a huge textual
> change that will make the code much less readable.
> Still, let me do it and send it here anyway, unless you tell me no to.
Of course readability is very subjective, so there really is no substitute
for each interested person just seeing how things look and giving their
opinion. The most trivial mechanical change might harm readability in ways
that can be improved with a little thought.