Date: Sat, 14 Sep 2013 20:23:37 -0400 From: Eitan Adler <lists@eitanadler.com> To: Adrian Chadd <adrian@freebsd.org> Cc: freebsd-sparc64 <freebsd-sparc64@freebsd.org>, "freebsd-arm@freebsd.org" <freebsd-arm@freebsd.org>, Ian Lepore <ian@freebsd.org>, FreeBSD PowerPC ML <freebsd-ppc@freebsd.org> Subject: Re: Patch for Cross-Reference Phandles Message-ID: <CAF6rxgnDd3uYg7dJHg4AcG_Y1S2dTH_i-ZcsWuFZ%2BJKFT7WW6w@mail.gmail.com> In-Reply-To: <CAJ-Vmo=_7oBjyrkAw=-mYzOgQr_GL0K1C%2B78u%2BmQRbxm6Nrjtw@mail.gmail.com> References: <522CC7E0.9060508@freebsd.org> <5231D8A6.9080501@freebsd.org> <20130914133155.GA32845@alchemy.franken.de> <52346D71.4080407@freebsd.org> <1379173405.1197.7.camel@revolution.hippie.lan> <20130914172735.GX68682@funkthat.com> <1379181341.1197.26.camel@revolution.hippie.lan> <CAJ-Vmonk_cjc%2BwfCqsHdVESMvRMQTP4c2myK9jZr_%2B59GmWLQQ@mail.gmail.com> <CAF6rxgmiYxEEBtK2Rt0-PoKfu8MP6gow%2B-Ncx7Ma3LFzGd8e9w@mail.gmail.com> <CAJ-Vmo=_7oBjyrkAw=-mYzOgQr_GL0K1C%2B78u%2BmQRbxm6Nrjtw@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, Sep 14, 2013 at 8:06 PM, Adrian Chadd <adrian@freebsd.org> wrote:
> On 14 September 2013 17:04, Eitan Adler <lists@eitanadler.com> wrote:
>
>>
>> With an explicit zeroing:
>>
>> text data bss dec hex filename
>> 1572 484 16 2072 818 a.out
>>
>> Without:
>>
>> text data bss dec hex filename
>> 1572 484 16 2072 818 a.out
>
>
> This is exactly the opposite of what I see with kernel code though. Eg,
> kern_umtx.c and it's very large up-front allocated array. Same as the
> upfront array of KTR entries for early boot tracing.
>
> Try bumping it to something stupid like 65536 and retest?
>
> Is this a gcc-4.2 versus clang difference?
#ifdef EXPLICIT
int foo[65536] = { 0 };
#else
int foo[65536];
#endif
int main()
{
int i;
for (i = 0; i < 2; i++) {
foo[i] = i;
}
}
No optimizations.
text data bss dec hex filename
1572 484 262152 264208 40810 clang-explicit
1572 484 262160 264216 40818 clang-implicit
1508 456 262184 264148 407d4 gcc46-explicit
1508 456 262176 264140 407cc gcc46-implicit
1652 456 262184 264292 40864 gcc49-explicit
1652 456 262176 264284 4085c gcc49-implicit
I'm not sure of the cause of the 8 byte difference.
--
Eitan Adler
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAF6rxgnDd3uYg7dJHg4AcG_Y1S2dTH_i-ZcsWuFZ%2BJKFT7WW6w>