Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 24 Aug 2017 13:27:14 +0900 (JST)
From:      Mori Hiroki <yamori813@yahoo.co.jp>
To:        Adrian Chadd <adrian.chadd@gmail.com>
Cc:        "freebsd-mips@freebsd.org" <freebsd-mips@freebsd.org>
Subject:   Re: performance by endian
Message-ID:  <2796.80438.qm@web101711.mail.ssk.yahoo.co.jp>
In-Reply-To: <CAJ-Vmo=fkYZ=FZrisrjZ9W=MFJd85z=UiEQBBF-j8SVx=PLnvA@mail.gmail.com>
References:  <319927.83012.qm@web101708.mail.ssk.yahoo.co.jp> <CAJ-VmomwtjgPjR-1GLsbXm7g%2Bxo5Mf1CMnXcqXED_-_vgWdsWA@mail.gmail.com> <539086.28647.qm@web101720.mail.ssk.yahoo.co.jp> <CAJ-Vmo=fkYZ=FZrisrjZ9W=MFJd85z=UiEQBBF-j8SVx=PLnvA@mail.gmail.com>

next in thread | previous in thread | raw e-mail | index | archive | help
Hi=0A=0AAR9331 is ONION OMEGA. This module have Winbond W9751G6KB=0ADDR2 me=
mory.=0A=0A=0AI checked=A0AR9331 DDR TAP register at boot time.=0A=0ATAP_CO=
NTROL_0(0x1800001c) =3D 0xf=0ATAP_CONTROL_1(0x18000020) =3D 0x10 of 0xf=0A=
=0AI think this is=A0ar7240_ddr_ap_init() result in=A0u-boot code.=0A=0AI d=
on't know what is TAP. But is this relation to this=0Aproblem ?=0A=0AHiroki=
 Mori=0A=0A----- Original Message -----=0A> From: Adrian Chadd <adrian.chad=
d@gmail.com>=0A> To: Mori Hiroki <yamori813@yahoo.co.jp>=0A> Cc: "freebsd-m=
ips@freebsd.org" <freebsd-mips@freebsd.org>=0A> Date: 2017/8/15, Tue 06:10=
=0A> Subject: Re: performance by endian=0A> =0A> hm, which DDR bus does it =
have?=0A> =0A> =0A> =0A> -adrian=0A> =0A> =0A> On 25 July 2017 at 19:09, Mo=
ri Hiroki <yamori813@yahoo.co.jp> wrote:=0A>>  Hi=0A>> =0A>> =0A>>  Sorry I=
 talk about AR9331 soc.=0A>> =0A>>  I check boot log.=0A>> =0A>>  RT3050 is=
 this.=0A>> =0A>> =0A>>  Cache info:=0A>> =A0  picache_stride=A0 =A0 =3D 40=
96=0A>> =A0  picache_loopcount =3D 8=0A>> =A0  pdcache_stride=A0 =A0 =3D 40=
96=0A>> =A0  pdcache_loopcount =3D 4=0A>>  cpu0: MIPS Technologies processo=
r v76.150=0A>> =A0  MMU: Standard TLB, 32 entries (4K 16K 64K 256K 1M 16M 6=
4M 256M pg sizes)=0A>> =A0  L1 i-cache: 4 ways of 256 sets, 32 bytes per li=
ne=0A>> =A0  L1 d-cache: 4 ways of 128 sets, 32 bytes per line=0A>> =A0  L2=
 cache: disabled =0A> Config1=3D0xbea3319e<PerfCount,WatchRegs,MIPS16,EJTAG=
>=0A>> =A0  Config2=3D0x80000000=0A>> =A0  Config3=3D0x420=0A>> =0A>>  AR93=
31 is this.=0A>> =0A>>  Cache info:=0A>> =A0  picache_stride=A0 =A0 =3D 409=
6=0A>> =A0  picache_loopcount =3D 16=0A>> =A0  pdcache_stride=A0 =A0 =3D 40=
96=0A>> =A0  pdcache_loopcount =3D 8=0A>> =A0  max line size=A0 =A0  =3D 32=
=0A>>  cpu0: MIPS Technologies processor v116.147=0A>> =A0  MMU: Standard T=
LB, 16 entries (4K 16K 64K 256K 1M 16M 64M 256M pg sizes)=0A>> =A0  L1 i-ca=
che: 4 ways of 512 sets, 32 bytes per line=0A>> =A0  L1 d-cache: 4 ways of =
256 sets, 32 bytes per line=0A>> =A0  L2 cache: disabled=0A>> =A0  Config1=
=3D0x9ee3519e<PerfCount,WatchRegs,MIPS16,EJTAG>=0A>> =A0  Config2=3D0x80000=
000=0A>> =A0  Config3=3D0x20=0A>> =A0  Config7=3D0x0=0A>> =0A>>  I seem AR9=
331 have more cache than RT3050.=0A>> =0A>>  awk script is this.=0A>> =0A>>=
  BEGIN {=0A>> =A0 =A0  printf "%d=A5n", fib(32)=0A>>  }=0A>> =0A>>  functi=
on fib(n) {=0A>> =A0 =A0  if (n < 2) return n=0A>> =A0 =A0  return fib(n - =
2) + fib(n - 1)=0A>>  }=0A>> =0A>>  ----- Original Message -----=0A>>>  Fro=
m: Adrian Chadd <adrian.chadd@gmail.com>=0A>>>  To: Mori Hiroki <yamori813@=
yahoo.co.jp>=0A>>>  Cc: "freebsd-mips@freebsd.org" =0A> <freebsd-mips@freeb=
sd.org>=0A>>>  Date: 2017/7/26, Wed 02:14=0A>>>  Subject: Re: performance b=
y endian=0A>>> =0A>>>  well, how much cache, how much out-of-order is it? T=
he AR9330 SoC has=0A>>>  very limited cache :(=0A>>> =0A>>> =0A>>> =0A>>>  =
-adrian=0A>>> =0A>>> =0A>>>  On 24 July 2017 at 00:53, Mori Hiroki <yamori8=
13@yahoo.co.jp> =0A> wrote:=0A>>>> =A0 Hi.=0A>>>> =0A>>>> =A0 I checked fib=
(32) by awk script.=0A>>>> =0A>>>> =A0 RT3050 is 80 sec=0A>>>> =A0 AR9330 i=
s 99 sec=0A>>>> =0A>>>> =0A>>>> =A0 AR9330 is newer and clock is hight. But=
 slow.=0A>>>> =0A>>>> =A0 Why this result ?=0A>>>> =0A>>>> =A0 I suspect th=
is is endian issue.=0A>>>> =0A>>>> =A0 I have problem mruby performance at =
Atheros SOC.=0A>>>> =A0 It is slow almost double from same clock little=0A>=
>>> =A0 endian mips soc.=0A>>>> =0A>>>> =A0 Regards=0A>>>> =0A>>>> =A0 Hiro=
ki Mori=0A>>>> =A0 _______________________________________________=0A>>>> =
=A0 freebsd-mips@freebsd.org mailing list=0A>>>> =A0 https://lists.freebsd.=
org/mailman/listinfo/freebsd-mips=0A>>>> =A0 To unsubscribe, send any mail =
to=0A>>>  "freebsd-mips-unsubscribe@freebsd.org"=0A>>> =0A> 



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?2796.80438.qm>