Date: Sat, 12 Mar 2005 02:23:24 +0000 From: David O'Brien <obrien@FreeBSD.ORG> To: Bartosz Fabianowski <freebsd@chillt.de> Cc: freebsd-mobile@freebsd.org Subject: Re: CPUTYPE=pentium-m Message-ID: <20050312022324.GA4899@hub.freebsd.org> In-Reply-To: <422B502C.6090003@chillt.de> References: <422ACE0B.4060503@lumachedelcervello.cjb.net> <422B502C.6090003@chillt.de>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sun, Mar 06, 2005 at 07:47:08PM +0100, Bartosz Fabianowski wrote: > >i have read that were some problems compiling the kernel and the > >loader with "pentium-m" in CPUTYPE. are they fixed now? > > I'm the one who filed the original bug report: > http://www.freebsd.org/cgi/query-pr.cgi?pr=75898 .. > I'd assume the answer to this one is yes - it's safe. Personally, I compile > world and kernel as pentium3, but that's just out of laziness. As you can > read in the bug report, the problem is that SSE2 instructions get used by > the kernel and loader before they are enabled. Once the boot gets to the > first userland programs, SSE2 is enabled so any world programs should run > just fine when compiled as pentium-m. Are you saying, all we need to do is commit this diff to make everyone's environment happy? -- -- David (obrien@FreeBSD.org) Index: kern.mk =================================================================== RCS file: /home/ncvs/src/sys/conf/kern.mk,v retrieving revision 1.42 diff -u -r1.42 kern.mk --- kern.mk 14 May 2004 13:35:46 -0000 1.42 +++ kern.mk 12 Mar 2005 02:03:47 -0000 @@ -28,7 +28,8 @@ # cache tag lines) # .if ${MACHINE_ARCH} == "i386" && ${CC} != "icc" -CFLAGS+= -mno-align-long-strings -mpreferred-stack-boundary=2 +CFLAGS+= -mno-align-long-strings -mpreferred-stack-boundary=2 \ + -mno-sse -mno-sse2 INLINE_LIMIT?= 8000 .endif
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20050312022324.GA4899>