Date: Sun, 01 Oct 2000 21:47:25 GMT From: Salvo Bartolotta <bartequi@inwind.it> To: Bill Fumerola <billf@chimesnet.com> Cc: BSD <bsd@shell-server.com>, "Jeffrey J. Mountin" <jeff-ml@mountin.net>, freebsd-chat@freebsd.org Subject: Re: Another 4.1-S panic (full report) Message-ID: <20001001.21472500@bartequi.ottodomain.org> References: <4.3.2.20001001132645.00cae340@207.227.119.2> <Pine.BSF.4.10.10010011401380.70294-100000@marvin.shell-server.com> <20001001154131.Y38472@jade.chc-chimes.com> <20001001.20514000@bartequi.ottodomain.org> <20001001155807.Z38472@jade.chc-chimes.com>
next in thread | previous in thread | raw e-mail | index | archive | help
[ redirected to -chat ] >>>>>>>>>>>>>>>>>> Original Message <<<<<<<<<<<<<<<<<< On 10/1/00, 8:58:07 PM, Bill Fumerola <billf@chimesnet.com> wrote regarding Re: Another 4.1-S panic (full report): > On Sun, Oct 01, 2000 at 08:51:40PM +0000, Salvo Bartolotta wrote: > > LINT (cvsup'ed yesterday) states: > > > > <blockquote> > > > > # NO_F00F_HACK disables the hack that prevents Pentiums (and ONLY > > # Pentiums) from locking up when a LOCK CMPXCHG8B instruction is > > # executed. This should be included for ALL kernels that won't run > > # on a Pentium. > > > > </blockquote> > > > > Either the statement in LINT is not correct, or your statement is no= t > > correct. Tertium non datur :-) > > > > Seriously: if NO_F00F_HACK is only used in conjunction with pentiums= > > (I586_CPU), then LINT should be modified accordingly. > ./i386/identcpu.c:#if defined(I586_CPU) && !defined(NO_F00F_HACK) > ./i386/identcpu.c:#if defined(I586_CPU) && !defined(NO_F00F_HACK) > ./i386/machdep.c:#if defined(I586_CPU) && !defined(NO_F00F_HACK) > ./i386/machdep.c:#if defined(I586_CPU) && !defined(NO_F00F_HACK) > ./i386/machdep.c:#endif /* defined(I586_CPU) && !NO_F00F_HACK */ > ./i386/mp_machdep.c:#if defined(I586_CPU) && !defined(NO_F00F_HACK) > ./i386/trap.c:#if defined(I586_CPU) && !defined(NO_F00F_HACK) > ./i386/trap.c:#if defined(I586_CPU) && !defined(NO_F00F_HACK) > ./i386/trap.c:#if defined(I586_CPU) && !defined(NO_F00F_HACK) > ./i386/trap.c:#if defined(I586_CPU) && !defined(NO_F00F_HACK) > So, because the microphone evidently wasn't on the first time: > If you don't define I586_CPU, NO_F00F_HACK does nothing. I am afraid you missed my point. And I am afraid the reverse is true: I did pay attention. In the last sentence of my previous letter (see above), I suggested modifying LINT if it was incorrect. I did NOT state that you were certainly wrong. You have just shown that LINT IS incorrect. By the way, Jeffrey Mountin has actually just suggested a possible modification. I should have had a look at the code first; this is certainly my only fault *blushing like a primary school pupil not having done his/her homework* :-) Best regards, Salvo To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-chat" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20001001.21472500>