Date: Fri, 21 Aug 1998 11:17:28 -0700 (PDT) From: Tom Bartol <bartol@salk.edu> To: Jonathan Lemon <jlemon@americantv.com> Cc: questions@FreeBSD.ORG Subject: Re: ThinkPad 770 hangs during boot -- available memory probe problem? (was: trouble with sio on Thinkpad 770) Message-ID: <Pine.BSF.3.96.980821102859.24886A-100000@cole.salk.edu> In-Reply-To: <19980819140829.08171@right.PCS>
next in thread | previous in thread | raw e-mail | index | archive | help
Hi, I tried adding options VM86 to a 3.0 kernel from 19980711-SNAP but this did not fix the problem. I noticed in the LINT config that npx0 now has an iosiz parameter that can be set to the memory size instead of MAXMEM and that this can be set with userconfig as well. This option does work when I set it in the kernel config file but does not work when using userconfig since it only takes effect on the next boot after the parameter is recorded in the kernel -- since the machine hangs during boot it never gets recorded -- a catch 22 scenario. This is all very useful for working around the problem and I thank you for it but is not getting at the deeper issue of what is going wrong in the first place and to determine the appropriate action to take -- either fix FreeBSD or give useful information to IBM so they can fix their BIOS. I'm looking for direction and assistance in what I should be looking at to get a handle on the deeper issue. Thanks again, Tom On Wed, 19 Aug 1998, Jonathan Lemon wrote: > On Aug 08, 1998 at 09:37:39AM -0700, Tom Bartol wrote: > > installed 3.0-970807-SNAP. I could now boot my system. I then configured > > a custom kernel with this system and after more investigative work found > > that if I set MAXMEM = (96*1024) in the kernel config file (recall that > > these older systems still required the MAXMEM parameter for >64MB to be > > recognized) then my system would hang on boot, but if I set MAXMEM = > > (95*1024) everything works perfectly, even the serial port. :-) I guess > > that more recent kernels can somehow get the correct available memory > > value on their own by some means (perhaps from the BIOS) and that this > > algorithm works for some BIOSes but not others and could be fixed to work > > with the new BIOS I installed on my 770, or perhaps the 770 BIOS is not > > reporting the correct value when probed by FreeBSD. So, the question is: > > what's going here and what should I do to help narrow it down further so > > that the appropriate action can be taken. > > If you have -current sources newer than 980324, then you might > want to try compiling a kernel with ``options "VM86"'', instead > of MAXMEM. This enables some functions which query the BIOS > directly for the memory size on bootup, and might fix your > problem. > -- > Jonathan > To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.3.96.980821102859.24886A-100000>