From mboxrd@z Thu Jan 1 00:00:00 1970 Message-ID: <38DF94A9.CDDBD6FF@cray.com> Date: Mon, 27 Mar 2000 11:04:41 -0600 From: Mike Wolf MIME-Version: 1.0 To: Ralph Blach , linuxppc-embedded@lists.linuxppc.org Subject: Re: 405gc problem with shared libraries References: <38DF707F.DF49A490@raleigh.ibm.com> Content-Type: text/plain; charset=us-ascii Sender: owner-linuxppc-embedded@lists.linuxppc.org List-Id: Ralph Blach wrote: > > I am attempting to run the latest 405gc port and am getting error. > > I boot up to sash, and then run the command ls > everthing seems to run fine untill I get the message > > ./ls: error in loading shared libraries: libc.so.6: failed to map > segment > from shared object: Error 19 > > This message is from ld.so.1. What is ld.so.1 trying to do? Any > suggestions. > > Chip > First check the archives. There is some mail from Marcus Sundberg describing how to set up the compiler so it will not use floating point. Make sure to compile everything with the -msoft-float option. My guess from your error is you have already done this. Also I got this response from Ben Bodley when I was having a similiar problem "as far as my progress on shared libraries goes, currently: - i have checked the reason for failure, and it occurs in the 'dl-load.c' file from glibc, the initial __mmap() function is failing the reason is that the kernel mmap_pgoff() is getting passed a null 'file->f_op' pointer, and is exiting the __mmap() function with -ENODEV.. it's strange though, as even though file->f_op is at 0x00000000, the function file->f_op->mmap() points to a valid address! so if the file_operations struct is getting allocated at 0x00000000 this will cause all sorts of problems.. the major one being 0x00000000 is null, even though it is a valid logical memory address.." -- Mike Wolf Cray Research 655-F Lone Oak Drive, Eagan, MN 55121 mwolf@cray.com (651)683-5313 ** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/