Matthew, uname reads /proc/version
The only think /proc/version has that uname does not is the name of the compiler that was used to compile the kernel. --brian On Fri, 06 Jun 2003, Matthew Gierlach wrote: > Hi Dave: > > I understand the point about /proc/version. In reading the > installation script, the value of uname -r is used to > define the /lib/modules/<version>/misc location for the > streams.o binary instead of the version extracted from > /proc/version. This may be a non-issue since both are > supposed to represent the version of the executing OS. > > Regardless of whether uname -r or /proc/version is used, > neither is audited against the version of the kernel in the > source tree. The script proceeds to build and install LiS > without the audit. > > In prior incantations of the install script, the audit was > performed and the install was aborted if the versions came up > different. LiS instructed the user to ensure the executable > and the source versions were the same before reattempting the > install. > > That seems much safer than proceeding without the audit and > then having to field support calls for LiS Oops because of > incompatible source and binary. > > Am I missing something that destroys the assumption that mis- > matching kernel binary and source versions are not good, especially > for the SMP case? > > Thanks, Matt Gierlach > > On Fri, 6 Jun 2003, David Grothe wrote: > > > As I recall the idea was to place more stock in the kernel version > > information found in /proc when building for the running kernel. If > > version.h differs from what /proc says then I guess it means that the > > kernel source directory is not for the running kernel. > > > > If I put that check back in, I wonder what else would break.... > > > > We seem to keep going round and round on these issues. It doesn't help, of > > course, that the distributions introduce chaos into these efforts with > > every new release. What are the units for entropy? > > > > -- Dave > > > > At 08:04 PM 6/4/2003 Wednesday, you wrote: > > >Hi Dave: > > > > > > I notice that the current LiS installation script > > > (LiS-2.16.8 for example) no longer audits the > > > kernel version from the version.h file with that > > > of the running kernel (via uname -r). So it > > > appears conceivable that a version.h file exists > > > that is different from the running kernel and LiS > > > will proceed to install regardless. This seems > > > risky. One could end up with an LiS compiled with > > > header files that do not correlate to the running > > > kernel. No? > > > > > > Thanks, Matt Gierlach > > > > > > > > _______________________________________________ > Linux-streams mailing list > [EMAIL PROTECTED] > http://gsyc.escet.urjc.es/mailman/listinfo/linux-streams -- Brian F. G. Bidulock � The reasonable man adapts himself to the � [EMAIL PROTECTED] � world; the unreasonable one persists in � http://www.openss7.org/ � trying to adapt the world to himself. � � Therefore all progress depends on the � � unreasonable man. -- George Bernard Shaw � _______________________________________________ Linux-streams mailing list [EMAIL PROTECTED] http://gsyc.escet.urjc.es/mailman/listinfo/linux-streams
