Kyle Burton on Mon, 7 Jun 1999 16:32:41 -0400 (EDT) |
for the brave... I've seen 'make -j' work on a box with 128MB of ram + ~100mb of swap. The load did jump to something like 48, but it didn't get any errors...it just bogged the living daylights out of the system. X got a gooey %-) (pun intended) k ------------------------------------------------------------------------------ All theoretical chemistry is really physics; and all theoretical chemists know it. -- Richard P. Feynman mortis@voicenet.com http://www.voicenet.com/~mortis ------------------------------------------------------------------------------ On Mon, 7 Jun 1999, Tracy Nelson wrote: > Being ignorant (a suitable substitute for brave in this case) I issued a > "make -j" to build my 2.2.7 kernel. What a trip watching xosview -- load > average up over 48.0, CPU pegged at 100%, watching memory used wipe out > memory buffered/cached...until I ran out of memory, then CPU usage dropped > to ~50% and swap space started climbing. I noticed that I started getting > lots of compiler errors at that point -- parse errors and the like. Anyone > want to guess why? I'm wondering if it was because I ran out of file > handles and gcc couldn't open all the include files or temporary files it > needed. I didn't see any "xxxx.c: cannot open" errors, though. Might have > to do that again and pipe the output to a file... > > Just junk food for thought! > -- Tracy Nelson > > -----Original Message----- > From: Kyle Burton [SMTP:mortis@voicenet.com] > Sent: Monday, June 07, 1999 14:55 > To: plug@lists.nothinbut.net > Subject: RE: [Plug] Kernel recompile... > > make -j 5 > > speed up your builds > > > make dep && make clean && make -j 5 bzImage && make -j 5 modules && make -j > 5 modules_install > > zoom zoom zoom > > > Of course, you can do -j with no parameter if you're brave...but it should > shave a few more seconds off your kernel compiles (ever seen a linux box > bog down? -- just do 'make -j bzImage', it's like swimming in molasses). > > k > > ------------------------------------------------------------------------ > ------ > All theoretical chemistry is really physics; and all theoretical chemists > know > it. > -- Richard P. Feynman > mortis@voicenet.com > http://www.voicenet.com/~mortis > ------------------------------------------------------------------------ > ------ > > On Mon, 7 Jun 1999 Vale_Kenny@vanguard.com wrote: > > > I'll try it tonight, I've gotten a couple of emails saying there is not a > > problem w/ bzImage, on modern machines w/ lotsa RAM. Therefore, I'll be > > recompiling tonight w/ bzImage, and put all of the cool goodies back in.. > > Thanks, > > I'll keep the list informed if I have any hassles.. > > Peace, > > Vale > > > > > > > > _______________________________________________ > > Plug maillist - Plug@lists.nothinbut.net > > http://lists.nothinbut.net/mail/listinfo/plug > > > > > _______________________________________________ > Plug maillist - Plug@lists.nothinbut.net > http://lists.nothinbut.net/mail/listinfo/plug > > > _______________________________________________ > Plug maillist - Plug@lists.nothinbut.net > http://lists.nothinbut.net/mail/listinfo/plug > _______________________________________________ Plug maillist - Plug@lists.nothinbut.net http://lists.nothinbut.net/mail/listinfo/plug
|
|