I've been running Gentoo for a few months on one of my other boxes and it's been solid as can be. Last night I decided to do away with my Red Hat install and move to Gentoo on my workstation. I initially began with a complete source install -- I began the bootstrap script and went to bed and awoke to find the process halted in the middle of the build. The build system it seems was simply totally halted, the last line was a GCC build line. Moving over to another virtual console I was presented with the 'Hit enter to activate this console' message, but upon hitting enter the cursor simply stared back at me and nothing happened. I tried again and the build died elsewhere. This morning I decided to fall back and punt and try to get up and running with the stage2 tarball. The bootstrap completed successfully, but the emerge system call resulted in another outright hang -- right now my console has printed the following: cd ../obj_s; gcc -I../ncurses -I. -DNDEBUG -I. -I../include - DHAVE_CONFIG_H -march=i686 -O3 -pipe -fPIC -c ../ncurses/./base/lib_vline.c Again, virtual consoles respond but fail to activate. For reference, my system specs follow. Also, Portage is receiving the kill signal, but fails to unload me back into a shell. Blake Athlon 1.4GHz Abit KG7-RAID 1GB DDR RAM 2 x 30GB IBM DeskStar ATA/100 HD 1 x 60GB WD ATA/100 HD Sound Blaster Live Pinnacle FireWire Adapter 3Com 359x NIC 10X ATAPI DVD ROM 40x12x40A Plextor Atapi CD-RW 64MB GeForce2 MX-400 AGP Video Adapter (ViewSonic G810 21" monitor) 32MB SiS 300 PCI Video Adapter (NEC MultiSync XP21 21" monitor) ATI TV Wonder PCI Onboard HPT-370 IDE RAID Controller Build attempted with Gentoo Linux 1.1A ISO 30GB IBM Drive dedicated to install XFS partitions (/boot and /) Dual booted with Windows XP Professional
This sounds like a hardware problem. Test your memory with memtest86 and ensure that your CPU is being adequately cooled. If it dies in different places every time you try the install, then it is almost certainly a hardware problem.
OK, we are getting additional info on this bug. Try passing "mem=512m" as a kernel boot parameter and the CD kernel should work. It appears to be a kernel problem with machines with 1Gb+ of RAM -- and it appears that it *may* be caused by XFS being incompatible with highmem. You may want to retry with ext3 partitions and see if you can get through the build.
We at least have a workaround for this problem, so I'm marking it as fixed. We'll fix the issue for good in our next kernel releases.
