I believe I've discovered an odd discrepancy between what is in the official Linux BitKeeper repository, and what is on ftp.kernel.org. According to BitKeeper, the last time linux/arch/s390/config.in and linux/arch/s390x/config.in were changed is 17 months ago. What is in ftp.kernel.org/pub/linux/kernel/v2.4/linux-2.4.28.tar.bz2 was last modified on November 17, 2004. The difference between the two versions is this: @@ -57,7 +57,7 @@ bool 'BSD Process Accounting' CONFIG_BSD_PROCESS_ACCT bool 'Sysctl support' CONFIG_SYSCTL define_bool CONFIG_KCORE_ELF y -tristate 'Kernel support for ELF binaries' CONFIG_BINFMT_ELF +bool 'Kernel support for ELF binaries' CONFIG_BINFMT_ELF tristate 'Kernel support for MISC binaries' CONFIG_BINFMT_MISC bool 'Show crashed user process info' CONFIG_PROCESS_DEBUG bool 'Pseudo page fault support' CONFIG_PFAULT
Admittedly, pretty small, but still disturbing to me, at least. What's interesting is that this change seems to have been made to a _lot_ of architectures (14 in all) in patch-2.4.28-pre2.bz2, which is now in ftp.kernel.org/pub/linux/kernel/v2.4/testing/old/ That file is dated August 26, 2004. This same change was included in -pre3, all the way through -rc4, and then the final 2.4.28. The entry in the patch-2.4.28.log looks like this: Adrian Bunk: o 2.4.28-pre1: add two SATA Configure.help entries o disallow modular BINFMT_ELF Does anyone have any idea why this didn't make it into BitKeeper? Should it be in BitKeeper or not? This looks like some sort of process failure (or failure to follow the process), which is what concerns me the most. Mark Post - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
