Greetings. Linux kernel module installation (2.4.x) fails because cp from fileutils 4.1.1 is super protective of the user. See forwarded message below which is a response from Keith Owens (one of those kernel guys) on the linux kernel mailing list.
The point here is: is the observed behavior unnecessary and if not will such behavior become a problem down the road? Granted I'd say it's more like the particular kernel module Makefile in question is broken rather than cp, but Keith (or anyone for that matter) was unwilling to fix it. sigh. Regards, kw PS -- There is a comment in fileutils-4.1.1/src/copy.c which explains the logic behind the behavior and I forwarded said comment to the kernel list but it mysteriously never made it to the list. :?/ ---------- Forwarded message ---------- Date: Fri, 14 Dec 2001 15:31:00 +1100 From: Keith Owens <[EMAIL PROTECTED]> To: Keith Warno <[EMAIL PROTECTED]> Cc: Linux Kernel List <[EMAIL PROTECTED]> Subject: Re: make modules_install: chokes on emu10k1 module install On Thu, 13 Dec 2001 18:41:13 -0500 (EST), Keith Warno <[EMAIL PROTECTED]> wrote: >make[2]: Entering directory `/usr/src/linux-2.4.17-rc1/drivers/sound' >mkdir -p /lib/modules/2.4.17-rc1/kernel/drivers/sound/ >cp soundcore.o ac97_codec.o ac97_codec.o >/lib/modules/2.4.17-rc1/kernel/drivers/sound/ >cp: will not overwrite just-created >`/lib/modules/2.4.17-rc1/kernel/drivers/sound/ac97_codec.o' with >`ac97_codec.o' Broken version of cp. Get a version of fileutils that does not make unnecessary changes to existing behaviour. _______________________________________________ Bug-fileutils mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/bug-fileutils
