Date: Mon, 9 Sep 2013 13:37:29 -0400 From: John Baldwin <jhb@freebsd.org> To: Adrian Chadd <adrian@freebsd.org> Cc: svn-src-head@freebsd.org, svn-src-all@freebsd.org, src-committers@freebsd.org Subject: Re: svn commit: r254199 - in head/sys: contrib/dev/iwn modules/iwnfw/iwn6000g2a Message-ID: <201309091337.29326.jhb@freebsd.org> In-Reply-To: <201308110103.r7B13WJn070719@svn.freebsd.org> References: <201308110103.r7B13WJn070719@svn.freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Saturday, August 10, 2013 9:03:32 pm Adrian Chadd wrote: > Author: adrian > Date: Sun Aug 11 01:03:32 2013 > New Revision: 254199 > URL: http://svnweb.freebsd.org/changeset/base/254199 > > Log: > Update the 6000g2a image. > > Obtained from: Linux, Intel > > Added: > head/sys/contrib/dev/iwn/iwlwifi-6000g2a-18.168.6.1.fw.uu > Modified: > head/sys/modules/iwnfw/iwn6000g2a/Makefile This exposed a bug in sys/conf/kmod.mk for me. I use NO_KERNELCLEAN by default when upgrading my kernels (and have for many years), but the version information in the 'FIRMWS' variable used to generate the C stub is not treated as a dependency of the generated stub. Thus, when I upgraded past this, the firmware module reused the previous C stub which used a symbol that hardcoded the previous firmware version number, so the linker saw it as an unresolved symbol and failed to load it. I don't think anything is wrong with this commit at all. I just wonder if there's a good way to make the C stub depend on the value of FIRMWS and have it be regenerated if FIRMWS changes. -- John Baldwin
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201309091337.29326.jhb>