Subject : build issue #549 for v2.6.27-rc2 :ixj.c:(.text+0x172b89): undefined reference to `pnp_get_resource' Submitter : Toralf Förster <toralf.foerster@gmx.de> Date : 2008-08-06 17:18 References : http://marc.info/?l=linux-kernel&m=121804329014332&w=4 This entry is being used for tracking a regression from 2.6.26. Please don't close it until the problem is fixed in the mainline.
Submitter : Randy Dunlap <randy.dunlap@oracle.com> References : http://lkml.org/lkml/2008/7/22/353 http://lkml.org/lkml/2008/7/30/189 Handled-By : Bjorn Helgaas <bjorn.helgaas@hp.com> Patch : http://lkml.org/lkml/2008/7/22/364 Caused by: commit 3f9b5cc018566ad9562df0648395649aebdbc5e0 Author: Ingo Molnar <mingo@elte.hu> Date: Fri Jul 18 16:30:05 2008 +0200 x86: re-enable OPTIMIZE_INLINING re-enable OPTIMIZE_INLINING more widely. Jeff Dike fixed the remaining outstanding issue in this commit:
On Friday, 12 of September 2008, Randy Dunlap wrote: > Rafael J. Wysocki wrote: > > This message has been generated automatically as a part of a report > > of recent regressions. > > > > The following bug entry is on the current list of known regressions > > from 2.6.26. Please verify if it still should be listed and let me know > > (either way). > > > > > > Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=11276 > > Subject : build error: CONFIG_OPTIMIZE_INLINING=y causes gcc > 4.2 to do stupid things > > Submitter : Randy Dunlap <randy.dunlap@oracle.com> > > Date : 2008-08-06 17:18 (38 days old) > > References : http://marc.info/?l=linux-kernel&m=121804329014332&w=4 > > http://lkml.org/lkml/2008/7/22/353 > > Handled-By : Bjorn Helgaas <bjorn.helgaas@hp.com> > > Patch : http://lkml.org/lkml/2008/7/22/364 > > Yes, I still see this build error.
The way these PNP drivers deal with the CONFIG_PNP dependency (they essentialy don't deal with it and depend upon some levels of optimization when compiling the inline functions in linux/pnp.h) is way too fragile to live. Either we should add proper selects in the Kconfig files for these drivers using the pnp_*() interfaces, or pnp.h should be refactored so that it works no matter how the compiler decides to optimize or not optimize constant compile time conditions.
Created attachment 17816 [details] Fix for build problem. This is the only sane way to fix this problem, if we are going to provide dummy interfaces for PNP even when CONFIG_PNP is disabled, we have to do so for all such referenced interfaces.
fixed by commit ef3d7714f6b75b51825ad0384b5ce48358427e50