Subject: port-sparc64/23342: sparc64 clock.c depends on mc146818
To: None <gnats-bugs@gnats.netbsd.org>
From: None <he@netbsd.org>
List: netbsd-bugs
Date: 11/02/2003 11:08:10
>Number:         23342
>Category:       port-sparc64
>Synopsis:       sparc64 clock.c depends on mc146818
>Confidential:   no
>Severity:       non-critical
>Priority:       medium
>Responsible:    port-sparc64-maintainer
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Sun Nov 02 10:15:00 UTC 2003
>Closed-Date:
>Last-Modified:
>Originator:     Havard Eidnes
>Release:        NetBSD 1.6ZE
>Organization:
>Environment:
System: NetBSD mizar.urc.uninett.no 1.6ZE NetBSD 1.6ZE (MIZAR) #17: Sun Nov 2 10:31:32 CET 2003 he@mizar.urc.uninett.no:/usr/obj/sys/arch/sparc64/compile/MIZAR sparc64
Architecture: sparc64
Machine: sparc64
>Description:
	After the recent reshuffling of the clock chip drivers,
	the sparc64 clock.c driver depends on the presence of
	the mc146818 driver.  This caused a build problem for
	kernels configured for e.g. Ultra2 systems, where 'rtc at ebus'
	is not required (which otherwise pulls in the mc146818 driver).

	The build problem was fixed by files.sparc64 revision 1.86,
	but the clock driver should probably be somewhat restructured.

>How-To-Repeat:
	Try to revert files.sparc64 to 1.85, and watch the linking of
	the kernel fail with "undefined refrerence to `mc146818_attach'
	when 'rtc* at ebus?' is not present in the configuration file.

>Fix:
	Worked around by files.sparc64 revision 1.86, but this should
	probably be fixed in a better way.
>Release-Note:
>Audit-Trail:
>Unformatted: