Date: Sat, 20 Sep 2003 16:30:23 -0500 From: "Alan L. Cox" <alc@imimic.com> To: Jake Burkholder <jake@locore.ca> Cc: cvs-all@FreeBSD.org Subject: Re: cvs commit: src/sys/sparc64/sparc64 pmap.c Message-ID: <3F6CC6EF.52583B04@imimic.com> References: <200309201700.h8KH0xN0092997@repoman.freebsd.org> <3F6C9DD5.289EBC11@imimic.com> <20030920211524.GB15802@locore.ca>
next in thread | previous in thread | raw e-mail | index | archive | help
Jake Burkholder wrote: > > Apparently, On Sat, Sep 20, 2003 at 01:35:01PM -0500, > Alan L. Cox said words to the effect of; > > > Jake Burkholder wrote: > > > > > > jake 2003/09/20 10:00:59 PDT > > > > > > FreeBSD src repository > > > > > > Modified files: > > > sys/sparc64/sparc64 pmap.c > > > Log: > > > Remove an invalid KASSERT. Apparently pmap_remove_all gets called on > > > unmanaged pages. > > > > > > Revision Changes Path > > > 1.124 +0 -3 src/sys/sparc64/sparc64/pmap.c > > > > FYI, this is, perhaps, triggered by a recent change that I made to > > kmem_malloc(), which declares the allocated pages as unmanaged. > > Ahh, ok, thanks, I wondered. The i386 pmap has similar assertions but > they seem to never be compiled. > In case it's not obvious, my rationale was to stop the creation of "pv" entries for kmem_map (including mbuf-related) pages. They serve no purpose. Regards, Alan P.S. In fact, I would argue that changing the type of the kernel and kmem vm objects to type PHYS would make a lot of sense. Today, it wouldn't have any functional effect, but it better describes the kind of object they are.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3F6CC6EF.52583B04>