Date: Tue, 11 Feb 2003 22:33:18 -0700 From: "Justin T. Gibbs" <gibbs@scsiguy.com> To: Scott Long <scottl@FreeBSD.org>, cvs-committers@FreeBSD.org, cvs-all@FreeBSD.org Subject: Re: cvs commit: src/sys/i386/i386 busdma_machdep.c src/sys/i386/include bus_dma.h src/sys/alpha/alpha busdma_machdep.c src/sys/alpha/include bus.h src/sys/sparc64/include bus.h iommuvar.h src/sys/sparc64/pci psycho.c src/sys/sparc64/sbus ... Message-ID: <2218062704.1045027998@aslan.scsiguy.com> In-Reply-To: <200301290725.h0T7PRK6066218@repoman.freebsd.org> References: <200301290725.h0T7PRK6066218@repoman.freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
> scottl 2003/01/28 23:25:27 PST > > Modified files: > sys/i386/i386 busdma_machdep.c > sys/i386/include bus_dma.h > sys/alpha/alpha busdma_machdep.c > sys/alpha/include bus.h > sys/sparc64/include bus.h iommuvar.h > sys/sparc64/pci psycho.c > sys/sparc64/sbus sbus.c > sys/sparc64/sparc64 bus_machdep.c iommu.c > Log: > Implement bus_dmamem_alloc_size() and bus_dmamem_free_size() as ... The commit message should have also indicated that bus_dmamem_*_size() are mearly crutches until the standard bus_dmamap_load() api on all architectures learns how to compact mappings to fit within the S/G requirements of the caller. Having an S/G list size of 1 is just a special case of this requirement. For the common case where more "static" mappings are required, bus_dmamem_alloc()/free() should still be used. Hopefully this crutch won't last very long as it just perpetuates bad usage of the bus dma api that will have to be cleaned up by, in all likelyhood, someone other than the driver/module maintainer that uses this API. -- Justin To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-all" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?2218062704.1045027998>