Date: Wed, 29 Aug 2012 17:11:38 +0530 From: "Jayachandran C." <c.jayachandran@gmail.com> To: Alan Cox <alc@rice.edu> Cc: mips@freebsd.org Subject: Re: mips pmap patch Message-ID: <CA%2B7sy7C43zJxK8%2B_rMS%2BjMdRZYHqc2gSWxNAxg=ZSJwOoS3n1g@mail.gmail.com> In-Reply-To: <503AF695.9060405@rice.edu> References: <50228F5C.1000408@rice.edu> <CA%2B7sy7DxqhGhJt%2BwE3WW2-j4SxnPweULjYS5GQ=NgMYSrwJHtw@mail.gmail.com> <50269AD4.9050804@rice.edu> <CA%2B7sy7AZ-s2H6COfvz60N=kxw%2BWUiUC9diVfWg9aOzWSZKGWRQ@mail.gmail.com> <5029635A.4050209@rice.edu> <CA%2B7sy7Cnsy7Ag1iG=_Kj04gEXeYp7kZnpACQpD8THvkp0VKdcA@mail.gmail.com> <502D2271.6080105@rice.edu> <CA%2B7sy7CK=EXu88XKYYXDV1uf3U7eebq3e6rfwgHRhQyFTMv7dQ@mail.gmail.com> <503AF695.9060405@rice.edu>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, Aug 27, 2012 at 9:54 AM, Alan Cox <alc@rice.edu> wrote: > Could you please test the attached patch? For a change, this patch is an > optimization (as opposed to a bug fix). It carves off a third high-order > bit from the PTE in order to implement another software flag. For 64-bit > systems, this bit was previously unused. For 32-bit systems, this patch > effectively reduces the size of the PTE's page frame number (PFN) field by 1 > bit. However, the PFN still has more than enough bits to support the 32-bit > vm_paddr_t. > > This patch needs both 32-bit and 64-bit testing. I'm fairly confident of > the correctness of the 64-bit case, but less so of the 32-bit case. Tested this on both 32 and 64 bit configuration on XLP and no problems were seen. > Going forward, I expect to claim a fourth bit from the PTE in order to > implement a proper access bit emulation. Regards, JC
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CA%2B7sy7C43zJxK8%2B_rMS%2BjMdRZYHqc2gSWxNAxg=ZSJwOoS3n1g>