| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <003101c6e949$f018cf40$f500030a@rtlogic.com> Date: Fri, 6 Oct 2006 07:18:39 -0600 From: "Russell Johnson" <rjohnson@...ogic.com> To: <linux-kernel@...r.kernel.org> Subject: bigphysarea mmap and direct_io My driver is allocating memory via bigphysarea and my application mmap's the area for performance reasons. With linux kernels 2.6.14 and prior on i386 platforms, I have been able to issue disk reads using direct_io to fill this bigphysarea memory and prevent multiple copies of the data. The file system code for direct I/O disk reads checks for VM_IO and will not allow direct io in that case. So to do this I had my driver clear the VM_IO flag after calling remap_pfn_range. With kernel 2.6.16 and newer, remap_pfn_range sets a new flag VM_PFNMAP which direct_io also checks. If my driver clears VM_PFNMAP then the kernel will oops upon an munmap call because it will attempt to free the pages. I'm looking for the proper way to do what I want which is to allow the application to mmap and do direct io disk access to the bigphysarea memory. I've unsuccessfully tried writing a nopage handler and I still get an oops upon munmap. Can someone point me in the right direction? Please cc me as I'm not on this list. Thanks! rjohnson@...ogic.com Russell Johnson RT Logic! - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@...r.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists