Re: [PATCH 1/2] mm/vmalloc: export is_vmalloc_or_module_addr

On 02/08/2014 02:45 PM, Linus Torvalds wrote:

On Sat, Feb 8, 2014 at 11:12 AM, Richard Yao [off-list ref] wrote:
And what's the backtrace that gets mentioned - but not quoted - for
the horrible 9p crap? So that we can see who the guilty party is that
thinks that it's somehow ok to pass module addresses off to other
kernel users..

My apologies for that. Here is the backtrace:

[<ffffffff814878ce>] p9_virtio_zc_request+0x45e/0x510
[<ffffffff814814ed>] p9_client_zc_rpc.constprop.16+0xfd/0x4f0
[<ffffffff814839dd>] p9_client_read+0x15d/0x240
[<ffffffff811c8440>] v9fs_fid_readn+0x50/0xa0
[<ffffffff811c84a0>] v9fs_file_readn+0x10/0x20
[<ffffffff811c84e7>] v9fs_file_read+0x37/0x70
[<ffffffff8114e3fb>] vfs_read+0x9b/0x160
[<ffffffff81153571>] kernel_read+0x41/0x60
[<ffffffff810c83ab>] copy_module_from_fd.isra.34+0xfb/0x180
[<ffffffff810cc420>] SyS_finit_module+0x70/0xd0
[<ffffffff814a08fd>] system_call_fastpath+0x1a/0x1f
[<ffffffffffffffff>] 0xffffffffffffffff

The addresses are being passed to permit zero-copy between the guest and
the host. You can see this in the original thread back from December,
which is from before I knew how to submit patches properly:

http://www.spinics.net/lists/linux-virtualization/msg21734.html

Please let's just fix the real problem, don't add more horridness on
top of it because somebody messed up earlier.

This code avoids a copy between QEMU's 9p server and a guest Linux
kernel. Without it, the 9p-virtio code must allocate additional buffers
into which QEMU will write the data and then copy from those buffers
into the actual kernel buffers. How would you like 9p-virtio to do this?