summaryrefslogtreecommitdiffstats
path: root/lib/halfmd4.c
diff options
context:
space:
mode:
authorZoltan Kiss <zoltan.kiss@citrix.com>2013-09-04 21:11:05 +0100
committerKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>2013-10-02 12:53:26 -0400
commit2b2b614dd24e4e6474fcf2dcf69c95c908838959 (patch)
tree9874899c477d237e1ec207834fc730c1151088f5 /lib/halfmd4.c
parent04b772d2b819f0dda2163e3193fa7cd447a6245c (diff)
downloadlinux-2b2b614dd24e4e6474fcf2dcf69c95c908838959.tar.gz
linux-2b2b614dd24e4e6474fcf2dcf69c95c908838959.tar.bz2
linux-2b2b614dd24e4e6474fcf2dcf69c95c908838959.zip
tracing/events: Add bounce tracing to swiotbl
Ftrace is currently not able to detect when SWIOTLB has to do double buffering. Under Xen you can only see it indirectly in function_graph, when xen_swiotlb_map_page() doesn't stop after range_straddles_page_boundary(), but calls spinlock functions, memcpy() and xen_phys_to_bus() as well. This patch introduces the swiotlb:swiotlb_bounced event, which also prints out the following informations to help you find out why bouncing happened: dev_name: 0000:08:00.0 dma_mask=ffffffffffffffff dev_addr=9149f000 size=32768 swiotlb_force=0 If you use Xen, and (dev_addr + size + 1) > dma_mask, the buffer is out of the device's DMA range. If swiotlb_force == 1, you should really change the kernel parameters. Otherwise, the buffer is not contiguous in mfn space. Signed-off-by: Zoltan Kiss <zoltan.kiss@citrix.com> [v1: Don't print 'swiotlb_force=X', just print swiotlb_force if it is enabled] Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Diffstat (limited to 'lib/halfmd4.c')
0 files changed, 0 insertions, 0 deletions