bisect log

bisect.log (text/x-log), 2.91 KB, created by DaggyStyle on 2021-02-05 09:36:13 UTC

4fb044562641e742ff558811ffbd0c7fa8da8a05 was both good and bad
4fb044562641e742ff558811ffbd0c7fa8da8a05 was both good and bad
Bisecting: 102 revisions left to test after this (roughly 7 steps)
[ca431352900a2d0e86bc298551fb090a39e5cf42] gpio: sifive: select IRQ_DOMAIN_HIERARCHY rather than depend on it
Bisecting: 50 revisions left to test after this (roughly 6 steps)
[a8749dfcb892e7a42a9eeb8b1dc402a3d6188dbd] clk: tegra30: Add hda clock default rates to clock driver
Bisecting: 25 revisions left to test after this (roughly 5 steps)
[5a5095ac9e0b716ece6b13a6a760cef2396a4af7] dm: avoid filesystem lookup in dm_get_dev_t()
Bisecting: 12 revisions left to test after this (roughly 4 steps)
[e03bbc55b14905b6f1a11bd9ee5032e23963a692] HID: sony: select CONFIG_CRC32
Bisecting: 6 revisions left to test after this (roughly 3 steps)
[7f8049df7c5094f9913cd49a8b56999321ea3683] drm/amd/display: DCN2X Find Secondary Pipe properly in MPO + ODM Case
Bisecting: 3 revisions left to test after this (roughly 2 steps)
[14fe083fd052efed08a565bf79e177d6cd156135] ASoC: Intel: haswell: Add missing pm_ops
Bisecting: 1 revision left to test after this (roughly 1 step)
[00ee972739fb2526d3936f1e7ccfc8c91d250c60] SUNRPC: Handle TCP socket sends with kernel_sendpage() again
Bisecting: 0 revisions left to test after this (roughly 0 steps)
[ae3e2f34b30d6ec01e246aef06f1574ba002ec97] ASoC: rt711: mutex between calibration and power state changes
00ee972739fb2526d3936f1e7ccfc8c91d250c60 is the first bad commit
commit 00ee972739fb2526d3936f1e7ccfc8c91d250c60
Author: Chuck Lever <[e-mail]>
Date:   Fri Dec 18 12:28:41 2020 -0500

    SUNRPC: Handle TCP socket sends with kernel_sendpage() again
    
    [ Upstream commit 4a85a6a3320b4a622315d2e0ea91a1d2b013bce4 ]
    
    Daire Byrne reports a ~50% aggregrate throughput regression on his
    Linux NFS server after commit da1661b93bf4 ("SUNRPC: Teach server to
    use xprt_sock_sendmsg for socket sends"), which replaced
    kernel_send_page() calls in NFSD's socket send path with calls to
    sock_sendmsg() using iov_iter.
    
    Investigation showed that tcp_sendmsg() was not using zero-copy to
    send the xdr_buf's bvec pages, but instead was relying on memcpy.
    This means copying every byte of a large NFS READ payload.
    
    It looks like TLS sockets do indeed support a ->sendpage method,
    so it's really not necessary to use xprt_sock_sendmsg() to support
    TLS fully on the server. A mechanical reversion of da1661b93bf4 is
    not possible at this point, but we can re-implement the server's
    TCP socket sendmsg path using kernel_sendpage().
    
    Reported-by: Daire Byrne <[e-mail]>
    BugLink: https://bugzilla.kernel.org/show_bug.cgi?id=209439
    Signed-off-by: Chuck Lever <[e-mail]>
    Signed-off-by: Sasha Levin <[e-mail]>

 net/sunrpc/svcsock.c | 86 +++++++++++++++++++++++++++++++++++++++++++++++++++-
 1 file changed, 85 insertions(+), 1 deletion(-)