IB/ipath: Lower default number of kernel send buffers

The default calculation for the number of send buffers to allocate to
the kernel was too high for the PCIe version of the chip thus leaving
fewer than desired send buffers for user MPI applications.

Signed-off-by: Ralph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>

diff --git a/drivers/infiniband/hw/ipath/ipath_init_chip.c b/drivers/infiniband/hw/ipath/ipath_init_chip.c
index 1b1af34..fa98aab 100644 (file)
--- a/drivers/infiniband/hw/ipath/ipath_init_chip.c
+++ b/drivers/infiniband/hw/ipath/ipath_init_chip.c
@@ -737,7 +737,7 @@ int ipath_init_chip(struct ipath_devdata *dd, int reinit)
        uports = dd->ipath_cfgports ? dd->ipath_cfgports - 1 : 0;
        if (ipath_kpiobufs == 0) {
                /* not set by user (this is default) */
-               if (piobufs >= (uports * IPATH_MIN_USER_PORT_BUFCNT) + 32)
+               if (piobufs > 144)
                        kpiobufs = 32;
                else
                        kpiobufs = 16;