[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <DB1899F3-88A0-44A2-8F44-A380D625A98F@bytedance.com>
Date: Mon, 6 Sep 2021 15:53:26 +0800
From: Junji Wei <weijunji@...edance.com>
To: Zhu Yanjun <zyjzyj2000@...il.com>
Cc: Doug Ledford <dledford@...hat.com>, Jason Gunthorpe <jgg@...pe.ca>,
RDMA mailing list <linux-rdma@...r.kernel.org>,
LKML <linux-kernel@...r.kernel.org>, xieyongji@...edance.com
Subject: Re: [PATCH] RDMA/rxe: Fix wrong port_cap_flags
> On Sep 6, 2021, at 3:21 PM, Zhu Yanjun <zyjzyj2000@...il.com> wrote:
>
> On Tue, Aug 31, 2021 at 4:32 PM Junji Wei <weijunji@...edance.com> wrote:
>>
>> The port->attr.port_cap_flags should be set to enum
>> ib_port_capability_mask_bits in ib_mad.h,
>> not RDMA_CORE_CAP_PROT_ROCE_UDP_ENCAP.
>>
>> Signed-off-by: Junji Wei <weijunji@...edance.com>
>> ---
>> drivers/infiniband/sw/rxe/rxe_param.h | 2 +-
>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/drivers/infiniband/sw/rxe/rxe_param.h b/drivers/infiniband/sw/rxe/rxe_param.h
>> index 742e6ec93686..b5a70cbe94aa 100644
>> --- a/drivers/infiniband/sw/rxe/rxe_param.h
>> +++ b/drivers/infiniband/sw/rxe/rxe_param.h
>> @@ -113,7 +113,7 @@ enum rxe_device_param {
>> /* default/initial rxe port parameters */
>> enum rxe_port_param {
>> RXE_PORT_GID_TBL_LEN = 1024,
>> - RXE_PORT_PORT_CAP_FLAGS = RDMA_CORE_CAP_PROT_ROCE_UDP_ENCAP,
>> + RXE_PORT_PORT_CAP_FLAGS = IB_PORT_CM_SUP,
>
> RXE_PORT_PORT_CAP_FLAGS = IB_PORT_CM_SUP |
> RDMA_CORE_CAP_PROT_ROCE_UDP_ENCAP,
>
> Is it better?
>
> Zhu Yanjun
I don’t think so.
Because RDMA_CORE_CAP_PROT_ROCE_UDP_ENCAP(0x800000)
means IB_PORT_BOOT_MGMT_SUP(1 << 23) in ib_mad.h.
RDMA_CORE_CAP_PROT_ROCE_UDP_ENCAP should be used for
port’s core_cap_flags.
>
>> RXE_PORT_MAX_MSG_SZ = 0x800000,
>> RXE_PORT_BAD_PKEY_CNTR = 0,
>> RXE_PORT_QKEY_VIOL_CNTR = 0,
>> --
>> 2.30.1 (Apple Git-130)
>>
Powered by blists - more mailing lists