Re: [PATCH mlx5-next] RDMA/mlx5: Don't use cached IRQ affinity mask

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 7/24/2018 10:24 AM, Steve Wise wrote:
>
> On 7/19/2018 8:25 PM, Max Gurtovoy wrote:
>>>>> [ 2032.194376] nvme nvme0: failed to connect queue: 9 ret=-18
>>>> queue 9 is not mapped (overlap).
>>>> please try the bellow:
>>>>
>>> This seems to work.  Here are three mapping cases:  each vector on its
>>> own cpu, each vector on 1 cpu within the local numa node, and each
>>> vector having all cpus in its numa node.  The 2nd mapping looks kinda
>>> funny, but I think it achieved what you wanted?  And all the cases
>>> resulted in successful connections.
>>>
>> Thanks for testing this.
>> I slightly improved the setting of the left CPUs and actually used
>> Sagi's initial proposal.
>>
>> Sagi,
>> please review the attached patch and let me know if I should add your
>> signature on it.
>> I'll run some perf test early next week on it (meanwhile I run
>> login/logout with different num_queues successfully and irq settings).
>>
>> Steve,
>> It will be great if you can apply the attached in your system and send
>> your findings.
> Sorry, I got side tracked.  I'll try and test this today and report back.
>
> Steve.


###  each vector gets a unique cpu, starting with node-local:

[  754.976577] iw_cxgb4: comp_vector 0, irq 203 mask 0x100
[  754.982378] iw_cxgb4: comp_vector 1, irq 204 mask 0x200
[  754.988167] iw_cxgb4: comp_vector 2, irq 205 mask 0x400
[  754.993935] iw_cxgb4: comp_vector 3, irq 206 mask 0x800
[  754.999686] iw_cxgb4: comp_vector 4, irq 207 mask 0x1000
[  755.005509] iw_cxgb4: comp_vector 5, irq 208 mask 0x2000
[  755.011318] iw_cxgb4: comp_vector 6, irq 209 mask 0x4000
[  755.017124] iw_cxgb4: comp_vector 7, irq 210 mask 0x8000
[  755.022915] iw_cxgb4: comp_vector 8, irq 211 mask 0x1
[  755.028437] iw_cxgb4: comp_vector 9, irq 212 mask 0x2
[  755.033948] iw_cxgb4: comp_vector 10, irq 213 mask 0x4
[  755.039543] iw_cxgb4: comp_vector 11, irq 214 mask 0x8
[  755.045135] iw_cxgb4: comp_vector 12, irq 215 mask 0x10
[  755.050801] iw_cxgb4: comp_vector 13, irq 216 mask 0x20
[  755.056464] iw_cxgb4: comp_vector 14, irq 217 mask 0x40
[  755.062117] iw_cxgb4: comp_vector 15, irq 218 mask 0x80
[  755.067767] blk_mq_rdma_map_queues: set->mq_map[0] queue 8 vector 8
[  755.067767] blk_mq_rdma_map_queues: set->mq_map[1] queue 9 vector 9
[  755.067768] blk_mq_rdma_map_queues: set->mq_map[2] queue 10 vector 10
[  755.067769] blk_mq_rdma_map_queues: set->mq_map[3] queue 11 vector 11
[  755.067769] blk_mq_rdma_map_queues: set->mq_map[4] queue 12 vector 12
[  755.067770] blk_mq_rdma_map_queues: set->mq_map[5] queue 13 vector 13
[  755.067771] blk_mq_rdma_map_queues: set->mq_map[6] queue 14 vector 14
[  755.067772] blk_mq_rdma_map_queues: set->mq_map[7] queue 15 vector 15
[  755.067772] blk_mq_rdma_map_queues: set->mq_map[8] queue 0 vector 0
[  755.067773] blk_mq_rdma_map_queues: set->mq_map[9] queue 1 vector 1
[  755.067774] blk_mq_rdma_map_queues: set->mq_map[10] queue 2 vector 2
[  755.067774] blk_mq_rdma_map_queues: set->mq_map[11] queue 3 vector 3
[  755.067775] blk_mq_rdma_map_queues: set->mq_map[12] queue 4 vector 4
[  755.067775] blk_mq_rdma_map_queues: set->mq_map[13] queue 5 vector 5
[  755.067776] blk_mq_rdma_map_queues: set->mq_map[14] queue 6 vector 6
[  755.067777] blk_mq_rdma_map_queues: set->mq_map[15] queue 7 vector 7

###  each vector gets one cpu within the local node:

[  777.590913] iw_cxgb4: comp_vector 0, irq 203 mask 0x400
[  777.596588] iw_cxgb4: comp_vector 1, irq 204 mask 0x800
[  777.602249] iw_cxgb4: comp_vector 2, irq 205 mask 0x1000
[  777.607984] iw_cxgb4: comp_vector 3, irq 206 mask 0x2000
[  777.613708] iw_cxgb4: comp_vector 4, irq 207 mask 0x4000
[  777.619431] iw_cxgb4: comp_vector 5, irq 208 mask 0x8000
[  777.625142] iw_cxgb4: comp_vector 6, irq 209 mask 0x100
[  777.630762] iw_cxgb4: comp_vector 7, irq 210 mask 0x200
[  777.636373] iw_cxgb4: comp_vector 8, irq 211 mask 0x400
[  777.641982] iw_cxgb4: comp_vector 9, irq 212 mask 0x800
[  777.647583] iw_cxgb4: comp_vector 10, irq 213 mask 0x1000
[  777.653353] iw_cxgb4: comp_vector 11, irq 214 mask 0x2000
[  777.659119] iw_cxgb4: comp_vector 12, irq 215 mask 0x4000
[  777.664877] iw_cxgb4: comp_vector 13, irq 216 mask 0x8000
[  777.670628] iw_cxgb4: comp_vector 14, irq 217 mask 0x100
[  777.676289] iw_cxgb4: comp_vector 15, irq 218 mask 0x200
[  777.681946] blk_mq_rdma_map_queues: set->mq_map[0] queue 8 vector 8
[  777.681947] blk_mq_rdma_map_queues: set->mq_map[1] queue 9 vector 9
[  777.681947] blk_mq_rdma_map_queues: set->mq_map[2] queue 10 vector 10
[  777.681948] blk_mq_rdma_map_queues: set->mq_map[3] queue 11 vector 11
[  777.681948] blk_mq_rdma_map_queues: set->mq_map[4] queue 12 vector 12
[  777.681949] blk_mq_rdma_map_queues: set->mq_map[5] queue 13 vector 13
[  777.681950] blk_mq_rdma_map_queues: set->mq_map[6] queue 14 vector 14
[  777.681950] blk_mq_rdma_map_queues: set->mq_map[7] queue 15 vector 15
[  777.681951] blk_mq_rdma_map_queues: set->mq_map[8] queue 6 vector 6
[  777.681952] blk_mq_rdma_map_queues: set->mq_map[9] queue 7 vector 7
[  777.681952] blk_mq_rdma_map_queues: set->mq_map[10] queue 0 vector 0
[  777.681953] blk_mq_rdma_map_queues: set->mq_map[11] queue 1 vector 1
[  777.681953] blk_mq_rdma_map_queues: set->mq_map[12] queue 2 vector 2
[  777.681954] blk_mq_rdma_map_queues: set->mq_map[13] queue 3 vector 3
[  777.681955] blk_mq_rdma_map_queues: set->mq_map[14] queue 4 vector 4
[  777.681955] blk_mq_rdma_map_queues: set->mq_map[15] queue 5 vector 5


###  each vector gets all cpus within the local node:

[  838.251643] iw_cxgb4: comp_vector 0, irq 203 mask 0xff00
[  838.257346] iw_cxgb4: comp_vector 1, irq 204 mask 0xff00
[  838.263038] iw_cxgb4: comp_vector 2, irq 205 mask 0xff00
[  838.268710] iw_cxgb4: comp_vector 3, irq 206 mask 0xff00
[  838.274351] iw_cxgb4: comp_vector 4, irq 207 mask 0xff00
[  838.279985] iw_cxgb4: comp_vector 5, irq 208 mask 0xff00
[  838.285610] iw_cxgb4: comp_vector 6, irq 209 mask 0xff00
[  838.291234] iw_cxgb4: comp_vector 7, irq 210 mask 0xff00
[  838.296865] iw_cxgb4: comp_vector 8, irq 211 mask 0xff00
[  838.302484] iw_cxgb4: comp_vector 9, irq 212 mask 0xff00
[  838.308109] iw_cxgb4: comp_vector 10, irq 213 mask 0xff00
[  838.313827] iw_cxgb4: comp_vector 11, irq 214 mask 0xff00
[  838.319539] iw_cxgb4: comp_vector 12, irq 215 mask 0xff00
[  838.325250] iw_cxgb4: comp_vector 13, irq 216 mask 0xff00
[  838.330963] iw_cxgb4: comp_vector 14, irq 217 mask 0xff00
[  838.336674] iw_cxgb4: comp_vector 15, irq 218 mask 0xff00
[  838.342385] blk_mq_rdma_map_queues: set->mq_map[0] queue 8 vector 8
[  838.342385] blk_mq_rdma_map_queues: set->mq_map[1] queue 9 vector 9
[  838.342386] blk_mq_rdma_map_queues: set->mq_map[2] queue 10 vector 10
[  838.342387] blk_mq_rdma_map_queues: set->mq_map[3] queue 11 vector 11
[  838.342387] blk_mq_rdma_map_queues: set->mq_map[4] queue 12 vector 12
[  838.342388] blk_mq_rdma_map_queues: set->mq_map[5] queue 13 vector 13
[  838.342389] blk_mq_rdma_map_queues: set->mq_map[6] queue 14 vector 14
[  838.342390] blk_mq_rdma_map_queues: set->mq_map[7] queue 15 vector 15
[  838.342391] blk_mq_rdma_map_queues: set->mq_map[8] queue 0 vector 0
[  838.342391] blk_mq_rdma_map_queues: set->mq_map[9] queue 1 vector 1
[  838.342392] blk_mq_rdma_map_queues: set->mq_map[10] queue 2 vector 2
[  838.342392] blk_mq_rdma_map_queues: set->mq_map[11] queue 3 vector 3
[  838.342393] blk_mq_rdma_map_queues: set->mq_map[12] queue 4 vector 4
[  838.342394] blk_mq_rdma_map_queues: set->mq_map[13] queue 5 vector 5
[  838.342394] blk_mq_rdma_map_queues: set->mq_map[14] queue 6 vector 6
[  838.342395] blk_mq_rdma_map_queues: set->mq_map[15] queue 7 vector 7


--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html



[Index of Archives]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Photo]     [Yosemite News]     [Yosemite Photos]     [Linux Kernel]     [Linux SCSI]     [XFree86]

  Powered by Linux