Commit 0ebd0dce authored by Maciej Wawryk's avatar Maciej Wawryk Committed by Tomasz Zawadzki
Browse files

Add subsystem division in Kernel initiator 



This will allow us to use num_cores and cpus_allowed option also
in Kernel initiator side.

Signed-off-by: default avatarMaciej Wawryk <maciejx.wawryk@intel.com>
Change-Id: Iad4e8bf7cd8a614ada8f443a8b1036dc2c4803e4
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/1173


Tested-by: default avatarSPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: default avatarShuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: default avatarAleksey Marchuk <alexeymar@mellanox.com>
Reviewed-by: default avatarKarol Latecki <karol.latecki@intel.com>
parent 0eab3515

scripts/perf/nvmf/run_nvmf.py

+16 −4
Original line number Diff line number Diff line
@@ -665,10 +665,22 @@ class KernelInitiator(Initiator): 
        nvme_list = [x for x in out.split("\n") if "nvme" in x]



        filename_section = ""

        for i, nvme in enumerate(nvme_list):

            filename_section = "\n".join([filename_section,

                                          "[filename%s]" % i,

                                          "filename=%s" % nvme])

        filenames = ["nvme%sn1" % x for x in range(0, len(nvme_list))]

        nvme_per_split = int(len(nvme_list) / threads)

        remainder = len(nvme_list) % threads

        iterator = iter(filenames)

        result = []

        for i in range(threads):

            result.append([])

            for j in range(nvme_per_split):

                result[i].append(next(iterator))

                if remainder:

                    result[i].append(next(iterator))

                    remainder -= 1

        for i, r in enumerate(result):

            header = "[filename%s]" % i

            disks = "\n".join(["filename=/dev/%s" % x for x in r])

            filename_section = "\n".join([filename_section, header, disks])



        return filename_section