Maybe, Currently SPDK are BSD licensed. One question from me is that instead of wrote an simulated NVMe device, maybe you can run SPDK in Guest VM, we tried this in our autotest scenarios and it worked. *From:* SPDK [mailto:spdk-bounces@lists.01.org] *On Behalf Of *Crane Chu *Sent:* Monday, July 24, 2017 1:35 PM *To:* SPDK(a)lists.01.org *Subject:* [SPDK] simulated NVMe device Hi, Recently, I started to read and try SPDK source code, focusing on NVMe driver layer. However, my laptop does not have NVMe devices, and I think most of us have the same problem. So, I implemented a simulated NVMe device in env_dpdk, and now I can run some examples (hello_world, perf, identify) on it. If it is useful, I'd like to submit patches. The code is derived from QEMU NVMe device (e.g. nvme.c, nvme.h), which use GPL. Would it be a license problem in SPDK? Thanks, -Crane _______________________________________________ SPDK mailing list SPDK(a)lists.01.org https://lists.01.org/mailman/listinfo/spdk

Hi Crane, Do you still see the hang with the VirtualBox quirk we added a couple of months ago? We saw a similar issue to what you report below when we added our Vagrant setups – and this patch resolved those issues. https://github.com/spdk/spdk/commit/13f8cf1536ca5222878d2c03802fb3 67e1340ef9 Regarding your earlier question, unfortunately the GPL licensing on the NVMe emulation code would make it unsuitable for upstreaming into the SPDK code base. Thanks, -Jim *From: *SPDK <spdk-bounces(a)lists.01.org&gt; on behalf of Crane Chu < cranechu(a)gmail.com&gt; *Reply-To: *Storage Performance Development Kit <spdk(a)lists.01.org&gt; *Date: *Monday, July 24, 2017 at 1:47 AM *To: *Storage Performance Development Kit <spdk(a)lists.01.org&gt; *Subject: *Re: [SPDK] simulated NVMe device Hi, Changpeng, I used Virtualbox to test SPDK for months. It works, but not smooth. For example, in the following case, hello_word fails in the 2nd time of run, due to no CQ return. The symptom is not stable, but it always hangs sooner or later. Seems that I should file this issue to Virtualbox team. :) Personally, I prefer to code and debug in physical machine, and so ported QEMU's nvme into env_dpdk. Thanks, -Crane ===================== hello_word 2nd run fail ================================== [root@ ~/spdk] lspci 00:02.0 VGA compatible controller: InnoTek Systemberatung GmbH VirtualBox Graphics Adapter 00:03.0 Ethernet controller: Intel Corporation 82540EM Gigabit Ethernet Controller (rev 02) 00:04.0 System peripheral: InnoTek Systemberatung GmbH VirtualBox Guest Service 00:07.0 Bridge: Intel Corporation 82371AB/EB/MB PIIX4 ACPI (rev 08) 00:08.0 Ethernet controller: Intel Corporation 82540EM Gigabit Ethernet Controller (rev 02) 00:0e.0 Non-Volatile memory controller: InnoTek Systemberatung GmbH Device 4e56 <================ 00:18.0 PCI bridge: Intel Corporation 82801 Mobile PCI Bridge (rev f2) 00:19.0 PCI bridge: Intel Corporation 82801 Mobile PCI Bridge (rev f2) 00:1f.0 ISA bridge: Intel Corporation 82801GBM (ICH7-M) LPC Interface Bridge (rev 02) 00:1f.2 SATA controller: Intel Corporation 82801HM/HEM (ICH8M/ICH8M-E) SATA Controller [AHCI mode] (rev 02) 00:1f.4 USB controller: Apple Inc. KeyLargo/Intrepid USB [root@ ~/spdk] ./scripts/setup.sh 0000:00:0e.0 (80ee 4e56): nvme -> uio_pci_generic [root@ ~/spdk] cd examples/nvme/hello_world/ [root@ ~/spdk/examples/nvme/hello_world] ./hello_world Starting DPDK 17.05.0 initialization... [ DPDK EAL parameters: hello_world -c 0x1 --file-prefix=spdk0 --base-virtaddr=0x1000000000 --proc-type=auto ] EAL: Detected 2 lcore(s) EAL: Auto-detected process type: PRIMARY EAL: Probing VFIO support... Initializing NVMe Controllers EAL: PCI device 0000:00:0e.0 on NUMA socket 0 EAL: probe driver: 80ee:4e56 spdk_nvme Attaching to 0000:00:0e.0 Attached to 0000:00:0e.0 Using controller ORCL-VBOX-NVME-VER12 (VB1234-56789 ) with 1 namespaces. Namespace ID: 1 size: 8GB Initialization complete. Hello world! [root@ ~/spdk/examples/nvme/hello_world] ./hello_world Starting DPDK 17.05.0 initialization... [ DPDK EAL parameters: hello_world -c 0x1 --file-prefix=spdk0 --base-virtaddr=0x1000000000 --proc-type=auto ] EAL: Detected 2 lcore(s) EAL: Auto-detected process type: PRIMARY EAL: Probing VFIO support... Initializing NVMe Controllers EAL: PCI device 0000:00:0e.0 on NUMA socket 0 EAL: probe driver: 80ee:4e56 spdk_nvme Attaching to 0000:00:0e.0 Attached to 0000:00:0e.0 Using controller ORCL-VBOX-NVME-VER12 (VB1234-56789 ) with 1 namespaces. Namespace ID: 1 size: 8GB Initialization complete. ===================== test hangs here ================================== On Mon, Jul 24, 2017 at 1:41 PM, Liu, Changpeng <changpeng.liu(a)intel.com&gt; wrote: Maybe, Currently SPDK are BSD licensed. One question from me is that instead of wrote an simulated NVMe device, maybe you can run SPDK in Guest VM, we tried this in our autotest scenarios and it worked. *From:* SPDK [mailto:spdk-bounces@lists.01.org] *On Behalf Of *Crane Chu *Sent:* Monday, July 24, 2017 1:35 PM *To:* SPDK(a)lists.01.org *Subject:* [SPDK] simulated NVMe device Hi, Recently, I started to read and try SPDK source code, focusing on NVMe driver layer. However, my laptop does not have NVMe devices, and I think most of us have the same problem. So, I implemented a simulated NVMe device in env_dpdk, and now I can run some examples (hello_world, perf, identify) on it. If it is useful, I'd like to submit patches. The code is derived from QEMU NVMe device (e.g. nvme.c, nvme.h), which use GPL. Would it be a license problem in SPDK? Thanks, -Crane _______________________________________________ SPDK mailing list SPDK(a)lists.01.org https://lists.01.org/mailman/listinfo/spdk _______________________________________________ SPDK mailing list SPDK(a)lists.01.org https://lists.01.org/mailman/listinfo/spdk

Yes – thanks Crane! *From: *SPDK <spdk-bounces(a)lists.01.org&gt; on behalf of Paul E Luse < paul.e.luse(a)intel.com&gt; *Reply-To: *Storage Performance Development Kit <spdk(a)lists.01.org&gt; *Date: *Wednesday, July 26, 2017 at 4:13 PM *To: *Storage Performance Development Kit <spdk(a)lists.01.org&gt; *Subject: *Re: [SPDK] simulated NVMe device Awesome! *From:* SPDK [mailto:spdk-bounces@lists.01.org] *On Behalf Of *Crane Chu *Sent:* Tuesday, July 25, 2017 7:55 PM *To:* Storage Performance Development Kit <spdk(a)lists.01.org&gt; *Subject:* Re: [SPDK] simulated NVMe device With 1000us delay, the hello_world example run for 50000+ times in my Virtualbox. I have submitted a ticket to Virtualbox: https://www. virtualbox.org/ticket/16945. On Tue, Jul 25, 2017 at 2:13 PM, Crane Chu <cranechu(a)gmail.com&gt; wrote: Hi, Jim, Thanks for the information. It seems that I do meet the same issue. However, I have to enlarge the delay to 1000us. Now, it runs for 100+ loops of hello_world example app. I have not tested other delay time. I will keep 1000us delay running to meet 50000 times, and update the status later. My Virtualbox is installed in Win10 with my 5-year old laptop. So, my VirtualBox may need more time to get IO queues prepared. For the GPL license consideration, got it and understood. Thanks, -Crane On Mon, Jul 24, 2017 at 11:54 PM, Harris, James R < james.r.harris(a)intel.com&gt; wrote: Hi Crane, Do you still see the hang with the VirtualBox quirk we added a couple of months ago? We saw a similar issue to what you report below when we added our Vagrant setups – and this patch resolved those issues. https://github.com/spdk/spdk/commit/13f8cf1536ca5222878d2c03802fb3 67e1340ef9 Regarding your earlier question, unfortunately the GPL licensing on the NVMe emulation code would make it unsuitable for upstreaming into the SPDK code base. Thanks, -Jim *From: *SPDK <spdk-bounces(a)lists.01.org&gt; on behalf of Crane Chu < cranechu(a)gmail.com&gt; *Reply-To: *Storage Performance Development Kit <spdk(a)lists.01.org&gt; *Date: *Monday, July 24, 2017 at 1:47 AM *To: *Storage Performance Development Kit <spdk(a)lists.01.org&gt; *Subject: *Re: [SPDK] simulated NVMe device Hi, Changpeng, I used Virtualbox to test SPDK for months. It works, but not smooth. For example, in the following case, hello_word fails in the 2nd time of run, due to no CQ return. The symptom is not stable, but it always hangs sooner or later. Seems that I should file this issue to Virtualbox team. :) Personally, I prefer to code and debug in physical machine, and so ported QEMU's nvme into env_dpdk. Thanks, -Crane ===================== hello_word 2nd run fail ================================== [root@ ~/spdk] lspci 00:02.0 VGA compatible controller: InnoTek Systemberatung GmbH VirtualBox Graphics Adapter 00:03.0 Ethernet controller: Intel Corporation 82540EM Gigabit Ethernet Controller (rev 02) 00:04.0 System peripheral: InnoTek Systemberatung GmbH VirtualBox Guest Service 00:07.0 Bridge: Intel Corporation 82371AB/EB/MB PIIX4 ACPI (rev 08) 00:08.0 Ethernet controller: Intel Corporation 82540EM Gigabit Ethernet Controller (rev 02) 00:0e.0 Non-Volatile memory controller: InnoTek Systemberatung GmbH Device 4e56 <================ 00:18.0 PCI bridge: Intel Corporation 82801 Mobile PCI Bridge (rev f2) 00:19.0 PCI bridge: Intel Corporation 82801 Mobile PCI Bridge (rev f2) 00:1f.0 ISA bridge: Intel Corporation 82801GBM (ICH7-M) LPC Interface Bridge (rev 02) 00:1f.2 SATA controller: Intel Corporation 82801HM/HEM (ICH8M/ICH8M-E) SATA Controller [AHCI mode] (rev 02) 00:1f.4 USB controller: Apple Inc. KeyLargo/Intrepid USB [root@ ~/spdk] ./scripts/setup.sh 0000:00:0e.0 (80ee 4e56): nvme -> uio_pci_generic [root@ ~/spdk] cd examples/nvme/hello_world/ [root@ ~/spdk/examples/nvme/hello_world] ./hello_world Starting DPDK 17.05.0 initialization... [ DPDK EAL parameters: hello_world -c 0x1 --file-prefix=spdk0 --base-virtaddr=0x1000000000 --proc-type=auto ] EAL: Detected 2 lcore(s) EAL: Auto-detected process type: PRIMARY EAL: Probing VFIO support... Initializing NVMe Controllers EAL: PCI device 0000:00:0e.0 on NUMA socket 0 EAL: probe driver: 80ee:4e56 spdk_nvme Attaching to 0000:00:0e.0 Attached to 0000:00:0e.0 Using controller ORCL-VBOX-NVME-VER12 (VB1234-56789 ) with 1 namespaces. Namespace ID: 1 size: 8GB Initialization complete. Hello world! [root@ ~/spdk/examples/nvme/hello_world] ./hello_world Starting DPDK 17.05.0 initialization... [ DPDK EAL parameters: hello_world -c 0x1 --file-prefix=spdk0 --base-virtaddr=0x1000000000 --proc-type=auto ] EAL: Detected 2 lcore(s) EAL: Auto-detected process type: PRIMARY EAL: Probing VFIO support... Initializing NVMe Controllers EAL: PCI device 0000:00:0e.0 on NUMA socket 0 EAL: probe driver: 80ee:4e56 spdk_nvme Attaching to 0000:00:0e.0 Attached to 0000:00:0e.0 Using controller ORCL-VBOX-NVME-VER12 (VB1234-56789 ) with 1 namespaces. Namespace ID: 1 size: 8GB Initialization complete. ===================== test hangs here ================================== On Mon, Jul 24, 2017 at 1:41 PM, Liu, Changpeng <changpeng.liu(a)intel.com&gt; wrote: Maybe, Currently SPDK are BSD licensed. One question from me is that instead of wrote an simulated NVMe device, maybe you can run SPDK in Guest VM, we tried this in our autotest scenarios and it worked. *From:* SPDK [mailto:spdk-bounces@lists.01.org] *On Behalf Of *Crane Chu *Sent:* Monday, July 24, 2017 1:35 PM *To:* SPDK(a)lists.01.org *Subject:* [SPDK] simulated NVMe device Hi, Recently, I started to read and try SPDK source code, focusing on NVMe driver layer. However, my laptop does not have NVMe devices, and I think most of us have the same problem. So, I implemented a simulated NVMe device in env_dpdk, and now I can run some examples (hello_world, perf, identify) on it. If it is useful, I'd like to submit patches. The code is derived from QEMU NVMe device (e.g. nvme.c, nvme.h), which use GPL. Would it be a license problem in SPDK? Thanks, -Crane _______________________________________________ SPDK mailing list SPDK(a)lists.01.org https://lists.01.org/mailman/listinfo/spdk _______________________________________________ SPDK mailing list SPDK(a)lists.01.org https://lists.01.org/mailman/listinfo/spdk _______________________________________________ SPDK mailing list SPDK(a)lists.01.org https://lists.01.org/mailman/listinfo/spdk