QA output created by 745 Testing behaviour with no nvidia library available === std out === nvidia.numcards PMID: 120.0.0 [Number of Graphics Cards] Data Type: 32-bit unsigned int InDom: PM_INDOM_NULL 0xffffffff Semantics: instant Units: none Help: The number of NVIDIA Graphics cards installed in this system value 0 nvidia.gpuid PMID: 120.0.1 [GPU ID] Data Type: 32-bit unsigned int InDom: 120.0 0x1e000000 Semantics: instant Units: none Help: Zero indexed id of this NVIDIA card No value(s) available! nvidia.cardname PMID: 120.0.2 [GPU Name] Data Type: string InDom: 120.0 0x1e000000 Semantics: instant Units: none Help: The name of the graphics card No value(s) available! nvidia.busid PMID: 120.0.3 [Card Bus ID] Data Type: string InDom: 120.0 0x1e000000 Semantics: instant Units: none Help: The Bus ID as reported by the NVIDIA tools, not lspci No value(s) available! nvidia.temp PMID: 120.0.4 [The temperature of the card] Data Type: 32-bit unsigned int InDom: 120.0 0x1e000000 Semantics: instant Units: none Help: The Temperature of the GPU on the NVIDIA card in degrees celcius. No value(s) available! nvidia.fanspeed PMID: 120.0.5 [Fanspeed] Data Type: 32-bit unsigned int InDom: 120.0 0x1e000000 Semantics: instant Units: none Help: Speed of the GPU fan as a percentage of the maximum No value(s) available! nvidia.perfstate PMID: 120.0.6 [NVIDIA performance state] Data Type: 32-bit unsigned int InDom: 120.0 0x1e000000 Semantics: instant Units: none Help: The PX performance state as reported from NVML. Value is an integer which should range from 0 (maximum performance) to 15 (minimum). If the state is unknown the reported value will be 32, however. No value(s) available! nvidia.gpuactive PMID: 120.0.7 [Percentage of GPU utilization] Data Type: 32-bit unsigned int InDom: 120.0 0x1e000000 Semantics: instant Units: none Help: Percentage of time over the past sample period during which one or more kernels was executing on the GPU. No value(s) available! nvidia.memactive PMID: 120.0.8 [Percentage of time spent accessing memory] Data Type: 32-bit unsigned int InDom: 120.0 0x1e000000 Semantics: instant Units: none Help: Percent of time over the past sample period during which global (device) memory was being read or written. This metric shows if the memory is actively being accesed, and is not correlated to storage amount used. No value(s) available! nvidia.memused PMID: 120.0.9 [Allocated FB memory] Data Type: 64-bit unsigned int InDom: 120.0 0x1e000000 Semantics: instant Units: byte Help: Amount of GPU FB memory that has currently been allocated, in bytes. Note that the driver/GPU always sets aside a small amount of memory for bookkeeping. No value(s) available! nvidia.memtotal PMID: 120.0.10 [Total FB memory available] Data Type: 64-bit unsigned int InDom: 120.0 0x1e000000 Semantics: discrete Units: byte Help: The total amount of GPU FB memory available on the card, in bytes. No value(s) available! nvidia.memfree PMID: 120.0.11 [Unallocated FB memory] Data Type: 64-bit unsigned int InDom: 120.0 0x1e000000 Semantics: instant Units: byte Help: Amount of GPU FB memory that is not currently allocated, in bytes. No value(s) available! === std err === pminfo[PID] Info: NVIDIA NVML library currently unavailable === filtered valgrind report === Memcheck, a memory error detector Command: pminfo -L -K clear -K add,120,nvidia/pmda_nvidia.DSO,nvidia_init -dfmtT -n nvidia/root nvidia LEAK SUMMARY: definitely lost: 0 bytes in 0 blocks indirectly lost: 0 bytes in 0 blocks ERROR SUMMARY: 0 errors from 0 contexts ... Testing behaviour with QA wrapper nvidia library === std out === nvidia.numcards PMID: 120.0.0 [Number of Graphics Cards] Data Type: 32-bit unsigned int InDom: PM_INDOM_NULL 0xffffffff Semantics: instant Units: none Help: The number of NVIDIA Graphics cards installed in this system value 2 nvidia.gpuid PMID: 120.0.1 [GPU ID] Data Type: 32-bit unsigned int InDom: 120.0 0x1e000000 Semantics: instant Units: none Help: Zero indexed id of this NVIDIA card inst [0 or "gpu0"] value 0 inst [1 or "gpu1"] value 1 nvidia.cardname PMID: 120.0.2 [GPU Name] Data Type: string InDom: 120.0 0x1e000000 Semantics: instant Units: none Help: The name of the graphics card inst [0 or "gpu0"] value "GeForce 100M Series" inst [1 or "gpu1"] value "Quadro FX 200M Series" nvidia.busid PMID: 120.0.3 [Card Bus ID] Data Type: string InDom: 120.0 0x1e000000 Semantics: instant Units: none Help: The Bus ID as reported by the NVIDIA tools, not lspci inst [0 or "gpu0"] value "0:1:0x2:3:4" inst [1 or "gpu1"] value "20:21:0x2:23:24" nvidia.temp PMID: 120.0.4 [The temperature of the card] Data Type: 32-bit unsigned int InDom: 120.0 0x1e000000 Semantics: instant Units: none Help: The Temperature of the GPU on the NVIDIA card in degrees celcius. inst [0 or "gpu0"] value 6 inst [1 or "gpu1"] value 26 nvidia.fanspeed PMID: 120.0.5 [Fanspeed] Data Type: 32-bit unsigned int InDom: 120.0 0x1e000000 Semantics: instant Units: none Help: Speed of the GPU fan as a percentage of the maximum inst [0 or "gpu0"] value 5 inst [1 or "gpu1"] value 25 nvidia.perfstate PMID: 120.0.6 [NVIDIA performance state] Data Type: 32-bit unsigned int InDom: 120.0 0x1e000000 Semantics: instant Units: none Help: The PX performance state as reported from NVML. Value is an integer which should range from 0 (maximum performance) to 15 (minimum). If the state is unknown the reported value will be 32, however. inst [0 or "gpu0"] value 9 inst [1 or "gpu1"] value 29 nvidia.gpuactive PMID: 120.0.7 [Percentage of GPU utilization] Data Type: 32-bit unsigned int InDom: 120.0 0x1e000000 Semantics: instant Units: none Help: Percentage of time over the past sample period during which one or more kernels was executing on the GPU. inst [0 or "gpu0"] value 7 inst [1 or "gpu1"] value 27 nvidia.memactive PMID: 120.0.8 [Percentage of time spent accessing memory] Data Type: 32-bit unsigned int InDom: 120.0 0x1e000000 Semantics: instant Units: none Help: Percent of time over the past sample period during which global (device) memory was being read or written. This metric shows if the memory is actively being accesed, and is not correlated to storage amount used. inst [0 or "gpu0"] value 8 inst [1 or "gpu1"] value 28 nvidia.memused PMID: 120.0.9 [Allocated FB memory] Data Type: 64-bit unsigned int InDom: 120.0 0x1e000000 Semantics: instant Units: byte Help: Amount of GPU FB memory that has currently been allocated, in bytes. Note that the driver/GPU always sets aside a small amount of memory for bookkeeping. inst [0 or "gpu0"] value 104857600 inst [1 or "gpu1"] value 6442450944 nvidia.memtotal PMID: 120.0.10 [Total FB memory available] Data Type: 64-bit unsigned int InDom: 120.0 0x1e000000 Semantics: discrete Units: byte Help: The total amount of GPU FB memory available on the card, in bytes. inst [0 or "gpu0"] value 268435456 inst [1 or "gpu1"] value 8589934592 nvidia.memfree PMID: 120.0.11 [Unallocated FB memory] Data Type: 64-bit unsigned int InDom: 120.0 0x1e000000 Semantics: instant Units: byte Help: Amount of GPU FB memory that is not currently allocated, in bytes. inst [0 or "gpu0"] value 163577856 inst [1 or "gpu1"] value 2147483648 === std err === pminfo[PID] Info: Successfully loaded NVIDIA NVML library === filtered valgrind report === Memcheck, a memory error detector Command: pminfo -L -K clear -K add,120,nvidia/pmda_nvidia.DSO,nvidia_init -dfmtT -n nvidia/root nvidia LEAK SUMMARY: definitely lost: 0 bytes in 0 blocks indirectly lost: 0 bytes in 0 blocks ERROR SUMMARY: 0 errors from 0 contexts ...