| /spdk/doc/ |
| H A D | usdt.md | 6 and Fedora 33. So bpftrace should be built and installed from source. 21 replace that string with the full path name using information from procfs. 66 You will see output from the second terminal that looks like this: 73 2111.253560: nqn.2014-08.org.nvmexpress.discovery change state from INACTIVE to ACTIVE start 76 2111.284083: nqn.2014-08.org.nvmexpress.discovery change state from INACTIVE to ACTIVE done 80 2189.921492: nqn.2016-06.io.spdk:cnode1 change state from INACTIVE to ACTIVE start 83 2190.005832: nqn.2016-06.io.spdk:cnode1 change state from INACTIVE to ACTIVE done 86 2197.859104: nqn.2016-06.io.spdk:cnode1 change state from ACTIVE to PAUSED start 89 2197.902291: nqn.2016-06.io.spdk:cnode1 change state from ACTIVE to PAUSED done 90 2197.908939: nqn.2016-06.io.spdk:cnode1 change state from PAUSED to ACTIVE start [all …]
|
| H A D | backporting.md | 26 from a newer version of the repository and uploaded on top of an older one. Patch 33 * `Reviewed-on` line should contain, in parentheses, name of the branch it was pulled from, 34 * a line containing `(cherry picked from commit<original_commit_hash>)` should be added, 50 progress to a checkpoint file, from which user might resume after they are done fixing merge 60 The following example shows instructions on how DPDK would be updated from version 23.03 to 23.07, 89 ### 4. Cherry-pick patches from previous submodule fork and verify them 92 and verifying whether all patches from previous fork are necessary and if further changes should 93 be introduced to ensure compatibility. In this case, patches from DPDK submodule branch 101 * Patches from previous submodule fork are insufficient and further changes to the DPDK code are
|
| H A D | sma.md | 44 crypto engines by inheriting from the `CryptoEngine` class. 59 - `handle`: Device handle obtained from `CreateDevice`. 78 by inheriting from the `CryptoEngine` class. 79 - `device_handle`: Device handle obtained from `CreateDevice`. 83 This method detaches a volume from a device making it unavailable to the host. 89 - `device_handle`: Device handle obtained from `CreateDevice`. 99 - `handle`: Device handle obtained from `CreateDevice`. 163 - name: 'device-from-plugin1' 164 - name: 'device-from-plugin2' 172 of classes deriving from `spdk.sma.DeviceManager`. This base class define the
|
| H A D | compression.md | 20 to realize savings from the compression. In these cases, the data may be stored uncompressed on 24 device. The backing storage device must be thin-provisioned to realize any savings from 37 Storage from the backing storage device will be allocated, read, and written to in 4KB units for 38 best NVMe performance. These 4KB units are called "backing IO units". They are indexed from 0 to … 117 * Allocate 2 blocks (0 and 1) from the free backing IO unit list. Always use the lowest numbered 121 * Allocate a chunk map (0) from the free chunk map list. 154 * Allocate 1 block (2) from the free backing IO unit list. 156 * Allocate a chunk map (1) from the free chunk map list. 193 * Allocate a 16KB buffer for the uncompressed data for this chunk. Decompress the data from 198 * Allocate blocks 3 and 4 from the free backing IO unit list. [all …]
|
| H A D | userspace.md | 20 the kernel driver from the device and on Linux is done by 36 Once the device is unbound from the operating system kernel, the operating 76 requests to the device from multiple threads of execution in parallel without 78 from lots of different places either in the operating system or in various 84 temporarily disable interrupts to guard against preemption from threads 86 from older hardware interfaces that only had a single queue or no queue at 93 that a hardware queue is only ever accessed from one thread at a time. In
|
| H A D | blob.md | 42 * **Logical Block**: Logical blocks are exposed by the disk itself, which are numbered from 0 to N,… 45 …blocks that compose a page are always contiguous. Pages are also numbered from the beginning of th… 50 …are always contiguous. Clusters are also numbered from the beginning of the disk, where cluster 0 … 54 …Blobs are read and written in units of pages by specifying an offset from the start of the blob. A… 96 …all is completed. The callback will be made on the same thread that the call was made from, more on 132 With external snapshots (see @ref blob_pg_esnap_and_esnap_clone), a read from a blob may lead to 133 reading from the device containing the blobstore or an external snapshot device. To support this, 134 each blobstore IO channel maintains a tree of channels to be used when reading from external 182 * Metadata operations (API with MD in the name) should be isolated from each other as there is no i… 184 * Metadata operations should be isolated from conflicting IO operations (an example of a conflictin… [all …]
|
| H A D | concurrency.md | 6 hardware. This can mean many things in practice. For instance, moving from one 10 software's threads of execution must be independent from one another as much as 20 because you don't have to change the data model from the single-threaded 23 that you read from top to bottom. 31 the program will not benefit from more CPU cores. 121 is safe to use in SPDK libraries and applications. This safety comes from 150 it's easiest if you write the functions from bottom to top. By that we mean if 169 Don't split these functions up - keep them as a nice unit that can be read from bottom to top. 265 from top to bottom to get a clear overview of what's happening in the code
|
| H A D | event.md | 23 In the ideal case where each connection is independent from all other 38 from a queue. Each event consists of a bundled function pointer and its 51 threads from any core may insert events into the queue of any other core. The 55 directly from the event loop on the destination core.
|
| H A D | shfmt.md | 25 all .sh present in the staging area when run locally from our pre-commit 31 # Run from the root of the SPDK repo 35 The name of the patch is derived from the version of shfmt that is
|
| H A D | jsonrpc_proxy.md | 4 objects from users. It uses HTTP POST method to receive JSON objects including methods and paramete… 32 parameter and prints JSON object returned from remote_rpc script.
|
| H A D | nvmf_tgt_pg.md | 92 upon creation on the thread from which they were created. Most importantly, *a 93 poll group may only be accessed from the thread on which it was created.* 100 connections from any host or over any established listen address. Listeners 118 The SPDK NVMe-oF target is designed to be able to process I/O from multiple 132 processed on the initial thread from start to the point where it is submitted 151 For the RDMA transport, data is transferred from the RDMA NIC to host memory 153 copies. Data is never moved from one location in host memory to another. Other
|
| H A D | blobfs.md | 16 Clone the RocksDB repository from the SPDK GitHub fork into a separate directory. 44 If there are any NVMe SSDs you do not wish to use for RocksDB/SPDK testing, remove them from the co… 84 * BlobFS has primarily been tested with RocksDB so far, so any use cases different from how RocksDB…
|
| H A D | about.md | 9 and enables zero-copy access from the application. 16 parallel access directly to an SSD from a user space application. The driver is
|
| /spdk/lib/nvme/ |
| H A D | spdk_nvme.map | 4 # public functions from nvme.h 232 # functions from spdk_internal/nvme.h 239 # public functions from nvme_zns.h 261 # public functions from nvme_ocssd.h 271 # public functions from opal.h
|
| /spdk/scripts/bpf/ |
| H A D | nvmf.bt | 8 printf("%d.%06d: %s change state from %s to %s start\n", 14 printf("%d.%06d: %s change state from %s to %s %s\n", 56 printf("%d.%06d: remove qpair: %p from poll group on thread %d\n", 62 printf("%d.%06d: disconnect qpair: %p from poll group on thread %d\n", 74 printf("%d.%06d: drain qpair: %p from poll group on thread %d\n", 152 printf("%d.%06d: remove qpair: %p with id: %d from poll group on thread: %lu\n",
|
| /spdk/module/bdev/ocf/ |
| H A D | ctx.c | 220 uint64_t from, uint64_t bytes) in vbdev_ocf_ctx_data_cpy() argument 228 bytes = MIN(bytes, s->size - from); in vbdev_ocf_ctx_data_cpy() 232 while (from || bytes) { in vbdev_ocf_ctx_data_cpy() 239 if (from) { in vbdev_ocf_ctx_data_cpy() 240 n = MIN(from, s->iovs[it_iov].iov_len); in vbdev_ocf_ctx_data_cpy() 241 from -= n; in vbdev_ocf_ctx_data_cpy()
|
| /spdk/scripts/env_dpdk/22.11/ |
| H A D | 23.11-rte_bus_pci.h.patch | 24 - * Capability to be found, which can be any from 37 - * Find device's PCI capability starting from a previous offset in PCI 44 - * Capability to be found, which can be any from 47 - * An offset in the PCI configuration space from which the capability is 105 - * Read from a MMIO PCI resource.
|
| /spdk/test/app/fuzz/iscsi_fuzz/ |
| H A D | README.md | 13 Then iSCSI initiator will receive all kinds of response opcodes from iSCSI Target. 20 1. Complete successfully back from the target, or
|
| /spdk/ |
| H A D | CODE_OF_CONDUCT.md | 24 and learning from the experience 78 **Consequence**: A private, written warning from SPDK core maintainers, providing 99 **Consequence**: A temporary ban from any sort of interaction or public 111 **Consequence**: A permanent ban from any sort of public interaction within the 116 This Code of Conduct is adapted from the [Contributor Covenant][homepage],
|
| /spdk/lib/vfio_user/host/ |
| H A D | spdk_vfio_user.map | 4 # public functions from vfio_user_pci.h
|
| /spdk/lib/vfu_tgt/ |
| H A D | spdk_vfu_tgt.map | 4 # public functions from vfu_target.h
|
| /spdk/test/app/fuzz/nvme_fuzz/ |
| H A D | README.md | 19 1. Complete successfully back from the target, or 37 dumped as valid json. You can combine the dumped commands from the fuzzer into a json 43 from the fuzzer as a template.
|
| /spdk/docker/build_base/ |
| H A D | Dockerfile | 62 COPY --from=base /tmp/*.rpm /tmp/ 63 COPY --from=base /tmp/fio /tmp/
|
| /spdk/scripts/ |
| H A D | backport.sh | 30 Merged commits from 'master' branch with provided hashtag (eg. 23.05) are cherry-picked 43 local from="v${HASHTAG}-rc1" 47 git -C "$SPDK_DIR" show -s --format="%H" "$from^..HEAD" | tac > "$TMP_DIR/ordered_commits" 74 echo "(cherry picked from commit $hash)" >> "$SPDK_DIR/msg" 191 # Removing new commit to cherry-pick from the list
|
| /spdk/test/common/config/pkgdep/ |
| H A D | yum | 72 packages+=(avahi-tools) # from the devel repo 88 # FIXME: This does not fit here, but there's no sane alternative. Using RPMs from other sources will fail to meet
|