- 28 3月, 2020 20 次提交
-
-
由 Trond Myklebust 提交于
If the FLUSH_SYNC flag is set, nfs_initiate_pgio() will currently wait for completion, and then return the status of the I/O operation. What we actually want to report in nfs_pageio_doio() is whether or not the RPC call was launched successfully, whereas actual I/O status is intended handled in the reply callbacks. Since FLUSH_SYNC is never set by any of the callers anyway, let's just remove that code altogether. Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
Move from requesting only full file layout segments, to requesting layout segments that match our I/O size. This means the server is still free to return a full file layout, but we will no longer error out if it does not. Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
Remove the requirement that the server always sends whole file layouts. Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
When starting to read or write with a layout segment, check that the range matches our request. Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
Check that the number of mirrors, and the mirror information matches before deciding to merge layout segments in pNFS/flexfiles. Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
Fix up pnfs_layout_mark_request_commit() to alway reschedule the write if the layout segment is invalid. Also minor cleanup. Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
Move the pNFS commit related operations into a separate structure that can be carried by the pnfs_ds_commit_info. Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
Remove the unused bucket array in struct pnfs_ds_commit_info. Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
Lift filelayout_search_commit_reqs() into the generic pnfs/nfs code, and add support for commit arrays. Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
Enable adding and lookup of per-layout segment commits in filelayout and flexfilelayout. Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
Ensure that both the file and flexfiles layout types clean up when freeing the layout segments. Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
Add support for scanning the full list of per-layout segment commit arrays to nfs_clear_pnfs_ds_commit_verifiers(). Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
Instead of trying to save the commit verifiers and checking them against previous writes, adopt the same strategy as for buffered writes, of just checking the verifiers at commit time. Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
Fix the O_DIRECT code to avoid retries if the COMMIT fails with a fatal error. Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
Add a pNFS callback to allow the O_DIRECT code to release the DS commitinfo when freeing the dreq. Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
Add support for scanning the full list of per-layout segment commit arrays to pnfs_generic_commit_pagelist(). Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
Add support for scanning the full list of per-layout segment commit arrays to pnfs_generic_recover_commit_reqs(). Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
Add support for scanning the full list of per-layout segment commit arrays to pnfs_generic_scan_commit_lists() Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
When we have multiple layout segments with different lists of mirrored data, we need to track the commits on a per layout segment basis. This patch adds a list to support this tracking in struct pnfs_ds_commit_info. Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
- 26 3月, 2020 3 次提交
-
-
由 Trond Myklebust 提交于
Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
Refactor pnfs_generic_commit_pagelist() to simplify the conversion to layout segment based commit lists. Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
Just allocate the array at the end of the layout segment structure, instead of allocating it as a separate array of pointers. Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
- 25 3月, 2020 2 次提交
-
-
由 Petr Vorel 提交于
UDP was originally disabled in 6da1a034 for NFSv4. Later in b24ee6c6 UDP is by default disabled by NFS_DISABLE_UDP_SUPPORT=y for all NFS versions. Therefore remove v4 from error message. Fixes: b24ee6c6 ("NFS: allow deprecation of NFS UDP protocol") Signed-off-by: NPetr Vorel <pvorel@suse.cz> Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Liwei Song 提交于
UDP is disabled by default in commit b24ee6c6 ("NFS: allow deprecation of NFS UDP protocol"), but the default mount options is still udp, change it to tcp to avoid the "Unsupported transport protocol udp" error if no protocol is specified when mount nfs. Fixes: b24ee6c6 ("NFS: allow deprecation of NFS UDP protocol") Signed-off-by: NLiwei Song <liwei.song@windriver.com> Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
- 23 3月, 2020 1 次提交
-
-
由 Misono Tomohiro 提交于
When dreq is allocated by nfs_direct_req_alloc(), dreq->kref is initialized to 2. Therefore we need to call nfs_direct_req_release() twice to release the allocated dreq. Usually it is called in nfs_file_direct_{read, write}() and nfs_direct_complete(). However, current code only calls nfs_direct_req_relese() once if nfs_get_lock_context() fails in nfs_file_direct_{read, write}(). So, that case would result in memory leak. Fix this by adding the missing call. Signed-off-by: NMisono Tomohiro <misono.tomohiro@jp.fujitsu.com> Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
- 18 3月, 2020 1 次提交
-
-
由 Trond Myklebust 提交于
Fallout from the mount patches. Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
- 16 3月, 2020 13 次提交
-
-
由 Gustavo A. R. Silva 提交于
The current codebase makes use of the zero-length array language extension to the C90 standard, but the preferred mechanism to declare variable-length types such as these ones is a flexible array member[1][2], introduced in C99: struct foo { int stuff; struct boo array[]; }; By making use of the mechanism above, we will get a compiler warning in case the flexible array does not occur last in the structure, which will help us prevent some kind of undefined behavior bugs from being inadvertently introduced[3] to the codebase from now on. Also, notice that, dynamic memory allocations won't be affected by this change: "Flexible array members have incomplete type, and so the sizeof operator may not be applied. As a quirk of the original implementation of zero-length arrays, sizeof evaluates to zero."[1] This issue was found with the help of Coccinelle. [1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html [2] https://github.com/KSPP/linux/issues/21 [3] commit 76497732 ("cxgb3/l2t: Fix undefined behaviour") Signed-off-by: NGustavo A. R. Silva <gustavo@embeddedor.com> Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Murphy Zhou 提交于
This fixes xfstests generic/356 failure on NFSv4.2. Signed-off-by: NMurphy Zhou <jencce.kernel@gmail.com> Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Zhouyi Zhou 提交于
In function nfs_permission: 1. the rcu_read_lock and rcu_read_unlock around nfs_do_access is unnecessary because the rcu critical data structure is already protected in subsidiary function nfs_access_get_cached_rcu. No other data structure needs rcu_read_lock in nfs_do_access. 2. call nfs_do_access once is enough, because: 2-1. when mask has MAY_NOT_BLOCK bit The second call to nfs_do_access will not happen. 2-2. when mask has no MAY_NOT_BLOCK bit The second call to nfs_do_access will happen if res == -ECHILD, which means the first nfs_do_access goes out after statement if (!may_block). The second call to nfs_do_access will go through this procedure once again except continue the work after if (!may_block). But above work can be performed by only one call to nfs_do_access without mangling the mask flag. Tested in x86_64 Signed-off-by: NZhouyi Zhou <zhouzhouyi@gmail.com> Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
When we receive a CB_RECALL_ANY that asks us to return flexfiles layouts, we iterate through all the layouts and look at whether or not there are active open file descriptors that might need them for I/O. If there are no such descriptors, we return the layouts. Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
Convert to use nfs_client_for_each_server() for efficiency. Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
Convert nfs_delegation_reap_unclaimed() to use nfs_client_for_each_server() for efficiency. Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
Convert it to use the nfs_client_for_each_server() helper, and make it more efficient by skipping delegations for inodes we know are in the process of being freed. Also improve the efficiency of the cursor by skipping delegations that are being freed. Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
Add a helper nfs_client_for_each_server() to iterate through all the filesystems that are attached to a struct nfs_client, and apply a function to all the active ones. Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
Now that we can rely on just the rcu_read_lock(), remove the clp->cl_lock and clean up. Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
Make sure to test the stateid for validity so that we catch instances where the server may have been reusing stateids in nfs_layout_find_inode_by_stateid(). Fixes: 7b410d9c ("pNFS: Delay getting the layout header in CB_LAYOUTRECALL handlers") Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
Ensure that if the DS is returning too many DELAY and GRACE errors, we also report that to the MDS through the layouterror mechanism. Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-
由 Trond Myklebust 提交于
Currently, we have no real limit on the access cache size (we set it to ULONG_MAX). That can lead to credentials getting pinned for a very long time on lots of files if you have a system with a lot of memory. Signed-off-by: NTrond Myklebust <trond.myklebust@hammerspace.com>
-