Whamcloud - gitweb
LU-9679 modules: convert MIN/MAX to kernel style
[fs/lustre-release.git] / lustre / mdc / mdc_locks.c
index 86b0e67..574d09b 100644 (file)
@@ -23,7 +23,7 @@
  * Copyright (c) 2003, 2010, Oracle and/or its affiliates. All rights reserved.
  * Use is subject to license terms.
  *
- * Copyright (c) 2011, 2016, Intel Corporation.
+ * Copyright (c) 2011, 2017, Intel Corporation.
  */
 /*
  * This file is part of Lustre, http://www.lustre.org/
@@ -43,6 +43,7 @@
 #include <lustre_net.h>
 #include <lustre_req_layout.h>
 #include <lustre_swab.h>
+#include <lustre_acl.h>
 
 #include "mdc_internal.h"
 
@@ -196,11 +197,12 @@ static inline void mdc_clear_replay_flag(struct ptlrpc_request *req, int rc)
                spin_lock(&req->rq_lock);
                req->rq_replay = 0;
                spin_unlock(&req->rq_lock);
-        }
-        if (rc && req->rq_transno != 0) {
-                DEBUG_REQ(D_ERROR, req, "transno returned on error rc %d", rc);
-                LBUG();
-        }
+       }
+       if (rc && req->rq_transno != 0) {
+               DEBUG_REQ(D_ERROR, req, "transno returned on error: rc = %d",
+                         rc);
+               LBUG();
+       }
 }
 
 /* Save a large LOV EA into the request buffer so that it is available
@@ -219,7 +221,7 @@ int mdc_save_lovea(struct ptlrpc_request *req,
                   void *data, u32 size)
 {
        struct req_capsule *pill = &req->rq_pill;
-       void *lmm;
+       struct lov_user_md *lmm;
        int rc = 0;
 
        if (req_capsule_get_size(pill, field, RCL_CLIENT) < size) {
@@ -236,39 +238,48 @@ int mdc_save_lovea(struct ptlrpc_request *req,
 
        req_capsule_set_size(pill, field, RCL_CLIENT, size);
        lmm = req_capsule_client_get(pill, field);
-       if (lmm)
+       if (lmm) {
                memcpy(lmm, data, size);
+               /* overwrite layout generation returned from the MDS */
+               lmm->lmm_stripe_offset =
+                 (typeof(lmm->lmm_stripe_offset))LOV_OFFSET_DEFAULT;
+       }
 
        return rc;
 }
 
 static struct ptlrpc_request *
 mdc_intent_open_pack(struct obd_export *exp, struct lookup_intent *it,
-                    struct md_op_data *op_data)
+                    struct md_op_data *op_data, __u32 acl_bufsize)
 {
        struct ptlrpc_request   *req;
        struct obd_device       *obddev = class_exp2obd(exp);
        struct ldlm_intent      *lit;
        const void              *lmm = op_data->op_data;
        __u32                    lmmsize = op_data->op_data_size;
-       struct list_head         cancels = LIST_HEAD_INIT(cancels);
+       __u32                    mdt_md_capsule_size;
+       LIST_HEAD(cancels);
        int                      count = 0;
        enum ldlm_mode           mode;
        int                      rc;
+       int repsize, repsize_estimate;
+
        ENTRY;
 
+       mdt_md_capsule_size = obddev->u.cli.cl_default_mds_easize;
+
        it->it_create_mode = (it->it_create_mode & ~S_IFMT) | S_IFREG;
 
        /* XXX: openlock is not cancelled for cross-refs. */
        /* If inode is known, cancel conflicting OPEN locks. */
        if (fid_is_sane(&op_data->op_fid2)) {
                if (it->it_flags & MDS_OPEN_LEASE) { /* try to get lease */
-                       if (it->it_flags & FMODE_WRITE)
+                       if (it->it_flags & MDS_FMODE_WRITE)
                                mode = LCK_EX;
                        else
                                mode = LCK_PR;
                } else {
-                       if (it->it_flags & (FMODE_WRITE|MDS_OPEN_TRUNC))
+                       if (it->it_flags & (MDS_FMODE_WRITE | MDS_OPEN_TRUNC))
                                mode = LCK_CW;
 #ifdef FMODE_EXEC
                        else if (it->it_flags & FMODE_EXEC)
@@ -300,16 +311,32 @@ mdc_intent_open_pack(struct obd_export *exp, struct lookup_intent *it,
 
         req_capsule_set_size(&req->rq_pill, &RMF_NAME, RCL_CLIENT,
                              op_data->op_namelen + 1);
-       req_capsule_set_size(&req->rq_pill, &RMF_EADATA, RCL_CLIENT,
+       if (cl_is_lov_delay_create(it->it_flags)) {
+               /* open(O_LOV_DELAY_CREATE) won't pack lmm */
+               LASSERT(lmmsize == 0);
+               req_capsule_set_size(&req->rq_pill, &RMF_EADATA, RCL_CLIENT, 0);
+       } else {
+               req_capsule_set_size(&req->rq_pill, &RMF_EADATA, RCL_CLIENT,
                             max(lmmsize, obddev->u.cli.cl_default_mds_easize));
+       }
 
        req_capsule_set_size(&req->rq_pill, &RMF_FILE_SECCTX_NAME,
                             RCL_CLIENT, op_data->op_file_secctx_name != NULL ?
-                            strlen(op_data->op_file_secctx_name) + 1 : 0);
+                            op_data->op_file_secctx_name_size : 0);
 
        req_capsule_set_size(&req->rq_pill, &RMF_FILE_SECCTX, RCL_CLIENT,
                             op_data->op_file_secctx_size);
 
+       /* get SELinux policy info if any */
+       rc = sptlrpc_get_sepol(req);
+       if (rc < 0) {
+               ptlrpc_request_free(req);
+               RETURN(ERR_PTR(rc));
+       }
+       req_capsule_set_size(&req->rq_pill, &RMF_SELINUX_POL, RCL_CLIENT,
+                            strlen(req->rq_sepol) ?
+                            strlen(req->rq_sepol) + 1 : 0);
+
        rc = ldlm_prep_enqueue_req(exp, req, &cancels, count);
        if (rc < 0) {
                ptlrpc_request_free(req);
@@ -329,11 +356,72 @@ mdc_intent_open_pack(struct obd_export *exp, struct lookup_intent *it,
                       lmmsize);
 
        req_capsule_set_size(&req->rq_pill, &RMF_MDT_MD, RCL_SERVER,
-                            obddev->u.cli.cl_max_mds_easize);
-       req_capsule_set_size(&req->rq_pill, &RMF_ACL, RCL_SERVER,
-                            req->rq_import->imp_connect_data.ocd_max_easize);
-        ptlrpc_request_set_replen(req);
-        return req;
+                            mdt_md_capsule_size);
+       req_capsule_set_size(&req->rq_pill, &RMF_ACL, RCL_SERVER, acl_bufsize);
+
+       if (!(it->it_op & IT_CREAT) && it->it_op & IT_OPEN &&
+           req_capsule_has_field(&req->rq_pill, &RMF_FILE_SECCTX_NAME,
+                                 RCL_CLIENT) &&
+           op_data->op_file_secctx_name_size > 0 &&
+           op_data->op_file_secctx_name != NULL) {
+               char *secctx_name;
+
+               secctx_name = req_capsule_client_get(&req->rq_pill,
+                                                    &RMF_FILE_SECCTX_NAME);
+               memcpy(secctx_name, op_data->op_file_secctx_name,
+                      op_data->op_file_secctx_name_size);
+               req_capsule_set_size(&req->rq_pill, &RMF_FILE_SECCTX,
+                                    RCL_SERVER,
+                                    obddev->u.cli.cl_max_mds_easize);
+
+               CDEBUG(D_SEC, "packed '%.*s' as security xattr name\n",
+                      op_data->op_file_secctx_name_size,
+                      op_data->op_file_secctx_name);
+
+       } else {
+               req_capsule_set_size(&req->rq_pill, &RMF_FILE_SECCTX,
+                                    RCL_SERVER, 0);
+       }
+
+       /**
+        * Inline buffer for possible data from Data-on-MDT files.
+        */
+       req_capsule_set_size(&req->rq_pill, &RMF_NIOBUF_INLINE, RCL_SERVER,
+                            sizeof(struct niobuf_remote));
+       ptlrpc_request_set_replen(req);
+
+       /* Get real repbuf allocated size as rounded up power of 2 */
+       repsize = size_roundup_power2(req->rq_replen +
+                                     lustre_msg_early_size());
+       /* Estimate free space for DoM files in repbuf */
+       repsize_estimate = repsize - (req->rq_replen -
+                          mdt_md_capsule_size +
+                          sizeof(struct lov_comp_md_v1) +
+                          sizeof(struct lov_comp_md_entry_v1) +
+                          lov_mds_md_size(0, LOV_MAGIC_V3));
+
+       if (repsize_estimate < obddev->u.cli.cl_dom_min_inline_repsize) {
+               repsize = obddev->u.cli.cl_dom_min_inline_repsize -
+                         repsize_estimate + sizeof(struct niobuf_remote);
+               req_capsule_set_size(&req->rq_pill, &RMF_NIOBUF_INLINE,
+                                    RCL_SERVER,
+                                    sizeof(struct niobuf_remote) + repsize);
+               ptlrpc_request_set_replen(req);
+               CDEBUG(D_INFO, "Increase repbuf by %d bytes, total: %d\n",
+                      repsize, req->rq_replen);
+               repsize = size_roundup_power2(req->rq_replen +
+                                             lustre_msg_early_size());
+       }
+       /* The only way to report real allocated repbuf size to the server
+        * is the lm_repsize but it must be set prior buffer allocation itself
+        * due to security reasons - it is part of buffer used in signature
+        * calculation (see LU-11414). Therefore the saved size is predicted
+        * value as rq_replen rounded to the next higher power of 2.
+        * Such estimation is safe. Though the final allocated buffer might
+        * be even larger, it is not possible to know that at this point.
+        */
+       req->rq_reqmsg->lm_repsize = repsize;
+       RETURN(req);
 }
 
 #define GA_DEFAULT_EA_NAME_LEN 20
@@ -348,7 +436,8 @@ mdc_intent_getxattr_pack(struct obd_export *exp,
        struct ptlrpc_request   *req;
        struct ldlm_intent      *lit;
        int                     rc, count = 0;
-       struct list_head        cancels = LIST_HEAD_INIT(cancels);
+       LIST_HEAD(cancels);
+       u32 ea_vals_buf_size = GA_DEFAULT_EA_VAL_LEN * GA_DEFAULT_EA_NUM;
 
        ENTRY;
 
@@ -357,6 +446,16 @@ mdc_intent_getxattr_pack(struct obd_export *exp,
        if (req == NULL)
                RETURN(ERR_PTR(-ENOMEM));
 
+       /* get SELinux policy info if any */
+       rc = sptlrpc_get_sepol(req);
+       if (rc < 0) {
+               ptlrpc_request_free(req);
+               RETURN(ERR_PTR(rc));
+       }
+       req_capsule_set_size(&req->rq_pill, &RMF_SELINUX_POL, RCL_CLIENT,
+                            strlen(req->rq_sepol) ?
+                            strlen(req->rq_sepol) + 1 : 0);
+
        rc = ldlm_prep_enqueue_req(exp, req, &cancels, count);
        if (rc) {
                ptlrpc_request_free(req);
@@ -366,20 +465,38 @@ mdc_intent_getxattr_pack(struct obd_export *exp,
        /* pack the intent */
        lit = req_capsule_client_get(&req->rq_pill, &RMF_LDLM_INTENT);
        lit->opc = IT_GETXATTR;
+       CDEBUG(D_INFO, "%s: get xattrs for "DFID"\n",
+              exp->exp_obd->obd_name, PFID(&op_data->op_fid1));
+
+#if LUSTRE_VERSION_CODE < OBD_OCD_VERSION(3, 0, 53, 0)
+       /* If the supplied buffer is too small then the server will
+        * return -ERANGE and llite will fallback to using non cached
+        * xattr operations. On servers before 2.10.1 a (non-cached)
+        * listxattr RPC for an orphan or dead file causes an oops. So
+        * let's try to avoid sending too small a buffer to too old a
+        * server. This is effectively undoing the memory conservation
+        * of LU-9417 when it would be *more* likely to crash the
+        * server. See LU-9856. */
+       if (exp->exp_connect_data.ocd_version < OBD_OCD_VERSION(2, 10, 1, 0))
+               ea_vals_buf_size = max_t(u32, ea_vals_buf_size,
+                                        exp->exp_connect_data.ocd_max_easize);
+#endif
 
        /* pack the intended request */
        mdc_pack_body(req, &op_data->op_fid1, op_data->op_valid,
-                     GA_DEFAULT_EA_NAME_LEN * GA_DEFAULT_EA_NUM,
-                     -1, 0);
+                     ea_vals_buf_size, -1, 0);
+
+       /* get SELinux policy info if any */
+       mdc_file_sepol_pack(req);
 
        req_capsule_set_size(&req->rq_pill, &RMF_EADATA, RCL_SERVER,
                             GA_DEFAULT_EA_NAME_LEN * GA_DEFAULT_EA_NUM);
 
        req_capsule_set_size(&req->rq_pill, &RMF_EAVALS, RCL_SERVER,
-                            GA_DEFAULT_EA_VAL_LEN * GA_DEFAULT_EA_NUM);
+                            ea_vals_buf_size);
 
        req_capsule_set_size(&req->rq_pill, &RMF_EAVALS_LENS, RCL_SERVER,
-                            sizeof(__u32) * GA_DEFAULT_EA_NUM);
+                            sizeof(u32) * GA_DEFAULT_EA_NUM);
 
        req_capsule_set_size(&req->rq_pill, &RMF_ACL, RCL_SERVER, 0);
 
@@ -388,86 +505,81 @@ mdc_intent_getxattr_pack(struct obd_export *exp,
        RETURN(req);
 }
 
-static struct ptlrpc_request *mdc_intent_unlink_pack(struct obd_export *exp,
-                                                     struct lookup_intent *it,
-                                                     struct md_op_data *op_data)
+static struct ptlrpc_request *
+mdc_intent_getattr_pack(struct obd_export *exp, struct lookup_intent *it,
+                       struct md_op_data *op_data, __u32 acl_bufsize)
 {
-        struct ptlrpc_request *req;
-        struct obd_device     *obddev = class_exp2obd(exp);
-        struct ldlm_intent    *lit;
-        int                    rc;
-        ENTRY;
+       struct ptlrpc_request *req;
+       struct obd_device *obddev = class_exp2obd(exp);
+       u64 valid = OBD_MD_FLGETATTR | OBD_MD_FLEASIZE | OBD_MD_FLMODEASIZE |
+                   OBD_MD_FLDIREA | OBD_MD_MEA | OBD_MD_FLACL |
+                   OBD_MD_DEFAULT_MEA;
+       struct ldlm_intent *lit;
+       __u32 easize;
+       bool have_secctx = false;
+       int rc;
 
-        req = ptlrpc_request_alloc(class_exp2cliimp(exp),
-                                   &RQF_LDLM_INTENT_UNLINK);
-        if (req == NULL)
-                RETURN(ERR_PTR(-ENOMEM));
+       ENTRY;
 
-        req_capsule_set_size(&req->rq_pill, &RMF_NAME, RCL_CLIENT,
-                             op_data->op_namelen + 1);
+       req = ptlrpc_request_alloc(class_exp2cliimp(exp),
+                                  &RQF_LDLM_INTENT_GETATTR);
+       if (req == NULL)
+               RETURN(ERR_PTR(-ENOMEM));
 
-        rc = ldlm_prep_enqueue_req(exp, req, NULL, 0);
-        if (rc) {
-                ptlrpc_request_free(req);
-                RETURN(ERR_PTR(rc));
-        }
+       /* send name of security xattr to get upon intent */
+       if (it->it_op & (IT_LOOKUP | IT_GETATTR) &&
+           req_capsule_has_field(&req->rq_pill, &RMF_FILE_SECCTX_NAME,
+                                 RCL_CLIENT) &&
+           op_data->op_file_secctx_name_size > 0 &&
+           op_data->op_file_secctx_name != NULL) {
+               have_secctx = true;
+               req_capsule_set_size(&req->rq_pill, &RMF_FILE_SECCTX_NAME,
+                                    RCL_CLIENT,
+                                    op_data->op_file_secctx_name_size);
+       }
 
-        /* pack the intent */
-        lit = req_capsule_client_get(&req->rq_pill, &RMF_LDLM_INTENT);
-        lit->opc = (__u64)it->it_op;
+       req_capsule_set_size(&req->rq_pill, &RMF_NAME, RCL_CLIENT,
+                            op_data->op_namelen + 1);
 
-        /* pack the intended request */
-        mdc_unlink_pack(req, op_data);
+       rc = ldlm_prep_enqueue_req(exp, req, NULL, 0);
+       if (rc) {
+               ptlrpc_request_free(req);
+               RETURN(ERR_PTR(rc));
+       }
 
-       req_capsule_set_size(&req->rq_pill, &RMF_MDT_MD, RCL_SERVER,
-                            obddev->u.cli.cl_default_mds_easize);
-       ptlrpc_request_set_replen(req);
-       RETURN(req);
-}
+        /* pack the intent */
+       lit = req_capsule_client_get(&req->rq_pill, &RMF_LDLM_INTENT);
+       lit->opc = (__u64)it->it_op;
 
-static struct ptlrpc_request *mdc_intent_getattr_pack(struct obd_export *exp,
-                                                      struct lookup_intent *it,
-                                                      struct md_op_data *op_data)
-{
-       struct ptlrpc_request   *req;
-       struct obd_device       *obddev = class_exp2obd(exp);
-       u64                      valid = OBD_MD_FLGETATTR | OBD_MD_FLEASIZE |
-                                        OBD_MD_FLMODEASIZE | OBD_MD_FLDIREA |
-                                        OBD_MD_MEA | OBD_MD_FLACL;
-       struct ldlm_intent      *lit;
-       int                      rc;
-       __u32                    easize;
-       ENTRY;
+       easize = obddev->u.cli.cl_default_mds_easize;
 
-        req = ptlrpc_request_alloc(class_exp2cliimp(exp),
-                                   &RQF_LDLM_INTENT_GETATTR);
-        if (req == NULL)
-                RETURN(ERR_PTR(-ENOMEM));
+       /* pack the intended request */
+       mdc_getattr_pack(req, valid, it->it_flags, op_data, easize);
 
-        req_capsule_set_size(&req->rq_pill, &RMF_NAME, RCL_CLIENT,
-                             op_data->op_namelen + 1);
+       req_capsule_set_size(&req->rq_pill, &RMF_MDT_MD, RCL_SERVER, easize);
+       req_capsule_set_size(&req->rq_pill, &RMF_ACL, RCL_SERVER, acl_bufsize);
+       req_capsule_set_size(&req->rq_pill, &RMF_DEFAULT_MDT_MD, RCL_SERVER,
+                            sizeof(struct lmv_user_md));
 
-        rc = ldlm_prep_enqueue_req(exp, req, NULL, 0);
-        if (rc) {
-                ptlrpc_request_free(req);
-                RETURN(ERR_PTR(rc));
-        }
+       if (have_secctx) {
+               char *secctx_name;
 
-        /* pack the intent */
-        lit = req_capsule_client_get(&req->rq_pill, &RMF_LDLM_INTENT);
-        lit->opc = (__u64)it->it_op;
+               secctx_name = req_capsule_client_get(&req->rq_pill,
+                                                    &RMF_FILE_SECCTX_NAME);
+               memcpy(secctx_name, op_data->op_file_secctx_name,
+                      op_data->op_file_secctx_name_size);
 
-       if (obddev->u.cli.cl_default_mds_easize > 0)
-               easize = obddev->u.cli.cl_default_mds_easize;
-       else
-               easize = obddev->u.cli.cl_max_mds_easize;
+               req_capsule_set_size(&req->rq_pill, &RMF_FILE_SECCTX,
+                                    RCL_SERVER, easize);
 
-       /* pack the intended request */
-       mdc_getattr_pack(req, valid, it->it_flags, op_data, easize);
+               CDEBUG(D_SEC, "packed '%.*s' as security xattr name\n",
+                      op_data->op_file_secctx_name_size,
+                      op_data->op_file_secctx_name);
+       } else {
+               req_capsule_set_size(&req->rq_pill, &RMF_FILE_SECCTX,
+                                    RCL_SERVER, 0);
+       }
 
-       req_capsule_set_size(&req->rq_pill, &RMF_MDT_MD, RCL_SERVER, easize);
-       req_capsule_set_size(&req->rq_pill, &RMF_ACL, RCL_SERVER,
-                            req->rq_import->imp_connect_data.ocd_max_easize);
        ptlrpc_request_set_replen(req);
        RETURN(req);
 }
@@ -477,10 +589,11 @@ static struct ptlrpc_request *mdc_intent_layout_pack(struct obd_export *exp,
                                                     struct md_op_data *op_data)
 {
        struct obd_device     *obd = class_exp2obd(exp);
+       LIST_HEAD(cancels);
        struct ptlrpc_request *req;
        struct ldlm_intent    *lit;
        struct layout_intent  *layout;
-       int rc;
+       int count = 0, rc;
        ENTRY;
 
        req = ptlrpc_request_alloc(class_exp2cliimp(exp),
@@ -488,8 +601,15 @@ static struct ptlrpc_request *mdc_intent_layout_pack(struct obd_export *exp,
        if (req == NULL)
                RETURN(ERR_PTR(-ENOMEM));
 
+       if (fid_is_sane(&op_data->op_fid2) && (it->it_op & IT_LAYOUT) &&
+           (it->it_flags & FMODE_WRITE)) {
+               count = mdc_resource_get_unused(exp, &op_data->op_fid2,
+                                               &cancels, LCK_EX,
+                                               MDS_INODELOCK_LAYOUT);
+       }
+
        req_capsule_set_size(&req->rq_pill, &RMF_EADATA, RCL_CLIENT, 0);
-       rc = ldlm_prep_enqueue_req(exp, req, NULL, 0);
+       rc = ldlm_prep_enqueue_req(exp, req, &cancels, count);
        if (rc) {
                ptlrpc_request_free(req);
                RETURN(ERR_PTR(rc));
@@ -597,14 +717,14 @@ static int mdc_finish_enqueue(struct obd_export *exp,
          * It's important that we do this first!  Otherwise we might exit the
          * function without doing so, and try to replay a failed create
          * (bug 3440) */
-        if (it->it_op & IT_OPEN && req->rq_replay &&
+       if (it->it_op & IT_OPEN && req->rq_replay &&
            (!it_disposition(it, DISP_OPEN_OPEN) || it->it_status != 0))
                mdc_clear_replay_flag(req, it->it_status);
 
-       DEBUG_REQ(D_RPCTRACE, req, "op: %d disposition: %x, status: %d",
+       DEBUG_REQ(D_RPCTRACE, req, "op=%x disposition=%x, status=%d",
                  it->it_op, it->it_disposition, it->it_status);
 
-        /* We know what to expect, so we do any byte flipping required here */
+       /* We know what to expect, so we do any byte flipping required here */
        if (it_has_reply_body(it)) {
                 body = req_capsule_server_get(pill, &RMF_MDT_BODY);
                 if (body == NULL) {
@@ -623,6 +743,12 @@ static int mdc_finish_enqueue(struct obd_export *exp,
                        mdc_set_open_replay_data(NULL, NULL, it);
                }
 
+               if (it_disposition(it, DISP_OPEN_CREATE) &&
+                   !it_open_error(DISP_OPEN_CREATE, it)) {
+                       lprocfs_counter_incr(exp->exp_obd->obd_md_stats,
+                                            LPROC_MD_CREATE);
+               }
+
                if (body->mbo_valid & (OBD_MD_FLDIREA | OBD_MD_FLEASIZE)) {
                         void *eadata;
 
@@ -665,6 +791,8 @@ static int mdc_finish_enqueue(struct obd_export *exp,
                /* maybe the lock was granted right away and layout
                 * is packed into RMF_DLM_LVB of req */
                lvb_len = req_capsule_get_size(pill, &RMF_DLM_LVB, RCL_SERVER);
+               CDEBUG(D_INFO, "%s: layout return lvb %d transno %lld\n",
+                      class_exp2obd(exp)->obd_name, lvb_len, req->rq_transno);
                if (lvb_len > 0) {
                        lvb_data = req_capsule_server_sized_get(pill,
                                                        &RMF_DLM_LVB, lvb_len);
@@ -722,7 +850,7 @@ static int mdc_finish_enqueue(struct obd_export *exp,
 
                body = req_capsule_server_get(pill, &RMF_MDT_BODY);
                if (!(body->mbo_valid & OBD_MD_DOM_SIZE)) {
-                       LDLM_ERROR(lock, "%s: DoM lock without size.\n",
+                       LDLM_ERROR(lock, "%s: DoM lock without size.",
                                   exp->exp_obd->obd_name);
                        GOTO(out_lock, rc = -EPROTO);
                }
@@ -738,6 +866,16 @@ out_lock:
        RETURN(rc);
 }
 
+static inline bool mdc_skip_mod_rpc_slot(const struct lookup_intent *it)
+{
+       if (it != NULL &&
+           (it->it_op == IT_GETATTR || it->it_op == IT_LOOKUP ||
+            it->it_op == IT_READDIR ||
+            (it->it_op == IT_LAYOUT && !(it->it_flags & MDS_FMODE_WRITE))))
+               return true;
+       return false;
+}
+
 /* We always reserve enough space in the reply packet for a stripe MD, because
  * we don't know in advance the file type. */
 static int mdc_enqueue_base(struct obd_export *exp,
@@ -749,7 +887,7 @@ static int mdc_enqueue_base(struct obd_export *exp,
                            __u64 extra_lock_flags)
 {
        struct obd_device *obddev = class_exp2obd(exp);
-       struct ptlrpc_request *req = NULL;
+       struct ptlrpc_request *req;
        __u64 flags, saved_flags = extra_lock_flags;
        struct ldlm_res_id res_id;
        static const union ldlm_policy_data lookup_policy = {
@@ -762,6 +900,8 @@ static int mdc_enqueue_base(struct obd_export *exp,
                                  .l_inodebits = { MDS_INODELOCK_XATTR } };
        int generation, resends = 0;
        struct ldlm_reply *lockrep;
+       struct obd_import *imp = class_exp2cliimp(exp);
+       __u32 acl_bufsize;
        enum lvb_type lvb_type = 0;
        int rc;
        ENTRY;
@@ -774,34 +914,39 @@ static int mdc_enqueue_base(struct obd_export *exp,
                LASSERT(policy == NULL);
 
                saved_flags |= LDLM_FL_HAS_INTENT;
-               if (it->it_op & (IT_UNLINK | IT_GETATTR | IT_READDIR))
+               if (it->it_op & (IT_GETATTR | IT_READDIR))
                        policy = &update_policy;
                else if (it->it_op & IT_LAYOUT)
                        policy = &layout_policy;
-               else if (it->it_op & (IT_GETXATTR | IT_SETXATTR))
+               else if (it->it_op & IT_GETXATTR)
                        policy = &getxattr_policy;
                else
                        policy = &lookup_policy;
        }
 
-        generation = obddev->u.cli.cl_import->imp_generation;
+       generation = obddev->u.cli.cl_import->imp_generation;
+       if (!it || (it->it_op & (IT_OPEN | IT_CREAT)))
+               acl_bufsize = min_t(__u32, imp->imp_connect_data.ocd_max_easize,
+                                   XATTR_SIZE_MAX);
+       else
+               acl_bufsize = LUSTRE_POSIX_ACL_MAX_SIZE_OLD;
+
 resend:
-        flags = saved_flags;
+       flags = saved_flags;
        if (it == NULL) {
                /* The only way right now is FLOCK. */
                LASSERTF(einfo->ei_type == LDLM_FLOCK, "lock type %d\n",
                         einfo->ei_type);
                res_id.name[3] = LDLM_FLOCK;
+               req = ldlm_enqueue_pack(exp, 0);
        } else if (it->it_op & IT_OPEN) {
-               req = mdc_intent_open_pack(exp, it, op_data);
-       } else if (it->it_op & IT_UNLINK) {
-               req = mdc_intent_unlink_pack(exp, it, op_data);
+               req = mdc_intent_open_pack(exp, it, op_data, acl_bufsize);
        } else if (it->it_op & (IT_GETATTR | IT_LOOKUP)) {
-               req = mdc_intent_getattr_pack(exp, it, op_data);
+               req = mdc_intent_getattr_pack(exp, it, op_data, acl_bufsize);
        } else if (it->it_op & IT_READDIR) {
                req = mdc_enqueue_pack(exp, 0);
        } else if (it->it_op & IT_LAYOUT) {
-               if (!imp_connect_lvb_type(class_exp2cliimp(exp)))
+               if (!imp_connect_lvb_type(imp))
                        RETURN(-EOPNOTSUPP);
                req = mdc_intent_layout_pack(exp, it, op_data);
                lvb_type = LVB_T_LAYOUT;
@@ -821,20 +966,7 @@ resend:
                req->rq_sent = ktime_get_real_seconds() + resends;
         }
 
-       /* It is important to obtain modify RPC slot first (if applicable), so
-        * that threads that are waiting for a modify RPC slot are not polluting
-        * our rpcs in flight counter.
-        * We do not do flock request limiting, though */
-       if (it) {
-               mdc_get_mod_rpc_slot(req, it);
-               rc = obd_get_request_slot(&obddev->u.cli);
-               if (rc != 0) {
-                       mdc_put_mod_rpc_slot(req, it);
-                       mdc_clear_replay_flag(req, 0);
-                       ptlrpc_req_finished(req);
-                       RETURN(rc);
-               }
-       }
+       einfo->ei_enq_slot = !mdc_skip_mod_rpc_slot(it);
 
        /* With Data-on-MDT the glimpse callback is needed too.
         * It is set here in advance but not in mdc_finish_enqueue()
@@ -845,6 +977,7 @@ resend:
 
        rc = ldlm_cli_enqueue(exp, &req, einfo, &res_id, policy, &flags, NULL,
                              0, lvb_type, lockh, 0);
+
        if (!it) {
                /* For flock requests we immediatelly return without further
                   delay and let caller deal with the rest, since rest of
@@ -858,12 +991,10 @@ resend:
                    (einfo->ei_type == LDLM_FLOCK) &&
                    (einfo->ei_mode == LCK_NL))
                        goto resend;
+               ptlrpc_req_finished(req);
                RETURN(rc);
        }
 
-       obd_put_request_slot(&obddev->u.cli);
-       mdc_put_mod_rpc_slot(req, it);
-
        if (rc < 0) {
                CDEBUG(D_INFO,
                      "%s: ldlm_cli_enqueue "DFID":"DFID"=%s failed: rc = %d\n",
@@ -885,21 +1016,33 @@ resend:
         * intent operation, when server returns -EINPROGRESS for acquiring
         * intent lock, we'll retry in after_reply(). */
        if (it && (int)lockrep->lock_policy_res2 == -EINPROGRESS) {
-                mdc_clear_replay_flag(req, rc);
-                ptlrpc_req_finished(req);
-                resends++;
-
-                CDEBUG(D_HA, "%s: resend:%d op:%d "DFID"/"DFID"\n",
-                       obddev->obd_name, resends, it->it_op,
-                       PFID(&op_data->op_fid1), PFID(&op_data->op_fid2));
-
-                if (generation == obddev->u.cli.cl_import->imp_generation) {
-                        goto resend;
-                } else {
+               mdc_clear_replay_flag(req, rc);
+               ptlrpc_req_finished(req);
+               if (generation == obddev->u.cli.cl_import->imp_generation) {
+                       if (signal_pending(current))
+                               RETURN(-EINTR);
+
+                       resends++;
+                       CDEBUG(D_HA, "%s: resend:%d op:%d "DFID"/"DFID"\n",
+                              obddev->obd_name, resends, it->it_op,
+                              PFID(&op_data->op_fid1),
+                              PFID(&op_data->op_fid2));
+                       goto resend;
+               } else {
                        CDEBUG(D_HA, "resend cross eviction\n");
-                        RETURN(-EIO);
-                }
-        }
+                       RETURN(-EIO);
+               }
+       }
+
+       if ((int)lockrep->lock_policy_res2 == -ERANGE &&
+           it->it_op & (IT_OPEN | IT_GETATTR | IT_LOOKUP) &&
+           acl_bufsize == LUSTRE_POSIX_ACL_MAX_SIZE_OLD) {
+               mdc_clear_replay_flag(req, -ERANGE);
+               ptlrpc_req_finished(req);
+               acl_bufsize = min_t(__u32, imp->imp_connect_data.ocd_max_easize,
+                                   XATTR_SIZE_MAX);
+               goto resend;
+       }
 
        rc = mdc_finish_enqueue(exp, req, einfo, it, lockh, rc);
        if (rc < 0) {
@@ -1188,16 +1331,15 @@ int mdc_intent_lock(struct obd_export *exp, struct md_op_data *op_data,
 }
 
 static int mdc_intent_getattr_async_interpret(const struct lu_env *env,
-                                              struct ptlrpc_request *req,
-                                              void *args, int rc)
+                                             struct ptlrpc_request *req,
+                                             void *args, int rc)
 {
        struct mdc_getattr_args  *ga = args;
-       struct obd_export        *exp = ga->ga_exp;
-       struct md_enqueue_info   *minfo = ga->ga_minfo;
+       struct obd_export *exp = ga->ga_exp;
+       struct md_enqueue_info *minfo = ga->ga_minfo;
        struct ldlm_enqueue_info *einfo = &minfo->mi_einfo;
        struct lookup_intent     *it;
        struct lustre_handle     *lockh;
-       struct obd_device        *obddev;
        struct ldlm_reply        *lockrep;
        __u64                     flags = LDLM_FL_HAS_INTENT;
        ENTRY;
@@ -1205,9 +1347,6 @@ static int mdc_intent_getattr_async_interpret(const struct lu_env *env,
         it    = &minfo->mi_it;
         lockh = &minfo->mi_lockh;
 
-        obddev = class_exp2obd(exp);
-
-       obd_put_request_slot(&obddev->u.cli);
         if (OBD_FAIL_CHECK(OBD_FAIL_MDC_GETATTR_ENQUEUE))
                 rc = -ETIMEDOUT;
 
@@ -1244,7 +1383,6 @@ int mdc_intent_getattr_async(struct obd_export *exp,
        struct lookup_intent    *it = &minfo->mi_it;
        struct ptlrpc_request   *req;
        struct mdc_getattr_args *ga;
-       struct obd_device       *obddev = class_exp2obd(exp);
        struct ldlm_res_id       res_id;
        union ldlm_policy_data policy = {
                                .l_inodebits = { MDS_INODELOCK_LOOKUP |
@@ -1258,16 +1396,13 @@ int mdc_intent_getattr_async(struct obd_export *exp,
                PFID(&op_data->op_fid1), ldlm_it2str(it->it_op), it->it_flags);
 
        fid_build_reg_res_name(&op_data->op_fid1, &res_id);
-       req = mdc_intent_getattr_pack(exp, it, op_data);
+       /* If the MDT return -ERANGE because of large ACL, then the sponsor
+        * of the async getattr RPC will handle that by itself. */
+       req = mdc_intent_getattr_pack(exp, it, op_data,
+                                     LUSTRE_POSIX_ACL_MAX_SIZE_OLD);
        if (IS_ERR(req))
                RETURN(PTR_ERR(req));
 
-       rc = obd_get_request_slot(&obddev->u.cli);
-       if (rc != 0) {
-               ptlrpc_req_finished(req);
-               RETURN(rc);
-       }
-
        /* With Data-on-MDT the glimpse callback is needed too.
         * It is set here in advance but not in mdc_finish_enqueue()
         * to avoid possible races. It is safe to have glimpse handler
@@ -1278,13 +1413,11 @@ int mdc_intent_getattr_async(struct obd_export *exp,
        rc = ldlm_cli_enqueue(exp, &req, &minfo->mi_einfo, &res_id, &policy,
                              &flags, NULL, 0, LVB_T_NONE, &minfo->mi_lockh, 1);
        if (rc < 0) {
-               obd_put_request_slot(&obddev->u.cli);
                ptlrpc_req_finished(req);
                RETURN(rc);
        }
 
-       CLASSERT(sizeof(*ga) <= sizeof(req->rq_async_args));
-       ga = ptlrpc_req_async_args(req);
+       ga = ptlrpc_req_async_args(ga, req);
        ga->ga_exp = exp;
        ga->ga_minfo = minfo;