diff options
Diffstat (limited to 'net/sunrpc')
32 files changed, 682 insertions, 685 deletions
diff --git a/net/sunrpc/addr.c b/net/sunrpc/addr.c index 8b4d72b1a066..010dcb876f9d 100644 --- a/net/sunrpc/addr.c +++ b/net/sunrpc/addr.c @@ -82,11 +82,11 @@ static size_t rpc_ntop6(const struct sockaddr *sap, rc = snprintf(scopebuf, sizeof(scopebuf), "%c%u", IPV6_SCOPE_DELIMITER, sin6->sin6_scope_id); - if (unlikely((size_t)rc > sizeof(scopebuf))) + if (unlikely((size_t)rc >= sizeof(scopebuf))) return 0; len += rc; - if (unlikely(len > buflen)) + if (unlikely(len >= buflen)) return 0; strcat(buf, scopebuf); diff --git a/net/sunrpc/auth.c b/net/sunrpc/auth.c index 5748ad0ba1bd..a9f0d17fdb0d 100644 --- a/net/sunrpc/auth.c +++ b/net/sunrpc/auth.c @@ -81,7 +81,7 @@ static int param_get_hashtbl_sz(char *buffer, const struct kernel_param *kp) unsigned int nbits; nbits = *(unsigned int *)kp->arg; - return sprintf(buffer, "%u", 1U << nbits); + return sprintf(buffer, "%u\n", 1U << nbits); } #define param_check_hashtbl_sz(name, p) __param_check(name, p, unsigned int); diff --git a/net/sunrpc/auth_gss/auth_gss.c b/net/sunrpc/auth_gss/auth_gss.c index 25fbd8d9de74..4ecc2a959567 100644 --- a/net/sunrpc/auth_gss/auth_gss.c +++ b/net/sunrpc/auth_gss/auth_gss.c @@ -254,7 +254,7 @@ gss_fill_context(const void *p, const void *end, struct gss_cl_ctx *ctx, struct if (IS_ERR(p)) goto err; done: - trace_rpcgss_context(ctx->gc_expiry, now, timeout, + trace_rpcgss_context(window_size, ctx->gc_expiry, now, timeout, ctx->gc_acceptor.len, ctx->gc_acceptor.data); err: return p; @@ -697,10 +697,12 @@ retry: } schedule(); } - if (gss_msg->ctx) + if (gss_msg->ctx) { + trace_rpcgss_ctx_init(gss_cred); gss_cred_set_ctx(cred, gss_msg->ctx); - else + } else { err = gss_msg->msg.errno; + } spin_unlock(&pipe->lock); out_intr: finish_wait(&gss_msg->waitqueue, &wait); @@ -1054,11 +1056,11 @@ gss_create_new(const struct rpc_auth_create_args *args, struct rpc_clnt *clnt) auth->au_rslack = GSS_KRB5_MAX_SLACK_NEEDED >> 2; auth->au_verfsize = GSS_VERF_SLACK >> 2; auth->au_ralign = GSS_VERF_SLACK >> 2; - auth->au_flags = 0; + __set_bit(RPCAUTH_AUTH_UPDATE_SLACK, &auth->au_flags); auth->au_ops = &authgss_ops; auth->au_flavor = flavor; if (gss_pseudoflavor_to_datatouch(gss_auth->mech, flavor)) - auth->au_flags |= RPCAUTH_AUTH_DATATOUCH; + __set_bit(RPCAUTH_AUTH_DATATOUCH, &auth->au_flags); refcount_set(&auth->au_count, 1); kref_init(&gss_auth->kref); @@ -1284,8 +1286,9 @@ gss_send_destroy_context(struct rpc_cred *cred) if (new) { ctx->gc_proc = RPC_GSS_PROC_DESTROY; + trace_rpcgss_ctx_destroy(gss_cred); task = rpc_call_null(gss_auth->client, &new->gc_base, - RPC_TASK_ASYNC|RPC_TASK_SOFT); + RPC_TASK_ASYNC); if (!IS_ERR(task)) rpc_put_task(task); @@ -1349,7 +1352,6 @@ gss_destroy_nullcred(struct rpc_cred *cred) static void gss_destroy_cred(struct rpc_cred *cred) { - if (test_and_clear_bit(RPCAUTH_CRED_UPTODATE, &cred->cr_flags) != 0) gss_send_destroy_context(cred); gss_destroy_nullcred(cred); @@ -1613,6 +1615,7 @@ static int gss_renew_cred(struct rpc_task *task) new = gss_lookup_cred(auth, &acred, RPCAUTH_LOOKUP_NEW); if (IS_ERR(new)) return PTR_ERR(new); + task->tk_rqstp->rq_cred = new; put_rpccred(oldcred); return 0; @@ -1709,7 +1712,8 @@ gss_validate(struct rpc_task *task, struct xdr_stream *xdr) /* We leave it to unwrap to calculate au_rslack. For now we just * calculate the length of the verifier: */ - cred->cr_auth->au_verfsize = XDR_QUADLEN(len) + 2; + if (test_bit(RPCAUTH_AUTH_UPDATE_SLACK, &cred->cr_auth->au_flags)) + cred->cr_auth->au_verfsize = XDR_QUADLEN(len) + 2; status = 0; out: gss_put_ctx(ctx); @@ -1927,13 +1931,30 @@ out: return status; } -static int -gss_unwrap_resp_auth(struct rpc_cred *cred) +/** + * gss_update_rslack - Possibly update RPC receive buffer size estimates + * @task: rpc_task for incoming RPC Reply being unwrapped + * @cred: controlling rpc_cred for @task + * @before: XDR words needed before each RPC Reply message + * @after: XDR words needed following each RPC Reply message + * + */ +static void gss_update_rslack(struct rpc_task *task, struct rpc_cred *cred, + unsigned int before, unsigned int after) { struct rpc_auth *auth = cred->cr_auth; - auth->au_rslack = auth->au_verfsize; - auth->au_ralign = auth->au_verfsize; + if (test_and_clear_bit(RPCAUTH_AUTH_UPDATE_SLACK, &auth->au_flags)) { + auth->au_ralign = auth->au_verfsize + before; + auth->au_rslack = auth->au_verfsize + after; + trace_rpcgss_update_slack(task, auth); + } +} + +static int +gss_unwrap_resp_auth(struct rpc_task *task, struct rpc_cred *cred) +{ + gss_update_rslack(task, cred, 0, 0); return 0; } @@ -1956,7 +1977,6 @@ gss_unwrap_resp_integ(struct rpc_task *task, struct rpc_cred *cred, struct xdr_stream *xdr) { struct xdr_buf gss_data, *rcv_buf = &rqstp->rq_rcv_buf; - struct rpc_auth *auth = cred->cr_auth; u32 len, offset, seqno, maj_stat; struct xdr_netobj mic; int ret; @@ -2005,8 +2025,7 @@ gss_unwrap_resp_integ(struct rpc_task *task, struct rpc_cred *cred, if (maj_stat != GSS_S_COMPLETE) goto bad_mic; - auth->au_rslack = auth->au_verfsize + 2 + 1 + XDR_QUADLEN(mic.len); - auth->au_ralign = auth->au_verfsize + 2; + gss_update_rslack(task, cred, 2, 2 + 1 + XDR_QUADLEN(mic.len)); ret = 0; out: @@ -2031,8 +2050,6 @@ gss_unwrap_resp_priv(struct rpc_task *task, struct rpc_cred *cred, { struct xdr_buf *rcv_buf = &rqstp->rq_rcv_buf; struct kvec *head = rqstp->rq_rcv_buf.head; - struct rpc_auth *auth = cred->cr_auth; - unsigned int savedlen = rcv_buf->len; u32 offset, opaque_len, maj_stat; __be32 *p; @@ -2043,9 +2060,9 @@ gss_unwrap_resp_priv(struct rpc_task *task, struct rpc_cred *cred, offset = (u8 *)(p) - (u8 *)head->iov_base; if (offset + opaque_len > rcv_buf->len) goto unwrap_failed; - rcv_buf->len = offset + opaque_len; - maj_stat = gss_unwrap(ctx->gc_gss_ctx, offset, rcv_buf); + maj_stat = gss_unwrap(ctx->gc_gss_ctx, offset, + offset + opaque_len, rcv_buf); if (maj_stat == GSS_S_CONTEXT_EXPIRED) clear_bit(RPCAUTH_CRED_UPTODATE, &cred->cr_flags); if (maj_stat != GSS_S_COMPLETE) @@ -2059,10 +2076,9 @@ gss_unwrap_resp_priv(struct rpc_task *task, struct rpc_cred *cred, */ xdr_init_decode(xdr, rcv_buf, p, rqstp); - auth->au_rslack = auth->au_verfsize + 2 + - XDR_QUADLEN(savedlen - rcv_buf->len); - auth->au_ralign = auth->au_verfsize + 2 + - XDR_QUADLEN(savedlen - rcv_buf->len); + gss_update_rslack(task, cred, 2 + ctx->gc_gss_ctx->align, + 2 + ctx->gc_gss_ctx->slack); + return 0; unwrap_failed: trace_rpcgss_unwrap_failed(task); @@ -2132,7 +2148,7 @@ gss_unwrap_resp(struct rpc_task *task, struct xdr_stream *xdr) goto out_decode; switch (gss_cred->gc_service) { case RPC_GSS_SVC_NONE: - status = gss_unwrap_resp_auth(cred); + status = gss_unwrap_resp_auth(task, cred); break; case RPC_GSS_SVC_INTEGRITY: status = gss_unwrap_resp_integ(task, cred, ctx, rqstp, xdr); diff --git a/net/sunrpc/auth_gss/gss_krb5_crypto.c b/net/sunrpc/auth_gss/gss_krb5_crypto.c index 6f2d30d7b766..e7180da1fc6a 100644 --- a/net/sunrpc/auth_gss/gss_krb5_crypto.c +++ b/net/sunrpc/auth_gss/gss_krb5_crypto.c @@ -851,8 +851,8 @@ out_err: } u32 -gss_krb5_aes_decrypt(struct krb5_ctx *kctx, u32 offset, struct xdr_buf *buf, - u32 *headskip, u32 *tailskip) +gss_krb5_aes_decrypt(struct krb5_ctx *kctx, u32 offset, u32 len, + struct xdr_buf *buf, u32 *headskip, u32 *tailskip) { struct xdr_buf subbuf; u32 ret = 0; @@ -881,7 +881,7 @@ gss_krb5_aes_decrypt(struct krb5_ctx *kctx, u32 offset, struct xdr_buf *buf, /* create a segment skipping the header and leaving out the checksum */ xdr_buf_subsegment(buf, &subbuf, offset + GSS_KRB5_TOK_HDR_LEN, - (buf->len - offset - GSS_KRB5_TOK_HDR_LEN - + (len - offset - GSS_KRB5_TOK_HDR_LEN - kctx->gk5e->cksumlength)); nblocks = (subbuf.len + blocksize - 1) / blocksize; @@ -926,7 +926,7 @@ gss_krb5_aes_decrypt(struct krb5_ctx *kctx, u32 offset, struct xdr_buf *buf, goto out_err; /* Get the packet's hmac value */ - ret = read_bytes_from_xdr_buf(buf, buf->len - kctx->gk5e->cksumlength, + ret = read_bytes_from_xdr_buf(buf, len - kctx->gk5e->cksumlength, pkt_hmac, kctx->gk5e->cksumlength); if (ret) goto out_err; diff --git a/net/sunrpc/auth_gss/gss_krb5_wrap.c b/net/sunrpc/auth_gss/gss_krb5_wrap.c index 6c1920eed771..cf0fd170ac18 100644 --- a/net/sunrpc/auth_gss/gss_krb5_wrap.c +++ b/net/sunrpc/auth_gss/gss_krb5_wrap.c @@ -261,7 +261,9 @@ gss_wrap_kerberos_v1(struct krb5_ctx *kctx, int offset, } static u32 -gss_unwrap_kerberos_v1(struct krb5_ctx *kctx, int offset, struct xdr_buf *buf) +gss_unwrap_kerberos_v1(struct krb5_ctx *kctx, int offset, int len, + struct xdr_buf *buf, unsigned int *slack, + unsigned int *align) { int signalg; int sealalg; @@ -279,12 +281,13 @@ gss_unwrap_kerberos_v1(struct krb5_ctx *kctx, int offset, struct xdr_buf *buf) u32 conflen = kctx->gk5e->conflen; int crypt_offset; u8 *cksumkey; + unsigned int saved_len = buf->len; dprintk("RPC: gss_unwrap_kerberos\n"); ptr = (u8 *)buf->head[0].iov_base + offset; if (g_verify_token_header(&kctx->mech_used, &bodysize, &ptr, - buf->len - offset)) + len - offset)) return GSS_S_DEFECTIVE_TOKEN; if ((ptr[0] != ((KG_TOK_WRAP_MSG >> 8) & 0xff)) || @@ -324,6 +327,7 @@ gss_unwrap_kerberos_v1(struct krb5_ctx *kctx, int offset, struct xdr_buf *buf) (!kctx->initiate && direction != 0)) return GSS_S_BAD_SIG; + buf->len = len; if (kctx->enctype == ENCTYPE_ARCFOUR_HMAC) { struct crypto_sync_skcipher *cipher; int err; @@ -376,11 +380,15 @@ gss_unwrap_kerberos_v1(struct krb5_ctx *kctx, int offset, struct xdr_buf *buf) data_len = (buf->head[0].iov_base + buf->head[0].iov_len) - data_start; memmove(orig_start, data_start, data_len); buf->head[0].iov_len -= (data_start - orig_start); - buf->len -= (data_start - orig_start); + buf->len = len - (data_start - orig_start); if (gss_krb5_remove_padding(buf, blocksize)) return GSS_S_DEFECTIVE_TOKEN; + /* slack must include room for krb5 padding */ + *slack = XDR_QUADLEN(saved_len - buf->len); + /* The GSS blob always precedes the RPC message payload */ + *align = *slack; return GSS_S_COMPLETE; } @@ -486,7 +494,9 @@ gss_wrap_kerberos_v2(struct krb5_ctx *kctx, u32 offset, } static u32 -gss_unwrap_kerberos_v2(struct krb5_ctx *kctx, int offset, struct xdr_buf *buf) +gss_unwrap_kerberos_v2(struct krb5_ctx *kctx, int offset, int len, + struct xdr_buf *buf, unsigned int *slack, + unsigned int *align) { time64_t now; u8 *ptr; @@ -532,7 +542,7 @@ gss_unwrap_kerberos_v2(struct krb5_ctx *kctx, int offset, struct xdr_buf *buf) if (rrc != 0) rotate_left(offset + 16, buf, rrc); - err = (*kctx->gk5e->decrypt_v2)(kctx, offset, buf, + err = (*kctx->gk5e->decrypt_v2)(kctx, offset, len, buf, &headskip, &tailskip); if (err) return GSS_S_FAILURE; @@ -542,7 +552,7 @@ gss_unwrap_kerberos_v2(struct krb5_ctx *kctx, int offset, struct xdr_buf *buf) * it against the original */ err = read_bytes_from_xdr_buf(buf, - buf->len - GSS_KRB5_TOK_HDR_LEN - tailskip, + len - GSS_KRB5_TOK_HDR_LEN - tailskip, decrypted_hdr, GSS_KRB5_TOK_HDR_LEN); if (err) { dprintk("%s: error %u getting decrypted_hdr\n", __func__, err); @@ -568,18 +578,19 @@ gss_unwrap_kerberos_v2(struct krb5_ctx *kctx, int offset, struct xdr_buf *buf) * Note that buf->head[0].iov_len may indicate the available * head buffer space rather than that actually occupied. */ - movelen = min_t(unsigned int, buf->head[0].iov_len, buf->len); + movelen = min_t(unsigned int, buf->head[0].iov_len, len); movelen -= offset + GSS_KRB5_TOK_HDR_LEN + headskip; - if (offset + GSS_KRB5_TOK_HDR_LEN + headskip + movelen > - buf->head[0].iov_len) - return GSS_S_FAILURE; + BUG_ON(offset + GSS_KRB5_TOK_HDR_LEN + headskip + movelen > + buf->head[0].iov_len); memmove(ptr, ptr + GSS_KRB5_TOK_HDR_LEN + headskip, movelen); buf->head[0].iov_len -= GSS_KRB5_TOK_HDR_LEN + headskip; - buf->len -= GSS_KRB5_TOK_HDR_LEN + headskip; + buf->len = len - GSS_KRB5_TOK_HDR_LEN + headskip; /* Trim off the trailing "extra count" and checksum blob */ - buf->len -= ec + GSS_KRB5_TOK_HDR_LEN + tailskip; + xdr_buf_trim(buf, ec + GSS_KRB5_TOK_HDR_LEN + tailskip); + *align = XDR_QUADLEN(GSS_KRB5_TOK_HDR_LEN + headskip); + *slack = *align + XDR_QUADLEN(ec + GSS_KRB5_TOK_HDR_LEN + tailskip); return GSS_S_COMPLETE; } @@ -603,7 +614,8 @@ gss_wrap_kerberos(struct gss_ctx *gctx, int offset, } u32 -gss_unwrap_kerberos(struct gss_ctx *gctx, int offset, struct xdr_buf *buf) +gss_unwrap_kerberos(struct gss_ctx *gctx, int offset, + int len, struct xdr_buf *buf) { struct krb5_ctx *kctx = gctx->internal_ctx_id; @@ -613,9 +625,11 @@ gss_unwrap_kerberos(struct gss_ctx *gctx, int offset, struct xdr_buf *buf) case ENCTYPE_DES_CBC_RAW: case ENCTYPE_DES3_CBC_RAW: case ENCTYPE_ARCFOUR_HMAC: - return gss_unwrap_kerberos_v1(kctx, offset, buf); + return gss_unwrap_kerberos_v1(kctx, offset, len, buf, + &gctx->slack, &gctx->align); case ENCTYPE_AES128_CTS_HMAC_SHA1_96: case ENCTYPE_AES256_CTS_HMAC_SHA1_96: - return gss_unwrap_kerberos_v2(kctx, offset, buf); + return gss_unwrap_kerberos_v2(kctx, offset, len, buf, + &gctx->slack, &gctx->align); } } diff --git a/net/sunrpc/auth_gss/gss_mech_switch.c b/net/sunrpc/auth_gss/gss_mech_switch.c index db550bfc2642..fae632da1058 100644 --- a/net/sunrpc/auth_gss/gss_mech_switch.c +++ b/net/sunrpc/auth_gss/gss_mech_switch.c @@ -37,6 +37,8 @@ gss_mech_free(struct gss_api_mech *gm) for (i = 0; i < gm->gm_pf_num; i++) { pf = &gm->gm_pfs[i]; + if (pf->domain) + auth_domain_put(pf->domain); kfree(pf->auth_domain_name); pf->auth_domain_name = NULL; } @@ -59,6 +61,7 @@ make_auth_domain_name(char *name) static int gss_mech_svc_setup(struct gss_api_mech *gm) { + struct auth_domain *dom; struct pf_desc *pf; int i, status; @@ -68,10 +71,13 @@ gss_mech_svc_setup(struct gss_api_mech *gm) status = -ENOMEM; if (pf->auth_domain_name == NULL) goto out; - status = svcauth_gss_register_pseudoflavor(pf->pseudoflavor, - pf->auth_domain_name); - if (status) + dom = svcauth_gss_register_pseudoflavor( + pf->pseudoflavor, pf->auth_domain_name); + if (IS_ERR(dom)) { + status = PTR_ERR(dom); goto out; + } + pf->domain = dom; } return 0; out: @@ -411,10 +417,11 @@ gss_wrap(struct gss_ctx *ctx_id, u32 gss_unwrap(struct gss_ctx *ctx_id, int offset, + int len, struct xdr_buf *buf) { return ctx_id->mech_type->gm_ops - ->gss_unwrap(ctx_id, offset, buf); + ->gss_unwrap(ctx_id, offset, len, buf); } diff --git a/net/sunrpc/auth_gss/gss_rpc_upcall.c b/net/sunrpc/auth_gss/gss_rpc_upcall.c index 0349f455a862..af9c7f43859c 100644 --- a/net/sunrpc/auth_gss/gss_rpc_upcall.c +++ b/net/sunrpc/auth_gss/gss_rpc_upcall.c @@ -223,7 +223,7 @@ static int gssp_alloc_receive_pages(struct gssx_arg_accept_sec_context *arg) static char *gssp_stringify(struct xdr_netobj *netobj) { - return kstrndup(netobj->data, netobj->len, GFP_KERNEL); + return kmemdup_nul(netobj->data, netobj->len, GFP_KERNEL); } static void gssp_hostbased_service(char **principal) diff --git a/net/sunrpc/auth_gss/svcauth_gss.c b/net/sunrpc/auth_gss/svcauth_gss.c index 54ae5be62f6a..46027d0c903f 100644 --- a/net/sunrpc/auth_gss/svcauth_gss.c +++ b/net/sunrpc/auth_gss/svcauth_gss.c @@ -809,7 +809,7 @@ u32 svcauth_gss_flavor(struct auth_domain *dom) EXPORT_SYMBOL_GPL(svcauth_gss_flavor); -int +struct auth_domain * svcauth_gss_register_pseudoflavor(u32 pseudoflavor, char * name) { struct gss_domain *new; @@ -826,21 +826,23 @@ svcauth_gss_register_pseudoflavor(u32 pseudoflavor, char * name) new->h.flavour = &svcauthops_gss; new->pseudoflavor = pseudoflavor; - stat = 0; test = auth_domain_lookup(name, &new->h); - if (test != &new->h) { /* Duplicate registration */ + if (test != &new->h) { + pr_warn("svc: duplicate registration of gss pseudo flavour %s.\n", + name); + stat = -EADDRINUSE; auth_domain_put(test); - kfree(new->h.name); - goto out_free_dom; + goto out_free_name; } - return 0; + return test; +out_free_name: + kfree(new->h.name); out_free_dom: kfree(new); out: - return stat; + return ERR_PTR(stat); } - EXPORT_SYMBOL_GPL(svcauth_gss_register_pseudoflavor); static inline int @@ -906,7 +908,7 @@ unwrap_integ_data(struct svc_rqst *rqstp, struct xdr_buf *buf, u32 seq, struct g if (svc_getnl(&buf->head[0]) != seq) goto out; /* trim off the mic and padding at the end before returning */ - buf->len -= 4 + round_up_to_quad(mic.len); + xdr_buf_trim(buf, round_up_to_quad(mic.len) + 4); stat = 0; out: kfree(mic.data); @@ -934,7 +936,7 @@ static int unwrap_priv_data(struct svc_rqst *rqstp, struct xdr_buf *buf, u32 seq, struct gss_ctx *ctx) { u32 priv_len, maj_stat; - int pad, saved_len, remaining_len, offset; + int pad, remaining_len, offset; clear_bit(RQ_SPLICE_OK, &rqstp->rq_flags); @@ -954,12 +956,8 @@ unwrap_priv_data(struct svc_rqst *rqstp, struct xdr_buf *buf, u32 seq, struct gs buf->len -= pad; fix_priv_head(buf, pad); - /* Maybe it would be better to give gss_unwrap a length parameter: */ - saved_len = buf->len; - buf->len = priv_len; - maj_stat = gss_unwrap(ctx, 0, buf); + maj_stat = gss_unwrap(ctx, 0, priv_len, buf); pad = priv_len - buf->len; - buf->len = saved_len; buf->len -= pad; /* The upper layers assume the buffer is aligned on 4-byte boundaries. * In the krb5p case, at least, the data ends up offset, so we need to diff --git a/net/sunrpc/auth_gss/trace.c b/net/sunrpc/auth_gss/trace.c index 5576f1e66de9..49fa583d7f91 100644 --- a/net/sunrpc/auth_gss/trace.c +++ b/net/sunrpc/auth_gss/trace.c @@ -6,6 +6,7 @@ #include <linux/sunrpc/clnt.h> #include <linux/sunrpc/sched.h> #include <linux/sunrpc/gss_err.h> +#include <linux/sunrpc/auth_gss.h> #define CREATE_TRACE_POINTS #include <trace/events/rpcgss.h> diff --git a/net/sunrpc/cache.c b/net/sunrpc/cache.c index af0ddd28b081..baef5ee43dbb 100644 --- a/net/sunrpc/cache.c +++ b/net/sunrpc/cache.c @@ -529,7 +529,6 @@ void cache_purge(struct cache_detail *detail) { struct cache_head *ch = NULL; struct hlist_head *head = NULL; - struct hlist_node *tmp = NULL; int i = 0; spin_lock(&detail->hash_lock); @@ -541,7 +540,9 @@ void cache_purge(struct cache_detail *detail) dprintk("RPC: %d entries in %s cache\n", detail->entries, detail->name); for (i = 0; i < detail->hash_size; i++) { head = &detail->hash_table[i]; - hlist_for_each_entry_safe(ch, tmp, head, cache_list) { + while (!hlist_empty(head)) { + ch = hlist_entry(head->first, struct cache_head, + cache_list); sunrpc_begin_cache_remove_entry(ch, detail); spin_unlock(&detail->hash_lock); sunrpc_end_cache_remove_entry(ch, detail); diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 325a0858700f..a91d1cdad9d7 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -370,10 +370,6 @@ static struct rpc_clnt * rpc_new_client(const struct rpc_create_args *args, const char *nodename = args->nodename; int err; - /* sanity check the name before trying to print it */ - dprintk("RPC: creating %s client for %s (xprt %p)\n", - program->name, args->servername, xprt); - err = rpciod_up(); if (err) goto out_no_rpciod; @@ -436,6 +432,8 @@ static struct rpc_clnt * rpc_new_client(const struct rpc_create_args *args, goto out_no_path; if (parent) atomic_inc(&parent->cl_count); + + trace_rpc_clnt_new(clnt, xprt, program->name, args->servername); return clnt; out_no_path: @@ -450,6 +448,7 @@ out_err: out_no_rpciod: xprt_switch_put(xps); xprt_put(xprt); + trace_rpc_clnt_new_err(program->name, args->servername, err); return ERR_PTR(err); } @@ -634,10 +633,8 @@ static struct rpc_clnt *__rpc_clone_client(struct rpc_create_args *args, args->nodename = clnt->cl_nodename; new = rpc_new_client(args, xps, xprt, clnt); - if (IS_ERR(new)) { - err = PTR_ERR(new); - goto out_err; - } + if (IS_ERR(new)) + return new; /* Turn off autobind on clones */ new->cl_autobind = 0; @@ -650,7 +647,7 @@ static struct rpc_clnt *__rpc_clone_client(struct rpc_create_args *args, return new; out_err: - dprintk("RPC: %s: returned error %d\n", __func__, err); + trace_rpc_clnt_clone_err(clnt, err); return ERR_PTR(err); } @@ -723,11 +720,8 @@ int rpc_switch_client_transport(struct rpc_clnt *clnt, int err; xprt = xprt_create_transport(args); - if (IS_ERR(xprt)) { - dprintk("RPC: failed to create new xprt for clnt %p\n", - clnt); + if (IS_ERR(xprt)) return PTR_ERR(xprt); - } xps = xprt_switch_alloc(xprt, GFP_KERNEL); if (xps == NULL) { @@ -767,7 +761,7 @@ int rpc_switch_client_transport(struct rpc_clnt *clnt, rpc_release_client(parent); xprt_switch_put(oldxps); xprt_put(old); - dprintk("RPC: replaced xprt for clnt %p\n", clnt); + trace_rpc_clnt_replace_xprt(clnt); return 0; out_revert: @@ -777,7 +771,7 @@ out_revert: rpc_client_register(clnt, pseudoflavor, NULL); xprt_switch_put(xps); xprt_put(xprt); - dprintk("RPC: failed to switch xprt for clnt %p\n", clnt); + trace_rpc_clnt_replace_xprt_err(clnt); return err; } EXPORT_SYMBOL_GPL(rpc_switch_client_transport); @@ -844,10 +838,11 @@ void rpc_killall_tasks(struct rpc_clnt *clnt) if (list_empty(&clnt->cl_tasks)) return; - dprintk("RPC: killing all tasks for client %p\n", clnt); + /* * Spin lock all_tasks to prevent changes... */ + trace_rpc_clnt_killall(clnt); spin_lock(&clnt->cl_lock); list_for_each_entry(rovr, &clnt->cl_tasks, tk_task) rpc_signal_task(rovr); @@ -863,9 +858,7 @@ void rpc_shutdown_client(struct rpc_clnt *clnt) { might_sleep(); - dprintk_rcu("RPC: shutting down %s client for %s\n", - clnt->cl_program->name, - rcu_dereference(clnt->cl_xprt)->servername); + trace_rpc_clnt_shutdown(clnt); while (!list_empty(&clnt->cl_tasks)) { rpc_killall_tasks(clnt); @@ -880,27 +873,40 @@ EXPORT_SYMBOL_GPL(rpc_shutdown_client); /* * Free an RPC client */ +static void rpc_free_client_work(struct work_struct *work) +{ + struct rpc_clnt *clnt = container_of(work, struct rpc_clnt, cl_work); + + trace_rpc_clnt_free(clnt); + + /* These might block on processes that might allocate memory, + * so they cannot be called in rpciod, so they are handled separately + * here. + */ + rpc_clnt_debugfs_unregister(clnt); + rpc_free_clid(clnt); + rpc_clnt_remove_pipedir(clnt); + xprt_put(rcu_dereference_raw(clnt->cl_xprt)); + + kfree(clnt); + rpciod_down(); +} static struct rpc_clnt * rpc_free_client(struct rpc_clnt *clnt) { struct rpc_clnt *parent = NULL; - dprintk_rcu("RPC: destroying %s client for %s\n", - clnt->cl_program->name, - rcu_dereference(clnt->cl_xprt)->servername); + trace_rpc_clnt_release(clnt); if (clnt->cl_parent != clnt) parent = clnt->cl_parent; - rpc_clnt_debugfs_unregister(clnt); - rpc_clnt_remove_pipedir(clnt); rpc_unregister_client(clnt); rpc_free_iostats(clnt->cl_metrics); clnt->cl_metrics = NULL; - xprt_put(rcu_dereference_raw(clnt->cl_xprt)); xprt_iter_destroy(&clnt->cl_xpi); - rpciod_down(); put_cred(clnt->cl_cred); - rpc_free_clid(clnt); - kfree(clnt); + + INIT_WORK(&clnt->cl_work, rpc_free_client_work); + schedule_work(&clnt->cl_work); return parent; } @@ -932,8 +938,6 @@ rpc_free_auth(struct rpc_clnt *clnt) void rpc_release_client(struct rpc_clnt *clnt) { - dprintk("RPC: rpc_release_client(%p)\n", clnt); - do { if (list_empty(&clnt->cl_tasks)) wake_up(&destroy_wait); @@ -1257,7 +1261,7 @@ void rpc_prepare_reply_pages(struct rpc_rqst *req, struct page **pages, hdrsize += RPC_REPHDRSIZE + req->rq_cred->cr_auth->au_ralign - 1; xdr_inline_pages(&req->rq_rcv_buf, hdrsize << 2, pages, base, len); - trace_rpc_reply_pages(req); + trace_rpc_xdr_reply_pages(req->rq_task, &req->rq_rcv_buf); } EXPORT_SYMBOL_GPL(rpc_prepare_reply_pages); @@ -1611,6 +1615,7 @@ const char static void __rpc_call_rpcerror(struct rpc_task *task, int tk_status, int rpc_status) { + trace_rpc_call_rpcerror(task, tk_status, rpc_status); task->tk_rpc_status = rpc_status; rpc_exit(task, tk_status); } @@ -2420,6 +2425,11 @@ rpc_check_timeout(struct rpc_task *task) { struct rpc_clnt *clnt = task->tk_client; + if (RPC_SIGNALLED(task)) { + rpc_call_rpcerror(task, -ERESTARTSYS); + return; + } + if (xprt_adjust_timeout(task->tk_rqstp) == 0) return; @@ -2513,7 +2523,7 @@ call_decode(struct rpc_task *task) goto out; req->rq_rcv_buf.len = req->rq_private_buf.len; - trace_xprt_recvfrom(&req->rq_rcv_buf); + trace_rpc_xdr_recvfrom(task, &req->rq_rcv_buf); /* Check that the softirq receive buffer is valid */ WARN_ON(memcmp(&req->rq_rcv_buf, &req->rq_private_buf, @@ -2742,7 +2752,8 @@ struct rpc_task *rpc_call_null_helper(struct rpc_clnt *clnt, .rpc_op_cred = cred, .callback_ops = (ops != NULL) ? ops : &rpc_default_ops, .callback_data = data, - .flags = flags | RPC_TASK_NULLCREDS, + .flags = flags | RPC_TASK_SOFT | RPC_TASK_SOFTCONN | + RPC_TASK_NULLCREDS, }; return rpc_run_task(&task_setup_data); @@ -2805,11 +2816,9 @@ int rpc_clnt_test_and_add_xprt(struct rpc_clnt *clnt, goto success; } - task = rpc_call_null_helper(clnt, xprt, NULL, - RPC_TASK_SOFT|RPC_TASK_SOFTCONN|RPC_TASK_ASYNC|RPC_TASK_NULLCREDS, + task = rpc_call_null_helper(clnt, xprt, NULL, RPC_TASK_ASYNC, &rpc_cb_add_xprt_call_ops, data); - if (IS_ERR(task)) - return PTR_ERR(task); + rpc_put_task(task); success: return 1; @@ -2850,9 +2859,7 @@ int rpc_clnt_setup_test_and_add_xprt(struct rpc_clnt *clnt, goto out_err; /* Test the connection */ - task = rpc_call_null_helper(clnt, xprt, NULL, - RPC_TASK_SOFT | RPC_TASK_SOFTCONN | RPC_TASK_NULLCREDS, - NULL, NULL); + task = rpc_call_null_helper(clnt, xprt, NULL, 0, NULL, NULL); if (IS_ERR(task)) { status = PTR_ERR(task); goto out_err; diff --git a/net/sunrpc/rpcb_clnt.c b/net/sunrpc/rpcb_clnt.c index 4a020b688860..c27123e6ba80 100644 --- a/net/sunrpc/rpcb_clnt.c +++ b/net/sunrpc/rpcb_clnt.c @@ -795,12 +795,6 @@ void rpcb_getport_async(struct rpc_task *task) child = rpcb_call_async(rpcb_clnt, map, proc); rpc_release_client(rpcb_clnt); - if (IS_ERR(child)) { - /* rpcb_map_release() has freed the arguments */ - dprintk("RPC: %5u %s: rpc_run_task failed\n", - task->tk_pid, __func__); - return; - } xprt->stat.bind_count++; rpc_put_task(child); diff --git a/net/sunrpc/sunrpc.h b/net/sunrpc/sunrpc.h index 47a756503d11..f6fe2e6cd65a 100644 --- a/net/sunrpc/sunrpc.h +++ b/net/sunrpc/sunrpc.h @@ -52,4 +52,5 @@ static inline int sock_is_loopback(struct sock *sk) int rpc_clients_notifier_register(void); void rpc_clients_notifier_unregister(void); +void auth_domain_cleanup(void); #endif /* _NET_SUNRPC_SUNRPC_H */ diff --git a/net/sunrpc/sunrpc_syms.c b/net/sunrpc/sunrpc_syms.c index f9edaa9174a4..236fadc4a439 100644 --- a/net/sunrpc/sunrpc_syms.c +++ b/net/sunrpc/sunrpc_syms.c @@ -23,6 +23,7 @@ #include <linux/sunrpc/rpc_pipe_fs.h> #include <linux/sunrpc/xprtsock.h> +#include "sunrpc.h" #include "netns.h" unsigned int sunrpc_net_id; @@ -131,6 +132,7 @@ cleanup_sunrpc(void) unregister_rpc_pipefs(); rpc_destroy_mempool(); unregister_pernet_subsys(&sunrpc_net_ops); + auth_domain_cleanup(); #if IS_ENABLED(CONFIG_SUNRPC_DEBUG) rpc_unregister_sysctl(); #endif diff --git a/net/sunrpc/svc.c b/net/sunrpc/svc.c index 9ed3126600ce..c211b607239e 100644 --- a/net/sunrpc/svc.c +++ b/net/sunrpc/svc.c @@ -88,15 +88,15 @@ param_get_pool_mode(char *buf, const struct kernel_param *kp) switch (*ip) { case SVC_POOL_AUTO: - return strlcpy(buf, "auto", 20); + return strlcpy(buf, "auto\n", 20); case SVC_POOL_GLOBAL: - return strlcpy(buf, "global", 20); + return strlcpy(buf, "global\n", 20); case SVC_POOL_PERCPU: - return strlcpy(buf, "percpu", 20); + return strlcpy(buf, "percpu\n", 20); case SVC_POOL_PERNODE: - return strlcpy(buf, "pernode", 20); + return strlcpy(buf, "pernode\n", 20); default: - return sprintf(buf, "%d", *ip); + return sprintf(buf, "%d\n", *ip); } } @@ -991,6 +991,7 @@ static int __svc_register(struct net *net, const char *progname, #endif } + trace_svc_register(progname, version, protocol, port, family, error); return error; } @@ -1000,11 +1001,6 @@ int svc_rpcbind_set_version(struct net *net, unsigned short proto, unsigned short port) { - dprintk("svc: svc_register(%sv%d, %s, %u, %u)\n", - progp->pg_name, version, - proto == IPPROTO_UDP? "udp" : "tcp", - port, family); - return __svc_register(net, progp->pg_name, progp->pg_prog, version, family, proto, port); @@ -1024,11 +1020,8 @@ int svc_generic_rpcbind_set(struct net *net, return 0; if (vers->vs_hidden) { - dprintk("svc: svc_register(%sv%d, %s, %u, %u)" - " (but not telling portmap)\n", - progp->pg_name, version, - proto == IPPROTO_UDP? "udp" : "tcp", - port, family); + trace_svc_noregister(progp->pg_name, version, proto, + port, family, 0); return 0; } @@ -1106,8 +1099,7 @@ static void __svc_unregister(struct net *net, const u32 program, const u32 versi if (error == -EPROTONOSUPPORT) error = rpcb_register(net, program, version, 0, 0); - dprintk("svc: %s(%sv%u), error %d\n", - __func__, progname, version, error); + trace_svc_unregister(progname, version, error); } /* @@ -1132,9 +1124,6 @@ static void svc_unregister(const struct svc_serv *serv, struct net *net) continue; if (progp->pg_vers[i]->vs_hidden) continue; - - dprintk("svc: attempting to unregister %sv%u\n", - progp->pg_name, i); __svc_unregister(net, progp->pg_prog, i, progp->pg_name); } } diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c index e27e3532ec75..43cf8dbde898 100644 --- a/net/sunrpc/svc_xprt.c +++ b/net/sunrpc/svc_xprt.c @@ -153,6 +153,7 @@ static void svc_xprt_free(struct kref *kref) xprt_put(xprt->xpt_bc_xprt); if (xprt->xpt_bc_xps) xprt_switch_put(xprt->xpt_bc_xps); + trace_svc_xprt_free(xprt); xprt->xpt_ops->xpo_free(xprt); module_put(owner); } @@ -206,6 +207,7 @@ static struct svc_xprt *__svc_xpo_create(struct svc_xprt_class *xcl, .sin6_port = htons(port), }; #endif + struct svc_xprt *xprt; struct sockaddr *sap; size_t len; @@ -224,7 +226,11 @@ static struct svc_xprt *__svc_xpo_create(struct svc_xprt_class *xcl, return ERR_PTR(-EAFNOSUPPORT); } - return xcl->xcl_ops->xpo_create(serv, net, sap, len, flags); + xprt = xcl->xcl_ops->xpo_create(serv, net, sap, len, flags); + if (IS_ERR(xprt)) + trace_svc_xprt_create_err(serv->sv_program->pg_name, + xcl->xcl_name, sap, xprt); + return xprt; } /* @@ -304,15 +310,11 @@ int svc_create_xprt(struct svc_serv *serv, const char *xprt_name, { int err; - dprintk("svc: creating transport %s[%d]\n", xprt_name, port); err = _svc_create_xprt(serv, xprt_name, net, family, port, flags, cred); if (err == -EPROTONOSUPPORT) { request_module("svc%s", xprt_name); err = _svc_create_xprt(serv, xprt_name, net, family, port, flags, cred); } - if (err < 0) - dprintk("svc: transport %s not found, err %d\n", - xprt_name, -err); return err; } EXPORT_SYMBOL_GPL(svc_create_xprt); @@ -780,7 +782,6 @@ static int svc_handle_xprt(struct svc_rqst *rqstp, struct svc_xprt *xprt) int len = 0; if (test_bit(XPT_CLOSE, &xprt->xpt_flags)) { - dprintk("svc_recv: found XPT_CLOSE\n"); if (test_and_clear_bit(XPT_KILL_TEMP, &xprt->xpt_flags)) xprt->xpt_ops->xpo_kill_temp_xprt(xprt); svc_delete_xprt(xprt); @@ -799,6 +800,7 @@ static int svc_handle_xprt(struct svc_rqst *rqstp, struct svc_xprt *xprt) if (newxpt) { newxpt->xpt_cred = get_cred(xprt->xpt_cred); svc_add_new_temp_xprt(serv, newxpt); + trace_svc_xprt_accept(newxpt, serv->sv_name); } else module_put(xprt->xpt_class->xcl_owner); } else if (svc_xprt_reserve_slot(rqstp, xprt)) { @@ -812,7 +814,7 @@ static int svc_handle_xprt(struct svc_rqst *rqstp, struct svc_xprt *xprt) else len = xprt->xpt_ops->xpo_recvfrom(rqstp); if (len > 0) - trace_svc_recvfrom(&rqstp->rq_arg); + trace_svc_xdr_recvfrom(rqstp, &rqstp->rq_arg); rqstp->rq_stime = ktime_get(); rqstp->rq_reserved = serv->sv_max_mesg; atomic_add(rqstp->rq_reserved, &xprt->xpt_reserved); @@ -835,14 +837,6 @@ int svc_recv(struct svc_rqst *rqstp, long timeout) struct svc_serv *serv = rqstp->rq_server; int len, err; - dprintk("svc: server %p waiting for data (to = %ld)\n", - rqstp, timeout); - - if (rqstp->rq_xprt) - printk(KERN_ERR - "svc_recv: service %p, transport not NULL!\n", - rqstp); - err = svc_alloc_arg(rqstp); if (err) goto out; @@ -890,7 +884,6 @@ EXPORT_SYMBOL_GPL(svc_recv); void svc_drop(struct svc_rqst *rqstp) { trace_svc_drop(rqstp); - dprintk("svc: xprt %p dropped request\n", rqstp->rq_xprt); svc_xprt_release(rqstp); } EXPORT_SYMBOL_GPL(svc_drop); @@ -908,25 +901,16 @@ int svc_send(struct svc_rqst *rqstp) if (!xprt) goto out; - /* release the receive skb before sending the reply */ - xprt->xpt_ops->xpo_release_rqst(rqstp); - /* calculate over-all length */ xb = &rqstp->rq_res; xb->len = xb->head[0].iov_len + xb->page_len + xb->tail[0].iov_len; - trace_svc_sendto(xb); - - /* Grab mutex to serialize outgoing data. */ - mutex_lock(&xprt->xpt_mutex); + trace_svc_xdr_sendto(rqstp, xb); trace_svc_stats_latency(rqstp); - if (test_bit(XPT_DEAD, &xprt->xpt_flags) - || test_bit(XPT_CLOSE, &xprt->xpt_flags)) - len = -ENOTCONN; - else - len = xprt->xpt_ops->xpo_sendto(rqstp); - mutex_unlock(&xprt->xpt_mutex); + + len = xprt->xpt_ops->xpo_sendto(rqstp); + trace_svc_send(rqstp, len); svc_xprt_release(rqstp); @@ -1034,12 +1018,13 @@ static void svc_delete_xprt(struct svc_xprt *xprt) struct svc_serv *serv = xprt->xpt_server; struct svc_deferred_req *dr; - /* Only do this once */ if (test_and_set_bit(XPT_DEAD, &xprt->xpt_flags)) - BUG(); + return; - dprintk("svc: svc_delete_xprt(%p)\n", xprt); + trace_svc_xprt_detach(xprt); xprt->xpt_ops->xpo_detach(xprt); + if (xprt->xpt_bc_xprt) + xprt->xpt_bc_xprt->ops->close(xprt->xpt_bc_xprt); spin_lock_bh(&serv->sv_lock); list_del_init(&xprt->xpt_list); @@ -1057,6 +1042,7 @@ static void svc_delete_xprt(struct svc_xprt *xprt) void svc_close_xprt(struct svc_xprt *xprt) { + trace_svc_xprt_close(xprt); set_bit(XPT_CLOSE, &xprt->xpt_flags); if (test_and_set_bit(XPT_BUSY, &xprt->xpt_flags)) /* someone else will have to effect the close */ @@ -1159,16 +1145,15 @@ static void svc_revisit(struct cache_deferred_req *dreq, int too_many) set_bit(XPT_DEFERRED, &xprt->xpt_flags); if (too_many || test_bit(XPT_DEAD, &xprt->xpt_flags)) { spin_unlock(&xprt->xpt_lock); - dprintk("revisit canceled\n"); + trace_svc_defer_drop(dr); svc_xprt_put(xprt); - trace_svc_drop_deferred(dr); kfree(dr); return; } - dprintk("revisit queued\n"); dr->xprt = NULL; list_add(&dr->handle.recent, &xprt->xpt_deferred); spin_unlock(&xprt->xpt_lock); + trace_svc_defer_queue(dr); svc_xprt_enqueue(xprt); svc_xprt_put(xprt); } @@ -1214,22 +1199,24 @@ static struct cache_deferred_req *svc_defer(struct cache_req *req) memcpy(dr->args, rqstp->rq_arg.head[0].iov_base - skip, dr->argslen << 2); } + trace_svc_defer(rqstp); svc_xprt_get(rqstp->rq_xprt); dr->xprt = rqstp->rq_xprt; set_bit(RQ_DROPME, &rqstp->rq_flags); dr->handle.revisit = svc_revisit; - trace_svc_defer(rqstp); return &dr->handle; } /* * recv data from a deferred request into an active one */ -static int svc_deferred_recv(struct svc_rqst *rqstp) +static noinline int svc_deferred_recv(struct svc_rqst *rqstp) { struct svc_deferred_req *dr = rqstp->rq_deferred; + trace_svc_defer_recv(dr); + /* setup iov_base past transport header */ rqstp->rq_arg.head[0].iov_base = dr->args + (dr->xprt_hlen>>2); /* The iov_len does not include the transport header bytes */ @@ -1260,7 +1247,6 @@ static struct svc_deferred_req *svc_deferred_dequeue(struct svc_xprt *xprt) struct svc_deferred_req, handle.recent); list_del_init(&dr->handle.recent); - trace_svc_revisit_deferred(dr); } else clear_bit(XPT_DEFERRED, &xprt->xpt_flags); spin_unlock(&xprt->xpt_lock); diff --git a/net/sunrpc/svcauth.c b/net/sunrpc/svcauth.c index 552617e3467b..998b196b6176 100644 --- a/net/sunrpc/svcauth.c +++ b/net/sunrpc/svcauth.c @@ -21,6 +21,8 @@ #include <trace/events/sunrpc.h> +#include "sunrpc.h" + #define RPCDBG_FACILITY RPCDBG_AUTH @@ -205,3 +207,26 @@ struct auth_domain *auth_domain_find(char *name) return NULL; } EXPORT_SYMBOL_GPL(auth_domain_find); + +/** + * auth_domain_cleanup - check that the auth_domain table is empty + * + * On module unload the auth_domain_table must be empty. To make it + * easier to catch bugs which don't clean up domains properly, we + * warn if anything remains in the table at cleanup time. + * + * Note that we cannot proactively remove the domains at this stage. + * The ->release() function might be in a module that has already been + * unloaded. + */ + +void auth_domain_cleanup(void) +{ + int h; + struct auth_domain *hp; + + for (h = 0; h < DN_HASHMAX; h++) + hlist_for_each_entry(hp, &auth_domain_table[h], hash) + pr_warn("svc: domain %s still present at module unload.\n", + hp->name); +} diff --git a/net/sunrpc/svcauth_unix.c b/net/sunrpc/svcauth_unix.c index 6c8f802c4261..97c0bddba7a3 100644 --- a/net/sunrpc/svcauth_unix.c +++ b/net/sunrpc/svcauth_unix.c @@ -332,15 +332,6 @@ static int __ip_map_update(struct cache_detail *cd, struct ip_map *ipm, return 0; } -static inline int ip_map_update(struct net *net, struct ip_map *ipm, - struct unix_domain *udom, time64_t expiry) -{ - struct sunrpc_net *sn; - - sn = net_generic(net, sunrpc_net_id); - return __ip_map_update(sn->ip_map_cache, ipm, udom, expiry); -} - void svcauth_unix_purge(struct net *net) { struct sunrpc_net *sn; diff --git a/net/sunrpc/svcsock.c b/net/sunrpc/svcsock.c index 519cf9c4f8fd..5c4ec9386f81 100644 --- a/net/sunrpc/svcsock.c +++ b/net/sunrpc/svcsock.c @@ -45,7 +45,6 @@ #include <net/tcp_states.h> #include <linux/uaccess.h> #include <asm/ioctls.h> -#include <trace/events/skb.h> #include <linux/sunrpc/types.h> #include <linux/sunrpc/clnt.h> @@ -55,6 +54,8 @@ #include <linux/sunrpc/stats.h> #include <linux/sunrpc/xprt.h> +#include <trace/events/sunrpc.h> + #include "socklib.h" #include "sunrpc.h" @@ -108,31 +109,35 @@ static void svc_reclassify_socket(struct socket *sock) } #endif -/* - * Release an skbuff after use +/** + * svc_tcp_release_rqst - Release transport-related resources + * @rqstp: request structure with resources to be released + * */ -static void svc_release_skb(struct svc_rqst *rqstp) +static void svc_tcp_release_rqst(struct svc_rqst *rqstp) { struct sk_buff *skb = rqstp->rq_xprt_ctxt; if (skb) { struct svc_sock *svsk = container_of(rqstp->rq_xprt, struct svc_sock, sk_xprt); - rqstp->rq_xprt_ctxt = NULL; - dprintk("svc: service %p, releasing skb %p\n", rqstp, skb); + rqstp->rq_xprt_ctxt = NULL; skb_free_datagram_locked(svsk->sk_sk, skb); } } -static void svc_release_udp_skb(struct svc_rqst *rqstp) +/** + * svc_udp_release_rqst - Release transport-related resources + * @rqstp: request structure with resources to be released + * + */ +static void svc_udp_release_rqst(struct svc_rqst *rqstp) { struct sk_buff *skb = rqstp->rq_xprt_ctxt; if (skb) { rqstp->rq_xprt_ctxt = NULL; - - dprintk("svc: service %p, releasing skb %p\n", rqstp, skb); consume_skb(skb); } } @@ -218,34 +223,68 @@ static int svc_one_sock_name(struct svc_sock *svsk, char *buf, int remaining) return len; } +#if ARCH_IMPLEMENTS_FLUSH_DCACHE_PAGE +static void svc_flush_bvec(const struct bio_vec *bvec, size_t size, size_t seek) +{ + struct bvec_iter bi = { + .bi_size = size, + }; + struct bio_vec bv; + + bvec_iter_advance(bvec, &bi, seek & PAGE_MASK); + for_each_bvec(bv, bvec, bi, bi) + flush_dcache_page(bv.bv_page); +} +#else +static inline void svc_flush_bvec(const struct bio_vec *bvec, size_t size, + size_t seek) +{ +} +#endif + /* - * Generic recvfrom routine. + * Read from @rqstp's transport socket. The incoming message fills whole + * pages in @rqstp's rq_pages array until the last page of the message + * has been received into a partial page. */ -static ssize_t svc_recvfrom(struct svc_rqst *rqstp, struct kvec *iov, - unsigned int nr, size_t buflen, unsigned int base) +static ssize_t svc_tcp_read_msg(struct svc_rqst *rqstp, size_t buflen, + size_t seek) { struct svc_sock *svsk = container_of(rqstp->rq_xprt, struct svc_sock, sk_xprt); + struct bio_vec *bvec = rqstp->rq_bvec; struct msghdr msg = { NULL }; + unsigned int i; ssize_t len; + size_t t; rqstp->rq_xprt_hlen = 0; clear_bit(XPT_DATA, &svsk->sk_xprt.xpt_flags); - iov_iter_kvec(&msg.msg_iter, READ, iov, nr, buflen); - if (base != 0) { - iov_iter_advance(&msg.msg_iter, base); - buflen -= base; + + for (i = 0, t = 0; t < buflen; i++, t += PAGE_SIZE) { + bvec[i].bv_page = rqstp->rq_pages[i]; + bvec[i].bv_len = PAGE_SIZE; + bvec[i].bv_offset = 0; + } + rqstp->rq_respages = &rqstp->rq_pages[i]; + rqstp->rq_next_page = rqstp->rq_respages + 1; + + iov_iter_bvec(&msg.msg_iter, READ, bvec, i, buflen); + if (seek) { + iov_iter_advance(&msg.msg_iter, seek); + buflen -= seek; } len = sock_recvmsg(svsk->sk_sock, &msg, MSG_DONTWAIT); + if (len > 0) + svc_flush_bvec(bvec, len, seek); + /* If we read a full record, then assume there may be more * data to read (stream based sockets only!) */ if (len == buflen) set_bit(XPT_DATA, &svsk->sk_xprt.xpt_flags); - dprintk("svc: socket %p recvfrom(%p, %zu) = %zd\n", - svsk, iov[0].iov_base, iov[0].iov_len, len); return len; } @@ -282,13 +321,10 @@ static void svc_data_ready(struct sock *sk) struct svc_sock *svsk = (struct svc_sock *)sk->sk_user_data; if (svsk) { - dprintk("svc: socket %p(inet %p), busy=%d\n", - svsk, sk, - test_bit(XPT_BUSY, &svsk->sk_xprt.xpt_flags)); - /* Refer to svc_setup_socket() for details. */ rmb(); svsk->sk_odata(sk); + trace_svcsock_data_ready(&svsk->sk_xprt, 0); if (!test_and_set_bit(XPT_DATA, &svsk->sk_xprt.xpt_flags)) svc_xprt_enqueue(&svsk->sk_xprt); } @@ -302,11 +338,9 @@ static void svc_write_space(struct sock *sk) struct svc_sock *svsk = (struct svc_sock *)(sk->sk_user_data); if (svsk) { - dprintk("svc: socket %p(inet %p), write_space busy=%d\n", - svsk, sk, test_bit(XPT_BUSY, &svsk->sk_xprt.xpt_flags)); - /* Refer to svc_setup_socket() for details. */ rmb(); + trace_svcsock_write_space(&svsk->sk_xprt, 0); svsk->sk_owspace(sk); svc_xprt_enqueue(&svsk->sk_xprt); } @@ -323,17 +357,9 @@ static int svc_tcp_has_wspace(struct svc_xprt *xprt) static void svc_tcp_kill_temp_xprt(struct svc_xprt *xprt) { - struct svc_sock *svsk; - struct socket *sock; - struct linger no_linger = { - .l_onoff = 1, - .l_linger = 0, - }; + struct svc_sock *svsk = container_of(xprt, struct svc_sock, sk_xprt); - svsk = container_of(xprt, struct svc_sock, sk_xprt); - sock = svsk->sk_sock; - kernel_setsockopt(sock, SOL_SOCKET, SO_LINGER, - (char *)&no_linger, sizeof(no_linger)); + sock_no_linger(svsk->sk_sock->sk); } /* @@ -391,8 +417,15 @@ static int svc_udp_get_dest_address(struct svc_rqst *rqstp, return 0; } -/* - * Receive a datagram from a UDP socket. +/** + * svc_udp_recvfrom - Receive a datagram from a UDP socket. + * @rqstp: request structure into which to receive an RPC Call + * + * Called in a loop when XPT_DATA has been set. + * + * Returns: + * On success, the number of bytes in a received RPC Call, or + * %0 if a complete RPC Call message was not ready to return */ static int svc_udp_recvfrom(struct svc_rqst *rqstp) { @@ -426,20 +459,14 @@ static int svc_udp_recvfrom(struct svc_rqst *rqstp) svc_sock_setbufsize(svsk, serv->sv_nrthreads + 3); clear_bit(XPT_DATA, &svsk->sk_xprt.xpt_flags); - skb = NULL; err = kernel_recvmsg(svsk->sk_sock, &msg, NULL, 0, 0, MSG_PEEK | MSG_DONTWAIT); - if (err >= 0) - skb = skb_recv_udp(svsk->sk_sk, 0, 1, &err); - - if (skb == NULL) { - if (err != -EAGAIN) { - /* possibly an icmp error */ - dprintk("svc: recvfrom returned error %d\n", -err); - set_bit(XPT_DATA, &svsk->sk_xprt.xpt_flags); - } - return 0; - } + if (err < 0) + goto out_recv_err; + skb = skb_recv_udp(svsk->sk_sk, 0, 1, &err); + if (!skb) + goto out_recv_err; + len = svc_addr_len(svc_addr(rqstp)); rqstp->rq_addrlen = len; if (skb->tstamp == 0) { @@ -450,26 +477,21 @@ static int svc_udp_recvfrom(struct svc_rqst *rqstp) sock_write_timestamp(svsk->sk_sk, skb->tstamp); set_bit(XPT_DATA, &svsk->sk_xprt.xpt_flags); /* there may be more data... */ - len = skb->len; + len = skb->len; rqstp->rq_arg.len = len; + trace_svcsock_udp_recv(&svsk->sk_xprt, len); rqstp->rq_prot = IPPROTO_UDP; - if (!svc_udp_get_dest_address(rqstp, cmh)) { - net_warn_ratelimited("svc: received unknown control message %d/%d; dropping RPC reply datagram\n", - cmh->cmsg_level, cmh->cmsg_type); - goto out_free; - } + if (!svc_udp_get_dest_address(rqstp, cmh)) + goto out_cmsg_err; rqstp->rq_daddrlen = svc_addr_len(svc_daddr(rqstp)); if (skb_is_nonlinear(skb)) { /* we have to copy */ local_bh_disable(); - if (csum_partial_copy_to_xdr(&rqstp->rq_arg, skb)) { - local_bh_enable(); - /* checksum error */ - goto out_free; - } + if (csum_partial_copy_to_xdr(&rqstp->rq_arg, skb)) + goto out_bh_enable; local_bh_enable(); consume_skb(skb); } else { @@ -497,6 +519,20 @@ static int svc_udp_recvfrom(struct svc_rqst *rqstp) serv->sv_stats->netudpcnt++; return len; + +out_recv_err: + if (err != -EAGAIN) { + /* possibly an icmp error */ + set_bit(XPT_DATA, &svsk->sk_xprt.xpt_flags); + } + trace_svcsock_udp_recv_err(&svsk->sk_xprt, err); + return 0; +out_cmsg_err: + net_warn_ratelimited("svc: received unknown control message %d/%d; dropping RPC reply datagram\n", + cmh->cmsg_level, cmh->cmsg_type); + goto out_free; +out_bh_enable: + local_bh_enable(); out_free: kfree_skb(skb); return 0; @@ -506,6 +542,9 @@ out_free: * svc_udp_sendto - Send out a reply on a UDP socket * @rqstp: completed svc_rqst * + * xpt_mutex ensures @rqstp's whole message is written to the socket + * without interruption. + * * Returns the number of bytes sent, or a negative errno. */ static int svc_udp_sendto(struct svc_rqst *rqstp) @@ -527,8 +566,15 @@ static int svc_udp_sendto(struct svc_rqst *rqstp) unsigned int uninitialized_var(sent); int err; + svc_udp_release_rqst(rqstp); + svc_set_cmsg_data(rqstp, cmh); + mutex_lock(&xprt->xpt_mutex); + + if (svc_xprt_is_dead(xprt)) + goto out_notconn; + err = xprt_sock_sendmsg(svsk->sk_sock, &msg, xdr, 0, 0, &sent); xdr_free_bvec(xdr); if (err == -ECONNREFUSED) { @@ -536,9 +582,16 @@ static int svc_udp_sendto(struct svc_rqst *rqstp) err = xprt_sock_sendmsg(svsk->sk_sock, &msg, xdr, 0, 0, &sent); xdr_free_bvec(xdr); } + trace_svcsock_udp_send(xprt, err); + + mutex_unlock(&xprt->xpt_mutex); if (err < 0) return err; return sent; + +out_notconn: + mutex_unlock(&xprt->xpt_mutex); + return -ENOTCONN; } static int svc_udp_has_wspace(struct svc_xprt *xprt) @@ -582,7 +635,7 @@ static const struct svc_xprt_ops svc_udp_ops = { .xpo_recvfrom = svc_udp_recvfrom, .xpo_sendto = svc_udp_sendto, .xpo_read_payload = svc_sock_read_payload, - .xpo_release_rqst = svc_release_udp_skb, + .xpo_release_rqst = svc_udp_release_rqst, .xpo_detach = svc_sock_detach, .xpo_free = svc_sock_free, .xpo_has_wspace = svc_udp_has_wspace, @@ -601,8 +654,6 @@ static struct svc_xprt_class svc_udp_class = { static void svc_udp_init(struct svc_sock *svsk, struct svc_serv *serv) { - int err, level, optname, one = 1; - svc_xprt_init(sock_net(svsk->sk_sock->sk), &svc_udp_class, &svsk->sk_xprt, serv); clear_bit(XPT_CACHE_AUTH, &svsk->sk_xprt.xpt_flags); @@ -622,19 +673,14 @@ static void svc_udp_init(struct svc_sock *svsk, struct svc_serv *serv) /* make sure we get destination address info */ switch (svsk->sk_sk->sk_family) { case AF_INET: - level = SOL_IP; - optname = IP_PKTINFO; + ip_sock_set_pktinfo(svsk->sk_sock->sk); break; case AF_INET6: - level = SOL_IPV6; - optname = IPV6_RECVPKTINFO; + ip6_sock_set_recvpktinfo(svsk->sk_sock->sk); break; default: BUG(); } - err = kernel_setsockopt(svsk->sk_sock, level, optname, - (char *)&one, sizeof(one)); - dprintk("svc: kernel_setsockopt returned %d\n", err); } /* @@ -645,9 +691,6 @@ static void svc_tcp_listen_data_ready(struct sock *sk) { struct svc_sock *svsk = (struct svc_sock *)sk->sk_user_data; - dprintk("svc: socket %p TCP (listen) state change %d\n", - sk, sk->sk_state); - if (svsk) { /* Refer to svc_setup_socket() for details. */ rmb(); @@ -668,8 +711,7 @@ static void svc_tcp_listen_data_ready(struct sock *sk) if (svsk) { set_bit(XPT_CONN, &svsk->sk_xprt.xpt_flags); svc_xprt_enqueue(&svsk->sk_xprt); - } else - printk("svc: socket %p: no user data\n", sk); + } } } @@ -680,15 +722,11 @@ static void svc_tcp_state_change(struct sock *sk) { struct svc_sock *svsk = (struct svc_sock *)sk->sk_user_data; - dprintk("svc: socket %p TCP (connected) state change %d (svsk %p)\n", - sk, sk->sk_state, sk->sk_user_data); - - if (!svsk) - printk("svc: socket %p: no user data\n", sk); - else { + if (svsk) { /* Refer to svc_setup_socket() for details. */ rmb(); svsk->sk_ostate(sk); + trace_svcsock_tcp_state(&svsk->sk_xprt, svsk->sk_sock); if (sk->sk_state != TCP_ESTABLISHED) { set_bit(XPT_CLOSE, &svsk->sk_xprt.xpt_flags); svc_xprt_enqueue(&svsk->sk_xprt); @@ -709,9 +747,7 @@ static struct svc_xprt *svc_tcp_accept(struct svc_xprt *xprt) struct socket *newsock; struct svc_sock *newsvsk; int err, slen; - RPC_IFDEBUG(char buf[RPC_MAX_ADDRBUFLEN]); - dprintk("svc: tcp_accept %p sock %p\n", svsk, sock); if (!sock) return NULL; @@ -724,30 +760,18 @@ static struct svc_xprt *svc_tcp_accept(struct svc_xprt *xprt) else if (err != -EAGAIN) net_warn_ratelimited("%s: accept failed (err %d)!\n", serv->sv_name, -err); + trace_svcsock_accept_err(xprt, serv->sv_name, err); return NULL; } set_bit(XPT_CONN, &svsk->sk_xprt.xpt_flags); err = kernel_getpeername(newsock, sin); if (err < 0) { - net_warn_ratelimited("%s: peername failed (err %d)!\n", - serv->sv_name, -err); + trace_svcsock_getpeername_err(xprt, serv->sv_name, err); goto failed; /* aborted connection or whatever */ } slen = err; - /* Ideally, we would want to reject connections from unauthorized - * hosts here, but when we get encryption, the IP of the host won't - * tell us anything. For now just warn about unpriv connections. - */ - if (!svc_port_is_privileged(sin)) { - dprintk("%s: connect from unprivileged port: %s\n", - serv->sv_name, - __svc_print_addr(sin, buf, sizeof(buf))); - } - dprintk("%s: connect from %s\n", serv->sv_name, - __svc_print_addr(sin, buf, sizeof(buf))); - /* Reset the inherited callbacks before calling svc_setup_socket */ newsock->sk->sk_state_change = svsk->sk_ostate; newsock->sk->sk_data_ready = svsk->sk_odata; @@ -765,10 +789,8 @@ static struct svc_xprt *svc_tcp_accept(struct svc_xprt *xprt) svc_xprt_set_remote(&newsvsk->sk_xprt, sin, slen); err = kernel_getsockname(newsock, sin); slen = err; - if (unlikely(err < 0)) { - dprintk("svc_tcp_accept: kernel_getsockname error %d\n", -err); + if (unlikely(err < 0)) slen = offsetof(struct sockaddr, sa_data); - } svc_xprt_set_local(&newsvsk->sk_xprt, sin, slen); if (sock_is_loopback(newsock->sk)) @@ -785,13 +807,14 @@ failed: return NULL; } -static unsigned int svc_tcp_restore_pages(struct svc_sock *svsk, struct svc_rqst *rqstp) +static size_t svc_tcp_restore_pages(struct svc_sock *svsk, + struct svc_rqst *rqstp) { - unsigned int i, len, npages; + size_t len = svsk->sk_datalen; + unsigned int i, npages; - if (svsk->sk_datalen == 0) + if (!len) return 0; - len = svsk->sk_datalen; npages = (len + PAGE_SIZE - 1) >> PAGE_SHIFT; for (i = 0; i < npages; i++) { if (rqstp->rq_pages[i] != NULL) @@ -840,47 +863,45 @@ out: } /* - * Receive fragment record header. - * If we haven't gotten the record length yet, get the next four bytes. + * Receive fragment record header into sk_marker. */ -static int svc_tcp_recv_record(struct svc_sock *svsk, struct svc_rqst *rqstp) +static ssize_t svc_tcp_read_marker(struct svc_sock *svsk, + struct svc_rqst *rqstp) { - struct svc_serv *serv = svsk->sk_xprt.xpt_server; - unsigned int want; - int len; + ssize_t want, len; + /* If we haven't gotten the record length yet, + * get the next four bytes. + */ if (svsk->sk_tcplen < sizeof(rpc_fraghdr)) { + struct msghdr msg = { NULL }; struct kvec iov; want = sizeof(rpc_fraghdr) - svsk->sk_tcplen; - iov.iov_base = ((char *) &svsk->sk_reclen) + svsk->sk_tcplen; + iov.iov_base = ((char *)&svsk->sk_marker) + svsk->sk_tcplen; iov.iov_len = want; - len = svc_recvfrom(rqstp, &iov, 1, want, 0); + iov_iter_kvec(&msg.msg_iter, READ, &iov, 1, want); + len = sock_recvmsg(svsk->sk_sock, &msg, MSG_DONTWAIT); if (len < 0) - goto error; + return len; svsk->sk_tcplen += len; - if (len < want) { - dprintk("svc: short recvfrom while reading record " - "length (%d of %d)\n", len, want); - return -EAGAIN; + /* call again to read the remaining bytes */ + goto err_short; } - - dprintk("svc: TCP record, %d bytes\n", svc_sock_reclen(svsk)); + trace_svcsock_marker(&svsk->sk_xprt, svsk->sk_marker); if (svc_sock_reclen(svsk) + svsk->sk_datalen > - serv->sv_max_mesg) { - net_notice_ratelimited("RPC: fragment too large: %d\n", - svc_sock_reclen(svsk)); - goto err_delete; - } + svsk->sk_xprt.xpt_server->sv_max_mesg) + goto err_too_large; } - return svc_sock_reclen(svsk); -error: - dprintk("RPC: TCP recv_record got %d\n", len); - return len; -err_delete: + +err_too_large: + net_notice_ratelimited("svc: %s %s RPC fragment too large: %d\n", + __func__, svsk->sk_xprt.xpt_server->sv_name, + svc_sock_reclen(svsk)); set_bit(XPT_CLOSE, &svsk->sk_xprt.xpt_flags); +err_short: return -EAGAIN; } @@ -929,87 +950,58 @@ unlock_eagain: return -EAGAIN; } -static int copy_pages_to_kvecs(struct kvec *vec, struct page **pages, int len) -{ - int i = 0; - int t = 0; - - while (t < len) { - vec[i].iov_base = page_address(pages[i]); - vec[i].iov_len = PAGE_SIZE; - i++; - t += PAGE_SIZE; - } - return i; -} - static void svc_tcp_fragment_received(struct svc_sock *svsk) { /* If we have more data, signal svc_xprt_enqueue() to try again */ - dprintk("svc: TCP %s record (%d bytes)\n", - svc_sock_final_rec(svsk) ? "final" : "nonfinal", - svc_sock_reclen(svsk)); svsk->sk_tcplen = 0; - svsk->sk_reclen = 0; + svsk->sk_marker = xdr_zero; } -/* - * Receive data from a TCP socket. +/** + * svc_tcp_recvfrom - Receive data from a TCP socket + * @rqstp: request structure into which to receive an RPC Call + * + * Called in a loop when XPT_DATA has been set. + * + * Read the 4-byte stream record marker, then use the record length + * in that marker to set up exactly the resources needed to receive + * the next RPC message into @rqstp. + * + * Returns: + * On success, the number of bytes in a received RPC Call, or + * %0 if a complete RPC Call message was not ready to return + * + * The zero return case handles partial receives and callback Replies. + * The state of a partial receive is preserved in the svc_sock for + * the next call to svc_tcp_recvfrom. */ static int svc_tcp_recvfrom(struct svc_rqst *rqstp) { struct svc_sock *svsk = container_of(rqstp->rq_xprt, struct svc_sock, sk_xprt); struct svc_serv *serv = svsk->sk_xprt.xpt_server; - int len; - struct kvec *vec; - unsigned int want, base; + size_t want, base; + ssize_t len; __be32 *p; __be32 calldir; - int pnum; - - dprintk("svc: tcp_recv %p data %d conn %d close %d\n", - svsk, test_bit(XPT_DATA, &svsk->sk_xprt.xpt_flags), - test_bit(XPT_CONN, &svsk->sk_xprt.xpt_flags), - test_bit(XPT_CLOSE, &svsk->sk_xprt.xpt_flags)); - len = svc_tcp_recv_record(svsk, rqstp); + clear_bit(XPT_DATA, &svsk->sk_xprt.xpt_flags); + len = svc_tcp_read_marker(svsk, rqstp); if (len < 0) goto error; base = svc_tcp_restore_pages(svsk, rqstp); - want = svc_sock_reclen(svsk) - (svsk->sk_tcplen - sizeof(rpc_fraghdr)); - - vec = rqstp->rq_vec; - - pnum = copy_pages_to_kvecs(&vec[0], &rqstp->rq_pages[0], base + want); - - rqstp->rq_respages = &rqstp->rq_pages[pnum]; - rqstp->rq_next_page = rqstp->rq_respages + 1; - - /* Now receive data */ - len = svc_recvfrom(rqstp, vec, pnum, base + want, base); + want = len - (svsk->sk_tcplen - sizeof(rpc_fraghdr)); + len = svc_tcp_read_msg(rqstp, base + want, base); if (len >= 0) { + trace_svcsock_tcp_recv(&svsk->sk_xprt, len); svsk->sk_tcplen += len; svsk->sk_datalen += len; } - if (len != want || !svc_sock_final_rec(svsk)) { - svc_tcp_save_pages(svsk, rqstp); - if (len < 0 && len != -EAGAIN) - goto err_delete; - if (len == want) - svc_tcp_fragment_received(svsk); - else - dprintk("svc: incomplete TCP record (%d of %d)\n", - (int)(svsk->sk_tcplen - sizeof(rpc_fraghdr)), - svc_sock_reclen(svsk)); - goto err_noclose; - } - - if (svsk->sk_datalen < 8) { - svsk->sk_datalen = 0; - goto err_delete; /* client is nuts. */ - } + if (len != want || !svc_sock_final_rec(svsk)) + goto err_incomplete; + if (svsk->sk_datalen < 8) + goto err_nuts; rqstp->rq_arg.len = svsk->sk_datalen; rqstp->rq_arg.page_base = 0; @@ -1044,14 +1036,26 @@ static int svc_tcp_recvfrom(struct svc_rqst *rqstp) return rqstp->rq_arg.len; +err_incomplete: + svc_tcp_save_pages(svsk, rqstp); + if (len < 0 && len != -EAGAIN) + goto err_delete; + if (len == want) + svc_tcp_fragment_received(svsk); + else + trace_svcsock_tcp_recv_short(&svsk->sk_xprt, + svc_sock_reclen(svsk), + svsk->sk_tcplen - sizeof(rpc_fraghdr)); + goto err_noclose; error: if (len != -EAGAIN) goto err_delete; - dprintk("RPC: TCP recvfrom got EAGAIN\n"); + trace_svcsock_tcp_recv_eagain(&svsk->sk_xprt, 0); return 0; +err_nuts: + svsk->sk_datalen = 0; err_delete: - printk(KERN_NOTICE "%s: recvfrom returned errno %d\n", - svsk->sk_xprt.xpt_server->sv_name, -len); + trace_svcsock_tcp_recv_err(&svsk->sk_xprt, len); set_bit(XPT_CLOSE, &svsk->sk_xprt.xpt_flags); err_noclose: return 0; /* record not complete */ @@ -1061,6 +1065,9 @@ err_noclose: * svc_tcp_sendto - Send out a reply on a TCP socket * @rqstp: completed svc_rqst * + * xpt_mutex ensures @rqstp's whole message is written to the socket + * without interruption. + * * Returns the number of bytes sent, or a negative errno. */ static int svc_tcp_sendto(struct svc_rqst *rqstp) @@ -1076,12 +1083,22 @@ static int svc_tcp_sendto(struct svc_rqst *rqstp) unsigned int uninitialized_var(sent); int err; + svc_tcp_release_rqst(rqstp); + + mutex_lock(&xprt->xpt_mutex); + if (svc_xprt_is_dead(xprt)) + goto out_notconn; err = xprt_sock_sendmsg(svsk->sk_sock, &msg, xdr, 0, marker, &sent); xdr_free_bvec(xdr); + trace_svcsock_tcp_send(xprt, err < 0 ? err : sent); if (err < 0 || sent != (xdr->len + sizeof(marker))) goto out_close; + mutex_unlock(&xprt->xpt_mutex); return sent; +out_notconn: + mutex_unlock(&xprt->xpt_mutex); + return -ENOTCONN; out_close: pr_notice("rpc-srv/tcp: %s: %s %d when sending %d bytes - shutting down socket\n", xprt->xpt_server->sv_name, @@ -1089,6 +1106,7 @@ out_close: (err < 0) ? err : sent, xdr->len); set_bit(XPT_CLOSE, &xprt->xpt_flags); svc_xprt_enqueue(xprt); + mutex_unlock(&xprt->xpt_mutex); return -EAGAIN; } @@ -1105,7 +1123,7 @@ static const struct svc_xprt_ops svc_tcp_ops = { .xpo_recvfrom = svc_tcp_recvfrom, .xpo_sendto = svc_tcp_sendto, .xpo_read_payload = svc_sock_read_payload, - .xpo_release_rqst = svc_release_skb, + .xpo_release_rqst = svc_tcp_release_rqst, .xpo_detach = svc_tcp_sock_detach, .xpo_free = svc_sock_free, .xpo_has_wspace = svc_tcp_has_wspace, @@ -1143,18 +1161,16 @@ static void svc_tcp_init(struct svc_sock *svsk, struct svc_serv *serv) set_bit(XPT_CACHE_AUTH, &svsk->sk_xprt.xpt_flags); set_bit(XPT_CONG_CTRL, &svsk->sk_xprt.xpt_flags); if (sk->sk_state == TCP_LISTEN) { - dprintk("setting up TCP socket for listening\n"); strcpy(svsk->sk_xprt.xpt_remotebuf, "listener"); set_bit(XPT_LISTENER, &svsk->sk_xprt.xpt_flags); sk->sk_data_ready = svc_tcp_listen_data_ready; set_bit(XPT_CONN, &svsk->sk_xprt.xpt_flags); } else { - dprintk("setting up TCP socket for reading\n"); sk->sk_state_change = svc_tcp_state_change; sk->sk_data_ready = svc_data_ready; sk->sk_write_space = svc_write_space; - svsk->sk_reclen = 0; + svsk->sk_marker = xdr_zero; svsk->sk_tcplen = 0; svsk->sk_datalen = 0; memset(&svsk->sk_pages[0], 0, sizeof(svsk->sk_pages)); @@ -1199,7 +1215,6 @@ static struct svc_sock *svc_setup_socket(struct svc_serv *serv, int pmap_register = !(flags & SVC_SOCK_ANONYMOUS); int err = 0; - dprintk("svc: svc_setup_socket %p\n", sock); svsk = kzalloc(sizeof(*svsk), GFP_KERNEL); if (!svsk) return ERR_PTR(-ENOMEM); @@ -1236,12 +1251,7 @@ static struct svc_sock *svc_setup_socket(struct svc_serv *serv, else svc_tcp_init(svsk, serv); - dprintk("svc: svc_setup_socket created %p (inet %p), " - "listen %d close %d\n", - svsk, svsk->sk_sk, - test_bit(XPT_LISTENER, &svsk->sk_xprt.xpt_flags), - test_bit(XPT_CLOSE, &svsk->sk_xprt.xpt_flags)); - + trace_svcsock_new_socket(sock); return svsk; } @@ -1333,12 +1343,6 @@ static struct svc_xprt *svc_create_socket(struct svc_serv *serv, struct sockaddr *newsin = (struct sockaddr *)&addr; int newlen; int family; - int val; - RPC_IFDEBUG(char buf[RPC_MAX_ADDRBUFLEN]); - - dprintk("svc: svc_create_socket(%s, %d, %s)\n", - serv->sv_program->pg_name, protocol, - __svc_print_addr(sin, buf, sizeof(buf))); if (protocol != IPPROTO_UDP && protocol != IPPROTO_TCP) { printk(KERN_WARNING "svc: only UDP and TCP " @@ -1369,11 +1373,8 @@ static struct svc_xprt *svc_create_socket(struct svc_serv *serv, * getting requests from IPv4 remotes. Those should * be shunted to a PF_INET listener via rpcbind. */ - val = 1; if (family == PF_INET6) - kernel_setsockopt(sock, SOL_IPV6, IPV6_V6ONLY, - (char *)&val, sizeof(val)); - + ip6_sock_set_v6only(sock->sk); if (type == SOCK_STREAM) sock->sk->sk_reuse = SK_CAN_REUSE; /* allow address reuse */ error = kernel_bind(sock, sin, len); @@ -1398,7 +1399,6 @@ static struct svc_xprt *svc_create_socket(struct svc_serv *serv, svc_xprt_set_local(&svsk->sk_xprt, newsin, newlen); return (struct svc_xprt *)svsk; bummer: - dprintk("svc: svc_create_socket error = %d\n", -error); sock_release(sock); return ERR_PTR(error); } @@ -1412,8 +1412,6 @@ static void svc_sock_detach(struct svc_xprt *xprt) struct svc_sock *svsk = container_of(xprt, struct svc_sock, sk_xprt); struct sock *sk = svsk->sk_sk; - dprintk("svc: svc_sock_detach(%p)\n", svsk); - /* put back the old socket callbacks */ lock_sock(sk); sk->sk_state_change = svsk->sk_ostate; @@ -1430,8 +1428,6 @@ static void svc_tcp_sock_detach(struct svc_xprt *xprt) { struct svc_sock *svsk = container_of(xprt, struct svc_sock, sk_xprt); - dprintk("svc: svc_tcp_sock_detach(%p)\n", svsk); - svc_sock_detach(xprt); if (!test_bit(XPT_LISTENER, &xprt->xpt_flags)) { @@ -1446,7 +1442,6 @@ static void svc_tcp_sock_detach(struct svc_xprt *xprt) static void svc_sock_free(struct svc_xprt *xprt) { struct svc_sock *svsk = container_of(xprt, struct svc_sock, sk_xprt); - dprintk("svc: svc_sock_free(%p)\n", svsk); if (svsk->sk_sock->file) sockfd_put(svsk->sk_sock); diff --git a/net/sunrpc/sysctl.c b/net/sunrpc/sysctl.c index d75f17b56f0e..999eee1ed61c 100644 --- a/net/sunrpc/sysctl.c +++ b/net/sunrpc/sysctl.c @@ -60,7 +60,7 @@ rpc_unregister_sysctl(void) } static int proc_do_xprt(struct ctl_table *table, int write, - void __user *buffer, size_t *lenp, loff_t *ppos) + void *buffer, size_t *lenp, loff_t *ppos) { char tmpbuf[256]; size_t len; @@ -70,15 +70,15 @@ static int proc_do_xprt(struct ctl_table *table, int write, return 0; } len = svc_print_xprts(tmpbuf, sizeof(tmpbuf)); - return simple_read_from_buffer(buffer, *lenp, ppos, tmpbuf, len); + return memory_read_from_buffer(buffer, *lenp, ppos, tmpbuf, len); } static int -proc_dodebug(struct ctl_table *table, int write, - void __user *buffer, size_t *lenp, loff_t *ppos) +proc_dodebug(struct ctl_table *table, int write, void *buffer, size_t *lenp, + loff_t *ppos) { - char tmpbuf[20], c, *s = NULL; - char __user *p; + char tmpbuf[20], *s = NULL; + char *p; unsigned int value; size_t left, len; @@ -90,18 +90,17 @@ proc_dodebug(struct ctl_table *table, int write, left = *lenp; if (write) { - if (!access_ok(buffer, left)) - return -EFAULT; p = buffer; - while (left && __get_user(c, p) >= 0 && isspace(c)) - left--, p++; + while (left && isspace(*p)) { + left--; + p++; + } if (!left) goto done; if (left > sizeof(tmpbuf) - 1) return -EINVAL; - if (copy_from_user(tmpbuf, p, left)) - return -EFAULT; + memcpy(tmpbuf, p, left); tmpbuf[left] = '\0'; value = simple_strtol(tmpbuf, &s, 0); @@ -121,11 +120,9 @@ proc_dodebug(struct ctl_table *table, int write, len = sprintf(tmpbuf, "0x%04x", *(unsigned int *) table->data); if (len > left) len = left; - if (copy_to_user(buffer, tmpbuf, len)) - return -EFAULT; + memcpy(buffer, tmpbuf, len); if ((left -= len) > 0) { - if (put_user('\n', (char __user *)buffer + len)) - return -EFAULT; + *((char *)buffer + len) = '\n'; left--; } } diff --git a/net/sunrpc/xdr.c b/net/sunrpc/xdr.c index 15b58c5144f9..6f7d82fb1eb0 100644 --- a/net/sunrpc/xdr.c +++ b/net/sunrpc/xdr.c @@ -1150,6 +1150,47 @@ xdr_buf_subsegment(struct xdr_buf *buf, struct xdr_buf *subbuf, } EXPORT_SYMBOL_GPL(xdr_buf_subsegment); +/** + * xdr_buf_trim - lop at most "len" bytes off the end of "buf" + * @buf: buf to be trimmed + * @len: number of bytes to reduce "buf" by + * + * Trim an xdr_buf by the given number of bytes by fixing up the lengths. Note + * that it's possible that we'll trim less than that amount if the xdr_buf is + * too small, or if (for instance) it's all in the head and the parser has + * already read too far into it. + */ +void xdr_buf_trim(struct xdr_buf *buf, unsigned int len) +{ + size_t cur; + unsigned int trim = len; + + if (buf->tail[0].iov_len) { + cur = min_t(size_t, buf->tail[0].iov_len, trim); + buf->tail[0].iov_len -= cur; + trim -= cur; + if (!trim) + goto fix_len; + } + + if (buf->page_len) { + cur = min_t(unsigned int, buf->page_len, trim); + buf->page_len -= cur; + trim -= cur; + if (!trim) + goto fix_len; + } + + if (buf->head[0].iov_len) { + cur = min_t(size_t, buf->head[0].iov_len, trim); + buf->head[0].iov_len -= cur; + trim -= cur; + } +fix_len: + buf->len -= (len - trim); +} +EXPORT_SYMBOL_GPL(xdr_buf_trim); + static void __read_bytes_from_xdr_buf(struct xdr_buf *subbuf, void *obj, unsigned int len) { unsigned int this_len; diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c index 493a30a296fc..d5cc5db9dbf3 100644 --- a/net/sunrpc/xprt.c +++ b/net/sunrpc/xprt.c @@ -663,6 +663,7 @@ static void xprt_autoclose(struct work_struct *work) container_of(work, struct rpc_xprt, task_cleanup); unsigned int pflags = memalloc_nofs_save(); + trace_xprt_disconnect_auto(xprt); clear_bit(XPRT_CLOSE_WAIT, &xprt->state); xprt->ops->close(xprt); xprt_release_write(xprt, NULL); @@ -677,7 +678,7 @@ static void xprt_autoclose(struct work_struct *work) */ void xprt_disconnect_done(struct rpc_xprt *xprt) { - dprintk("RPC: disconnected transport %p\n", xprt); + trace_xprt_disconnect_done(xprt); spin_lock(&xprt->transport_lock); xprt_clear_connected(xprt); xprt_clear_write_space_locked(xprt); @@ -694,6 +695,8 @@ EXPORT_SYMBOL_GPL(xprt_disconnect_done); */ void xprt_force_disconnect(struct rpc_xprt *xprt) { + trace_xprt_disconnect_force(xprt); + /* Don't race with the test_bit() in xprt_clear_locked() */ spin_lock(&xprt->transport_lock); set_bit(XPRT_CLOSE_WAIT, &xprt->state); @@ -832,8 +835,10 @@ void xprt_connect(struct rpc_task *task) if (!xprt_lock_write(xprt, task)) return; - if (test_and_clear_bit(XPRT_CLOSE_WAIT, &xprt->state)) + if (test_and_clear_bit(XPRT_CLOSE_WAIT, &xprt->state)) { + trace_xprt_disconnect_cleanup(xprt); xprt->ops->close(xprt); + } if (!xprt_connected(xprt)) { task->tk_rqstp->rq_connect_cookie = xprt->connect_cookie; @@ -1460,7 +1465,7 @@ xprt_request_transmit(struct rpc_rqst *req, struct rpc_task *snd_task) */ req->rq_ntrans++; - trace_xprt_sendto(&req->rq_snd_buf); + trace_rpc_xdr_sendto(task, &req->rq_snd_buf); connect_cookie = xprt->connect_cookie; status = xprt->ops->send_request(req); if (status != 0) { @@ -1903,11 +1908,8 @@ struct rpc_xprt *xprt_create_transport(struct xprt_create *args) found: xprt = t->setup(args); - if (IS_ERR(xprt)) { - dprintk("RPC: xprt_create_transport: failed, %ld\n", - -PTR_ERR(xprt)); + if (IS_ERR(xprt)) goto out; - } if (args->flags & XPRT_CREATE_NO_IDLE_TIMEOUT) xprt->idle_timeout = 0; INIT_WORK(&xprt->task_cleanup, xprt_autoclose); @@ -1928,8 +1930,7 @@ found: rpc_xprt_debugfs_register(xprt); - dprintk("RPC: created transport %p with %u slots\n", xprt, - xprt->max_reqs); + trace_xprt_create(xprt); out: return xprt; } @@ -1939,6 +1940,8 @@ static void xprt_destroy_cb(struct work_struct *work) struct rpc_xprt *xprt = container_of(work, struct rpc_xprt, task_cleanup); + trace_xprt_destroy(xprt); + rpc_xprt_debugfs_unregister(xprt); rpc_destroy_wait_queue(&xprt->binding); rpc_destroy_wait_queue(&xprt->pending); @@ -1963,8 +1966,6 @@ static void xprt_destroy_cb(struct work_struct *work) */ static void xprt_destroy(struct rpc_xprt *xprt) { - dprintk("RPC: destroying transport %p\n", xprt); - /* * Exclude transport connect/disconnect handlers and autoclose */ diff --git a/net/sunrpc/xprtrdma/rpc_rdma.c b/net/sunrpc/xprtrdma/rpc_rdma.c index 4a81e6995d3e..2081c8fbfa48 100644 --- a/net/sunrpc/xprtrdma/rpc_rdma.c +++ b/net/sunrpc/xprtrdma/rpc_rdma.c @@ -388,7 +388,9 @@ static int rpcrdma_encode_read_list(struct rpcrdma_xprt *r_xprt, } while (nsegs); done: - return xdr_stream_encode_item_absent(xdr); + if (xdr_stream_encode_item_absent(xdr) < 0) + return -EMSGSIZE; + return 0; } /* Register and XDR encode the Write list. Supports encoding a list @@ -454,7 +456,9 @@ static int rpcrdma_encode_write_list(struct rpcrdma_xprt *r_xprt, *segcount = cpu_to_be32(nchunks); done: - return xdr_stream_encode_item_absent(xdr); + if (xdr_stream_encode_item_absent(xdr) < 0) + return -EMSGSIZE; + return 0; } /* Register and XDR encode the Reply chunk. Supports encoding an array @@ -480,8 +484,11 @@ static int rpcrdma_encode_reply_chunk(struct rpcrdma_xprt *r_xprt, int nsegs, nchunks; __be32 *segcount; - if (wtype != rpcrdma_replych) - return xdr_stream_encode_item_absent(xdr); + if (wtype != rpcrdma_replych) { + if (xdr_stream_encode_item_absent(xdr) < 0) + return -EMSGSIZE; + return 0; + } seg = req->rl_segments; nsegs = rpcrdma_convert_iovs(r_xprt, &rqst->rq_rcv_buf, 0, wtype, seg); @@ -885,8 +892,8 @@ rpcrdma_marshal_req(struct rpcrdma_xprt *r_xprt, struct rpc_rqst *rqst) * or privacy, direct data placement of individual data items * is not allowed. */ - ddp_allowed = !(rqst->rq_cred->cr_auth->au_flags & - RPCAUTH_AUTH_DATATOUCH); + ddp_allowed = !test_bit(RPCAUTH_AUTH_DATATOUCH, + &rqst->rq_cred->cr_auth->au_flags); /* * Chunks needed for results? diff --git a/net/sunrpc/xprtrdma/svc_rdma.c b/net/sunrpc/xprtrdma/svc_rdma.c index 97bca509a391..526da5d4710b 100644 --- a/net/sunrpc/xprtrdma/svc_rdma.c +++ b/net/sunrpc/xprtrdma/svc_rdma.c @@ -80,8 +80,7 @@ atomic_t rdma_stat_sq_prod; * current value. */ static int read_reset_stat(struct ctl_table *table, int write, - void __user *buffer, size_t *lenp, - loff_t *ppos) + void *buffer, size_t *lenp, loff_t *ppos) { atomic_t *stat = (atomic_t *)table->data; @@ -103,8 +102,8 @@ static int read_reset_stat(struct ctl_table *table, int write, len -= *ppos; if (len > *lenp) len = *lenp; - if (len && copy_to_user(buffer, str_buf, len)) - return -EFAULT; + if (len) + memcpy(buffer, str_buf, len); *lenp = len; *ppos += len; } diff --git a/net/sunrpc/xprtrdma/svc_rdma_backchannel.c b/net/sunrpc/xprtrdma/svc_rdma_backchannel.c index d510a3a15d4b..1ee73f7cf931 100644 --- a/net/sunrpc/xprtrdma/svc_rdma_backchannel.c +++ b/net/sunrpc/xprtrdma/svc_rdma_backchannel.c @@ -10,59 +10,34 @@ #include "xprt_rdma.h" #include <trace/events/rpcrdma.h> -#define RPCDBG_FACILITY RPCDBG_SVCXPRT - -#undef SVCRDMA_BACKCHANNEL_DEBUG - /** - * svc_rdma_handle_bc_reply - Process incoming backchannel reply - * @xprt: controlling backchannel transport - * @rdma_resp: pointer to incoming transport header - * @rcvbuf: XDR buffer into which to decode the reply + * svc_rdma_handle_bc_reply - Process incoming backchannel Reply + * @rqstp: resources for handling the Reply + * @rctxt: Received message * - * Returns: - * %0 if @rcvbuf is filled in, xprt_complete_rqst called, - * %-EAGAIN if server should call ->recvfrom again. */ -int svc_rdma_handle_bc_reply(struct rpc_xprt *xprt, __be32 *rdma_resp, - struct xdr_buf *rcvbuf) +void svc_rdma_handle_bc_reply(struct svc_rqst *rqstp, + struct svc_rdma_recv_ctxt *rctxt) { + struct svc_xprt *sxprt = rqstp->rq_xprt; + struct rpc_xprt *xprt = sxprt->xpt_bc_xprt; struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(xprt); + struct xdr_buf *rcvbuf = &rqstp->rq_arg; struct kvec *dst, *src = &rcvbuf->head[0]; + __be32 *rdma_resp = rctxt->rc_recv_buf; struct rpc_rqst *req; u32 credits; - size_t len; - __be32 xid; - __be32 *p; - int ret; - - p = (__be32 *)src->iov_base; - len = src->iov_len; - xid = *rdma_resp; - -#ifdef SVCRDMA_BACKCHANNEL_DEBUG - pr_info("%s: xid=%08x, length=%zu\n", - __func__, be32_to_cpu(xid), len); - pr_info("%s: RPC/RDMA: %*ph\n", - __func__, (int)RPCRDMA_HDRLEN_MIN, rdma_resp); - pr_info("%s: RPC: %*ph\n", - __func__, (int)len, p); -#endif - - ret = -EAGAIN; - if (src->iov_len < 24) - goto out_shortreply; spin_lock(&xprt->queue_lock); - req = xprt_lookup_rqst(xprt, xid); + req = xprt_lookup_rqst(xprt, *rdma_resp); if (!req) - goto out_notfound; + goto out_unlock; dst = &req->rq_private_buf.head[0]; memcpy(&req->rq_private_buf, &req->rq_rcv_buf, sizeof(struct xdr_buf)); - if (dst->iov_len < len) + if (dst->iov_len < src->iov_len) goto out_unlock; - memcpy(dst->iov_base, p, len); + memcpy(dst->iov_base, src->iov_base, src->iov_len); xprt_pin_rqst(req); spin_unlock(&xprt->queue_lock); @@ -71,31 +46,17 @@ int svc_rdma_handle_bc_reply(struct rpc_xprt *xprt, __be32 *rdma_resp, credits = 1; /* don't deadlock */ else if (credits > r_xprt->rx_buf.rb_bc_max_requests) credits = r_xprt->rx_buf.rb_bc_max_requests; - spin_lock(&xprt->transport_lock); xprt->cwnd = credits << RPC_CWNDSHIFT; spin_unlock(&xprt->transport_lock); spin_lock(&xprt->queue_lock); - ret = 0; xprt_complete_rqst(req->rq_task, rcvbuf->len); xprt_unpin_rqst(req); rcvbuf->len = 0; out_unlock: spin_unlock(&xprt->queue_lock); -out: - return ret; - -out_shortreply: - dprintk("svcrdma: short bc reply: xprt=%p, len=%zu\n", - xprt, src->iov_len); - goto out; - -out_notfound: - dprintk("svcrdma: unrecognized bc reply: xprt=%p, xid=%08x\n", - xprt, be32_to_cpu(xid)); - goto out_unlock; } /* Send a backwards direction RPC call. @@ -192,10 +153,6 @@ rpcrdma_bc_send_request(struct svcxprt_rdma *rdma, struct rpc_rqst *rqst) *p++ = xdr_zero; *p = xdr_zero; -#ifdef SVCRDMA_BACKCHANNEL_DEBUG - pr_info("%s: %*ph\n", __func__, 64, rqst->rq_buffer); -#endif - rqst->rq_xtime = ktime_get(); rc = svc_rdma_bc_sendto(rdma, rqst, ctxt); if (rc) @@ -206,52 +163,43 @@ put_ctxt: svc_rdma_send_ctxt_put(rdma, ctxt); drop_connection: - dprintk("svcrdma: failed to send bc call\n"); return -ENOTCONN; } -/* Send an RPC call on the passive end of a transport - * connection. +/** + * xprt_rdma_bc_send_request - Send a reverse-direction Call + * @rqst: rpc_rqst containing Call message to be sent + * + * Return values: + * %0 if the message was sent successfully + * %ENOTCONN if the message was not sent */ -static int -xprt_rdma_bc_send_request(struct rpc_rqst *rqst) +static int xprt_rdma_bc_send_request(struct rpc_rqst *rqst) { struct svc_xprt *sxprt = rqst->rq_xprt->bc_xprt; - struct svcxprt_rdma *rdma; + struct svcxprt_rdma *rdma = + container_of(sxprt, struct svcxprt_rdma, sc_xprt); int ret; - dprintk("svcrdma: sending bc call with xid: %08x\n", - be32_to_cpu(rqst->rq_xid)); - - mutex_lock(&sxprt->xpt_mutex); - - ret = -ENOTCONN; - rdma = container_of(sxprt, struct svcxprt_rdma, sc_xprt); - if (!test_bit(XPT_DEAD, &sxprt->xpt_flags)) { - ret = rpcrdma_bc_send_request(rdma, rqst); - if (ret == -ENOTCONN) - svc_close_xprt(sxprt); - } - - mutex_unlock(&sxprt->xpt_mutex); + if (test_bit(XPT_DEAD, &sxprt->xpt_flags)) + return -ENOTCONN; - if (ret < 0) - return ret; - return 0; + ret = rpcrdma_bc_send_request(rdma, rqst); + if (ret == -ENOTCONN) + svc_close_xprt(sxprt); + return ret; } static void xprt_rdma_bc_close(struct rpc_xprt *xprt) { - dprintk("svcrdma: %s: xprt %p\n", __func__, xprt); + xprt_disconnect_done(xprt); xprt->cwnd = RPC_CWNDSHIFT; } static void xprt_rdma_bc_put(struct rpc_xprt *xprt) { - dprintk("svcrdma: %s: xprt %p\n", __func__, xprt); - xprt_rdma_free_addresses(xprt); xprt_free(xprt); } @@ -286,19 +234,14 @@ xprt_setup_rdma_bc(struct xprt_create *args) struct rpc_xprt *xprt; struct rpcrdma_xprt *new_xprt; - if (args->addrlen > sizeof(xprt->addr)) { - dprintk("RPC: %s: address too large\n", __func__); + if (args->addrlen > sizeof(xprt->addr)) return ERR_PTR(-EBADF); - } xprt = xprt_alloc(args->net, sizeof(*new_xprt), RPCRDMA_MAX_BC_REQUESTS, RPCRDMA_MAX_BC_REQUESTS); - if (!xprt) { - dprintk("RPC: %s: couldn't allocate rpc_xprt\n", - __func__); + if (!xprt) return ERR_PTR(-ENOMEM); - } xprt->timeout = &xprt_rdma_bc_timeout; xprt_set_bound(xprt); diff --git a/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c b/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c index 54469b72b25f..e426fedb9524 100644 --- a/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c +++ b/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c @@ -223,6 +223,26 @@ void svc_rdma_recv_ctxt_put(struct svcxprt_rdma *rdma, svc_rdma_recv_ctxt_destroy(rdma, ctxt); } +/** + * svc_rdma_release_rqst - Release transport-specific per-rqst resources + * @rqstp: svc_rqst being released + * + * Ensure that the recv_ctxt is released whether or not a Reply + * was sent. For example, the client could close the connection, + * or svc_process could drop an RPC, before the Reply is sent. + */ +void svc_rdma_release_rqst(struct svc_rqst *rqstp) +{ + struct svc_rdma_recv_ctxt *ctxt = rqstp->rq_xprt_ctxt; + struct svc_xprt *xprt = rqstp->rq_xprt; + struct svcxprt_rdma *rdma = + container_of(xprt, struct svcxprt_rdma, sc_xprt); + + rqstp->rq_xprt_ctxt = NULL; + if (ctxt) + svc_rdma_recv_ctxt_put(rdma, ctxt); +} + static int __svc_rdma_post_recv(struct svcxprt_rdma *rdma, struct svc_rdma_recv_ctxt *ctxt) { @@ -645,23 +665,23 @@ static int svc_rdma_xdr_decode_req(struct xdr_buf *rq_arg, return hdr_len; out_short: - trace_svcrdma_decode_short(rq_arg->len); + trace_svcrdma_decode_short_err(rq_arg->len); return -EINVAL; out_version: - trace_svcrdma_decode_badvers(rdma_argp); + trace_svcrdma_decode_badvers_err(rdma_argp); return -EPROTONOSUPPORT; out_drop: - trace_svcrdma_decode_drop(rdma_argp); + trace_svcrdma_decode_drop_err(rdma_argp); return 0; out_proc: - trace_svcrdma_decode_badproc(rdma_argp); + trace_svcrdma_decode_badproc_err(rdma_argp); return -EINVAL; out_inval: - trace_svcrdma_decode_parse(rdma_argp); + trace_svcrdma_decode_parse_err(rdma_argp); return -EINVAL; } @@ -820,6 +840,8 @@ int svc_rdma_recvfrom(struct svc_rqst *rqstp) __be32 *p; int ret; + rqstp->rq_xprt_ctxt = NULL; + spin_lock(&rdma_xprt->sc_rq_dto_lock); ctxt = svc_rdma_next_recv_ctxt(&rdma_xprt->sc_read_complete_q); if (ctxt) { @@ -856,12 +878,9 @@ int svc_rdma_recvfrom(struct svc_rqst *rqstp) goto out_drop; rqstp->rq_xprt_hlen = ret; - if (svc_rdma_is_backchannel_reply(xprt, p)) { - ret = svc_rdma_handle_bc_reply(xprt->xpt_bc_xprt, p, - &rqstp->rq_arg); - svc_rdma_recv_ctxt_put(rdma_xprt, ctxt); - return ret; - } + if (svc_rdma_is_backchannel_reply(xprt, p)) + goto out_backchannel; + svc_rdma_get_inv_rkey(rdma_xprt, ctxt); p += rpcrdma_fixed_maxsz; @@ -891,6 +910,8 @@ out_postfail: svc_rdma_recv_ctxt_put(rdma_xprt, ctxt); return ret; +out_backchannel: + svc_rdma_handle_bc_reply(rqstp, ctxt); out_drop: svc_rdma_recv_ctxt_put(rdma_xprt, ctxt); return 0; diff --git a/net/sunrpc/xprtrdma/svc_rdma_rw.c b/net/sunrpc/xprtrdma/svc_rdma_rw.c index bd7c195d872e..5eb35309ecef 100644 --- a/net/sunrpc/xprtrdma/svc_rdma_rw.c +++ b/net/sunrpc/xprtrdma/svc_rdma_rw.c @@ -9,13 +9,10 @@ #include <linux/sunrpc/rpc_rdma.h> #include <linux/sunrpc/svc_rdma.h> -#include <linux/sunrpc/debug.h> #include "xprt_rdma.h" #include <trace/events/rpcrdma.h> -#define RPCDBG_FACILITY RPCDBG_SVCXPRT - static void svc_rdma_write_done(struct ib_cq *cq, struct ib_wc *wc); static void svc_rdma_wc_read_done(struct ib_cq *cq, struct ib_wc *wc); @@ -39,7 +36,7 @@ static void svc_rdma_wc_read_done(struct ib_cq *cq, struct ib_wc *wc); struct svc_rdma_rw_ctxt { struct list_head rw_list; struct rdma_rw_ctx rw_ctx; - int rw_nents; + unsigned int rw_nents; struct sg_table rw_sg_table; struct scatterlist rw_first_sgl[]; }; @@ -67,19 +64,22 @@ svc_rdma_get_rw_ctxt(struct svcxprt_rdma *rdma, unsigned int sges) ctxt = kmalloc(struct_size(ctxt, rw_first_sgl, SG_CHUNK_SIZE), GFP_KERNEL); if (!ctxt) - goto out; + goto out_noctx; INIT_LIST_HEAD(&ctxt->rw_list); } ctxt->rw_sg_table.sgl = ctxt->rw_first_sgl; if (sg_alloc_table_chained(&ctxt->rw_sg_table, sges, ctxt->rw_sg_table.sgl, - SG_CHUNK_SIZE)) { - kfree(ctxt); - ctxt = NULL; - } -out: + SG_CHUNK_SIZE)) + goto out_free; return ctxt; + +out_free: + kfree(ctxt); +out_noctx: + trace_svcrdma_no_rwctx_err(rdma, sges); + return NULL; } static void svc_rdma_put_rw_ctxt(struct svcxprt_rdma *rdma, @@ -107,6 +107,34 @@ void svc_rdma_destroy_rw_ctxts(struct svcxprt_rdma *rdma) } } +/** + * svc_rdma_rw_ctx_init - Prepare a R/W context for I/O + * @rdma: controlling transport instance + * @ctxt: R/W context to prepare + * @offset: RDMA offset + * @handle: RDMA tag/handle + * @direction: I/O direction + * + * Returns on success, the number of WQEs that will be needed + * on the workqueue, or a negative errno. + */ +static int svc_rdma_rw_ctx_init(struct svcxprt_rdma *rdma, + struct svc_rdma_rw_ctxt *ctxt, + u64 offset, u32 handle, + enum dma_data_direction direction) +{ + int ret; + + ret = rdma_rw_ctx_init(&ctxt->rw_ctx, rdma->sc_qp, rdma->sc_port_num, + ctxt->rw_sg_table.sgl, ctxt->rw_nents, + 0, offset, handle, direction); + if (unlikely(ret < 0)) { + svc_rdma_put_rw_ctxt(rdma, ctxt); + trace_svcrdma_dma_map_rw_err(rdma, ctxt->rw_nents, ret); + } + return ret; +} + /* A chunk context tracks all I/O for moving one Read or Write * chunk. This is a a set of rdma_rw's that handle data movement * for all segments of one chunk. @@ -323,8 +351,6 @@ static int svc_rdma_post_chunk_ctxt(struct svc_rdma_chunk_ctxt *cc) if (atomic_sub_return(cc->cc_sqecount, &rdma->sc_sq_avail) > 0) { ret = ib_post_send(rdma->sc_qp, first_wr, &bad_wr); - trace_svcrdma_post_rw(&cc->cc_cqe, - cc->cc_sqecount, ret); if (ret) break; return 0; @@ -337,6 +363,7 @@ static int svc_rdma_post_chunk_ctxt(struct svc_rdma_chunk_ctxt *cc) trace_svcrdma_sq_retry(rdma); } while (1); + trace_svcrdma_sq_post_err(rdma, ret); set_bit(XPT_CLOSE, &xprt->xpt_flags); /* If even one was posted, there will be a completion. */ @@ -429,15 +456,13 @@ svc_rdma_build_writes(struct svc_rdma_write_info *info, ctxt = svc_rdma_get_rw_ctxt(rdma, (write_len >> PAGE_SHIFT) + 2); if (!ctxt) - goto out_noctx; + return -ENOMEM; constructor(info, write_len, ctxt); - ret = rdma_rw_ctx_init(&ctxt->rw_ctx, rdma->sc_qp, - rdma->sc_port_num, ctxt->rw_sg_table.sgl, - ctxt->rw_nents, 0, seg_offset, - seg_handle, DMA_TO_DEVICE); + ret = svc_rdma_rw_ctx_init(rdma, ctxt, seg_offset, seg_handle, + DMA_TO_DEVICE); if (ret < 0) - goto out_initerr; + return -EIO; trace_svcrdma_send_wseg(seg_handle, write_len, seg_offset); @@ -456,18 +481,9 @@ svc_rdma_build_writes(struct svc_rdma_write_info *info, return 0; out_overflow: - dprintk("svcrdma: inadequate space in Write chunk (%u)\n", - info->wi_nsegs); + trace_svcrdma_small_wrch_err(rdma, remaining, info->wi_seg_no, + info->wi_nsegs); return -E2BIG; - -out_noctx: - dprintk("svcrdma: no R/W ctxs available\n"); - return -ENOMEM; - -out_initerr: - svc_rdma_put_rw_ctxt(rdma, ctxt); - trace_svcrdma_dma_map_rwctx(rdma, ret); - return -EIO; } /* Send one of an xdr_buf's kvecs by itself. To send a Reply @@ -617,7 +633,7 @@ static int svc_rdma_build_read_segment(struct svc_rdma_read_info *info, sge_no = PAGE_ALIGN(info->ri_pageoff + len) >> PAGE_SHIFT; ctxt = svc_rdma_get_rw_ctxt(cc->cc_rdma, sge_no); if (!ctxt) - goto out_noctx; + return -ENOMEM; ctxt->rw_nents = sge_no; sg = ctxt->rw_sg_table.sgl; @@ -647,29 +663,18 @@ static int svc_rdma_build_read_segment(struct svc_rdma_read_info *info, goto out_overrun; } - ret = rdma_rw_ctx_init(&ctxt->rw_ctx, cc->cc_rdma->sc_qp, - cc->cc_rdma->sc_port_num, - ctxt->rw_sg_table.sgl, ctxt->rw_nents, - 0, offset, rkey, DMA_FROM_DEVICE); + ret = svc_rdma_rw_ctx_init(cc->cc_rdma, ctxt, offset, rkey, + DMA_FROM_DEVICE); if (ret < 0) - goto out_initerr; + return -EIO; list_add(&ctxt->rw_list, &cc->cc_rwctxts); cc->cc_sqecount += ret; return 0; -out_noctx: - dprintk("svcrdma: no R/W ctxs available\n"); - return -ENOMEM; - out_overrun: - dprintk("svcrdma: request overruns rq_pages\n"); + trace_svcrdma_page_overrun_err(cc->cc_rdma, rqstp, info->ri_pageno); return -EINVAL; - -out_initerr: - trace_svcrdma_dma_map_rwctx(cc->cc_rdma, ret); - svc_rdma_put_rw_ctxt(cc->cc_rdma, ctxt); - return -EIO; } /* Walk the segments in the Read chunk starting at @p and construct diff --git a/net/sunrpc/xprtrdma/svc_rdma_sendto.c b/net/sunrpc/xprtrdma/svc_rdma_sendto.c index 90cba3058f04..38e7c3c8c4a9 100644 --- a/net/sunrpc/xprtrdma/svc_rdma_sendto.c +++ b/net/sunrpc/xprtrdma/svc_rdma_sendto.c @@ -322,15 +322,17 @@ int svc_rdma_send(struct svcxprt_rdma *rdma, struct ib_send_wr *wr) } svc_xprt_get(&rdma->sc_xprt); + trace_svcrdma_post_send(wr); ret = ib_post_send(rdma->sc_qp, wr, NULL); - trace_svcrdma_post_send(wr, ret); - if (ret) { - set_bit(XPT_CLOSE, &rdma->sc_xprt.xpt_flags); - svc_xprt_put(&rdma->sc_xprt); - wake_up(&rdma->sc_send_wait); - } - break; + if (ret) + break; + return 0; } + + trace_svcrdma_sq_post_err(rdma, ret); + set_bit(XPT_CLOSE, &rdma->sc_xprt.xpt_flags); + svc_xprt_put(&rdma->sc_xprt); + wake_up(&rdma->sc_send_wait); return ret; } @@ -866,12 +868,10 @@ int svc_rdma_sendto(struct svc_rqst *rqstp) __be32 *p; int ret; - /* Create the RDMA response header. xprt->xpt_mutex, - * acquired in svc_send(), serializes RPC replies. The - * code path below that inserts the credit grant value - * into each transport header runs only inside this - * critical section. - */ + ret = -ENOTCONN; + if (svc_xprt_is_dead(xprt)) + goto err0; + ret = -ENOMEM; sctxt = svc_rdma_send_ctxt_get(rdma); if (!sctxt) @@ -924,12 +924,7 @@ int svc_rdma_sendto(struct svc_rqst *rqstp) ret = svc_rdma_send_reply_msg(rdma, sctxt, rctxt, rqstp); if (ret < 0) goto err1; - ret = 0; - -out: - rqstp->rq_xprt_ctxt = NULL; - svc_rdma_recv_ctxt_put(rdma, rctxt); - return ret; + return 0; err2: if (ret != -E2BIG && ret != -EINVAL) @@ -938,16 +933,14 @@ out: ret = svc_rdma_send_error_msg(rdma, sctxt, rqstp); if (ret < 0) goto err1; - ret = 0; - goto out; + return 0; err1: svc_rdma_send_ctxt_put(rdma, sctxt); err0: trace_svcrdma_send_failed(rqstp, ret); set_bit(XPT_CLOSE, &xprt->xpt_flags); - ret = -ENOTCONN; - goto out; + return -ENOTCONN; } /** diff --git a/net/sunrpc/xprtrdma/svc_rdma_transport.c b/net/sunrpc/xprtrdma/svc_rdma_transport.c index 8bb99980ae85..d38be57b00ed 100644 --- a/net/sunrpc/xprtrdma/svc_rdma_transport.c +++ b/net/sunrpc/xprtrdma/svc_rdma_transport.c @@ -71,7 +71,6 @@ static struct svc_xprt *svc_rdma_create(struct svc_serv *serv, struct sockaddr *sa, int salen, int flags); static struct svc_xprt *svc_rdma_accept(struct svc_xprt *xprt); -static void svc_rdma_release_rqst(struct svc_rqst *); static void svc_rdma_detach(struct svc_xprt *xprt); static void svc_rdma_free(struct svc_xprt *xprt); static int svc_rdma_has_wspace(struct svc_xprt *xprt); @@ -212,7 +211,12 @@ static void handle_connect_req(struct rdma_cm_id *new_cma_id, newxprt->sc_ord = param->initiator_depth; sa = (struct sockaddr *)&newxprt->sc_cm_id->route.addr.dst_addr; - svc_xprt_set_remote(&newxprt->sc_xprt, sa, svc_addr_len(sa)); + newxprt->sc_xprt.xpt_remotelen = svc_addr_len(sa); + memcpy(&newxprt->sc_xprt.xpt_remote, sa, + newxprt->sc_xprt.xpt_remotelen); + snprintf(newxprt->sc_xprt.xpt_remotebuf, + sizeof(newxprt->sc_xprt.xpt_remotebuf) - 1, "%pISc", sa); + /* The remote port is arbitrary and not under the control of the * client ULP. Set it to a fixed value so that the DRC continues * to be effective after a reconnect. @@ -310,11 +314,8 @@ static struct svc_xprt *svc_rdma_create(struct svc_serv *serv, struct svcxprt_rdma *cma_xprt; int ret; - dprintk("svcrdma: Creating RDMA listener\n"); - if ((sa->sa_family != AF_INET) && (sa->sa_family != AF_INET6)) { - dprintk("svcrdma: Address family %d is not supported.\n", sa->sa_family); + if (sa->sa_family != AF_INET && sa->sa_family != AF_INET6) return ERR_PTR(-EAFNOSUPPORT); - } cma_xprt = svc_rdma_create_xprt(serv, net); if (!cma_xprt) return ERR_PTR(-ENOMEM); @@ -325,7 +326,6 @@ static struct svc_xprt *svc_rdma_create(struct svc_serv *serv, RDMA_PS_TCP, IB_QPT_RC); if (IS_ERR(listen_id)) { ret = PTR_ERR(listen_id); - dprintk("svcrdma: rdma_create_id failed = %d\n", ret); goto err0; } @@ -334,23 +334,17 @@ static struct svc_xprt *svc_rdma_create(struct svc_serv *serv, */ #if IS_ENABLED(CONFIG_IPV6) ret = rdma_set_afonly(listen_id, 1); - if (ret) { - dprintk("svcrdma: rdma_set_afonly failed = %d\n", ret); + if (ret) goto err1; - } #endif ret = rdma_bind_addr(listen_id, sa); - if (ret) { - dprintk("svcrdma: rdma_bind_addr failed = %d\n", ret); + if (ret) goto err1; - } cma_xprt->sc_cm_id = listen_id; ret = rdma_listen(listen_id, RPCRDMA_LISTEN_BACKLOG); - if (ret) { - dprintk("svcrdma: rdma_listen failed = %d\n", ret); + if (ret) goto err1; - } /* * We need to use the address from the cm_id in case the @@ -406,9 +400,6 @@ static struct svc_xprt *svc_rdma_accept(struct svc_xprt *xprt) if (!newxprt) return NULL; - dprintk("svcrdma: newxprt from accept queue = %p, cm_id=%p\n", - newxprt, newxprt->sc_cm_id); - dev = newxprt->sc_cm_id->device; newxprt->sc_port_num = newxprt->sc_cm_id->port_num; @@ -444,21 +435,17 @@ static struct svc_xprt *svc_rdma_accept(struct svc_xprt *xprt) newxprt->sc_pd = ib_alloc_pd(dev, 0); if (IS_ERR(newxprt->sc_pd)) { - dprintk("svcrdma: error creating PD for connect request\n"); + trace_svcrdma_pd_err(newxprt, PTR_ERR(newxprt->sc_pd)); goto errout; } newxprt->sc_sq_cq = ib_alloc_cq_any(dev, newxprt, newxprt->sc_sq_depth, IB_POLL_WORKQUEUE); - if (IS_ERR(newxprt->sc_sq_cq)) { - dprintk("svcrdma: error creating SQ CQ for connect request\n"); + if (IS_ERR(newxprt->sc_sq_cq)) goto errout; - } newxprt->sc_rq_cq = ib_alloc_cq_any(dev, newxprt, rq_depth, IB_POLL_WORKQUEUE); - if (IS_ERR(newxprt->sc_rq_cq)) { - dprintk("svcrdma: error creating RQ CQ for connect request\n"); + if (IS_ERR(newxprt->sc_rq_cq)) goto errout; - } memset(&qp_attr, 0, sizeof qp_attr); qp_attr.event_handler = qp_event_handler; @@ -482,7 +469,7 @@ static struct svc_xprt *svc_rdma_accept(struct svc_xprt *xprt) ret = rdma_create_qp(newxprt->sc_cm_id, newxprt->sc_pd, &qp_attr); if (ret) { - dprintk("svcrdma: failed to create QP, ret=%d\n", ret); + trace_svcrdma_qp_err(newxprt, ret); goto errout; } newxprt->sc_qp = newxprt->sc_cm_id->qp; @@ -490,8 +477,10 @@ static struct svc_xprt *svc_rdma_accept(struct svc_xprt *xprt) if (!(dev->attrs.device_cap_flags & IB_DEVICE_MEM_MGT_EXTENSIONS)) newxprt->sc_snd_w_inv = false; if (!rdma_protocol_iwarp(dev, newxprt->sc_port_num) && - !rdma_ib_or_roce(dev, newxprt->sc_port_num)) + !rdma_ib_or_roce(dev, newxprt->sc_port_num)) { + trace_svcrdma_fabric_err(newxprt, -EINVAL); goto errout; + } if (!svc_rdma_post_recvs(newxprt)) goto errout; @@ -513,15 +502,17 @@ static struct svc_xprt *svc_rdma_accept(struct svc_xprt *xprt) conn_param.initiator_depth = min_t(int, newxprt->sc_ord, dev->attrs.max_qp_init_rd_atom); if (!conn_param.initiator_depth) { - dprintk("svcrdma: invalid ORD setting\n"); ret = -EINVAL; + trace_svcrdma_initdepth_err(newxprt, ret); goto errout; } conn_param.private_data = &pmsg; conn_param.private_data_len = sizeof(pmsg); ret = rdma_accept(newxprt->sc_cm_id, &conn_param); - if (ret) + if (ret) { + trace_svcrdma_accept_err(newxprt, ret); goto errout; + } #if IS_ENABLED(CONFIG_SUNRPC_DEBUG) dprintk("svcrdma: new connection %p accepted:\n", newxprt); @@ -536,12 +527,9 @@ static struct svc_xprt *svc_rdma_accept(struct svc_xprt *xprt) dprintk(" ord : %d\n", conn_param.initiator_depth); #endif - trace_svcrdma_xprt_accept(&newxprt->sc_xprt); return &newxprt->sc_xprt; errout: - dprintk("svcrdma: failure accepting new connection rc=%d.\n", ret); - trace_svcrdma_xprt_fail(&newxprt->sc_xprt); /* Take a reference in case the DTO handler runs */ svc_xprt_get(&newxprt->sc_xprt); if (newxprt->sc_qp && !IS_ERR(newxprt->sc_qp)) @@ -552,10 +540,6 @@ static struct svc_xprt *svc_rdma_accept(struct svc_xprt *xprt) return NULL; } -static void svc_rdma_release_rqst(struct svc_rqst *rqstp) -{ -} - /* * When connected, an svc_xprt has at least two references: * @@ -583,8 +567,6 @@ static void __svc_rdma_free(struct work_struct *work) container_of(work, struct svcxprt_rdma, sc_work); struct svc_xprt *xprt = &rdma->sc_xprt; - trace_svcrdma_xprt_free(xprt); - if (rdma->sc_qp && !IS_ERR(rdma->sc_qp)) ib_drain_qp(rdma->sc_qp); diff --git a/net/sunrpc/xprtrdma/transport.c b/net/sunrpc/xprtrdma/transport.c index 659da37020a4..0c4af7f5e241 100644 --- a/net/sunrpc/xprtrdma/transport.c +++ b/net/sunrpc/xprtrdma/transport.c @@ -68,7 +68,7 @@ * tunables */ -unsigned int xprt_rdma_slot_table_entries = RPCRDMA_DEF_SLOT_TABLE; +static unsigned int xprt_rdma_slot_table_entries = RPCRDMA_DEF_SLOT_TABLE; unsigned int xprt_rdma_max_inline_read = RPCRDMA_DEF_INLINE; unsigned int xprt_rdma_max_inline_write = RPCRDMA_DEF_INLINE; unsigned int xprt_rdma_memreg_strategy = RPCRDMA_FRWR; @@ -281,8 +281,6 @@ xprt_rdma_destroy(struct rpc_xprt *xprt) { struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(xprt); - trace_xprtrdma_op_destroy(r_xprt); - cancel_delayed_work_sync(&r_xprt->rx_connect_worker); rpcrdma_xprt_disconnect(r_xprt); @@ -365,10 +363,6 @@ xprt_setup_rdma(struct xprt_create *args) xprt->max_payload = RPCRDMA_MAX_DATA_SEGS << PAGE_SHIFT; - dprintk("RPC: %s: %s:%s\n", __func__, - xprt->address_strings[RPC_DISPLAY_ADDR], - xprt->address_strings[RPC_DISPLAY_PORT]); - trace_xprtrdma_create(new_xprt); return xprt; } @@ -385,8 +379,6 @@ void xprt_rdma_close(struct rpc_xprt *xprt) { struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(xprt); - trace_xprtrdma_op_close(r_xprt); - rpcrdma_xprt_disconnect(r_xprt); xprt->reestablish_timeout = 0; diff --git a/net/sunrpc/xprtrdma/verbs.c b/net/sunrpc/xprtrdma/verbs.c index cdd84c09df10..2ae348377806 100644 --- a/net/sunrpc/xprtrdma/verbs.c +++ b/net/sunrpc/xprtrdma/verbs.c @@ -141,7 +141,6 @@ void rpcrdma_flush_disconnect(struct ib_cq *cq, struct ib_wc *wc) if (wc->status != IB_WC_SUCCESS && r_xprt->rx_ep->re_connect_status == 1) { r_xprt->rx_ep->re_connect_status = -ECONNABORTED; - trace_xprtrdma_flush_dct(r_xprt, wc->status); xprt_force_disconnect(xprt); } } @@ -289,6 +288,7 @@ rpcrdma_cm_event_handler(struct rdma_cm_id *id, struct rdma_cm_event *event) case RDMA_CM_EVENT_DISCONNECTED: ep->re_connect_status = -ECONNABORTED; disconnected: + xprt_force_disconnect(xprt); return rpcrdma_ep_destroy(ep); default: break; @@ -1355,8 +1355,8 @@ int rpcrdma_post_sends(struct rpcrdma_xprt *r_xprt, struct rpcrdma_req *req) --ep->re_send_count; } + trace_xprtrdma_post_send(req); rc = frwr_send(r_xprt, req); - trace_xprtrdma_post_send(req, rc); if (rc) return -ENOTCONN; return 0; diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index 0bda8a73e8a8..914508ea9b84 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -1594,21 +1594,6 @@ static int xs_get_random_port(void) return rand + min; } -/** - * xs_set_reuseaddr_port - set the socket's port and address reuse options - * @sock: socket - * - * Note that this function has to be called on all sockets that share the - * same port, and it must be called before binding. - */ -static void xs_sock_set_reuseport(struct socket *sock) -{ - int opt = 1; - - kernel_setsockopt(sock, SOL_SOCKET, SO_REUSEPORT, - (char *)&opt, sizeof(opt)); -} - static unsigned short xs_sock_getport(struct socket *sock) { struct sockaddr_storage buf; @@ -1801,7 +1786,7 @@ static struct socket *xs_create_sock(struct rpc_xprt *xprt, xs_reclassify_socket(family, sock); if (reuseport) - xs_sock_set_reuseport(sock); + sock_set_reuseport(sock->sk); err = xs_bind(transport, sock); if (err) { @@ -2110,7 +2095,6 @@ static void xs_tcp_set_socket_timeouts(struct rpc_xprt *xprt, struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt); unsigned int keepidle; unsigned int keepcnt; - unsigned int opt_on = 1; unsigned int timeo; spin_lock(&xprt->transport_lock); @@ -2122,18 +2106,13 @@ static void xs_tcp_set_socket_timeouts(struct rpc_xprt *xprt, spin_unlock(&xprt->transport_lock); /* TCP Keepalive options */ - kernel_setsockopt(sock, SOL_SOCKET, SO_KEEPALIVE, - (char *)&opt_on, sizeof(opt_on)); - kernel_setsockopt(sock, SOL_TCP, TCP_KEEPIDLE, - (char *)&keepidle, sizeof(keepidle)); - kernel_setsockopt(sock, SOL_TCP, TCP_KEEPINTVL, - (char *)&keepidle, sizeof(keepidle)); - kernel_setsockopt(sock, SOL_TCP, TCP_KEEPCNT, - (char *)&keepcnt, sizeof(keepcnt)); + sock_set_keepalive(sock->sk); + tcp_sock_set_keepidle(sock->sk, keepidle); + tcp_sock_set_keepintvl(sock->sk, keepidle); + tcp_sock_set_keepcnt(sock->sk, keepcnt); /* TCP user timeout (see RFC5482) */ - kernel_setsockopt(sock, SOL_TCP, TCP_USER_TIMEOUT, - (char *)&timeo, sizeof(timeo)); + tcp_sock_set_user_timeout(sock->sk, timeo); } static void xs_tcp_set_connect_timeout(struct rpc_xprt *xprt, @@ -2171,7 +2150,6 @@ static int xs_tcp_finish_connecting(struct rpc_xprt *xprt, struct socket *sock) if (!transport->inet) { struct sock *sk = sock->sk; - unsigned int addr_pref = IPV6_PREFER_SRC_PUBLIC; /* Avoid temporary address, they are bad for long-lived * connections such as NFS mounts. @@ -2180,8 +2158,10 @@ static int xs_tcp_finish_connecting(struct rpc_xprt *xprt, struct socket *sock) * knowledge about the normal duration of connections, * MAY override this as appropriate. */ - kernel_setsockopt(sock, SOL_IPV6, IPV6_ADDR_PREFERENCES, - (char *)&addr_pref, sizeof(addr_pref)); + if (xs_addr(xprt)->sa_family == PF_INET6) { + ip6_sock_set_addr_preferences(sk, + IPV6_PREFER_SRC_PUBLIC); + } xs_tcp_set_socket_timeouts(xprt, sock); @@ -2548,8 +2528,16 @@ static int bc_sendto(struct rpc_rqst *req) return sent; } -/* - * The send routine. Borrows from svc_send +/** + * bc_send_request - Send a backchannel Call on a TCP socket + * @req: rpc_rqst containing Call message to be sent + * + * xpt_mutex ensures @rqstp's whole message is written to the socket + * without interruption. + * + * Return values: + * %0 if the message was sent successfully + * %ENOTCONN if the message was not sent */ static int bc_send_request(struct rpc_rqst *req) { @@ -2584,6 +2572,7 @@ static int bc_send_request(struct rpc_rqst *req) static void bc_close(struct rpc_xprt *xprt) { + xprt_disconnect_done(xprt); } /* |