From patchwork Thu Jun 30 08:20:05 2011 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Mi Jinlong X-Patchwork-Id: 931902 Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by demeter1.kernel.org (8.14.4/8.14.4) with ESMTP id p5U8Grct019800 for ; Thu, 30 Jun 2011 08:16:55 GMT Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1757684Ab1F3IQy (ORCPT ); Thu, 30 Jun 2011 04:16:54 -0400 Received: from cn.fujitsu.com ([222.73.24.84]:52337 "EHLO song.cn.fujitsu.com" rhost-flags-OK-FAIL-OK-OK) by vger.kernel.org with ESMTP id S1758530Ab1F3IQw (ORCPT ); Thu, 30 Jun 2011 04:16:52 -0400 Received: from tang.cn.fujitsu.com (tang.cn.fujitsu.com [10.167.250.3]) by song.cn.fujitsu.com (Postfix) with ESMTP id 8F9AE17011A; Thu, 30 Jun 2011 16:16:48 +0800 (CST) Received: from mailserver.fnst.cn.fujitsu.com (tang.cn.fujitsu.com [127.0.0.1]) by tang.cn.fujitsu.com (8.14.3/8.13.1) with ESMTP id p5U8Gmsr009869; Thu, 30 Jun 2011 16:16:48 +0800 Received: from [127.0.0.1] ([10.167.225.24]) by mailserver.fnst.cn.fujitsu.com (Lotus Domino Release 8.5.1FP4) with ESMTP id 2011063016161656-711110 ; Thu, 30 Jun 2011 16:16:16 +0800 Message-ID: <4E0C31B5.9090302@cn.fujitsu.com> Date: Thu, 30 Jun 2011 16:20:05 +0800 From: Mi Jinlong User-Agent: Thunderbird 2.0.0.9 (Windows/20071031) MIME-Version: 1.0 To: "J. Bruce Fields" CC: NFS Subject: [RFC][PATCH 2/2] nfsd41: try to check reply size before operation X-MIMETrack: Itemize by SMTP Server on mailserver/fnst(Release 8.5.1FP4|July 25, 2010) at 2011-06-30 16:16:16, Serialize by Router on mailserver/fnst(Release 8.5.1FP4|July 25, 2010) at 2011-06-30 16:16:16, Serialize complete at 2011-06-30 16:16:16 Sender: linux-nfs-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-nfs@vger.kernel.org X-Greylist: IP, sender and recipient auto-whitelisted, not delayed by milter-greylist-4.2.6 (demeter1.kernel.org [140.211.167.41]); Thu, 30 Jun 2011 08:16:55 +0000 (UTC) Implementing each operation's enc_size, and try to check reply size before operation. Signed-off-by: Mi Jinlong --- fs/nfsd/nfs4proc.c | 2 +- fs/nfsd/nfs4xdr.c | 210 +++++++++++++++++++++++++++++++++++++++++++++++----- 2 files changed, 193 insertions(+), 19 deletions(-) diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index 3a6dbd7..06ca11c 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -1165,7 +1165,7 @@ nfsd4_proc_compound(struct svc_rqst *rqstp, goto encode_op; } - if (opdesc->op_func) + if (op->status == 0 && opdesc->op_func) op->status = opdesc->op_func(rqstp, cstate, &op->u); else BUG_ON(op->status == nfs_ok); diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index 1273661..ae35f7e 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -3153,6 +3153,89 @@ struct nfsd4_enc_op { u32 enc_size; }; +#define op_encode_hdr_size (2) + +#define encode_stateid_maxsz (XDR_QUADLEN(NFS4_STATEID_SIZE)) +#define encode_verifier_maxsz (XDR_QUADLEN(NFS4_VERIFIER_SIZE)) + +#define nfsd4_enc_access_sz (op_encode_hdr_size + 2) +#define nfsd4_enc_close_sz (op_encode_hdr_size + encode_stateid_maxsz) +#define nfsd4_enc_commit_sz (op_encode_hdr_size + encode_verifier_maxsz) +#define nfsd4_enc_create_sz (op_encode_hdr_size + encode_stateid_maxsz) + +#define nfs4_owner_maxsz (1 + XDR_QUADLEN(IDMAP_NAMESZ)) +#define nfs4_group_maxsz (1 + XDR_QUADLEN(IDMAP_NAMESZ)) +#define nfs4_fattr_bitmap_maxsz (3) +#define nfs4_fattr_value_maxsz (1 + (1 + 2 + 2 + 4 + 2 + 1 + 1 + 2 + 2 + \ + 3 + 3 + 3 + nfs4_owner_maxsz + \ + nfs4_group_maxsz)) +#define nfs4_fattr_maxsz (nfs4_fattr_bitmap_maxsz + \ + nfs4_fattr_value_maxsz) +#define nfsd4_enc_getattr_sz (op_encode_hdr_size + nfs4_fattr_maxsz) +#define nfsd4_enc_getfh_sz (op_encode_hdr_size + 1 + \ + ((3+NFS4_FHSIZE) >> 2)) + +#define encode_change_info_maxsz (5) +#define nfsd4_enc_link_sz (op_encode_hdr_size + encode_change_info_maxsz) + +#define encode_lockowner_maxsz (1 + XDR_QUADLEN(IDMAP_NAMESZ)) +#define encode_lock_denied_maxsz \ + (8 + encode_lockowner_maxsz) +#define nfsd4_enc_lock_sz (op_encode_hdr_size + encode_lock_denied_maxsz) +#define nfsd4_enc_lockt_sz (op_encode_hdr_size + encode_lock_denied_maxsz) +#define nfsd4_enc_locku_sz (op_encode_hdr_size + encode_stateid_maxsz) + +#define encode_ace_maxsz (3 + nfs4_owner_maxsz) +#define encode_delegation_maxsz (1 + encode_stateid_maxsz + 1 + \ + encode_ace_maxsz) +#define nfsd4_enc_open_sz (op_encode_hdr_size + encode_stateid_maxsz + \ + encode_change_info_maxsz + 1 + \ + nfs4_fattr_bitmap_maxsz + \ + encode_delegation_maxsz) +#define nfsd4_enc_open_confirm_sz \ + (op_encode_hdr_size + encode_stateid_maxsz) +#define nfsd4_enc_open_downgrade_sz \ + (op_encode_hdr_size + encode_stateid_maxsz) +#define nfsd4_enc_read_sz (op_encode_hdr_size + 2) +#define nfsd4_enc_readdir_sz (op_encode_hdr_size + encode_verifier_maxsz) +#define nfsd4_enc_readlink_sz (op_encode_hdr_size + 1) +#define nfsd4_enc_remove_sz (op_encode_hdr_size + encode_change_info_maxsz) +#define nfsd4_enc_rename_sz (op_encode_hdr_size + \ + encode_change_info_maxsz + \ + encode_change_info_maxsz) + +#define NFS_MAX_SECFLAVORS (12) +#define nfsd4_enc_secinfo_sz (op_encode_hdr_size + 4 + \ + (NFS_MAX_SECFLAVORS * (16 + GSS_OID_MAX_LEN))) + +#define nfsd4_enc_setattr_sz (op_encode_hdr_size + nfs4_fattr_bitmap_maxsz) +#define nfsd4_enc_setclientid_sz \ + (op_encode_hdr_size + 2 + 1024) +#define nfsd4_enc_write_sz (op_encode_hdr_size + 2 + encode_verifier_maxsz) + +/* For NFSv4.1*/ +#define nfsd4_enc_bind_conn_to_session_sz \ + (op_encode_hdr_size + \ + XDR_QUADLEN(NFS4_MAX_SESSIONID_LEN) + 2) +#define nfsd4_enc_exchange_id_sz \ + (op_encode_hdr_size + 2 + 1 + 1 + 1 + 0 + \ + 2 + XDR_QUADLEN(NFS4_OPAQUE_LIMIT) + 1 + \ + XDR_QUADLEN(NFS4_OPAQUE_LIMIT) + 1 + \ + 1 + 0) + +#define encode_channel_attrs_maxsz (6 + 1 + 1) +#define nfsd4_enc_create_session_sz \ + (op_encode_hdr_size + \ + XDR_QUADLEN(NFS4_MAX_SESSIONID_LEN) + \ + 1 + 1 + encode_channel_attrs_maxsz + \ + encode_channel_attrs_maxsz) + +#define nfsd4_enc_destroy_session_sz \ + (op_encode_hdr_size) +#define nfsd4_enc_secinfo_no_name_sz nfsd4_enc_secinfo_sz +#define nfsd4_enc_sequence_sz (op_encode_hdr_size + \ + XDR_QUADLEN(NFS4_MAX_SESSIONID_LEN) + 5) + /* * Note: nfsd4_enc_ops vector is shared for v4.0 and v4.1 * since we don't need to filter out obsolete ops as this is @@ -3161,222 +3244,313 @@ struct nfsd4_enc_op { static struct nfsd4_enc_op nfsd4_enc_ops[] = { [OP_ACCESS] = { .enc_func = (nfsd4_enc)nfsd4_encode_access, + .enc_size = nfsd4_enc_access_sz, }, [OP_CLOSE] = { .enc_func = (nfsd4_enc)nfsd4_encode_close, + .enc_size = nfsd4_enc_close_sz, }, [OP_COMMIT] = { .enc_func = (nfsd4_enc)nfsd4_encode_commit, + .enc_size = nfsd4_enc_commit_sz, }, [OP_CREATE] = { .enc_func = (nfsd4_enc)nfsd4_encode_create, + .enc_size = nfsd4_enc_create_sz, }, [OP_DELEGPURGE] = { .enc_func = (nfsd4_enc)nfsd4_encode_noop, + .enc_size = op_encode_hdr_size, }, [OP_DELEGRETURN] = { .enc_func = (nfsd4_enc)nfsd4_encode_noop, + .enc_size = op_encode_hdr_size, }, [OP_GETATTR] = { .enc_func = (nfsd4_enc)nfsd4_encode_getattr, + .enc_size = nfsd4_enc_getattr_sz, }, [OP_GETFH] = { .enc_func = (nfsd4_enc)nfsd4_encode_getfh, + .enc_size = nfsd4_enc_getfh_sz, }, [OP_LINK] = { .enc_func = (nfsd4_enc)nfsd4_encode_link, + .enc_size = nfsd4_enc_link_sz, }, [OP_LOCK] = { .enc_func = (nfsd4_enc)nfsd4_encode_lock, + .enc_size = nfsd4_enc_lock_sz, }, [OP_LOCKT] = { .enc_func = (nfsd4_enc)nfsd4_encode_lockt, + .enc_size = nfsd4_enc_lockt_sz, }, [OP_LOCKU] = { .enc_func = (nfsd4_enc)nfsd4_encode_locku, + .enc_size = nfsd4_enc_locku_sz, }, [OP_LOOKUP] = { .enc_func = (nfsd4_enc)nfsd4_encode_noop, + .enc_size = op_encode_hdr_size, }, [OP_LOOKUPP] = { .enc_func = (nfsd4_enc)nfsd4_encode_noop, + .enc_size = op_encode_hdr_size, }, [OP_NVERIFY] = { .enc_func = (nfsd4_enc)nfsd4_encode_noop, + .enc_size = op_encode_hdr_size, }, [OP_OPEN] = { .enc_func = (nfsd4_enc)nfsd4_encode_open, + .enc_size = nfsd4_enc_open_sz, }, [OP_OPENATTR] = { .enc_func = (nfsd4_enc)nfsd4_encode_noop, + .enc_size = op_encode_hdr_size, }, [OP_OPEN_CONFIRM] = { .enc_func = (nfsd4_enc)nfsd4_encode_open_confirm, + .enc_size = nfsd4_enc_open_confirm_sz, }, [OP_OPEN_DOWNGRADE] = { .enc_func = (nfsd4_enc)nfsd4_encode_open_downgrade, + .enc_size = nfsd4_enc_open_downgrade_sz, }, [OP_PUTFH] = { .enc_func = (nfsd4_enc)nfsd4_encode_noop, + .enc_size = op_encode_hdr_size, }, [OP_PUTPUBFH] = { .enc_func = (nfsd4_enc)nfsd4_encode_noop, + .enc_size = op_encode_hdr_size, }, [OP_PUTROOTFH] = { .enc_func = (nfsd4_enc)nfsd4_encode_noop, + .enc_size = op_encode_hdr_size, }, [OP_READ] = { .enc_func = (nfsd4_enc)nfsd4_encode_read, + .enc_size = nfsd4_enc_read_sz, }, [OP_READDIR] = { .enc_func = (nfsd4_enc)nfsd4_encode_readdir, + .enc_size = nfsd4_enc_readdir_sz, }, [OP_READLINK] = { .enc_func = (nfsd4_enc)nfsd4_encode_readlink, + .enc_size = nfsd4_enc_readlink_sz, }, [OP_REMOVE] = { .enc_func = (nfsd4_enc)nfsd4_encode_remove, + .enc_size = nfsd4_enc_remove_sz, }, [OP_RENAME] = { .enc_func = (nfsd4_enc)nfsd4_encode_rename, + .enc_size = nfsd4_enc_rename_sz, }, [OP_RENEW] = { .enc_func = (nfsd4_enc)nfsd4_encode_noop, + .enc_size = op_encode_hdr_size, }, [OP_RESTOREFH] = { .enc_func = (nfsd4_enc)nfsd4_encode_noop, + .enc_size = op_encode_hdr_size, }, [OP_SAVEFH] = { .enc_func = (nfsd4_enc)nfsd4_encode_noop, + .enc_size = op_encode_hdr_size, }, [OP_SECINFO] = { .enc_func = (nfsd4_enc)nfsd4_encode_secinfo, + .enc_size = nfsd4_enc_secinfo_sz, }, [OP_SETATTR] = { .enc_func = (nfsd4_enc)nfsd4_encode_setattr, + .enc_size = nfsd4_enc_setattr_sz, }, [OP_SETCLIENTID] = { .enc_func = (nfsd4_enc)nfsd4_encode_setclientid, + .enc_size = nfsd4_enc_setclientid_sz, }, [OP_SETCLIENTID_CONFIRM] = { .enc_func = (nfsd4_enc)nfsd4_encode_noop, + .enc_size = op_encode_hdr_size, }, [OP_VERIFY] = { .enc_func = (nfsd4_enc)nfsd4_encode_noop, + .enc_size = op_encode_hdr_size, }, [OP_WRITE] = { .enc_func = (nfsd4_enc)nfsd4_encode_write, + .enc_size = nfsd4_enc_write_sz, }, [OP_RELEASE_LOCKOWNER] = { .enc_func = (nfsd4_enc)nfsd4_encode_noop, + .enc_size = op_encode_hdr_size, }, /* NFSv4.1 operations */ [OP_BACKCHANNEL_CTL] = { .enc_func = (nfsd4_enc)nfsd4_encode_noop, + .enc_size = op_encode_hdr_size, }, [OP_BIND_CONN_TO_SESSION] = { .enc_func = (nfsd4_enc)nfsd4_encode_bind_conn_to_session, + .enc_size = nfsd4_enc_bind_conn_to_session_sz, }, [OP_EXCHANGE_ID] = { .enc_func = (nfsd4_enc)nfsd4_encode_exchange_id, + .enc_size = nfsd4_enc_exchange_id_sz, }, [OP_CREATE_SESSION] = { .enc_func = (nfsd4_enc)nfsd4_encode_create_session, + .enc_size = nfsd4_enc_create_session_sz, }, [OP_DESTROY_SESSION] = { .enc_func = (nfsd4_enc)nfsd4_encode_destroy_session, + .enc_size = nfsd4_enc_destroy_session_sz, }, [OP_FREE_STATEID] = { .enc_func = (nfsd4_enc)nfsd4_encode_noop, + .enc_size = op_encode_hdr_size, }, [OP_GET_DIR_DELEGATION] = { .enc_func = (nfsd4_enc)nfsd4_encode_noop, + .enc_size = op_encode_hdr_size, }, [OP_GETDEVICEINFO] = { .enc_func = (nfsd4_enc)nfsd4_encode_noop, + .enc_size = op_encode_hdr_size, }, [OP_GETDEVICELIST] = { .enc_func = (nfsd4_enc)nfsd4_encode_noop, + .enc_size = op_encode_hdr_size, }, [OP_LAYOUTCOMMIT] = { .enc_func = (nfsd4_enc)nfsd4_encode_noop, + .enc_size = op_encode_hdr_size, }, [OP_LAYOUTGET] = { .enc_func = (nfsd4_enc)nfsd4_encode_noop, + .enc_size = op_encode_hdr_size, }, [OP_LAYOUTRETURN] = { .enc_func = (nfsd4_enc)nfsd4_encode_noop, + .enc_size = op_encode_hdr_size, }, [OP_SECINFO_NO_NAME] = { .enc_func = (nfsd4_enc)nfsd4_encode_secinfo_no_name, + .enc_size = nfsd4_enc_secinfo_no_name_sz, }, [OP_SEQUENCE] = { .enc_func = (nfsd4_enc)nfsd4_encode_sequence, + .enc_size = nfsd4_enc_sequence_sz, }, [OP_SET_SSV] = { .enc_func = (nfsd4_enc)nfsd4_encode_noop, + .enc_size = op_encode_hdr_size, }, [OP_TEST_STATEID] = { .enc_func = (nfsd4_enc)nfsd4_encode_noop, + .enc_size = op_encode_hdr_size, }, [OP_WANT_DELEGATION] = { .enc_func = (nfsd4_enc)nfsd4_encode_noop, + .enc_size = op_encode_hdr_size, }, [OP_DESTROY_CLIENTID] = { .enc_func = (nfsd4_enc)nfsd4_encode_noop, + .enc_size = op_encode_hdr_size, }, [OP_RECLAIM_COMPLETE] = { .enc_func = (nfsd4_enc)nfsd4_encode_noop, + .enc_size = op_encode_hdr_size, }, }; +static u32 get_ops_max_rsz(struct nfsd4_compoundres *resp) +{ + struct nfsd4_compoundargs *args = resp->rqstp->rq_argp; + struct nfsd4_op op = args->ops[resp->opcnt]; + int length = 0, maxcount = 0; + + switch (op.opnum) { + case OP_READ: + maxcount = svc_max_payload(resp->rqstp); + if (maxcount > op.u.read.rd_length) + length = op.u.read.rd_length; + else + length = maxcount; + break; + + case OP_READDIR: + if (op.u.readdir.rd_maxcount < PAGE_SIZE) + length = op.u.readdir.rd_maxcount; + else + length = PAGE_SIZE; + break; + case OP_READLINK: + length = PATH_MAX; + break; + } + + length += nfsd4_enc_ops[op.opnum].enc_size * 4; + dprintk("%s opnum %u max reply %u\n", __func__, op.opnum, length); + + return length; +} + /* * Calculate the total amount of memory that the compound response has taken * after encoding the current operation. * - * pad: add on 8 bytes for the next operation's op_code and status so that - * there is room to cache a failure on the next operation. - * - * Compare this length to the session se_fmaxresp_cached. + * Compare this length to the session se_fmaxresp_sz and se_fmaxresp_cached. * * Our se_fmaxresp_cached will always be a multiple of PAGE_SIZE, and so * will be at least a page and will therefore hold the xdr_buf head. */ -static int nfsd4_check_drc_limit(struct nfsd4_compoundres *resp) +static int nfsd4_check_resp_limit(struct nfsd4_compoundres *resp) { int status = 0; struct xdr_buf *xb = &resp->rqstp->rq_res; struct nfsd4_compoundargs *args = resp->rqstp->rq_argp; struct nfsd4_session *session = NULL; struct nfsd4_slot *slot = resp->cstate.slot; - u32 length, tlen = 0, pad = 8; + u32 length = 0, tlen = 0, nlen = 0; if (!nfsd4_has_session(&resp->cstate)) return status; session = resp->cstate.session; - if (session == NULL || slot->sl_cachethis == 0) + if (session == NULL) return status; - if (resp->opcnt >= args->opcnt) - pad = 0; /* this is the last operation */ + if (resp->opcnt < args->opcnt) + nlen = get_ops_max_rsz(resp); + else + return status; if (xb->page_len == 0) { - length = (char *)resp->p - (char *)xb->head[0].iov_base + pad; + length = (char *)resp->p - (char *)xb->head[0].iov_base + nlen; } else { if (xb->tail[0].iov_base && xb->tail[0].iov_len > 0) tlen = (char *)resp->p - (char *)xb->tail[0].iov_base; - length = xb->head[0].iov_len + xb->page_len + tlen + pad; + length = xb->head[0].iov_len + xb->page_len + tlen + nlen; } - dprintk("%s length %u, xb->page_len %u tlen %u pad %u\n", __func__, - length, xb->page_len, tlen, pad); + dprintk("%s length %u, xb->page_len %u tlen %u nlen %u\n", __func__, + length, xb->page_len, tlen, nlen); - if (length <= session->se_fchannel.maxresp_cached) - return status; - else - return nfserr_rep_too_big_to_cache; + if (length > session->se_fchannel.maxresp_sz) + args->ops[resp->opcnt].status = nfserr_rep_too_big; + + if (slot->sl_cachethis == 1 && + length > session->se_fchannel.maxresp_cached) + args->ops[resp->opcnt].status = nfserr_rep_too_big_to_cache; + + return 0; } void @@ -3396,7 +3570,7 @@ nfsd4_encode_operation(struct nfsd4_compoundres *resp, struct nfsd4_op *op) !nfsd4_enc_ops[op->opnum].enc_func); op->status = nfsd4_enc_ops[op->opnum].enc_func(resp, op->status, &op->u); /* nfsd4_check_drc_limit guarantees enough room for error status */ - if (!op->status && nfsd4_check_drc_limit(resp)) + if (!op->status && nfsd4_check_resp_limit(resp)) op->status = nfserr_rep_too_big_to_cache; status: /*