From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756664Ab1CWTeJ (ORCPT ); Wed, 23 Mar 2011 15:34:09 -0400 Received: from fieldses.org ([174.143.236.118]:35416 "EHLO fieldses.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756647Ab1CWTeH (ORCPT ); Wed, 23 Mar 2011 15:34:07 -0400 Date: Wed, 23 Mar 2011 15:34:06 -0400 To: Linus Torvalds Cc: linux-nfs@vger.kernel.org, linux-kernel@vger.kernel.org Subject: nfsd changes for 2.6.39 Message-ID: <20110323193406.GB26306@fieldses.org> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline Content-Transfer-Encoding: 8bit User-Agent: Mutt/1.5.20 (2009-06-14) From: "J. Bruce Fields" Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org The following nfsd changes are available from the for-2.6.39 branch at: git://linux-nfs.org/~bfields/linux.git for-2.6.39 This was a quiet release cycle! Kevin Coffman added a new interface allowing userland to query our enctype support, fixing some problems with rpcsec_gss security negotiation; other than that, these are routine bugfixes. I do have one more set of server rpc patches that I may pass along in a few days depending on some test results. --b. J. Bruce Fields (7): svcrpc: fix rare race on unix_domain creation nfsd4: minor nfs4state.c reshuffling nfsd4: fix struct file leak svcrpc: fix bad argument in unix_domain_find nfsd: fix compile error nfsd4: fix comment and remove unused nfsd4_file fields Merge create_session decoding fix into for-2.6.39 Jesper Juhl (2): NFSD, VFS: Remove dead code in nfsd_rename() SUNRPC: Remove resource leak in svc_rdma_send_error() Kevin Coffman (2): gss:krb5 only include enctype numbers in gm_upcall_enctypes nfsd: add proc file listing kernel's gss_krb5 enctypes Mi Jinlong (3): nfsd41: modify the members value of nfsd4_op_flags nfs41: make sure nfs server return right ca_maxresponsesize_cached nfsd: wrong index used in inner loop Namhyung Kim (1): locks: use assign_type() Shan Wei (1): nfsd: kill unused macro definition commit 4be34b9d69c97211ff4eb00d79078f3c1593804d Author: Jesper Juhl Date: Sat Jan 22 22:40:20 2011 +0100 SUNRPC: Remove resource leak in svc_rdma_send_error() We leak the memory allocated to 'ctxt' when we return after 'ib_dma_mapping_error()' returns !=0. Signed-off-by: Jesper Juhl Signed-off-by: J. Bruce Fields commit 5a02ab7c3c4580f94d13c683721039855b67cda6 Author: Mi Jinlong Date: Fri Mar 11 12:13:55 2011 +0800 nfsd: wrong index used in inner loop We must not use dummy for index. After the first index, READ32(dummy) will change dummy!!!! Signed-off-by: Mi Jinlong [bfields@redhat.com: Trond points out READ_BUF alone is sufficient.] Cc: stable@kernel.org Signed-off-by: J. Bruce Fields commit cf507b6f8ebae89c0e7bbd00abc6d8887c9035b1 Merge: 9ae78bc 3ec07aa Author: J. Bruce Fields Date: Thu Mar 17 13:07:20 2011 -0400 Merge create_session decoding fix into for-2.6.39 This needs a further fixup! commit 9ae78bcc000168251f893b1bf92a848308187695 Author: J. Bruce Fields Date: Thu Mar 17 12:52:33 2011 -0400 nfsd4: fix comment and remove unused nfsd4_file fields A couple fields here were left over from a previous version of a patch, and are no longer used. Signed-off-by: J. Bruce Fields commit d2b217439fd1d8f7857175f063113f4d1d66306c Author: Mi Jinlong Date: Thu Mar 10 17:43:37 2011 +0800 nfs41: make sure nfs server return right ca_maxresponsesize_cached According to rfc5661, ca_maxresponsesize_cached: Like ca_maxresponsesize, but the maximum size of a reply that will be stored in the reply cache (Section 2.10.6.1). For each channel, the server MAY decrease this value, but MUST NOT increase it. the latest kernel(2.6.38-rc8) may increase the value for ignoring request's ca_maxresponsesize_cached value. We should not ignore it. Signed-off-by: Mi Jinlong Signed-off-by: J. Bruce Fields commit 0a5e5f122c756d1c1a6ca712eda76ea8664e5fd9 Author: J. Bruce Fields Date: Mon Mar 14 20:57:44 2011 -0400 nfsd: fix compile error "fs/built-in.o: In function `supported_enctypes_show': nfsctl.c:(.text+0x7beb0): undefined reference to `gss_mech_get_by_name' nfsctl.c:(.text+0x7bebc): undefined reference to `gss_mech_put' " Reported-by: Guennadi Liakhovetski Signed-off-by: J. Bruce Fields commit 352b5d13c0684ba8cd103aa20cb74f105334562a Author: J. Bruce Fields Date: Wed Mar 9 22:40:30 2011 -0500 svcrpc: fix bad argument in unix_domain_find "After merging the nfsd tree, today's linux-next build (powerpc ppc64_defconfig) produced this warning: net/sunrpc/svcauth_unix.c: In function 'unix_domain_find': net/sunrpc/svcauth_unix.c:58: warning: passing argument 1 of +'svcauth_unix_domain_release' from incompatible pointer type net/sunrpc/svcauth_unix.c:41: note: expected 'struct auth_domain *' but argument +is of type 'struct unix_domain *' Introduced by commit 8b3e07ac908d ("svcrpc: fix rare race on unix_domain creation")." Reported-by: Stephen Rothwell Signed-off-by: J. Bruce Fields commit 0997b173609b9229ece28941c118a2a9b278796e Author: J. Bruce Fields Date: Wed Mar 2 18:01:35 2011 -0500 nfsd4: fix struct file leak Make sure we properly reference count the struct files that a lock depends on, and release them when the lock stateid is released. This fixes a major leak of struct files when using locking over nfsv4. Cc: stable@kernel.org Reported-by: Rick Koshi Tested-by: Ivo Přikryl Signed-off-by: J. Bruce Fields commit 529d7b2a7fa31e9f7d08bc790d232c3cbe64fa24 Author: J. Bruce Fields Date: Wed Mar 2 23:48:33 2011 -0500 nfsd4: minor nfs4state.c reshuffling Minor cleanup in preparation for a bugfix--moving some code to avoid forward references, etc. No change in functionality. Cc: stable@kernel.org Signed-off-by: J. Bruce Fields commit 8b3e07ac908d005bb791410f594cce8744f6806a Author: J. Bruce Fields Date: Mon Mar 7 22:50:47 2011 -0500 svcrpc: fix rare race on unix_domain creation Note that "new" here is not yet fully initialized; auth_domain_put should be called only on auth_domains that have actually been added to the hash. Before this fix, two attempts to add the same domain at once could cause the hlist_del in auth_domain_put to fail. Signed-off-by: J. Bruce Fields commit 5ece3cafbd88d4da5c734e1810c4a2e6474b57b2 Author: Mi Jinlong Date: Fri Feb 18 09:08:31 2011 +0800 nfsd41: modify the members value of nfsd4_op_flags The members of nfsd4_op_flags, (ALLOWED_WITHOUT_FH | ALLOWED_ON_ABSENT_FS) equals to ALLOWED_AS_FIRST_OP, maybe that's not what we want. OP_PUTROOTFH with op_flags = ALLOWED_WITHOUT_FH | ALLOWED_ON_ABSENT_FS, can't appears as the first operation with out SEQUENCE ops. This patch modify the wrong value of ALLOWED_WITHOUT_FH etc which was introduced by f9bb94c4. Cc: stable@kernel.org Reviewed-by: Benny Halevy Signed-off-by: Mi Jinlong Signed-off-by: J. Bruce Fields commit b0b0c0a26e846ae6646af9f59a3d2ea06b49cbc7 Author: Kevin Coffman Date: Wed Mar 2 19:51:42 2011 -0500 nfsd: add proc file listing kernel's gss_krb5 enctypes Add a new proc file which lists the encryption types supported by the kernel's gss_krb5 code. Newer MIT Kerberos libraries support the assertion of acceptor subkeys. This enctype information allows user-land (svcgssd) to request that the Kerberos libraries limit the encryption types that it uses when generating the subkeys. Signed-off-by: Kevin Coffman Signed-off-by: J. Bruce Fields commit 540c8cb6a576f34a9a0b04467f46bb6e67a1f852 Author: Kevin Coffman Date: Wed Mar 2 19:51:41 2011 -0500 gss:krb5 only include enctype numbers in gm_upcall_enctypes Make the value in gm_upcall_enctypes just the enctype values. This allows the values to be used more easily elsewhere. Signed-off-by: Kevin Coffman Signed-off-by: J. Bruce Fields commit 46d4cef9cf54f2f8b15216e3f6dad69750c69e0c Author: Jesper Juhl Date: Sun Mar 6 00:30:35 2011 +0100 NFSD, VFS: Remove dead code in nfsd_rename() Currently we have the following code in fs/nfsd/vfs.c::nfsd_rename() : ... host_err = nfsd_break_lease(odentry->d_inode); if (host_err) goto out_drop_write; if (ndentry->d_inode) { host_err = nfsd_break_lease(ndentry->d_inode); if (host_err) goto out_drop_write; } if (host_err) goto out_drop_write; ... 'host_err' is guaranteed to be 0 by the time we test 'ndentry->d_inode'. If 'host_err' becomes != 0 inside the 'if' statement, then we goto 'out_drop_write'. So, after the 'if' statement there is no way that 'host_err' can be anything but 0, so the test afterwards is just dead code. This patch removes the dead code. Signed-off-by: Jesper Juhl Signed-off-by: J. Bruce Fields commit 35079582e72efcabf7c70f3d3ee4f96e6f196606 Author: Shan Wei Date: Fri Jan 14 17:35:59 2011 +0800 nfsd: kill unused macro definition These macros had never been used for several years. So, remove them. Signed-off-by: Shan Wei Signed-off-by: J. Bruce Fields commit f32cb53219a956b96a4cd1ee7c6b1b8a48d40e9f Author: Namhyung Kim Date: Mon Jan 17 15:45:59 2011 +0900 locks: use assign_type() Signed-off-by: Namhyung Kim Signed-off-by: J. Bruce Fields