Re: [SCM] Samba Shared Repository - branch master updated
Am 04.01.24 um 15:36 schrieb Brown, James William: On 1/4/2024 7:46 AM, Stefan Metzmacher wrote: @@ -3905,7 +3937,24 @@ static int moveip(TALLOC_CTX *mem_ctx, struct ctdb_context *ctdb, return ret; } - return 0; + /* +* It isn't strictly necessary to wait until takeover runs are +* re-enabled but doing so can't hurt. +*/ + ret = disable_takeover_runs(mem_ctx, + ctdb, + 0, + connected_pnn, + connected_count); + if (ret != 0) { + fprintf(stderr, "Failed to enable takeover runs\n"); + return ret; + } Message should be "Failed to disable takeover runs\n" like previous change at 3886. This is a bit confusing but correct, the function name is strange. We have "disable" with a timeout of 2*options.timelimit and "enable" with a timeout of 0. metze
Re: [SCM] Samba Shared Repository - branch master updated
On 1/4/2024 7:46 AM, Stefan Metzmacher wrote: @@ -3905,7 +3937,24 @@ static int moveip(TALLOC_CTX *mem_ctx, struct ctdb_context *ctdb, return ret; } - return 0; + /* +* It isn't strictly necessary to wait until takeover runs are +* re-enabled but doing so can't hurt. +*/ + ret = disable_takeover_runs(mem_ctx, + ctdb, + 0, + connected_pnn, + connected_count); + if (ret != 0) { + fprintf(stderr, "Failed to enable takeover runs\n"); + return ret; + } Message should be "Failed to disable takeover runs\n" like previous change at 3886.
Re: [SCM] Samba Shared Repository - branch master updated
Am 24.01.23 um 10:16 schrieb Volker Lendecke: The branch, master has been updated via f7b50bc059d smbd: Use smbXsrv_open_global_parse_record() in .._verify_record() via 132b83d0659 smbd: Simplify smbXsrv_open_global_parse_record() via 2f6776741dc smbd: Move smbXsrv_open_global_parse_record() up in smbXsrv_open.c via 3c779de8cf9 smbd: Simplify smbXsrv_open_global_verify_record() via f1a66267bcf smbd: Save a few lines in smb2srv_open_lookup_replay_cache() via 35a32171b50 smbd: Fix a typo from 253891032ee python: Don't use deprecated escape sequences https://git.samba.org/?p=samba.git;a=shortlog;h=master - Log - commit f7b50bc059d1b5c7e40cdc4e88ef5ee16f7db670 Author: Volker Lendecke Date: Thu Jan 19 12:29:20 2023 +0100 smbd: Use smbXsrv_open_global_parse_record() in .._verify_record() Signed-off-by: Volker Lendecke Autobuild-User(master): Volker Lendecke Autobuild-Date(master): Tue Jan 24 09:15:26 UTC 2023 on atb-devel-224 commit 132b83d0659ddc25a96327edc1c7dd23b17a56fd Author: Volker Lendecke Date: Thu Jan 19 12:25:21 2023 +0100 smbd: Simplify smbXsrv_open_global_parse_record() It does not need a db_record. Signed-off-by: Volker Lendecke commit 2f6776741dc6469d78b94da22d75f26a5fc9 Author: Volker Lendecke Date: Thu Jan 19 12:22:33 2023 +0100 smbd: Move smbXsrv_open_global_parse_record() up in smbXsrv_open.c Avoid a prototype in the next patches Signed-off-by: Volker Lendecke commit 3c779de8cf99d0936956a12484fd726d5be46c7e Author: Volker Lendecke Date: Fri Jan 6 16:25:03 2023 +0100 smbd: Simplify smbXsrv_open_global_verify_record() Don't depend on the record to be passed in, return NTSTATUS. The two flags were a bit confusing to me, now NT_STATUS_OK means "found a valid record with a live process", and NT_STATUS_FATAL_APP_EXIT means we found a stale record from a crashed smbd Signed-off-by: Volker Lendecke commit f1a66267bcfcd48f3c7ca2ada3f62d40209163e3 Author: Volker Lendecke Date: Wed Jan 11 11:44:29 2023 +0100 smbd: Save a few lines in smb2srv_open_lookup_replay_cache() Directly initialize variables, don't leave dangling pointers in TDB_DATA Signed-off-by: Volker Lendecke commit 35a32171b5067d5b80acffc99f8d43cdc7f5f9a7 Author: Volker Lendecke Date: Wed Jan 11 08:18:35 2023 +0100 smbd: Fix a typo Signed-off-by: Volker Lendecke Reviewed-by: Stefan Metzmacher These were all reviewd by me... metze
Re: s4:kdc: Set kerberos debug class for kdc service (Re: [SCM] Samba Shared Repository - branch master updated)
On Friday, 9 September 2022 10:53:51 CEST Stefan Metzmacher wrote: > Hi Andreas, > > > - Log - > > commit 761ce8cfe41139ab5656dec5cc05f2f576095216 > > Author: Andreas Schneider > > Date: Tue Sep 6 10:19:54 2022 +0200 > > > > s4:kdc: Set kerberos debug class for kdc service > > > > Signed-off-by: Andreas Schneider > > Reviewed-by: Douglas Bagnall > > Reviewed-by: Andrew Bartlett > > Can we please do this for all files under source4/kdc ? Hi metze, yes we can. https://gitlab.com/samba-team/samba/-/merge_requests/2715 > metze -- Andreas Schneider a...@samba.org Samba Team www.samba.org GPG-ID: 8DFF53E18F2ABC8D8F3C92237EE0FC4DCC014E3D
s4:kdc: Set kerberos debug class for kdc service (Re: [SCM] Samba Shared Repository - branch master updated)
Hi Andreas, - Log - commit 761ce8cfe41139ab5656dec5cc05f2f576095216 Author: Andreas Schneider Date: Tue Sep 6 10:19:54 2022 +0200 s4:kdc: Set kerberos debug class for kdc service Signed-off-by: Andreas Schneider Reviewed-by: Douglas Bagnall Reviewed-by: Andrew Bartlett Can we please do this for all files under source4/kdc ? metze
Re: [SCM] Samba Shared Repository - branch master updated
The delete should be "rm -f build/DTD/samba.build.pathconfig" to match other uses. On 6/22/2022 12:44 PM, Andreas Schneider wrote: > diff --git a/docs-xml/Makefile b/docs-xml/Makefile > index c614097f6c0..ae50b14d79a 100644 > --- a/docs-xml/Makefile > +++ b/docs-xml/Makefile > @@ -17,7 +17,7 @@ DBLATEX_OPTIONS = -p xslt/latex.xsl -i xslt/latex > > DATETIME := $(shell date +%Y%m%d%H%M%S) > > -XSLTPROC_DEPS = build/catalog.xml build/DTD/samba.build.version > +XSLTPROC_DEPS = build/catalog.xml build/DTD/samba.build.version > build/DTD/samba.build.pathconfig > > ifeq ($(PROFILE), Y) > XSLTPROC += --profile --load-trace --timing > @@ -57,6 +57,7 @@ clean:: > rm -f xslt/figures/*pdf > rm -f $(SMBDOTCONFDOC)/parameters.*.xml > rm -f build/catalog.xml > + rm -f DTD/samba.build.pathconfig
Re: Conflicting public symbols... (Re: [SCM] Samba Shared Repository - branch master updated)
On Wed, Feb 10, 2021 at 10:36:18PM +0100, Stefan Metzmacher wrote: Hi, - Log - commit 551532d007970ab11dca71b532754728a6a78496 Author: xzhao9 Date: Sun Jan 3 13:02:18 2021 -0500 s3:registry Renaming get_charset() to smbreg_get_charset() Rename to smbreg_get_charset() function to avoid naming conflict with MariaDB. Signed-off-by: xzhao9 Reviewed-by: Jeremy Allison Reviewed-by: David Mulder Autobuild-User(master): Jeremy Allison Autobuild-Date(master): Wed Feb 10 21:00:28 UTC 2021 on sn-devel-184 Do we need a bug report to backport this? I'll create one. There's a similar problem reported here: https://www.downtowndougbrown.com/2021/01/tracking-down-a-segfault-that-suddenly-started-happening/ Does anyone have time to have a deeper look? Looks like "-fvisibility=hidden" is the way to go here.
Conflicting public symbols... (Re: [SCM] Samba Shared Repository - branch master updated)
Hi, > - Log - > commit 551532d007970ab11dca71b532754728a6a78496 > Author: xzhao9 > Date: Sun Jan 3 13:02:18 2021 -0500 > > s3:registry Renaming get_charset() to smbreg_get_charset() > > Rename to smbreg_get_charset() function to avoid naming conflict > with MariaDB. > > Signed-off-by: xzhao9 > Reviewed-by: Jeremy Allison > Reviewed-by: David Mulder > > Autobuild-User(master): Jeremy Allison > Autobuild-Date(master): Wed Feb 10 21:00:28 UTC 2021 on sn-devel-184 Do we need a bug report to backport this? There's a similar problem reported here: https://www.downtowndougbrown.com/2021/01/tracking-down-a-segfault-that-suddenly-started-happening/ Does anyone have time to have a deeper look? Thanks! metze signature.asc Description: OpenPGP digital signature
Re: cli_credentials_parse_name... (Re: [SCM] Samba Shared Repository - branch master updated)
On to, 05 marras 2020, Andrew Bartlett wrote: > On Wed, 2020-11-04 at 19:23 +0200, Alexander Bokovoy wrote: > > On ke, 04 marras 2020, Stefan Metzmacher wrote: > > > Am 04.11.20 um 17:24 schrieb Alexander Bokovoy: > > > > The branch, master has been updated > > > >via f9016912098 lookup_name: allow lookup for own realm > > > >via 00f4262ed0b cli_credentials: add a helper to parse > > > > user or group names > > > >via eb0474d27ba cli_credentials_parse_string: fix parsing > > > > of principals > > > > from a1b021200e3 selftest: add test for new "samba-tool > > > > user unlock" command > > > > > > > > https://git.samba.org/?p=samba.git;a=shortlog;h=master > > > > > > > > > > > > - Log - > > > > > > > > commit f901691209867b32c2d7c5c9274eee196f541654 > > > > Author: Alexander Bokovoy > > > > Date: Wed Nov 4 14:21:33 2020 +0200 > > > > > > > > lookup_name: allow lookup for own realm > > > > > > > > When using a security tab in Windows Explorer, a lookup over > > > > a trusted > > > > forest might come as realm\name instead of NetBIOS domain > > > > name: > > > > > > > > - > > > > --- > > > > [2020/01/13 11:12:39.859134, 1, pid=33253, > > > > effective(1732401004, 1732401004), real(1732401004, 0), > > > > class=rpc_parse] > > > > ../../librpc/ndr/ndr.c:471(ndr_print_function_debug) > > > >lsa_LookupNames3: struct lsa_LookupNames3 > > > > in: struct lsa_LookupNames3 > > > > handle : * > > > > handle: struct policy_handle > > > > handle_type : 0x > > > > (0) > > > > uuid : 000e- > > > > --1c5e-a750e581 > > > > num_names: 0x0001 (1) > > > > names: ARRAY(1) > > > > names: struct lsa_String > > > > length : 0x001e (30) > > > > size : 0x0020 (32) > > > > string : * > > > > string : > > > > 'ipa.test\admins' > > > > sids : * > > > > sids: struct lsa_TransSidArray3 > > > > count: 0x > > > > (0) > > > > sids : NULL > > > > level: > > > > LSA_LOOKUP_NAMES_UPLEVEL_TRUSTS_ONLY2 (6) > > > > count: * > > > > count: 0x (0) > > > > lookup_options : > > > > LSA_LOOKUP_OPTION_SEARCH_ISOLATED_NAMES (0) > > > > client_revision : > > > > LSA_CLIENT_REVISION_2 (2) > > > > > > > > ... > > > > > > > > diff --git a/auth/credentials/tests/test_creds.c > > > > b/auth/credentials/tests/test_creds.c > > > > index d2d3d30d73d..38550d6ecf9 100644 > > > > --- a/auth/credentials/tests/test_creds.c > > > > +++ b/auth/credentials/tests/test_creds.c > > > > @@ -187,7 +187,7 @@ static void torture_creds_parse_string(void > > > > **state) > > > > assert_string_equal(creds->domain, ""); > > > > assert_int_equal(creds->domain_obtained, CRED_SPECIFIED); > > > > > > > > - assert_string_equal(creds->username, "wurst@brot.realm"); > > > > + assert_string_equal(creds->username, "wurst"); > > > > > > I'm sorry but this is wrong! > > > I'm wondering why this wasn't covered by any high level test. > > > > > > This needs to result in domain="" and username="wurst@brot.realm" > > > and that's exactly what we need to use for NTLMSSP. > > > Also note that "brot.realm" may not be a realm and "wurst" may not > > > be a sAMAccountName. A userPrincipalName can be > > > anything@anydomain-of-msDS-SPNSuffixes. > > cli_credentials_get_ntlm_username_domain() does this already. > > > > I fear we need to revert these changes. > > > From the merge request ( > > > https://gitlab.com/samba-team/samba/-/merge_requests/1658) > > > I didn't really look at the whole patchset (with behavior change) > > > I only focused on CRED_NO_PASSWORD. > > > > > > I think we need to logic we have in wb_irpc_lsa_LookupNames4_call() > > > and/or parse_domain_user() here. One thing I need for lookup_name() is to be able to lookup those realm/NetBIOS-qualified requests along with UPN-based queries through passdb when they are done with LookupNames3 using LSA_LOOKUP_NAMES_UPLEVEL_TRUSTS_ONLY2 level. We don't see the level passed, as lookup_name() operates on own flags only. Without lookup_name() on DC (FreeIPA DC) being able to lookup UPNs against passdb, we don't even get the
Re: cli_credentials_parse_name... (Re: [SCM] Samba Shared Repository - branch master updated)
On Wed, 2020-11-04 at 19:23 +0200, Alexander Bokovoy wrote: > On ke, 04 marras 2020, Stefan Metzmacher wrote: > > Am 04.11.20 um 17:24 schrieb Alexander Bokovoy: > > > The branch, master has been updated > > >via f9016912098 lookup_name: allow lookup for own realm > > >via 00f4262ed0b cli_credentials: add a helper to parse > > > user or group names > > >via eb0474d27ba cli_credentials_parse_string: fix parsing > > > of principals > > > from a1b021200e3 selftest: add test for new "samba-tool > > > user unlock" command > > > > > > https://git.samba.org/?p=samba.git;a=shortlog;h=master > > > > > > > > > - Log - > > > > > > commit f901691209867b32c2d7c5c9274eee196f541654 > > > Author: Alexander Bokovoy > > > Date: Wed Nov 4 14:21:33 2020 +0200 > > > > > > lookup_name: allow lookup for own realm > > > > > > When using a security tab in Windows Explorer, a lookup over > > > a trusted > > > forest might come as realm\name instead of NetBIOS domain > > > name: > > > > > > - > > > --- > > > [2020/01/13 11:12:39.859134, 1, pid=33253, > > > effective(1732401004, 1732401004), real(1732401004, 0), > > > class=rpc_parse] > > > ../../librpc/ndr/ndr.c:471(ndr_print_function_debug) > > >lsa_LookupNames3: struct lsa_LookupNames3 > > > in: struct lsa_LookupNames3 > > > handle : * > > > handle: struct policy_handle > > > handle_type : 0x > > > (0) > > > uuid : 000e- > > > --1c5e-a750e581 > > > num_names: 0x0001 (1) > > > names: ARRAY(1) > > > names: struct lsa_String > > > length : 0x001e (30) > > > size : 0x0020 (32) > > > string : * > > > string : > > > 'ipa.test\admins' > > > sids : * > > > sids: struct lsa_TransSidArray3 > > > count: 0x > > > (0) > > > sids : NULL > > > level: > > > LSA_LOOKUP_NAMES_UPLEVEL_TRUSTS_ONLY2 (6) > > > count: * > > > count: 0x (0) > > > lookup_options : > > > LSA_LOOKUP_OPTION_SEARCH_ISOLATED_NAMES (0) > > > client_revision : > > > LSA_CLIENT_REVISION_2 (2) > > > > > > ... > > > > > > diff --git a/auth/credentials/tests/test_creds.c > > > b/auth/credentials/tests/test_creds.c > > > index d2d3d30d73d..38550d6ecf9 100644 > > > --- a/auth/credentials/tests/test_creds.c > > > +++ b/auth/credentials/tests/test_creds.c > > > @@ -187,7 +187,7 @@ static void torture_creds_parse_string(void > > > **state) > > > assert_string_equal(creds->domain, ""); > > > assert_int_equal(creds->domain_obtained, CRED_SPECIFIED); > > > > > > - assert_string_equal(creds->username, "wurst@brot.realm"); > > > + assert_string_equal(creds->username, "wurst"); > > > > I'm sorry but this is wrong! > > I'm wondering why this wasn't covered by any high level test. > > > > This needs to result in domain="" and username="wurst@brot.realm" > > and that's exactly what we need to use for NTLMSSP. > > Also note that "brot.realm" may not be a realm and "wurst" may not > > be a sAMAccountName. A userPrincipalName can be > > anything@anydomain-of-msDS-SPNSuffixes. cli_credentials_get_ntlm_username_domain() does this already. > > I fear we need to revert these changes. > > From the merge request ( > > https://gitlab.com/samba-team/samba/-/merge_requests/1658) > > I didn't really look at the whole patchset (with behavior change) > > I only focused on CRED_NO_PASSWORD. > > > > I think we need to logic we have in wb_irpc_lsa_LookupNames4_call() > > and/or parse_domain_user() here. > > I'm pushing a revert for now and will look at those. I'm not so sure this is totally wrong. Can I have a look over these paths at the office? I need any possible distraction from US election results anyway... Andrew Bartlett -- Andrew Bartlett https://samba.org/~abartlet/ Authentication Developer, Samba Team https://samba.org Samba Developer, Catalyst IT https://catalyst.net.nz/services/samba
Re: cli_credentials_parse_name... (Re: [SCM] Samba Shared Repository - branch master updated)
On ke, 04 marras 2020, Stefan Metzmacher wrote: > Am 04.11.20 um 17:24 schrieb Alexander Bokovoy: > > The branch, master has been updated > >via f9016912098 lookup_name: allow lookup for own realm > >via 00f4262ed0b cli_credentials: add a helper to parse user or > > group names > >via eb0474d27ba cli_credentials_parse_string: fix parsing of > > principals > > from a1b021200e3 selftest: add test for new "samba-tool user unlock" > > command > > > > https://git.samba.org/?p=samba.git;a=shortlog;h=master > > > > > > - Log - > > commit f901691209867b32c2d7c5c9274eee196f541654 > > Author: Alexander Bokovoy > > Date: Wed Nov 4 14:21:33 2020 +0200 > > > > lookup_name: allow lookup for own realm > > > > When using a security tab in Windows Explorer, a lookup over a trusted > > forest might come as realm\name instead of NetBIOS domain name: > > > > > > [2020/01/13 11:12:39.859134, 1, pid=33253, effective(1732401004, > > 1732401004), real(1732401004, 0), class=rpc_parse] > > ../../librpc/ndr/ndr.c:471(ndr_print_function_debug) > >lsa_LookupNames3: struct lsa_LookupNames3 > > in: struct lsa_LookupNames3 > > handle : * > > handle: struct policy_handle > > handle_type : 0x (0) > > uuid : > > 000e---1c5e-a750e581 > > num_names: 0x0001 (1) > > names: ARRAY(1) > > names: struct lsa_String > > length : 0x001e (30) > > size : 0x0020 (32) > > string : * > > string : 'ipa.test\admins' > > sids : * > > sids: struct lsa_TransSidArray3 > > count: 0x (0) > > sids : NULL > > level: > > LSA_LOOKUP_NAMES_UPLEVEL_TRUSTS_ONLY2 (6) > > count: * > > count: 0x (0) > > lookup_options : > > LSA_LOOKUP_OPTION_SEARCH_ISOLATED_NAMES (0) > > client_revision : LSA_CLIENT_REVISION_2 (2) > > > > ... > > > > diff --git a/auth/credentials/tests/test_creds.c > > b/auth/credentials/tests/test_creds.c > > index d2d3d30d73d..38550d6ecf9 100644 > > --- a/auth/credentials/tests/test_creds.c > > +++ b/auth/credentials/tests/test_creds.c > > @@ -187,7 +187,7 @@ static void torture_creds_parse_string(void **state) > > assert_string_equal(creds->domain, ""); > > assert_int_equal(creds->domain_obtained, CRED_SPECIFIED); > > > > - assert_string_equal(creds->username, "wurst@brot.realm"); > > + assert_string_equal(creds->username, "wurst"); > > I'm sorry but this is wrong! > I'm wondering why this wasn't covered by any high level test. > > This needs to result in domain="" and username="wurst@brot.realm" > and that's exactly what we need to use for NTLMSSP. > Also note that "brot.realm" may not be a realm and "wurst" may not > be a sAMAccountName. A userPrincipalName can be > anything@anydomain-of-msDS-SPNSuffixes. > > I fear we need to revert these changes. > From the merge request > (https://gitlab.com/samba-team/samba/-/merge_requests/1658) > I didn't really look at the whole patchset (with behavior change) > I only focused on CRED_NO_PASSWORD. > > I think we need to logic we have in wb_irpc_lsa_LookupNames4_call() and/or > parse_domain_user() here. I'm pushing a revert for now and will look at those. -- / Alexander Bokovoy
cli_credentials_parse_name... (Re: [SCM] Samba Shared Repository - branch master updated)
Am 04.11.20 um 17:24 schrieb Alexander Bokovoy: > The branch, master has been updated >via f9016912098 lookup_name: allow lookup for own realm >via 00f4262ed0b cli_credentials: add a helper to parse user or group > names >via eb0474d27ba cli_credentials_parse_string: fix parsing of > principals > from a1b021200e3 selftest: add test for new "samba-tool user unlock" > command > > https://git.samba.org/?p=samba.git;a=shortlog;h=master > > > - Log - > commit f901691209867b32c2d7c5c9274eee196f541654 > Author: Alexander Bokovoy > Date: Wed Nov 4 14:21:33 2020 +0200 > > lookup_name: allow lookup for own realm > > When using a security tab in Windows Explorer, a lookup over a trusted > forest might come as realm\name instead of NetBIOS domain name: > > > [2020/01/13 11:12:39.859134, 1, pid=33253, effective(1732401004, > 1732401004), real(1732401004, 0), class=rpc_parse] > ../../librpc/ndr/ndr.c:471(ndr_print_function_debug) >lsa_LookupNames3: struct lsa_LookupNames3 > in: struct lsa_LookupNames3 > handle : * > handle: struct policy_handle > handle_type : 0x (0) > uuid : > 000e---1c5e-a750e581 > num_names: 0x0001 (1) > names: ARRAY(1) > names: struct lsa_String > length : 0x001e (30) > size : 0x0020 (32) > string : * > string : 'ipa.test\admins' > sids : * > sids: struct lsa_TransSidArray3 > count: 0x (0) > sids : NULL > level: > LSA_LOOKUP_NAMES_UPLEVEL_TRUSTS_ONLY2 (6) > count: * > count: 0x (0) > lookup_options : > LSA_LOOKUP_OPTION_SEARCH_ISOLATED_NAMES (0) > client_revision : LSA_CLIENT_REVISION_2 (2) > > ... > > diff --git a/auth/credentials/tests/test_creds.c > b/auth/credentials/tests/test_creds.c > index d2d3d30d73d..38550d6ecf9 100644 > --- a/auth/credentials/tests/test_creds.c > +++ b/auth/credentials/tests/test_creds.c > @@ -187,7 +187,7 @@ static void torture_creds_parse_string(void **state) > assert_string_equal(creds->domain, ""); > assert_int_equal(creds->domain_obtained, CRED_SPECIFIED); > > - assert_string_equal(creds->username, "wurst@brot.realm"); > + assert_string_equal(creds->username, "wurst"); I'm sorry but this is wrong! I'm wondering why this wasn't covered by any high level test. This needs to result in domain="" and username="wurst@brot.realm" and that's exactly what we need to use for NTLMSSP. Also note that "brot.realm" may not be a realm and "wurst" may not be a sAMAccountName. A userPrincipalName can be anything@anydomain-of-msDS-SPNSuffixes. I fear we need to revert these changes. From the merge request (https://gitlab.com/samba-team/samba/-/merge_requests/1658) I didn't really look at the whole patchset (with behavior change) I only focused on CRED_NO_PASSWORD. I think we need to logic we have in wb_irpc_lsa_LookupNames4_call() and/or parse_domain_user() here. metze signature.asc Description: OpenPGP digital signature
Re: [SCM] Samba Shared Repository - branch master updated
Amitay You did not add "|| true" to the statd-callout invocation of nfsconf like you did in nfs-linux-kernel-callout. Regards, Jim On 7/27/2020 3:07 AM, Amitay Isaacs wrote: +# If NFS_HOSTNAME not set then try to pull it out of /etc/nfs.conf +if [ -z "$NFS_HOSTNAME" ] && type nfsconf >/dev/null 2>&1 ; then + NFS_HOSTNAME=$(nfsconf --get statd name) +fi
ctdb_pkt_recv_recv regression (Re: [SCM] Samba Shared Repository - branch master updated)
Hi Noel, can you please have a look at the mail I wrote a month ago, I think we need to fix that up. Thanks! metze Am 26.07.19 um 21:31 schrieb Stefan Metzmacher: > Hi Noel, > > thanks for fixing clang warnings, but > I fear the following changes are wrong. > > Can we have SMB_ASSERT(hdr != NULL); after the if statements with a > comment explaining why it's there. > >> diff --git a/source3/lib/ctdbd_conn.c b/source3/lib/ctdbd_conn.c >> index d5fac572d3c..8a8fbec4552 100644 >> --- a/source3/lib/ctdbd_conn.c >> +++ b/source3/lib/ctdbd_conn.c >> @@ -398,15 +398,16 @@ static int ctdb_read_packet(int fd, int timeout, >> TALLOC_CTX *mem_ctx, >> static int ctdb_read_req(struct ctdbd_connection *conn, uint32_t reqid, >> TALLOC_CTX *mem_ctx, struct ctdb_req_header **result) >> { >> -struct ctdb_req_header *hdr; >> +struct ctdb_req_header *hdr = NULL; >> int ret; >> >> next_pkt: >> >> ret = ctdb_read_packet(conn->fd, conn->timeout, mem_ctx, ); >> -if (ret != 0) { >> +if (hdr == NULL || ret != 0) { >> DBG_ERR("ctdb_read_packet failed: %s\n", strerror(ret)); >> cluster_fatal("failed to read data from ctdbd\n"); >> +return -1; >> } > > This is not strictly a bug, but it's makes the code really unclear. > >> DEBUG(11, ("Received ctdb packet\n")); >> @@ -1038,7 +1039,7 @@ int ctdbd_traverse(struct ctdbd_connection *conn, >> uint32_t db_id, >> int ret; >> TDB_DATA key, data; >> struct ctdb_traverse_start t; >> -int32_t cstatus; >> +int32_t cstatus = 0; >> >> if (ctdbd_conn_has_async_reqs(conn)) { >> /* >> @@ -1945,7 +1946,7 @@ static void ctdbd_parse_done(struct tevent_req *subreq) >> >> ret = ctdb_pkt_recv_recv(subreq, state, ); >> TALLOC_FREE(subreq); >> -if (tevent_req_error(req, ret)) { >> +if ((hdr == NULL) || tevent_req_error(req, ret)) { >> DBG_ERR("ctdb_pkt_recv_recv returned %s\n", strerror(ret)); >> return; >> } > > This is actually really a bug, in case ctdb_pkt_recv_recv fails > we'll never run tevent_req_error() and the callers callback function, > so it will just hang as hdr is always NULL if ret is not 0. > > Can you revert this and add SMB_ASSERT(). > > Thanks! > metze > signature.asc Description: OpenPGP digital signature
Re: [SCM] Samba Shared Repository - branch master updated
Hi Noel, thanks for fixing clang warnings, but I fear the following changes are wrong. Can we have SMB_ASSERT(hdr != NULL); after the if statements with a comment explaining why it's there. > diff --git a/source3/lib/ctdbd_conn.c b/source3/lib/ctdbd_conn.c > index d5fac572d3c..8a8fbec4552 100644 > --- a/source3/lib/ctdbd_conn.c > +++ b/source3/lib/ctdbd_conn.c > @@ -398,15 +398,16 @@ static int ctdb_read_packet(int fd, int timeout, > TALLOC_CTX *mem_ctx, > static int ctdb_read_req(struct ctdbd_connection *conn, uint32_t reqid, >TALLOC_CTX *mem_ctx, struct ctdb_req_header **result) > { > - struct ctdb_req_header *hdr; > + struct ctdb_req_header *hdr = NULL; > int ret; > > next_pkt: > > ret = ctdb_read_packet(conn->fd, conn->timeout, mem_ctx, ); > - if (ret != 0) { > + if (hdr == NULL || ret != 0) { > DBG_ERR("ctdb_read_packet failed: %s\n", strerror(ret)); > cluster_fatal("failed to read data from ctdbd\n"); > + return -1; > } This is not strictly a bug, but it's makes the code really unclear. > DEBUG(11, ("Received ctdb packet\n")); > @@ -1038,7 +1039,7 @@ int ctdbd_traverse(struct ctdbd_connection *conn, > uint32_t db_id, > int ret; > TDB_DATA key, data; > struct ctdb_traverse_start t; > - int32_t cstatus; > + int32_t cstatus = 0; > > if (ctdbd_conn_has_async_reqs(conn)) { > /* > @@ -1945,7 +1946,7 @@ static void ctdbd_parse_done(struct tevent_req *subreq) > > ret = ctdb_pkt_recv_recv(subreq, state, ); > TALLOC_FREE(subreq); > - if (tevent_req_error(req, ret)) { > + if ((hdr == NULL) || tevent_req_error(req, ret)) { > DBG_ERR("ctdb_pkt_recv_recv returned %s\n", strerror(ret)); > return; > } This is actually really a bug, in case ctdb_pkt_recv_recv fails we'll never run tevent_req_error() and the callers callback function, so it will just hang as hdr is always NULL if ret is not 0. Can you revert this and add SMB_ASSERT(). Thanks! metze signature.asc Description: OpenPGP digital signature
sambadowngradedatabase (Re: [SCM] Samba Shared Repository - branch master updated)
Hi Andrew, can you please prepare backports of the sambadowngradedatabase changes? These are needed in the old branches. Thanks! metze Am 29.05.19 um 07:58 schrieb Andrew Bartlett: > The branch, master has been updated >via 8831b06d3d4 ldb: Release ldb 2.0.3 >via 4a95410a154 downgradedatabase: blackbox: MDB backend >via 0942a65b63c downgradedatabase: adding special case for MDB >via 4eee09a2c17 dsdb: disable ORDERED_INTEGER with MDB pack format v1 >via 6b4abb99521 ldb: pack_format_override option >via 68d99187cd5 downgradedatabase: blackbox: database repacked >via 8db1312b08e ldb: python test for repack >via 73763acf49c ldb: repack old format database if GUID indexing > enabled >via d6ded22cb61 downgradedatabase: blackbox: check ordered integer > removed >via 08b9d204b6e ldb: binding ordered indexes to GUID indexing >via 6005c8cbad7 ldb: only used a->syntax->index_format_fn if GUID > indexing is enabled >via 74d15c9bf76 downgradedatabase: blackbox test >via 09f2a187b3d sambadowngradedatabase: Add "or later" to warning > about using tools from Samba 4.8 >via c0b679f6a3f sambaundoguididx: renamed to downgradedatabase >via 40ca8ed5a15 sambaundoguididx: fix for -s >via a3101b9704f ldb: Fix segfault parsing new pack formats >via 2de0aebed60 ldb: test for parse errors > from 73080f255e3 smbd: Adapt brl_pending_overlap to README.Coding > https://git.samba.org/?p=samba.git;a=shortlog;h=master > > > - Log - > commit 8831b06d3d4c1cb5b7732e9863228f8f3aea4e36 > Author: Aaron Haslett > Date: Fri May 24 15:34:43 2019 +1200 > > ldb: Release ldb 2.0.3 > > * Default pack format changed to version 2 for GUID Indexed databases > (essentially Samba's sam.ldb). This means faster unpacking for records > with lots of attributes. > > * Enforce link between all new database features, following toggling of > GUID indexing. If user toggles GUID indexing on/off, target pack format > is changed. Likewise the special ORDERED_INTEGER index format is used > (when GUID indexing on) or behaves like INTEGER (when GUID indexing > off). > > * Database repacked if packing format not as expected. > > * In the case of MDB, since GUID indexing is mandatory, feature toggling > provided by pack_format_override LDB option. > > * Check for errors from ldb_unpack_data() in ldb_tdb > > Signed-off-by: Aaron Haslett > Signed-off-by: Andrew Bartlett > Reviewed-by: Garming Sam > Pair-Programmed-With: Andrew Bartlett > > Autobuild-User(master): Andrew Bartlett > Autobuild-Date(master): Wed May 29 05:56:59 UTC 2019 on sn-devel-184 > > commit 4a95410a1543f854743aac24c6c7dc5ffc423aae > Author: Aaron Haslett > Date: Fri May 24 14:54:03 2019 +1200 > > downgradedatabase: blackbox: MDB backend > > When Samba uses TDB as a backend store, all new database features are > toggled on/off when GUID indexing is toggled. But, GUID indexing can't > be disabled on MDB, so the other features are toggled separately. > Consequently, the downgradedatabase script behaves differently depending > on the database backend. This patch adds testing for the MDB behaviour. > > Signed-off-by: Aaron Haslett > Signed-off-by: Andrew Bartlett > Reviewed-by: Garming Sam > Pair-Programmed-With: Andrew Bartlett > > commit 0942a65b63cc99f36d3eba99e9c9551e10c5782e > Author: Aaron Haslett > Date: Thu May 23 20:06:56 2019 +1200 > > downgradedatabase: adding special case for MDB > > Though this script was initially written for undoing GUID indexing on > TDB databases, we're repurposing it to do a full downgrade of any > database. MDB databases can't be DN indexed, but they can have pack > format version 2 and ORDERED_INTEGER data types, which must be removed > during a downgrade. > > Signed-off-by: Aaron Haslett > Signed-off-by: Andrew Bartlett > Reviewed-by: Garming Sam > Pair-Programmed-With: Andrew Bartlett > > commit 4eee09a2c17d1276b1d0be9f26b23743eec485c2 > Author: Aaron Haslett > Date: Thu May 23 20:13:05 2019 +1200 > > dsdb: disable ORDERED_INTEGER with MDB pack format v1 > > For TDB databases, the new ORDERED_INTEGER type is disabled along with > repacking at format version 1 if GUID indexing is disabled, so all the new > database features are toggled together. This scheme doesn't work with > MDB because GUID indexing is mandatory when using MDB. However, a > downgrade path is still required so in a previous commit we added > a pack_format_override option which allows a downgrade script to force > the database to use an earlier packing format. But, the new >
Re: [SCM] Samba Shared Repository - branch master updated
On Thu, 2019-03-07 at 05:59 +0100, Andrew Bartlett wrote: > The branch, master has been updated >via 8b18da27cf2 s4-server: Open and close a transaction on sam.ldb at > startup >via 6f2558cab81 dsdb: Provide better error strings in rootdse GUID > attribute handling >via 30f93bc7ac7 kcc: Give a better error message when > samdb_ntds_objectGUID fails >via 0c52a6bee78 dsdb: Unify > samdb_{get,set}_ntds_{objectGUID,invocation_id} My apologies, the last three commits were reviewed by Douglas here: https://gitlab.com/samba-team/samba/merge_requests/273#note_148039507 Sorry, Andrew Bartlett -- Andrew Bartlett http://samba.org/~abartlet/ Authentication Developer, Samba Team http://samba.org Samba Developer, Catalyst IT http://catalyst.net.nz/services/samba
Re: [SCM] Samba Shared Repository - branch master updated
Hi Metze, I wanted to rename the python module because the naming didn't seem consistent with any of the other python bindings Samba has. I think 'libsmb_samba_internal' made sense when it was unused by any of the samba python code (except some test code). But now (well, soon) these Python bindings will be used by the samba-tool code. It looks a bit strange in samba-tool to have: conn = libsmb_samba_internal.Conn(server) E.g. my latest patches start to use the new bindings in more places: https://gitlab.com/catalyst-samba/samba/commits/timb-pysmb-latest But maybe I've misunderstood something. Why do you want to keep the libsmb_samba_internal name? Cheers, Tim On 8/01/19 5:54 PM, Stefan Metzmacher wrote: > Hi Tim, > >> diff --git a/source3/wscript_build b/source3/wscript_build >> index a8ea8e581df..9d188a8d36a 100644 >> --- a/source3/wscript_build >> +++ b/source3/wscript_build >> @@ -1323,7 +1323,7 @@ for env in bld.gen_python_environments(): >> bld.SAMBA3_PYTHON('pylibsmb', >>source='libsmb/pylibsmb.c', >>deps='smbclient samba-credentials %s' % pycredentials, >> - realname='samba/samba3/libsmb_samba_internal.so' >> + realname='samba/samba3/libsmb.so' >>) > What's wrong with libsmb_samba_internal? I'd keep the name. > > metze >
Re: [SCM] Samba Shared Repository - branch master updated
Hi Tim, > diff --git a/source3/wscript_build b/source3/wscript_build > index a8ea8e581df..9d188a8d36a 100644 > --- a/source3/wscript_build > +++ b/source3/wscript_build > @@ -1323,7 +1323,7 @@ for env in bld.gen_python_environments(): > bld.SAMBA3_PYTHON('pylibsmb', >source='libsmb/pylibsmb.c', >deps='smbclient samba-credentials %s' % pycredentials, > - realname='samba/samba3/libsmb_samba_internal.so' > + realname='samba/samba3/libsmb.so' >) What's wrong with libsmb_samba_internal? I'd keep the name. metze signature.asc Description: OpenPGP digital signature
Re: [SCM] Samba Shared Repository - branch master updated
On pe, 07 syys 2018, Andrew Bartlett wrote: > The branch, master has been updated >via 4847daf wafsamba/samba_waf18: redefine flex function >via 364077d lib/replace/wscript: fix detection of a fallthrough > attribute for clang >via 311e1eb wafsamba/samba_autoconf: when setting undefined result, > use empty tuple >via 193fdbf s3/wscript: fix flex detection >via e761271 s3/wscript: fix bison detection > from 4a63ab9 s4/selftest/tests: Enabled samba.tests.samba_tool.computer > > https://git.samba.org/?p=samba.git;a=shortlog;h=master While Andrew pushed the current patch set, Metze pointed out that it didn't solve the issue with CONFIG_SET() still reporting undefined defines are set if they were stored in the cache as '0'. Attached patch should fix the issue. On FreeBSD 11.2 I now have: $ egrep '(HAVE_BSD_STRTOLL|INCOHERENT_MMAP|WORKING_STRPTIME)' bin/default/include/config.h bin/c4che/default_cache.py bin/default/include/config.h:#define HAVE_BSD_STRTOLL 1 bin/default/include/config.h:/* #undef HAVE_WORKING_STRPTIME */ bin/default/include/config.h:/* #undef HAVE_INCOHERENT_MMAP */ bin/c4che/default_cache.py:HAVE_BSD_STRTOLL = 1 bin/c4che/default_cache.py:HAVE_INCOHERENT_MMAP = () bin/c4che/default_cache.py:HAVE_WORKING_STRPTIME = () -- / Alexander Bokovoy >From 92f2b9b2bac513412169de959285c35a40ea9d07 Mon Sep 17 00:00:00 2001 From: Alexander Bokovoy Date: Fri, 7 Sep 2018 14:34:48 + Subject: [PATCH] buildtools/wafsamba: undefine a define for a failed test unless asked to keep it If conf.CHECK_CODE() is called without `always=True` and the test has failed, undefine the define already set to '0' by conf.check_code(). This restores expectations that undefined symbols are not considered to be set by CONFIG_SET() method. Signed-off-by: Alexander Bokovoy --- buildtools/wafsamba/samba_autoconf.py | 2 ++ 1 file changed, 2 insertions(+) diff --git a/buildtools/wafsamba/samba_autoconf.py b/buildtools/wafsamba/samba_autoconf.py index 6b940e53c00..ced824d593d 100644 --- a/buildtools/wafsamba/samba_autoconf.py +++ b/buildtools/wafsamba/samba_autoconf.py @@ -447,6 +447,8 @@ def CHECK_CODE(conf, code, define, except Exception: if always: conf.DEFINE(define, 0) +else: +conf.undefine(define) conf.COMPOUND_END(False) if mandatory: raise -- 2.18.0
Re: [SCM] Samba Shared Repository - branch master updated
On ke, 05 syys 2018, Amitay Isaacs wrote: > > Thank you, Andrew, for the reviews and persistent push to complete this! > > > > We are using WAF 2.x now which should allow us to concentrate on > > switching to Python 3 for the build process as well for all our > > libraries and components. > > > > If you see any build issues related to the WAF, please report them. > > Autobuild succeeded but we might still have pieces of Python code that > > aren't excercised through the autobuild targets. > > Great job updating to the latest waf. > > Now fixing all the corner cases... ;-) :) > > On freebsd-11, waf does not want to use clang and complains that it > cannot find a C compiler. Thanks, I'll do a FreeBSD v11 VM later this week if nobody else gets it fixed earlier. The python exception you see is due to this one: def CHECK_STANDARD_LIBPATH(conf): # at least gcc and clang support this: try: cmd = conf.env.CC + ['-print-search-dirs'] out = Utils.cmd_output(cmd).split('\n') except ValueError: # option not supported by compiler - use a standard list of directories dirlist = [ '/usr/lib', '/usr/lib64' ] except: raise Errors.WafError('Unexpected error running "%s"' % (cmd)) . we don't even have conf.env.CC set and get an exception while defining 'cmd' value, thus it is not defined when the latter except: statement runs. I'll look at it. -- / Alexander Bokovoy
Re: [SCM] Samba Shared Repository - branch master updated
On Wed, Sep 5, 2018 at 6:52 PM, Amitay Isaacs wrote: > Hi Alexander, > > On Wed, Sep 5, 2018 at 6:41 PM, Alexander Bokovoy wrote: >> On ke, 05 syys 2018, Andrew Bartlett wrote: >>> The branch, master has been updated >>> >>> https://git.samba.org/?p=samba.git;a=shortlog;h=master >> Thank you, Andrew, for the reviews and persistent push to complete this! >> >> We are using WAF 2.x now which should allow us to concentrate on >> switching to Python 3 for the build process as well for all our >> libraries and components. >> >> If you see any build issues related to the WAF, please report them. >> Autobuild succeeded but we might still have pieces of Python code that >> aren't excercised through the autobuild targets. > > Great job updating to the latest waf. > > Now fixing all the corner cases... ;-) > Here's another one This time on AIX using xlc... $ ./configure --enable-developer Setting top to : /home/jenkins/workspace/samba-master/label/ppc64-aix7 Setting out to : /home/jenkins/workspace/samba-master/label/ppc64-aix7/ctdb/bin Checking for 'xlc' (C compiler) : Traceback (most recent call last): File "/home/jenkins/workspace/samba-master/label/ppc64-aix7/third_party/waf/waflib/Scripting.py", line 158, in waf_entry_point run_commands() File "/home/jenkins/workspace/samba-master/label/ppc64-aix7/third_party/waf/waflib/Scripting.py", line 251, in run_commands ctx = run_command(cmd_name) File "/home/jenkins/workspace/samba-master/label/ppc64-aix7/third_party/waf/waflib/Scripting.py", line 235, in run_command ctx.execute() File "/home/jenkins/workspace/samba-master/label/ppc64-aix7/third_party/waf/waflib/Configure.py", line 159, in execute super(ConfigurationContext, self).execute() File "/home/jenkins/workspace/samba-master/label/ppc64-aix7/third_party/waf/waflib/Context.py", line 204, in execute self.recurse([os.path.dirname(g_module.root_path)]) File "/home/jenkins/workspace/samba-master/label/ppc64-aix7/third_party/waf/waflib/Context.py", line 286, in recurse user_function(self) File "/home/jenkins/workspace/samba-master/label/ppc64-aix7/ctdb/wscript", line 118, in configure conf.RECURSE('lib/replace') File "./../buildtools/wafsamba/samba_utils.py", line 30, in fun return f(*k, **kw) File "./../buildtools/wafsamba/samba_utils.py", line 433, in RECURSE return ctx.recurse(relpath) File "/home/jenkins/workspace/samba-master/label/ppc64-aix7/third_party/waf/waflib/Context.py", line 286, in recurse user_function(self) File "/home/jenkins/workspace/samba-master/label/ppc64-aix7/third_party/waf/waflib/Utils.py", line 816, in wrap ret = fun(*k) File "/home/jenkins/workspace/samba-master/label/ppc64-aix7/lib/replace/wscript", line 30, in configure conf.RECURSE('buildtools/wafsamba') File "./../buildtools/wafsamba/samba_utils.py", line 30, in fun return f(*k, **kw) File "./../buildtools/wafsamba/samba_utils.py", line 433, in RECURSE return ctx.recurse(relpath) File "/home/jenkins/workspace/samba-master/label/ppc64-aix7/third_party/waf/waflib/Context.py", line 286, in recurse user_function(self) File "/home/jenkins/workspace/samba-master/label/ppc64-aix7/third_party/waf/waflib/Utils.py", line 816, in wrap ret = fun(*k) File "/home/jenkins/workspace/samba-master/label/ppc64-aix7/buildtools/wafsamba/wscript", line 234, in configure conf.load('compiler_cc') File "/home/jenkins/workspace/samba-master/label/ppc64-aix7/third_party/waf/waflib/Configure.py", line 270, in load func(self) File "/home/jenkins/workspace/samba-master/label/ppc64-aix7/third_party/waf/waflib/Tools/compiler_c.py", line 79, in configure conf.load(compiler) File "/home/jenkins/workspace/samba-master/label/ppc64-aix7/third_party/waf/waflib/Configure.py", line 270, in load func(self) File "/home/jenkins/workspace/samba-master/label/ppc64-aix7/third_party/waf/waflib/Tools/xlc.py", line 59, in configure conf.find_xlc() File "./../buildtools/wafsamba/samba_utils.py", line 30, in fun return f(*k, **kw) File "/home/jenkins/workspace/samba-master/label/ppc64-aix7/third_party/waf/waflib/Tools/xlc.py", line 17, in find_xlc conf.get_xlc_version(cc) File "/home/jenkins/workspace/samba-master/label/ppc64-aix7/third_party/waf/waflib/Configure.py", line 316, in fun return f(*k, **kw) File "/home/jenkins/workspace/samba-master/label/ppc64-aix7/third_party/waf/waflib/Tools/c_config.py", line 1097, in get_xlc_version cmd = cc + ['-qversion'] TypeError: unsupported operand type(s) for +: 'NoneType' and 'list' I don't have time to dig deeper this week. Will spend some time next week if this issue is not resolved by then. Amitay.
Re: [SCM] Samba Shared Repository - branch master updated
Hi Alexander, On Wed, Sep 5, 2018 at 6:41 PM, Alexander Bokovoy wrote: > On ke, 05 syys 2018, Andrew Bartlett wrote: >> The branch, master has been updated >>via 8de348e third_party: Import exact files from waf-2.0.8/waflib >>via 8f022a0 script/autobuild: Fix formatting in send_email >>via 67ed1ea script/autobuild: re-use CACHE_SUFFIX from waflib >>via 72c3ff9 buildtools/wafsamba: use CACHE_SUFFIX instead of a >> hard-coded name >>via 6fc9f1a lib/mscat: fix logging in wscript >>via 4c7c10b lib/audit_logging: update to waf 2.0 >>via f495f64 buildtools/wafsamba: remove ENFORCE_GROUP_ORDERING >>via 0a9d98b ctdb/wscript: rework how version number is retrieved >>via fdd89fe selftest/tests.py: update to support waf 2.0 >>via 72a9e6d wscript: port build_system_heimdal to waf 2.0 >>via 1efe689 wscript: port build_system_mitkrb5 to waf 2.0 >>via 7944ed6 buildtools/wafsamba: port stale_files to waf 2.0 >>via 2a3fcdf buildtools/wafsamba: generate build options output with >> waf 2.0 >>via 75d5bcb selftest/wscript: properly handle env.cwd which is a >> list, not a string >>via 31f8945 wafsamba: install Python modules back to bin/python, not >> bin/python_modules >>via 77f3a13 buildtools/wafsamba: use cflags instead of ccflags for >> waf 2.0 >>via 1d25ae5 buildtools/wafsamba: use context instead of options for >> cross-compile checks for waf 2.0 >>via aebcd69 buildtools/wafsamba: crosscompile should use >> Utils.subprocess in waf 2.0 >>via fb43723 buildtools/wafsamba: use top and out for waf 2.0 >>via 5c3d31e cdtb/wscript: use top and out for waf 2.0 >>via 699977a wscript: adopt to waf 2.0 >>via 0fdba4b buildtools/wafsamba: use top for waf 2.0 >>via cdda0d9 buildtools/wafsamba: change SAMBA_BUILD_ENV to use >> bldnode.abspath() >>via 35ebfd3 selftest/wscript: handle lists in environmental >> variables in waf >>via 2a63619 waf heimdal: use absolute path to compile_et >>via 1fdcbd0 wafsamba: use correct context for APPNAME >>via 850ceec buildtools/wafsamba: compile asn1 files by adding >> missing code from compat15 >>via 2e401d2 wscript: adopt to waf-2.0 >>via 175be93 ctdb/wscript: adopt to waf-2.0 >>via 7aaa1e6 buildtools/wafsamba/samba_abi: always_run helper was >> deprecated in waf 2.0 >>via 9f2f5b4 source3/libsmb/wscript: remove unneeded import >>via d00ba40 nsswitch/libwbclient/wscript: import from waflib >>via 944fce2 auth/wscript: import from waflib >>via 3fc4786 buildtools/wafsamba: add install_dir to build context >>via 69c655f buildtools/wafsamba: reduce imports >>via 7eab91a buildtools/wafsamba: adopt to waf 2.0.8 >>via 3b7dfc5 third_party/waf: upgrade to waf 2.0.8 >>via 79c4ba2 auth/wscript: fix options use >>via c5ab9ea heimdal wscript changes >>via ef4b2d4 wscript_configure_system_mitkrb5: update to handle waf >> 2.0.4 >>via e58ca30 wscript_build_embedded_heimdal: update to handle waf >> 2.0.4 >>via 1dc80eb wscript_build: update to handle waf 2.0.4 >>via c73b779 wscript: update to handle waf 2.0.4 >>via 0119a87 third_party/wscript: update to handle waf 2.0.4 >>via f14b8cb third_party/uid_wrapper/wscript: update to handle waf >> 2.0.4 >>via 051e7e7 third_party/socket_wrapper/wscript: update to handle waf >> 2.0.4 >>via 4c44153 third_party/resolv_wrapper/wscript: update to handle waf >> 2.0.4 >>via 7e2deed third_party/popt/wscript: update to handle waf 2.0.4 >>via 2847a38 third_party/pam_wrapper/wscript: update to handle waf >> 2.0.4 >>via 2b21e67 third_party/nss_wrapper/wscript: update to handle waf >> 2.0.4 >>via 3e71f06 third_party/cmocka/wscript: update to handle waf 2.0.4 >>via ba03d12 third_party/aesni-intel/wscript: update to handle waf >> 2.0.4 >>via ca5008a testsuite/headers/wscript_build: update to handle waf >> 2.0.4 >>via 0c423a3 source4/lib/tls/wscript: update to handle waf 2.0.4 >>via c797e92 source4/heimdal_build/wscript_configure: update to >> handle waf 2.0.4 >>via 055aae9 source4/heimdal_build/wscript_build: update to handle >> waf 2.0.4 >>via 029ac7d source4/dsdb/samdb/ldb_modules/wscript: update to handle >> waf 2.0.4 >>via 8b5ad02 source3/wscript_configure_system_ncurses: update to >> handle waf 2.0.4 >>via 91e099d source3/wscript: update to handle waf 2.0.4 >>via c898f18 source3/build/charset.py: update to handle waf 2.0.4 >>via 0de67cf selftest/wscript: update to handle waf 2.0.4 >>via fa1ca71 python/wscript: update to handle waf 2.0.4 >>via 7f6ce8f pidl/wscript: update to handle waf 2.0.4 >>via
Re: [SCM] Samba Shared Repository - branch master updated
On ke, 05 syys 2018, Andrew Bartlett wrote: > The branch, master has been updated >via 8de348e third_party: Import exact files from waf-2.0.8/waflib >via 8f022a0 script/autobuild: Fix formatting in send_email >via 67ed1ea script/autobuild: re-use CACHE_SUFFIX from waflib >via 72c3ff9 buildtools/wafsamba: use CACHE_SUFFIX instead of a > hard-coded name >via 6fc9f1a lib/mscat: fix logging in wscript >via 4c7c10b lib/audit_logging: update to waf 2.0 >via f495f64 buildtools/wafsamba: remove ENFORCE_GROUP_ORDERING >via 0a9d98b ctdb/wscript: rework how version number is retrieved >via fdd89fe selftest/tests.py: update to support waf 2.0 >via 72a9e6d wscript: port build_system_heimdal to waf 2.0 >via 1efe689 wscript: port build_system_mitkrb5 to waf 2.0 >via 7944ed6 buildtools/wafsamba: port stale_files to waf 2.0 >via 2a3fcdf buildtools/wafsamba: generate build options output with > waf 2.0 >via 75d5bcb selftest/wscript: properly handle env.cwd which is a > list, not a string >via 31f8945 wafsamba: install Python modules back to bin/python, not > bin/python_modules >via 77f3a13 buildtools/wafsamba: use cflags instead of ccflags for > waf 2.0 >via 1d25ae5 buildtools/wafsamba: use context instead of options for > cross-compile checks for waf 2.0 >via aebcd69 buildtools/wafsamba: crosscompile should use > Utils.subprocess in waf 2.0 >via fb43723 buildtools/wafsamba: use top and out for waf 2.0 >via 5c3d31e cdtb/wscript: use top and out for waf 2.0 >via 699977a wscript: adopt to waf 2.0 >via 0fdba4b buildtools/wafsamba: use top for waf 2.0 >via cdda0d9 buildtools/wafsamba: change SAMBA_BUILD_ENV to use > bldnode.abspath() >via 35ebfd3 selftest/wscript: handle lists in environmental variables > in waf >via 2a63619 waf heimdal: use absolute path to compile_et >via 1fdcbd0 wafsamba: use correct context for APPNAME >via 850ceec buildtools/wafsamba: compile asn1 files by adding missing > code from compat15 >via 2e401d2 wscript: adopt to waf-2.0 >via 175be93 ctdb/wscript: adopt to waf-2.0 >via 7aaa1e6 buildtools/wafsamba/samba_abi: always_run helper was > deprecated in waf 2.0 >via 9f2f5b4 source3/libsmb/wscript: remove unneeded import >via d00ba40 nsswitch/libwbclient/wscript: import from waflib >via 944fce2 auth/wscript: import from waflib >via 3fc4786 buildtools/wafsamba: add install_dir to build context >via 69c655f buildtools/wafsamba: reduce imports >via 7eab91a buildtools/wafsamba: adopt to waf 2.0.8 >via 3b7dfc5 third_party/waf: upgrade to waf 2.0.8 >via 79c4ba2 auth/wscript: fix options use >via c5ab9ea heimdal wscript changes >via ef4b2d4 wscript_configure_system_mitkrb5: update to handle waf > 2.0.4 >via e58ca30 wscript_build_embedded_heimdal: update to handle waf 2.0.4 >via 1dc80eb wscript_build: update to handle waf 2.0.4 >via c73b779 wscript: update to handle waf 2.0.4 >via 0119a87 third_party/wscript: update to handle waf 2.0.4 >via f14b8cb third_party/uid_wrapper/wscript: update to handle waf > 2.0.4 >via 051e7e7 third_party/socket_wrapper/wscript: update to handle waf > 2.0.4 >via 4c44153 third_party/resolv_wrapper/wscript: update to handle waf > 2.0.4 >via 7e2deed third_party/popt/wscript: update to handle waf 2.0.4 >via 2847a38 third_party/pam_wrapper/wscript: update to handle waf > 2.0.4 >via 2b21e67 third_party/nss_wrapper/wscript: update to handle waf > 2.0.4 >via 3e71f06 third_party/cmocka/wscript: update to handle waf 2.0.4 >via ba03d12 third_party/aesni-intel/wscript: update to handle waf > 2.0.4 >via ca5008a testsuite/headers/wscript_build: update to handle waf > 2.0.4 >via 0c423a3 source4/lib/tls/wscript: update to handle waf 2.0.4 >via c797e92 source4/heimdal_build/wscript_configure: update to handle > waf 2.0.4 >via 055aae9 source4/heimdal_build/wscript_build: update to handle waf > 2.0.4 >via 029ac7d source4/dsdb/samdb/ldb_modules/wscript: update to handle > waf 2.0.4 >via 8b5ad02 source3/wscript_configure_system_ncurses: update to > handle waf 2.0.4 >via 91e099d source3/wscript: update to handle waf 2.0.4 >via c898f18 source3/build/charset.py: update to handle waf 2.0.4 >via 0de67cf selftest/wscript: update to handle waf 2.0.4 >via fa1ca71 python/wscript: update to handle waf 2.0.4 >via 7f6ce8f pidl/wscript: update to handle waf 2.0.4 >via e1e5b57 packaging/wscript: update to handle waf 2.0.4 >via 8a74116 nsswitch/wscript_build: update to handle waf 2.0.4 >via e676183
Re: [SCM] Samba Shared Repository - branch master updated
On Sat, 2018-02-24 at 10:56 -0500, jim wrote: > The wrong version of this patch was pushed. > The argument to simpleclass_IClassFactory_CreateInstance must be > ...**ppv and the value assignment must be *ppv = (...)ret; > This was discussed on the list and the corrected patch was created. Sorry about that. I'm sure we can fix that up. Thankfully the code has no callers, the module is part of a subsystem that is disabled as not even compiling. DCOM is an area we explored but never got anywhere, see the thread on wmic on the samba list a few days ago. Andrew Bartlett -- Andrew Bartlett http://samba.org/~abartlet/ Authentication Developer, Samba Team http://samba.org Samba Developer, Catalyst IT http://catalyst.net.nz/services/samba
Re: [SCM] Samba Shared Repository - branch master updated
The wrong version of this patch was pushed. The argument to simpleclass_IClassFactory_CreateInstance must be ...**ppv and the value assignment must be *ppv = (...)ret; This was discussed on the list and the corrected patch was created. On 2/24/2018 9:51 AM, Andrew Bartlett wrote: -static WERROR simpleclass_IClassFactory_CreateInstance (struct IClassFactory *d, TALLOC_CTX *mem_ctx, struct IUnknown *iunk, struct GUID *iid, struct IUnknown **ppv) +static WERROR simpleclass_IClassFactory_CreateInstance(struct IClassFactory *d, + TALLOC_CTX *mem_ctx, + struct MInterfacePointer *pUnknown, + struct GUID *iid, + struct MInterfacePointer *ppv) { struct IStream *ret; /* FIXME: Check whether IID == ISTREAM_IID */ @@ -71,8 +84,8 @@ static WERROR simpleclass_IClassFactory_CreateInstance (struct IClassFactory *d, ret->vtable = _IStream_vtable; ret->object_data = NULL; - *ppv = (struct IUnknown *)ret; - + ppv = (struct MInterfacePointer *)ret; + return WERR_OK; }
Re: [SCM] Samba Shared Repository - branch master updated
On ti, 02 touko 2017, Andreas Schneider wrote: > The branch, master has been updated >via 09bc5b5 systemd: fix detection of libsystemd > from 2951c59 torture3: Make sure dbwrap_parse_record returns NOT_FOUND > for invalid watchers data > > https://git.samba.org/?p=samba.git;a=shortlog;h=master > > > - Log - > commit 09bc5b5374227a555f580c3d7c1d82f15bb818bc > Author: Alexander Bokovoy> Date: Thu Sep 29 00:09:12 2016 +0300 > > systemd: fix detection of libsystemd > > On Fedora 25 detection of libsystemd actually fails due to wrong > assumptions in the configure test. conf.CHECK_LIB returns a list > so 'not conf.CHECK_LIB(...)' is always False and we never get to check > libsystemd. > > Instead, remember result of checking pkg-config for separate > libsystemd-daemon and libsystemd-journal libraries. If they miss, > attempt to use libsystemd library instead. > > Signed-off-by: Alexander Bokovoy > Reviewed-by: Andreas Schneider > > Autobuild-User(master): Andreas Schneider > Autobuild-Date(master): Tue May 2 13:05:43 CEST 2017 on sn-devel-144 > > --- I created https://bugzilla.samba.org/show_bug.cgi?id=12764 for 4.6 backport. -- / Alexander Bokovoy
Re: [SCM] Samba Shared Repository - branch master updated
On pe, 10 maalis 2017, Andrew Bartlett wrote: > The branch, master has been updated >via 7142a88 waf: disable-python - don't include python.h in > test_headers.c >via 4455547 autobuild: Add nopython environment to test > --disable-python builds (but without tests) >via 7000fa0 waf: disable-python - don't build torture bits >via dbf97e8 waf: disable-python - don't build samba-policy >via fa82998 waf: disable-python - don't build samba-net >via 7eeda99 waf: disable-python - don't build pyrpc_util, dcerpc.py >via 696d2f7 waf: disable-python - don't build PROVISION, pyparam_util >via 676e80b waf: disable-python - don't build python/ >via 81ff0f3 waf: disable-python - align tdb's wscript >via b45ff0b waf: disable-python - align tevent wscript >via ce956be waf: disable-python - align ldb's wscript >via dcba0b1 waf: disable-python - align talloc's wscript >via 7d5db90 waf: disable-python - configuration adjustments >via 5bbcd09 waf: disable-python - add option globally to build system >via 9a9e342 waf: disable-python - fix ctdb configuration >via 8ba2fe5 python: Port the samba.net module to Python 3 >via 0d83cec python: Remove unused import PY3 >via 4ebd877 python: selftest: Add possibility to run old Python test > suites with Python 3 >via 64bc64c python: samba.gensec: Port module to Python 3 compatible > form >via 0672fc1 python: samba.gensec: Fix error handling in > set_credentials() function >via d0914e2 python: selftests: Enable samba.getopt tests execution > with Python 3 >via 9ada914 python: samba.getopt: Port module to Python 3 compatible > form >via da71c39 python: samba.tests.core: Port and enable core tests in > Python 3 >via afe1e83 python: samba.tests: Move import of ported modules out of > PY3 condition >via e9a4649 python: samba._ldb: Port of samba._ldb to Python 3 > compatible form >via 9d8bcea python: samba.tests.auth: Add tests for samba.auth module >via 1ac5bf2 python: samba.auth: Port samba.auth to Python 3 > compatible form >via 47b0148 python: wscript_build: Build some DCE/RPC modules with > Python 3 >via a40d7f1 python: samba.dcerpc: Port security module to Python 3 > comp. form >via e481aed dcerpc/misc tests: asset GUID ordering in python 2 and 3 >via 9843cce python: samba.tests.dcerpc.misc: Port and enable tests >via 3b1c0ba python: samba.dcerpc: Port RPC related stuff to Python 3 >via 2de5fa4 python: pidl: Port Python interface generator >via 5edb0b4 python: samba.tests: Enable Python 3 tests for ported > modules >via b7a5308 buildtools: Work around a . being in the target name when > building python3 helpers >via fe8bba5 python: wscript_build: Build some modules for Python 3 >via 6fa125e python: Make top-level samba modules Python 3 compatible >via 211df4a python: samba.tests.dcerpc: Move Class RawDCERPCTest to > separated file. >via 9c55bb9 python: samba.tests.glue: Add new tests for samba._glue. >via b454b09 python: samba._glue: Port samba._glue module to Python 3. >via 5123f15 python: samba.tests.param: Add missing tests >via 31cef92 python: samba.param: Port param module to Python 3 >via 035e6dc python: samba.tests.credentials: Python 3 compatible tests >via 1dab2b4 python: samba.credentials: Port pycredentials.c to > Python3-compatible form. >via 583ff0a lib/ldb: Enable use of a python3 pyldb-util system library >via f5cafee talloc: use the system pytalloc-util for python3 as well >via 132cbad scripts/traffic_summary: documentation typo >via 4485ba5 ./examples/scripts/SambaConfig.py: fix typo in "continue" >via d424c7d python/examples/winreg: two variable name typos on a > single line >via b7cd0f2 python sites/subnets: correctly spell variable name >via cedb991 python provision: FDSBackend takes forced uri >via b9c5614 python/remove_dc: avoid using non-existent variable >via 732233f samba-tool domain: correctly spell variable name >via 495383c python/join: correct spelling of "ctx.del_noerror" >via 0779d66 selftest: remove unused broken client.py >via 765275c gitignore: add some hidden files >via 17dce5d perftests/ad_dc_search: do less work in expensive member > searches >via bb4ad8d pyldb: p3k readiness: allow single unicode string in msg > element > from 2772fa4 talloc: fix doxygen of talloc_move > Thanks to everyone for this gigantic work on making Python use in Samba modern and more supportable going forward. -- / Alexander Bokovoy
Re: [SCM] Samba Shared Repository - branch master updated
On Fri, Dec 02, 2016 at 12:24:03AM +0100, Jeremy Allison wrote: > The branch, master has been updated >via 943c6ee s3-libads: Fix canonicalization support with MIT Kerberos >via 7ca7680 s3: libsmb: Setting the LIBSMBCLIENT_NO_CCACHE > environment variable doesn't turn off credential cache use. > from 8858657 replace: Include libgen.h if available > > https://git.samba.org/?p=samba.git;a=shortlog;h=master > > > - Log - > commit 943c6ee030a07f87545dcf72e3d1a5bc78eac701 > Author: Andreas Schneider> Date: Mon Feb 15 12:09:59 2016 +0100 > > s3-libads: Fix canonicalization support with MIT Kerberos > > This allows to authenticate using user@DOMAIN against an AD DC. > > https://bugzilla.samba.org/show_bug.cgi?id=12457 > > Signed-off-by: Andreas Schneider > Reviewed-by: Jeremy Allison > > Autobuild-User(master): Jeremy Allison > Autobuild-Date(master): Fri Dec 2 00:23:02 CET 2016 on sn-devel-144 > > commit 7ca7680c94f457dea8698124e2780efa55865ed3 > Author: Jeremy Allison > Date: Wed Nov 30 10:25:53 2016 -0800 > > s3: libsmb: Setting the LIBSMBCLIENT_NO_CCACHE environment variable > doesn't turn off credential cache use. > > Signed-off-by: Jeremy Allison Sorry, this was Reviewed-by: Ira but I forgot to add it in.
Re: [SCM] Samba Shared Repository - branch master updated
Hi Andrew, - Log - commit c8c2c850d47a6cdaee5afd84c8ebaa03586c1277 Author: Andrew Bartlett abart...@samba.org Date: Tue Feb 10 12:26:56 2015 +1300 Update mailing list references to point at lists.samba.org The mailing lists are on lists.samba.org, but there are many references that use the shorthand of samba.org Some references to samba@ have been changed to samba-technical@ where this make more sense. Signed-off-by: Andrew Bartlett abart...@samba.org Reviewed-by: Garming Sam garm...@catalyst.net.nz Autobuild-User(master): Andrew Bartlett abart...@samba.org Autobuild-Date(master): Tue Feb 10 07:08:28 CET 2015 on sn-devel-104 Can you prepare backports of this for our current release branches? v4-{0,1,2}-test? Thanks! metze signature.asc Description: OpenPGP digital signature
Re: [SCM] Samba Shared Repository - branch master updated
On Wed, Feb 19, 2014 at 5:13 PM, Jeremy Allison j...@samba.org wrote: On Wed, Feb 19, 2014 at 08:20:04PM +0100, Andreas Schneider wrote: The branch, master has been updated Andreas and David, via b8844fc clitar.c: check all allocations return value I *HATE* *HATE* *HATE* the idiom used here. Calling smb_panic on an allocation fail in the new tar code is manifestly the *WRONG* thing to do. +/* helper macro to die in case of NULL pointer */ +#define PANIC_IF_NULL(x) \ +_panic_if_null(x, __FILE__ : STR2(__LINE__) ( #x ) == NULL\n) + +/* prototype to silent gcc warning */ +static inline void* _panic_if_null(void *p, const char *expr); +static inline void* _panic_if_null(void *p, const char *expr) +{ +if (!p) { +smb_panic(expr); +} +return p; +} + Then later.. +TALLOC_CTX *ctx = PANIC_IF_NULL(talloc_new(NULL)); and: +fname = PANIC_IF_NULL(talloc_asprintf(ctx, + %s%s, + client_get_cur_dir(), + buf)); if (!fname) { err = 1; goto out; The above even left the (correct) handling of the NULL return alone ! Remember on common setups this will call a panic action script which will leave the client hanging on a sleep 999 call and leave the caller with a hung command and no indication of what failed. Can you fix this please to do proper error checking of a NULL return and to bail out of the tar command with a correct error indication that the command failed ? This really needs fixing properly before this code can go anywhere near a production branch. Personally, I would like to lobby for an extra parameter to smb_panic, called dump_core or something. Here is a situation where we don't need to dump core: source3/smbd/server.c#6 (text) @@ -640,6 +640,18 @@ because too many files are open\n)); goto exit; } + + /* +* If we get NT_STATUS_OPEN_FAILED, there was a problem with +* re-initing the TDBs, so complain, but do not dump core +* as this can exacerbate the problems. +*/ + if (NT_STATUS_EQUAL(status, NT_STATUS_OPEN_FAILED)) { + DEBUG(0, (FATAL: child process cannot initialize + because one or more TDBs could not be reopened\n)); + goto exit; + } + if (lp_clustering() NT_STATUS_EQUAL(status, NT_STATUS_INTERNAL_DB_ERROR)) { This happens after reinit_after_fork ... When it happens because tdb_reopen_all failed, we should probably not dump core, especially as that only exacerbates existing problems and the core file tells us absolutely nothing new (the error in tdb_reopen_all is long gone.) -- Regards, Richard Sharpe (何以解憂?唯有杜康。--曹操)
Re: [SCM] Samba Shared Repository - branch master updated
On Thu, Jun 12, 2014 at 04:02:04PM +0200, David Disseldorp wrote: The branch, master has been updated via 0057388 leases: Fix uint32-hyper via 2f15208 Remove .clang_complete. from 0b4af49 waf: fix the name of the WINBIND nss module on AIX http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master - Log - commit 00573880e5f0362c3c442e82501bb0efe731ce7b Author: Volker Lendecke v...@samba.org Date: Thu Jun 12 08:53:36 2014 + leases: Fix uint32-hyper Thanks to Noel Power noel.po...@suse.com for finding this! Signed-off-by: Volker Lendecke v...@samba.org Reviewed-by: Noel Power noel.po...@suse.com [dd...@samba.org: changed Noel's address, as requested] Reviewed-by: David Disseldorp dd...@samba.org Well, if I take a look at https://lists.samba.org/archive/samba-technical/2014-June/100408.html I see a Noel Power nopower at suse.com which to me is a pretty strong indication that Noel's address is indeed nopo...@suse.com. Where is the problem with that? Thanks, Volker -- SerNet GmbH, Bahnhofsallee 1b, 37081 Göttingen phone: +49-551-37-0, fax: +49-551-37-9 AG Göttingen, HRB 2816, GF: Dr. Johannes Loxen http://www.sernet.de, mailto:kont...@sernet.de
Re: [SCM] Samba Shared Repository - branch master updated
Hi Andreas, On Thu, 5 Jun 2014 13:19:04 +0200 (CEST), Andreas Schneider wrote: Author: Andreas Schneider a...@samba.org Date: Wed Apr 2 09:01:41 2014 +0200 Add YouCompleteMe config. Would have been good to see this go by the list beforehand... I'm not a fan of seeing developer environment tools, editor modelines, add-ons and other accessories go into the source repo. IMO this kind of cruft should remain on the developers' system - people have differing setups. Cheers, David
Re: [SCM] Samba Shared Repository - branch master updated
On Wed, Feb 19, 2014 at 08:20:04PM +0100, Andreas Schneider wrote: The branch, master has been updated Andreas and David, via b8844fc clitar.c: check all allocations return value I *HATE* *HATE* *HATE* the idiom used here. Calling smb_panic on an allocation fail in the new tar code is manifestly the *WRONG* thing to do. +/* helper macro to die in case of NULL pointer */ +#define PANIC_IF_NULL(x) \ +_panic_if_null(x, __FILE__ : STR2(__LINE__) ( #x ) == NULL\n) + +/* prototype to silent gcc warning */ +static inline void* _panic_if_null(void *p, const char *expr); +static inline void* _panic_if_null(void *p, const char *expr) +{ +if (!p) { +smb_panic(expr); +} +return p; +} + Then later.. +TALLOC_CTX *ctx = PANIC_IF_NULL(talloc_new(NULL)); and: +fname = PANIC_IF_NULL(talloc_asprintf(ctx, + %s%s, + client_get_cur_dir(), + buf)); if (!fname) { err = 1; goto out; The above even left the (correct) handling of the NULL return alone ! Remember on common setups this will call a panic action script which will leave the client hanging on a sleep 999 call and leave the caller with a hung command and no indication of what failed. Can you fix this please to do proper error checking of a NULL return and to bail out of the tar command with a correct error indication that the command failed ? This really needs fixing properly before this code can go anywhere near a production branch. Jeremy.
Re: [SCM] Samba Shared Repository - branch master updated
On Thu, Jan 23, 2014 at 11:52:04PM +0100, Andrew Bartlett wrote: The branch, master has been updated via 6104b1f samba-tool classicupgrade: Remove unsued upgrade_smbconf via 6c6c3fa samba-tool classicupgrade: Remove unsued reference to samba3sam from 7638f52 s3: Don't open the dir for stat/attr/acl related mask http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master - Log - commit 6104b1fe98798f275d0fe5f81a678941ab29b184 Author: Andrew Bartlett abart...@samba.org Date: Thu Jan 16 15:08:16 2014 +1300 samba-tool classicupgrade: Remove unsued upgrade_smbconf Signed-off-by: Andrew Bartlett abart...@samba.org Reviewed-By: Jelmer Vernooij jel...@samba.org Autobuild-User(master): Andrew Bartlett abart...@samba.org Autobuild-Date(master): Thu Jan 23 23:51:56 CET 2014 on sn-devel-104 commit 6c6c3fa7ccb1415addd9553c68fc21115b7c23da Author: Andrew Bartlett abart...@samba.org Date: Thu Jan 16 15:07:31 2014 +1300 samba-tool classicupgrade: Remove unsued reference to samba3sam Signed-off-by: Andrew Bartlett abart...@samba.org Reviewed-By: Jelmer Vernooij jel...@samba.org gwenhwyvar:~/src/samba% git log | grep -i unsued | wc -l 7 :-) signature.asc Description: Digital signature
Re: [SCM] Samba Shared Repository - branch master updated
On Wed, Jan 08, 2014 at 08:13:04AM +0100, Andrew Bartlett wrote: The branch, master has been updated via f8363dd crypto: fix build on OS X via af5102f build: test the generic md5 function after importing it from hashlib from 8e5f4ea libgpo: apply some const. http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master - Log - commit f8363dd22a67108022bc0fe95003450f79057f7d Author: Björn Jacke b...@sernet.de Date: Tue Jan 7 15:55:57 2014 +0100 crypto: fix build on OS X we also need to use the CC_MD5_CTX from CommonCrypto here instead of the MD5_CTX Signed-off-by: Bjoern Jacke b...@sernet.de Reviewed-by: Andrew Bartlett abart...@samba.org Autobuild-User(master): Andrew Bartlett abart...@samba.org Autobuild-Date(master): Wed Jan 8 08:12:29 CET 2014 on sn-devel-104 commit af5102fe0075520d62beba412c994d4a74f5446b Author: Björn Jacke b...@sernet.de Date: Tue Jan 7 15:55:56 2014 +0100 build: test the generic md5 function after importing it from hashlib otherwise we used the one from md5 which lead to the following warning on SerNet-imini: the md5 module is deprecated; use hashlib instead import md5 Do we still support Python 2.4 ? It does not have hashlib, just the md5 module. Cheers, jelmer
Re: [SCM] Samba Shared Repository - branch master updated
On Sat, Dec 14, 2013 at 01:20:04PM +0100, Stefan Metzmacher wrote: commit 5baa7402ba8eebd9b2ddc6b259ae9bb2852b4bb1 Author: Volker Lendecke v...@samba.org Date: Fri Dec 6 14:34:05 2013 + smbd: Implement and use full_path_tos Yes, this looks like a hack, but talloc_asprintf does show up high in profiles called from these routines Signed-off-by: Volker Lendecke v...@samba.org Reviewed-by: Stefan Metzmacher me...@samba.org Hahaha ! Coming this summer to a cinema near you : Samba 4 - Revenge of the pstrings :-) :-). Great work though. I take it this was a cachegrind find ? Cheers, Jeremy.
Re: [SCM] Samba Shared Repository - branch master updated
On Mon, 2013-12-16 at 10:34 -0800, Jeremy Allison wrote: On Sat, Dec 14, 2013 at 01:20:04PM +0100, Stefan Metzmacher wrote: commit 5baa7402ba8eebd9b2ddc6b259ae9bb2852b4bb1 Author: Volker Lendecke v...@samba.org Date: Fri Dec 6 14:34:05 2013 + smbd: Implement and use full_path_tos Yes, this looks like a hack, but talloc_asprintf does show up high in profiles called from these routines Signed-off-by: Volker Lendecke v...@samba.org Reviewed-by: Stefan Metzmacher me...@samba.org Hahaha ! Coming this summer to a cinema near you : Samba 4 - Revenge of the pstrings :-) :-). Uhm, given this string is never passed anywhere and you want to use the stack, why didn't you simply use alloca() + snprintf() ? That would seem a lot simpler and as effective to me, am I missing something ? Simo.
Re: [SCM] Samba Shared Repository - branch master updated
On Mon, Dec 16, 2013 at 03:56:16PM -0500, Simo wrote: On Mon, 2013-12-16 at 10:34 -0800, Jeremy Allison wrote: On Sat, Dec 14, 2013 at 01:20:04PM +0100, Stefan Metzmacher wrote: commit 5baa7402ba8eebd9b2ddc6b259ae9bb2852b4bb1 Author: Volker Lendecke v...@samba.org Date: Fri Dec 6 14:34:05 2013 + smbd: Implement and use full_path_tos Yes, this looks like a hack, but talloc_asprintf does show up high in profiles called from these routines Signed-off-by: Volker Lendecke v...@samba.org Reviewed-by: Stefan Metzmacher me...@samba.org Hahaha ! Coming this summer to a cinema near you : Samba 4 - Revenge of the pstrings :-) :-). Uhm, given this string is never passed anywhere and you want to use the stack, why didn't you simply use alloca() + snprintf() ? That would seem a lot simpler and as effective to me, am I missing something ? I don't think alloca is as portable as the code Volker wrote.
Re: [SCM] Samba Shared Repository - branch master updated
On Mon, 2013-12-16 at 13:05 -0800, Jeremy Allison wrote: On Mon, Dec 16, 2013 at 03:56:16PM -0500, Simo wrote: On Mon, 2013-12-16 at 10:34 -0800, Jeremy Allison wrote: On Sat, Dec 14, 2013 at 01:20:04PM +0100, Stefan Metzmacher wrote: commit 5baa7402ba8eebd9b2ddc6b259ae9bb2852b4bb1 Author: Volker Lendecke v...@samba.org Date: Fri Dec 6 14:34:05 2013 + smbd: Implement and use full_path_tos Yes, this looks like a hack, but talloc_asprintf does show up high in profiles called from these routines Signed-off-by: Volker Lendecke v...@samba.org Reviewed-by: Stefan Metzmacher me...@samba.org Hahaha ! Coming this summer to a cinema near you : Samba 4 - Revenge of the pstrings :-) :-). Uhm, given this string is never passed anywhere and you want to use the stack, why didn't you simply use alloca() + snprintf() ? That would seem a lot simpler and as effective to me, am I missing something ? I don't think alloca is as portable as the code Volker wrote. I see. Simo.
Re: [SCM] Samba Shared Repository - branch master updated
commit 12a2230581b3ff5c7a29819532652d7ddfe61521 Author: Andreas Schneider a...@samba.org Date: Fri Nov 8 16:14:35 2013 +0100 s4-smb_server: Fix a use after free. If we haven't allocated the smbsrv_session then we should not free it. Signed-off-by: Andreas Schneider a...@samba.org Reviewed-by: Jeremy Allison j...@samba.org diff --git a/source4/smb_server/smb/sesssetup.c b/source4/smb_server/smb/sesssetup.c index b26c128..4ebc0c4 100644 --- a/source4/smb_server/smb/sesssetup.c +++ b/source4/smb_server/smb/sesssetup.c @@ -415,6 +415,7 @@ static void sesssetup_spnego(struct smbsrv_request *req, union smb_sesssetup *se { NTSTATUS status; struct smbsrv_session *smb_sess = NULL; + bool is_smb_sess_new = false; struct sesssetup_spnego_state *s = NULL; uint16_t vuid; struct tevent_req *subreq; @@ -465,6 +466,7 @@ static void sesssetup_spnego(struct smbsrv_request *req, union smb_sesssetup *se status = NT_STATUS_INSUFFICIENT_RESOURCES; goto failed; } + is_smb_sess_new = true; } else { smb_sess = smbsrv_session_find_sesssetup(req-smb_conn, vuid); } @@ -510,7 +512,9 @@ static void sesssetup_spnego(struct smbsrv_request *req, union smb_sesssetup *se nomem: status = NT_STATUS_NO_MEMORY; failed: - talloc_free(smb_sess); + if (is_smb_sess_new) { + talloc_free(smb_sess); + } status = nt_status_squash(status); smbsrv_sesssetup_backend_send(req, sess, status); I think we need to talloc_steal(req, smb_sess) here. This is similar to https://git.samba.org/?p=samba.git;a=commitdiff;h=25494628a2e977568de0f634602ebe893d0a5b88 metze
Re: [SCM] Samba Shared Repository - branch master updated
On Wed, Sep 11, 2013 at 10:16:02AM +0200, Volker Lendecke wrote: The branch, master has been updated via 73278cd smbd: Convert br_lck-lock_data to talloc via 20cc710 smbd: Move struct byte_range_lock definition to brlock.c via 07948ef smbd: Add brl_fsp access function via 6bcfc1a smbd: Add brl_num_locks access function via 635c35d smbd: Use ZERO_STRUCT instead of memset via e2bb3b0 smbd: Fix a typo via 2aa6844 smbd: Make brl_lock_failed static via cca8faa smbd: Make brl_same_context static via e533bf3 smbd: Fix blank line endings from d809cf6 Raise the level of a debug. http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master - Log - commit 73278cd19bc80854a2b23ef787c3ebdc3fa56bfd Author: Volker Lendecke v...@samba.org Date: Tue Sep 10 21:04:47 2013 +0200 smbd: Convert br_lck-lock_data to talloc Autobuild-User(master): Volker Lendecke v...@samba.org Autobuild-Date(master): Wed Sep 11 10:15:38 CEST 2013 on sn-devel-104 Gna. Pushed from the wrong tree. This was meant to carry Reviewed-by: Jeremy Allison j...@samba.org tags. Push a revert? Sorry! Volker -- SerNet GmbH, Bahnhofsallee 1b, 37081 Göttingen phone: +49-551-37-0, fax: +49-551-37-9 AG Göttingen, HRB 2816, GF: Dr. Johannes Loxen http://www.sernet.de, mailto:kont...@sernet.de * visit us on it-sa:IT security exhibitions in Nürnberg, Germany October 8th - 10th 2013, hall 12, booth 333 free tickets available via code 270691 on: www.it-sa.de/gutschein **
Re: [SCM] Samba Shared Repository - branch master updated
On Wed, Sep 11, 2013 at 10:18:50AM +0200, Volker Lendecke wrote: On Wed, Sep 11, 2013 at 10:16:02AM +0200, Volker Lendecke wrote: The branch, master has been updated via 73278cd smbd: Convert br_lck-lock_data to talloc via 20cc710 smbd: Move struct byte_range_lock definition to brlock.c via 07948ef smbd: Add brl_fsp access function via 6bcfc1a smbd: Add brl_num_locks access function via 635c35d smbd: Use ZERO_STRUCT instead of memset via e2bb3b0 smbd: Fix a typo via 2aa6844 smbd: Make brl_lock_failed static via cca8faa smbd: Make brl_same_context static via e533bf3 smbd: Fix blank line endings from d809cf6 Raise the level of a debug. http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master - Log - commit 73278cd19bc80854a2b23ef787c3ebdc3fa56bfd Author: Volker Lendecke v...@samba.org Date: Tue Sep 10 21:04:47 2013 +0200 smbd: Convert br_lck-lock_data to talloc Autobuild-User(master): Volker Lendecke v...@samba.org Autobuild-Date(master): Wed Sep 11 10:15:38 CEST 2013 on sn-devel-104 Gna. Pushed from the wrong tree. This was meant to carry Reviewed-by: Jeremy Allison j...@samba.org tags. Push a revert? No, not worth it IMHO. These messages are archived so anyone can check this was correctly reviewed. Jeremy.
Re: [SCM] Samba Shared Repository - branch master updated
On 2013-09-11 at 10:18 +0200, Volker Lendecke wrote: On Wed, Sep 11, 2013 at 10:16:02AM +0200, Volker Lendecke wrote: Gna. Pushed from the wrong tree. This was meant to carry Reviewed-by: Jeremy Allison j...@samba.org tags. Push a revert? I'd say there is no point. It is in the tree and you would have to get review for the revert first... ;) The important point is that review has in fact happened. This would not have happened, if we had installed a mechanism that would reject a push if one at least of the commits does not contain at least two signoff/review tags by team members... (Hint, hint... ;-) Cheers - Michael signature.asc Description: Digital signature
Re: [SCM] Samba Shared Repository - branch master updated
On Friday 08 February 2013 17:06:12 Jeremy Allison wrote: On Thu, Feb 07, 2013 at 03:51:05PM +0100, Andreas Schneider wrote: The branch, master has been updated via 62e6ea4 s3-rpc_server: Fix password encoding in _netr_ServerGetTrustInfo(). via 823a25f s3-rpc_server: Dont wipe out ref pointers in _netr_ServerGetTrustInfo(). from ff496ef autoconf: rename pdb_ldap module to pdb_ldapsam commit 62e6ea408a20197420eadb133df6ba233c7874b4 Author: Sumit Bose sb...@redhat.com Date: Mon Feb 4 13:15:18 2013 +0100 s3-rpc_server: Fix password encoding in _netr_ServerGetTrustInfo(). Reviewed-by: Günther Deschner g...@samba.org Reviewed-by: Andreas Schneider a...@samba.org commit 823a25f6a47c311ab33874be8a889ffbaf8db527 Author: Günther Deschner g...@samba.org Date: Tue Dec 11 09:29:37 2012 +0100 s3-rpc_server: Dont wipe out ref pointers in _netr_ServerGetTrustInfo(). Guenther Signed-off-by: Günther Deschner g...@samba.org Reviewed-by: Andreas Schneider a...@samba.org Do we need these in 3.6.x and 4.0.x ? Jeremy. I think so, I thought Sumit or Günther created a bug for it. Günther? -- andreas -- Andreas Schneider GPG-ID: F33E3FC6 Samba Team a...@samba.org www.samba.org
Re: [SCM] Samba Shared Repository - branch master updated
On Thu, Feb 07, 2013 at 03:51:05PM +0100, Andreas Schneider wrote: The branch, master has been updated via 62e6ea4 s3-rpc_server: Fix password encoding in _netr_ServerGetTrustInfo(). via 823a25f s3-rpc_server: Dont wipe out ref pointers in _netr_ServerGetTrustInfo(). from ff496ef autoconf: rename pdb_ldap module to pdb_ldapsam commit 62e6ea408a20197420eadb133df6ba233c7874b4 Author: Sumit Bose sb...@redhat.com Date: Mon Feb 4 13:15:18 2013 +0100 s3-rpc_server: Fix password encoding in _netr_ServerGetTrustInfo(). Reviewed-by: Günther Deschner g...@samba.org Reviewed-by: Andreas Schneider a...@samba.org commit 823a25f6a47c311ab33874be8a889ffbaf8db527 Author: Günther Deschner g...@samba.org Date: Tue Dec 11 09:29:37 2012 +0100 s3-rpc_server: Dont wipe out ref pointers in _netr_ServerGetTrustInfo(). Guenther Signed-off-by: Günther Deschner g...@samba.org Reviewed-by: Andreas Schneider a...@samba.org Do we need these in 3.6.x and 4.0.x ? Jeremy.
Re: [SCM] Samba Shared Repository - branch master updated
On Sun, Nov 11, 2012 at 03:49:02PM +0100, Andrew Bartlett wrote: The branch, master has been updated via 1d81e52 selftest: Add tests for expected behaviour on directories as well as files via a6a0155 pysmbd: Add SMB_ACL_EXECUTE to the mask set by make_simple_acl() via 312f8dd selftest: Make samba.tests.ntacl also use TestCaseInTempDir via b4d8629 samba-tool: Rework ldap attribute fetch in classicupgrade for missing attributes from c06d602 s3:smbd: Fix typo in got_duplicate_group check http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master - Log - commit 1d81e52bba65f05378db7027537aa27eb5bfa70a Author: Andrew Bartlett abart...@samba.org Date: Sun Nov 11 21:33:41 2012 +1100 selftest: Add tests for expected behaviour on directories as well as files This is important because it covers the codepath which had the talloc error fixed by commit 60cf4cb5a630506747431ecbf00d890509baf2f3 (vfs_acl_common: In add_directory_inheritable_components allocate on psd as parent) Andrew Bartlett Signed-off-by: Andrew Bartlett abart...@samba.org Reviewed-by: Jelmer Vernooij jel...@samba.org Reviewed-by: Andrew Bartlett abart...@samba.org You've reviewed your own changes ? I'm shocked. :-) Is this a script gone crazy perhaps ? Cheers, Jelmer
Re: [SCM] Samba Shared Repository - branch master updated
On Tue, 2012-10-30 at 23:57 +0100, Andrew Bartlett wrote: commit cc6d0decc7980028293168aee267e7610752fc80 Author: Andrew Bartlett abart...@samba.org Date: Tue Oct 30 10:21:42 2012 +1100 ldb: Change ltdb_unpack_data to take an ldb_context It always de-references the module to find the ldb anyway. Andrew Bartlett Andrew, why are you messing over with these interface conventions ? I see no rationale for this change, can you please revert and learn a bit about consistent and predictable interfaces ? Thanks. Simo. -- Simo Sorce Samba Team GPL Compliance Officer s...@samba.org Principal Software Engineer at Red Hat, Inc. s...@redhat.com
Re: [SCM] Samba Shared Repository - branch master updated
On Fri, Oct 19, 2012 at 09:03:46PM +0200, Michael Adam wrote: On 2012-10-19 at 10:14 -0700, Jeremy Allison wrote: Can you guys split these checks out into a static function that you call once from vfs_default_durable_reconnect() ? Putting these checks inside that function makes it a million lines long and hard to read :-). /me goes counting lines and major culprits in various even longer functions like open_file_ntcreate()... ;-) Yes I know. And I got beaten up so much for that I never did it again :-). And vl and I have been slimming it down ever since :-). Seriously: Certainly, we can do that. Do you mean a single function of each of the checks, or would you be fine with one function for all the stat detail checks? One function for all of the stat checks works for me. That function just goes through a series of logical steps, to me it makes sense for it to be factored out into independent logical groups (i.e. the stat checks are one) that are put into separate static functions. Thanks ! Jeremy.
Re: [SCM] Samba Shared Repository - branch master updated
Hi Jeremy, On 2012-10-20 at 00:57 -0700, Jeremy Allison wrote: On Fri, Oct 19, 2012 at 09:03:46PM +0200, Michael Adam wrote: On 2012-10-19 at 10:14 -0700, Jeremy Allison wrote: Can you guys split these checks out into a static function that you call once from vfs_default_durable_reconnect() ? Putting these checks inside that function makes it a million lines long and hard to read :-). /me goes counting lines and major culprits in various even longer functions like open_file_ntcreate()... ;-) Yes I know. And I got beaten up so much for that I never did it again :-). :) And vl and I have been slimming it down ever since :-). True, and that is very helpful, indeed. Seriously: Certainly, we can do that. Do you mean a single function of each of the checks, or would you be fine with one function for all the stat detail checks? One function for all of the stat checks works for me. That function just goes through a series of logical steps, to me it makes sense for it to be factored out into independent logical groups (i.e. the stat checks are one) that are put into separate static functions. It definitley does make sense. I will factor it out and run the patch via you. Thanks for the hint. Cheers - Michael pgpQ8wf9yXElb.pgp Description: PGP signature
Re: [SCM] Samba Shared Repository - branch master updated
On Fri, Oct 19, 2012 at 06:57:03PM +0200, Stefan Metzmacher wrote: The branch, master has been updated via 7d05ff7 s3:smbd:durable: store stat info for disconnected durables and check at reconnect. via a6726f9 s3:open_files.idl: add stat-info to vfs_default_durable_cookie. via 660e90d s3:smbd/durable: remove an extra blank line from vfs_default_durable_reconnect() from 24d225f s3:smbd/durable: add write_time specific stuff to vfs_default_durable_cookie http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master - Log - commit 7d05ff77264cf48d1ee82ca19c8c7e6951e64cf5 Author: Michael Adam ob...@samba.org Date: Fri Oct 19 16:00:05 2012 +0200 s3:smbd:durable: store stat info for disconnected durables and check at reconnect. At durable reconnect, we check the stat info stored in the cookie against the current stat data from the file we just opened. If any detail differs, we deny the durable reconnect, because in that case it is very likely that someone opened the file while the handle was disconnected, which has to be interpreted as an oplock break. Can you guys split these checks out into a static function that you call once from vfs_default_durable_reconnect() ? Putting these checks inside that function makes it a million lines long and hard to read :-). Thanks, Jeremy.
Re: [SCM] Samba Shared Repository - branch master updated
On 2012-10-19 at 10:14 -0700, Jeremy Allison wrote: On Fri, Oct 19, 2012 at 06:57:03PM +0200, Stefan Metzmacher wrote: The branch, master has been updated via 7d05ff7 s3:smbd:durable: store stat info for disconnected durables and check at reconnect. via a6726f9 s3:open_files.idl: add stat-info to vfs_default_durable_cookie. via 660e90d s3:smbd/durable: remove an extra blank line from vfs_default_durable_reconnect() from 24d225f s3:smbd/durable: add write_time specific stuff to vfs_default_durable_cookie http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master - Log - commit 7d05ff77264cf48d1ee82ca19c8c7e6951e64cf5 Author: Michael Adam ob...@samba.org Date: Fri Oct 19 16:00:05 2012 +0200 s3:smbd:durable: store stat info for disconnected durables and check at reconnect. At durable reconnect, we check the stat info stored in the cookie against the current stat data from the file we just opened. If any detail differs, we deny the durable reconnect, because in that case it is very likely that someone opened the file while the handle was disconnected, which has to be interpreted as an oplock break. Can you guys split these checks out into a static function that you call once from vfs_default_durable_reconnect() ? Putting these checks inside that function makes it a million lines long and hard to read :-). /me goes counting lines and major culprits in various even longer functions like open_file_ntcreate()... ;-) Seriously: Certainly, we can do that. Do you mean a single function of each of the checks, or would you be fine with one function for all the stat detail checks? Cheers - Michael pgpUPLlVGDpLn.pgp Description: PGP signature
Re: [SCM] Samba Shared Repository - branch master updated
On 2012-10-11 21:23, Stefan (metze) Metzmacher wrote: Hi Jelmer, - Log - commit c2d14747d608d406de6410556807d467cd0b85ef Author: Jelmer Vernooij jel...@samba.org Date: Thu Oct 11 14:45:10 2012 +0200 provision: Always create DNS user. The DNS user is currently only used by the bind9 plugin. This makes it easier to later on switch between the builtin DNS server and bind backend. In addition, ideally the internal DNS server would use that (separate) user too. Why? Isn't that the job of samba_upgradedns? I removed this behavior because I want us to match windows as much as possible. +1 Cheers, Kai -- Kai Blin Worldforge developer http://www.worldforge.org/ Wine developer http://wiki.winehq.org/KaiBlin Samba team member http://www.samba.org/samba/team/
Re: [SCM] Samba Shared Repository - branch master updated
Hi Jelmer, - Log - commit c2d14747d608d406de6410556807d467cd0b85ef Author: Jelmer Vernooij jel...@samba.org Date: Thu Oct 11 14:45:10 2012 +0200 provision: Always create DNS user. The DNS user is currently only used by the bind9 plugin. This makes it easier to later on switch between the builtin DNS server and bind backend. In addition, ideally the internal DNS server would use that (separate) user too. Why? Isn't that the job of samba_upgradedns? I removed this behavior because I want us to match windows as much as possible. metze signature.asc Description: OpenPGP digital signature
Re: [SCM] Samba Shared Repository - branch master updated
On Wed, Aug 29, 2012 at 04:06:02PM +0200, Björn Jacke wrote: The branch, master has been updated via f31d0d0 vfs_media_harmony: fix some compile warnings with llvm from fb15e5a s3-printing: fix bug 9123 lprng job tracking errors http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master - Log - commit f31d0d0e30af6577de483ec4e811b9422b8ef39a Author: Björn Jacke b...@sernet.de Date: Wed Aug 29 13:37:05 2012 +0200 vfs_media_harmony: fix some compile warnings with llvm Autobuild-User(master): Björn Jacke b...@sernet.de Autobuild-Date(master): Wed Aug 29 16:05:10 CEST 2012 on sn-devel-104 --- Summary of changes: source3/modules/vfs_media_harmony.c | 132 +- 1 files changed, 66 insertions(+), 66 deletions(-) Changeset truncated at 500 lines: diff --git a/source3/modules/vfs_media_harmony.c b/source3/modules/vfs_media_harmony.c index e2cb407..53cb176 100644 --- a/source3/modules/vfs_media_harmony.c +++ b/source3/modules/vfs_media_harmony.c @@ -430,7 +430,7 @@ static int alloc_get_client_path(vfs_handle_struct *handle, (*newPath)[pathPtr - path + CREATING_DIRNAME_LEN] = '\0'; DEBUG(MH_INFO_DEBUG, (newPath #2 %s\n, *newPath)); - if (status = alloc_append_client_suffix(handle, newPath)) + if ((status = alloc_append_client_suffix(handle, newPath))) Can you make that into two lines? This form will create warnings on other compilers. status = ... if (status != NULL) { } is the one to use according to README.Coding. Thanks, Volker -- SerNet GmbH, Bahnhofsallee 1b, 37081 Göttingen phone: +49-551-37-0, fax: +49-551-37-9 AG Göttingen, HRB 2816, GF: Dr. Johannes Loxen http://www.sernet.de, mailto:kont...@sernet.de
Re: [SCM] Samba Shared Repository - branch master updated
On Thu, 2012-08-09 at 08:21 +0200, Andrew Bartlett wrote: The branch, master has been updated via a117fd6 s4-dsdb: Ensure we have indexing enabled during the provision via ef87b4e s4-pydsdb: Provide control of if we should write index attributes when reloading a schema from 1a1f01e s4-dsdb: Change talloc parent http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master - Log - commit a117fd6d11aef5360a1af1fc6da542829da8b7b0 Author: Andrew Bartlett abart...@samba.org Date: Thu Aug 9 14:33:49 2012 +1000 s4-dsdb: Ensure we have indexing enabled during the provision Because we set the schema before we connected the ldb to a file, the @INDEX records were not added until next startup. This cost 100% more time in running provision on my laptop. Andrew Bartlett Autobuild-User(master): Andrew Bartlett abart...@samba.org Autobuild-Date(master): Thu Aug 9 08:20:36 CEST 2012 on sn-devel-104 commit ef87b4e4f10eb7d5974cb0e0861648d537153a00 Author: Andrew Bartlett abart...@samba.org Date: Thu Aug 9 14:23:12 2012 +1000 s4-pydsdb: Provide control of if we should write index attributes when reloading a schema This allows us to carefully control the loading of the schema. Andrew Bartlett --- Summary of changes: source4/dsdb/pydsdb.c |6 -- .../scripting/python/samba/provision/__init__.py |7 ++- source4/scripting/python/samba/samdb.py|8 3 files changed, 14 insertions(+), 7 deletions(-) Changeset truncated at 500 lines: diff --git a/source4/dsdb/pydsdb.c b/source4/dsdb/pydsdb.c index 3558049..9023d69 100644 --- a/source4/dsdb/pydsdb.c +++ b/source4/dsdb/pydsdb.c @@ -873,7 +873,9 @@ static PyObject *py_dsdb_set_schema_from_ldb(PyObject *self, PyObject *args) struct ldb_context *from_ldb; struct dsdb_schema *schema; int ret; - if (!PyArg_ParseTuple(args, OO, py_ldb, py_from_ldb)) + char write_attributes = true; + if (!PyArg_ParseTuple(args, OO|b, + py_ldb, py_from_ldb, write_attributes)) return NULL; PyErr_LDB_OR_RAISE(py_ldb, ldb); @@ -886,7 +888,7 @@ static PyObject *py_dsdb_set_schema_from_ldb(PyObject *self, PyObject *args) return NULL; } - ret = dsdb_reference_schema(ldb, schema, true); + ret = dsdb_reference_schema(ldb, schema, write_attributes); PyErr_LDB_ERROR_IS_ERR_RAISE(py_ldb_get_exception(), ret, ldb); Py_RETURN_NONE; diff --git a/source4/scripting/python/samba/provision/__init__.py b/source4/scripting/python/samba/provision/__init__.py index 192130e..94e857e 100644 --- a/source4/scripting/python/samba/provision/__init__.py +++ b/source4/scripting/python/samba/provision/__init__.py @@ -1121,7 +1121,7 @@ def setup_samdb(path, session_info, provision_backend, lp, names, logger.info(Pre-loading the Samba 4 and AD schema) # Load the schema from the one we computed earlier -samdb.set_schema(schema) +samdb.set_schema(schema, write_attributes=False) # Set the NTDS settings DN manually - in order to have it already around # before the provisioned tree exists and we connect @@ -1131,6 +1131,11 @@ def setup_samdb(path, session_info, provision_backend, lp, names, # DB samdb.connect(path) +# But we have to give it one more kick to have it use the schema +# during provision - it needs, now that it is connected, to write +# the schema @INDEX records to the database. +samdb.set_schema(schema, write_attributes=True) + return samdb diff --git a/source4/scripting/python/samba/samdb.py b/source4/scripting/python/samba/samdb.py index 7451b6c..3355e9a 100644 --- a/source4/scripting/python/samba/samdb.py +++ b/source4/scripting/python/samba/samdb.py @@ -608,11 +608,11 @@ accountExpires: %u def load_partition_usn(self, base_dn): return dsdb._dsdb_load_partition_usn(self, base_dn) -def set_schema(self, schema): -self.set_schema_from_ldb(schema.ldb) +def set_schema(self, schema, write_attributes=True): +self.set_schema_from_ldb(schema.ldb, write_attributes=write_attributes) -def set_schema_from_ldb(self, ldb_conn): -dsdb._dsdb_set_schema_from_ldb(self, ldb_conn) +def set_schema_from_ldb(self, ldb_conn, write_attributes=True): +dsdb._dsdb_set_schema_from_ldb(self, ldb_conn, write_attributes) def dsdb_DsReplicaAttribute(self, ldb, ldap_display_name, ldif_elements): '''convert a list of attribute values to a DRSUAPI DsReplicaAttribute''' This code would be clearer if you called the variable write_index, rather then
Re: [SCM] Samba Shared Repository - branch master updated
On Sat, 2012-06-16 at 10:14 +0200, Andrew Bartlett wrote: The branch, master has been updated via 4edd8b8 s3-auth: Remove auth_netlogond via 9c715da s3-passdb: Remove pdb_ads Andrew, I would like you to revert these two commits ASAP. Simo. via d949736 s4-classicupgrade: Also ask testparm for 'smb passwd file' via a0a2f79 WHATSNEW: Bump the version and announce the s3fs default via d9f7195 s4-classicupgrade: Use samba classic description for samba3 NT4-like domains in samba3upgrade via 39766b7 s4-lib/param: FLAG DAY for the default FILE SERVER via b58dc18 s4-s3upgrade: Assert that administrator has a SID of -500, and only skip root if it is -500 via 61f7f01 s4-s3upgrade: Add my wins.dat and fix the parsing error via d0b60f0 s4-s3upgrade: improve idmap import to use posixAccount and posixGroup entries via 3c65bac s4-idmap: Add mapping using uidNumber and gidNumber like idmap_ad from bbb7cbf Same fix as bug 8989 - Samba 3.5.x (and probably all other versions of Samba) does not send correct responses to NT Transact Secondary when no data and no params http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master - Log - commit 4edd8b891a90a89a84fbfa3636cc568d247b04b2 Author: Andrew Bartlett abart...@samba.org Date: Sun Jun 3 10:56:46 2012 +1000 s3-auth: Remove auth_netlogond auth_netlogond was an important module in the development of the combined Samba 4.0, and was the first module to link smbd with the AD authentication store, showing that it was possible for NTLM authentication to be offloaded to the AD server components. We now have auth_samba4, which provides the full GENSEC stack to smbd, which also matches exactly the group membership and privileges assignment and which is supported and tested as part of the official Samba 4.0 release configuration. Andrew Bartlett Autobuild-User(master): Andrew Bartlett abart...@samba.org Autobuild-Date(master): Sat Jun 16 10:13:20 CEST 2012 on sn-devel-104 commit 9c715da1cbc256b9ae9298618c92807592607c9b Author: Andrew Bartlett abart...@samba.org Date: Sun Jun 3 10:54:06 2012 +1000 s3-passdb: Remove pdb_ads pdb_ads was an important module in the development of the combined Samba 4.0, and was the first module to show that standard samba3 tools such as smbpasswd can be made to operate on the sam.ldb. We now have pdb_samba4, which operates directly on the sam.ldb, rather than via ldapi://, which uses transactions and which is supported and tested as part of the official Samba 4.0 release configuration. This module is not as complete (for example, it does not honour the idmap configuration) and requires that the samba binary be running to operate. Andrew Bartlett commit d949736f8dc02eec180723a55f4604b7b3aa83d8 Author: Andrew Bartlett abart...@samba.org Date: Sat Jun 16 15:34:50 2012 +1000 s4-classicupgrade: Also ask testparm for 'smb passwd file' commit a0a2f7999e20ab64dcbfca8299dbf0adfba0dea3 Author: Andrew Bartlett abart...@samba.org Date: Sat Jun 16 13:12:50 2012 +1000 WHATSNEW: Bump the version and announce the s3fs default commit d9f7195a1f5a12d5dc8865aa5553b61a4f770e3d Author: Andrew Bartlett abart...@samba.org Date: Sat Jun 16 13:06:44 2012 +1000 s4-classicupgrade: Use samba classic description for samba3 NT4-like domains in samba3upgrade commit 39766b75a40fbab73fc23dd947de44f8349ed466 Author: Andrew Bartlett abart...@samba.org Date: Sat Jun 16 12:54:12 2012 +1000 s4-lib/param: FLAG DAY for the default FILE SERVER This commit changes the default file server to be s3fs. Existing installs wishing to keep the ntvfs file server need to set this in their smb.conf: server services = +smb -s3fs dcerpc endpoint services = +winreg +srvsvc Andrew Bartlett commit b58dc1826e69c61a30d38b05e7f451404670baef Author: Andrew Bartlett abart...@samba.org Date: Sat Jun 16 14:19:42 2012 +1000 s4-s3upgrade: Assert that administrator has a SID of -500, and only skip root if it is -500 Many upgraded installations have root as -1000, and so that account needs to be kept. Andrew Bartlett commit 61f7f0155465b14612f7ac29a12c442ff25031b4 Author: Andrew Bartlett abart...@samba.org Date: Sat Jun 16 13:58:06 2012 +1000 s4-s3upgrade: Add my wins.dat and fix the parsing error The issue was that the numbers at the end of the lines are space padded. Andrew Bartlett commit d0b60f02dd3c324d4c990dae7334b228dddba075 Author: Andrew Bartlett abart...@samba.org Date: Sun Jun 10 20:42:25 2012 +1000 s4-s3upgrade: improve idmap import to use posixAccount and
Remove unused finddcs_nbt (Re: [SCM] Samba Shared Repository - branch master updated)
Hi Andrew, commit 06c90cb6f55701effa4cbafaf189a4de8471949b Author: Andrew Bartlett abart...@samba.org Date: Thu Jun 7 14:21:25 2012 +1000 s4-libcli: Remove unused finddcs_nbt This would only do the NBT getdc lookup for a single DC (but would find multiple DCs at first stage), but more particular it of course uses Netbios rather than DNS names. In any case it was also unused, as we use CLDAP for reliable DC location these days. Found by callcatcher I think it's a bug that it's not used. Samba3 DCs doesn't support CLDAP. metze signature.asc Description: OpenPGP digital signature
Re: [SCM] Samba Shared Repository - branch master updated
On Wed, 2012-04-25 at 11:44 +0200, Andrew Bartlett wrote: commit b8055132b1c62dd19981fea2822ab9e1829a8ded Author: Andrew Bartlett abart...@samba.org Date: Wed Apr 25 17:53:18 2012 +1000 s4-messaging: Use generate_random() to get a unique ID for messaging clients The call to random() resulted in duplicate values for s3fs configurations which, due to the forked child, all started with the same random seed. A future improvement would be to move to a proven unique value. Andrew Bartlett Autobuild-User: Andrew Bartlett abart...@samba.org Autobuild-Date: Wed Apr 25 11:43:40 CEST 2012 on sn-devel-104 Andrew are you sure you need to use /dev/urandom here ? It doesn't look like you need absolutelu unpredictable numbers here, just non-colliding numbers. You changed the code to draw from urandom, and if it is used often it mean it will suck a lot of entropy out of the system, causing any application that need to use /dev/random to halt. Wouldn't it have been simpler to just run srand(time(NULL)*pid) to get a new seed for the process ? Simo. -- Simo Sorce Samba Team GPL Compliance Officer s...@samba.org Principal Software Engineer at Red Hat, Inc. s...@redhat.com
Re: [SCM] Samba Shared Repository - branch master updated
On Tue, 2012-04-10 at 11:15 +0200, Andrew Bartlett wrote: index cafc020..f07bbc9 100644 --- a/lib/ldb/include/ldb_private.h +++ b/lib/ldb/include/ldb_private.h @@ -181,4 +181,12 @@ struct ldb_val ldb_binary_decode(TALLOC_CTX *mem_ctx, const char *str); const char *ldb_options_find(struct ldb_context *ldb, const char *options[], const char *option_name); +struct ldif_read_file_state { + FILE *f; + size_t line_no; +}; Andrew, can you put this back in ldb_ldif.c and provide a getter function for the line number instead ? Let's avoid keeping spreading structures all over. +struct ldb_ldif *ldb_ldif_read_file_state(struct ldb_context *ldb, + struct ldif_read_file_state *state); + -- Simo Sorce Samba Team GPL Compliance Officer s...@samba.org Principal Software Engineer at Red Hat, Inc. s...@redhat.com
Re: [SCM] Samba Shared Repository - branch master updated
Hi Jeremy, Am 09/04/12 21:41, schrieb Jeremy Allison: The branch, master has been updated via 5701a4d Move blksize_t and blkcnt_t to replace.h from includes.h. Should help with platforms that don't have these. from 5184f41 libreplace: We have a poll replacement based on select http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master - Log - commit 5701a4d8611ad1e310b76b00b146b0902c9b34ff Author: Jeremy Allison j...@samba.org Date: Mon Apr 9 11:04:07 2012 -0700 Move blksize_t and blkcnt_t to replace.h from includes.h. Should help with platforms that don't have these. Thanks for moving these to libreplace. It'd be great if you can also move the relevant tests from configure.in / wscript to the equivalent files in lib/replace, so libreplace stays self-contained. Cheers, Jelmer signature.asc Description: OpenPGP digital signature
Re: [SCM] Samba Shared Repository - branch master updated
On Mon, Apr 09, 2012 at 09:53:25PM +0200, Jelmer Vernooij wrote: Hi Jeremy, Am 09/04/12 21:41, schrieb Jeremy Allison: The branch, master has been updated via 5701a4d Move blksize_t and blkcnt_t to replace.h from includes.h. Should help with platforms that don't have these. from 5184f41 libreplace: We have a poll replacement based on select http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master - Log - commit 5701a4d8611ad1e310b76b00b146b0902c9b34ff Author: Jeremy Allison j...@samba.org Date: Mon Apr 9 11:04:07 2012 -0700 Move blksize_t and blkcnt_t to replace.h from includes.h. Should help with platforms that don't have these. Thanks for moving these to libreplace. It'd be great if you can also move the relevant tests from configure.in / wscript to the equivalent files in lib/replace, so libreplace stays self-contained. Hmmm, ok. This is probably easy for wscript, but I'm not sure what to change for configure.in. Jeremy.
Re: [SCM] Samba Shared Repository - branch master updated
Hi Andreas, Björn, On 02/09/2012 04:40 PM, Andreas Schneider wrote: The branch, master has been updated via 5ec1273 s3-printing: Add new printers to registry. from 2464a7b s3-libsmb: Remove obsolete smb_krb5_locate_kdc. http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master - Log - commit 5ec12732c3092d248a374ae7af155a409c7ee88e Author: Björn Baumbachb...@sernet.de Date: Tue Feb 7 11:41:54 2012 +0100 s3-printing: Add new printers to registry. This fixes bug #8554, #8612 and #8748. Pair-Programmed-With: Stefan Metzmacherme...@samba.org Autobuild-User: Andreas Schneidera...@cryptomilk.org Autobuild-Date: Thu Feb 9 16:39:04 CET 2012 on sn-devel-104 After this change (or perhaps an earlier one?) I'm unable to push changes to autobuild. The last two builds fail with: [275/1226 in 17m41s] samba3.posix_s3.rpc.spoolss.printer (s3dc) Testing DsSpooler- SetPrinter relations Testing GetPrinter level 2 Testing SetPrinter level 2 Testing GetPrinterDataEx(DsSpooler - description) Testing SetPrinter level 2 Testing GetPrinterDataEx(DsSpooler - location) Testing GetPrinter level 2 Testing GetPrinterDataEx(DsSpooler - description) Testing GetPrinterDataEx(DsSpooler - driverName) WARNING!: ../source4/torture/rpc/spoolss.c:3648: r.out.result was WERR_BADFILE, expected WERR_OK: GetPrinterDataEx(DsSpooler - driverName) failed WARNING!: ../source4/torture/rpc/spoolss.c:5509: Expression `test_GetPrinterDataEx(tctx, p, handle, DsSpooler, driverName,type,data,needed)' failed: failed to query UNEXPECTED(failure): samba3.posix_s3.rpc.spoolss.printer .addprinterex.printerdata_dsspooler REASON: _StringException: _StringException: ../source4/torture/rpc/spoolss.c:7879: Expression `test_PrinterData_DsSpooler(tctx, p,t-handle, t-info2.printername)' failed: failed to test printer data winreg dsspooler FAILED (1 failures, 0 errors and 0 unexpected successes in 0 testsuites) A summary with detailed information can be found in: ./bin/ab/summary test: running (/usr/bin/perl /memdisk/autobuild/flakey/b12981/samba4/selftest/selftest.pl --target=samba --prefix=./bin/ab --srcdir=/memdisk/autobuild/flakey/b12981/samba4 --exclude=/memdisk/autobuild/flakey/b12981/samba4/selftest/skip --testlist=/usr/bin/python /memdisk/autobuild/flakey/b12981/samba4/source3/selftest/tests.py| --testlist=/usr/bin/python /memdisk/autobuild/flakey/b12981/samba4/source4/selftest/tests.py| --binary-mapping=nmblookup3:nmblookup3,smbclient3:smbclient3,smbtorture4:smbtorture,ntlm_auth3:ntlm_auth3 --exclude=/memdisk/autobuild/flakey/b12981/samba4/selftest/slow --socket-wrapper touch ./bin/ab/st_done) | /usr/bin/python -u /memdisk/autobuild/flakey/b12981/samba4/selftest/filter-subunit --expected-failures=/memdisk/autobuild/flakey/b12981/samba4/selftest/knownfail --flapping=/memdisk/autobuild/flakey/b12981/samba4/selftest/flapping --fail-immediately | tee ./bin/ab/subunit | /usr/bin/python -u /memdisk/autobuild/flakey/b12981/samba4/selftest/format-subunit --prefix=./bin/ab --immediate Cheers, Jelmer
Re: [SCM] Samba Shared Repository - branch master updated
Hi Tridge, Amitay, On 10/11/11 06:05, Andrew Tridgell wrote: The branch, master has been updated via ac0bba2 test: added -D option to subunitrun subunitrun is to be deprecated, python -m subunit.run is the preferred way to run tests. Can we revert this? via 333248f pyldb: fixed ldb API test This makes LDB depend on Samba, because it imports from samba.tests. This makes it impossible to run the pyldb testsuite without having Samba installed. I'll submit an alternative fix. Cheers, Jelmer
Re: [SCM] Samba Shared Repository - branch master updated
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hi Matthieu, On 02/11/11 15:15, Matthieu Patou wrote: +#tests on sites +class SimpleSitesTests(SitesBaseTests): + + + def test_create(self): + test creation of 1 site + + self.ldb_admin.transaction_start() + ok = sites.create_site(self.ldb_admin, self.ldb_admin.get_config_basedn(), + testsamba) + self.ldb_admin.transaction_commit() + self.assertTrue(ok) + ok = False + try: + ok = sites.create_site(self.ldb_admin, self.ldb_admin.get_config_basedn(), + testsamba) + self.assertFalse(ok) + except: + self.assertFalse(ok) This looks dodgy. If sites.create_site raises *any* exception (even a SyntaxError!), then this test still passes since ok is False by default. It's almost never right to use except: - and if you do use it, please use raise to re-raise the exception. except: catches every kind of exception, including KeyboardInterrupt and SyntaxError. diff --git a/source4/scripting/python/samba/netcmd/sites.py b/source4/scripting/python/samba/netcmd/sites.py new file mode 100644 index 000..a63b524 --- /dev/null +++ b/source4/scripting/python/samba/netcmd/sites.py @@ -0,0 +1,96 @@ + +#!/usr/bin/env python +# +# sites management +# +# Copyright Matthieu Patou m...@matws.net 2011 +# +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by +# the Free Software Foundation; either version 3 of the License, or +# (at your option) any later version. +# +# This program is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +# GNU General Public License for more details. +# +# You should have received a copy of the GNU General Public License +# along with this program. If not, see http://www.gnu.org/licenses/. +# + + + +import os +from samba import sites +from samba import Ldb +from samba.auth import system_session +from samba.netcmd import ( + Command, + CommandError, + SuperCommand + ) + + +class cmd_sites_create(Command): + Create a new site + + synopsis = %prog site [options] + + takes_args = [sitename] + + def run(self, sitename, sambaopts=None, credopts=None, versionopts=None): + lp = sambaopts.get_loadparm() + creds = credopts.get_credentials(lp, fallback_machine=True) + name = sam.ldb + path = lp.get(private dir) + url = os.path.join(path, name) It should be shorter to use lp.private_path(sam.ldb). + if not os.path.exists(url): + raise CommandError(secret database not found at %s % url) This says secrets database, but it looks like it's actually about the sam ? + samdb = Ldb(url=url, session_info=system_session(), + credentials=creds, lp=lp) You probably want to use SamDB rather than Ldb here. + + samdb.transaction_start() + ok = sites.create_site(samdb, samdb.get_config_basedn(), sitename) + samdb.transaction_commit() This needs a try/except/finally otherwise we'll get a nasty warning about transactions that weren't closed. +class cmd_sites_delete(Command): + Delete a new site + The same comments as mentioned above apply to cmd_sites_delete. diff --git a/source4/scripting/python/samba/sites.py b/source4/scripting/python/samba/sites.py new file mode 100644 index 000..d1d0e75 --- /dev/null +++ b/source4/scripting/python/samba/sites.py @@ -0,0 +1,63 @@ +#!/usr/bin/env python +# +# python site manipulation code +# Copyright Matthieu Patou m...@matws.net 2011 +# +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by +# the Free Software Foundation; either version 3 of the License, or +# (at your option) any later version. +# +# This program is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +# GNU General Public License for more details. +# +# You should have received a copy of the GNU General Public License +# along with this program. If not, see http://www.gnu.org/licenses/. +# + +Manipulating sites. + +import ldb +from ldb import FLAG_MOD_ADD + +def create_site(samdb, configDn, siteName): It would be nice to have a docstring on this method. It seems to always return True, but what does that actually indicate? It might be nicer to not return anything specific but just raise exceptions in case problems occur. Cheers, Jelmer -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.11 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iQIcBAEBAgAGBQJOsVKTAAoJEACAbyvXKaRXewoP/inmGReeG3tiz8pt9kXcrLNh dyFQEJf/8ckiGpgqZRT8wbqUgnq83VdXdg+6k0x8aPH4aNR+kvZ3v9iBSMtM9Uwa dl5Z7yl7mglNqU8xawB7QiPWvhElJYrkrKSR8fqgTGbAsbsRNYU7joBH+NiWo2iK MZNcq7oVr5+H02wHyzh+jY7x0Iadfids3MxO7mh5cvIX4auUlTjklQ+kEKwHEENN
Re: [SCM] Samba Shared Repository - branch master updated
On Fri, 2011-10-28 at 11:35 +0200, Andreas Schneider wrote: +Requires=smb.service nmb.service +After=syslog.target network.target smb.service nmb.service This looks wrong, winbind does not require smb or nmb to run, if you have a laptop and do not offer shares you may want to run just winbind and no smb or nmb service. And actually I think you probably want to start winbind before smb *if* you are using winbind on a samba file server. Simo. -- Simo Sorce Samba Team GPL Compliance Officer s...@samba.org Principal Software Engineer at Red Hat, Inc. s...@redhat.com
Re: [SCM] Samba Shared Repository - branch master updated
Hello Andrew, On two of my machines with this patchset I have this error: samba version 4.0.0alpha18-DEVELOPERBUILD started. Copyright Andrew Tridgell and the Samba Team 1992-2011 Called with maxruntime 7500 - current ts 1318577424 samba: using 'standard' process model Failed to start service 'rpc' - NT_STATUS_INVALID_SYSTEM_SERVICE Starting Services failed - NT_STATUS_INVALID_SYSTEM_SERVICE ./bin/samba failed with status 1! Lookup failed - NT_STATUS_IO_TIMEOUT Lookup failed - NT_STATUS_HOST_UNREACHABLE Lookup failed - NT_STATUS_IO_TIMEOUT Lookup failed - NT_STATUS_HOST_UNREACHABLE Lookup failed - NT_STATUS_IO_TIMEOUT Lookup failed - NT_STATUS_HOST_UNREACHABLE Lookup failed - NT_STATUS_IO_TIMEOUT Lookup failed - NT_STATUS_HOST_UNREACHABLE Lookup failed - NT_STATUS_IO_TIMEOUT Lookup failed - NT_STATUS_HOST_UNREACHABLE Lookup failed - NT_STATUS_IO_TIMEOUT Lookup failed - NT_STATUS_HOST_UNREACHABLE SAMBA LOG of: LOCALDC samba version 4.0.0alpha18-DEVELOPERBUILD started. Copyright Andrew Tridgell and the Samba Team 1992-2011 Called with maxruntime 7500 - current ts 1318577424 samba: using 'standard' process model Failed to start service 'rpc' - NT_STATUS_INVALID_SYSTEM_SERVICE Starting Services failed - NT_STATUS_INVALID_SYSTEM_SERVICE ./bin/samba failed with status 1! I bissect and found that it's the first patch b21b012756dbb9e7022280b34d7103a5dcbea6d6 that is causing the error. Can you have a look ? On 11/10/2011 06:14, Andrew Bartlett wrote: The branch, master has been updated via 4549862 gensec: trim header includes back to what is actually required via 534355f auth/credentials Declare remaining functions are public interfaces and put into credentials.h via fe02752 auth: move gensec_start.c to the top level via 561d834 auth: move credentials layer to the top level via 1255383 s4-s3-upgrade: Allow import (just without a uid mapping) where getpwnam fails via c9bb497 lib/param Avoid dyn_ defines that are not provided by the autoconf build via 1e5af9e lib/param Use strcasecmp_m rather than strcasecmp as this is banned in the s3 includes.h via 8f2741b lib/param Use talloc_strdup rather than strdup as strdup is banned in the s3 includes.h via 1b81af0 lib/param Avoid the name string_set as this is already used in the s3 param code via 15c97a8 lib/param Avoid the use of the name service_ok() which is used in the s3 param code via d0ecd1a lib/param: Remove unused #include of lib/socket/socket.h via 26de383 libcli/smb Move CSC_POLICY_* definition to smb_constants.h via b21b012 lib/param move source4 param code to the top level from 6bed577 pac: Fix wrong memory allocation check http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master - Log - commit 454986298aa5696b0b029e2feba0109617aaf968 Author: Andrew Bartlettabart...@samba.org Date: Sun Oct 9 23:28:15 2011 +1100 gensec: trim header includes back to what is actually required Autobuild-User: Andrew Bartlettabart...@samba.org Autobuild-Date: Tue Oct 11 06:13:08 CEST 2011 on sn-devel-104 commit 534355fecf5a14a36ec5a3d643bcf2140df2da4e Author: Andrew Bartlettabart...@samba.org Date: Sun Oct 9 23:27:44 2011 +1100 auth/credentials Declare remaining functions are public interfaces and put into credentials.h This is in preperation for this file being used by s3, and recognises that these are all reasonable, public interfaces but were not declared as such in the past. Andrew Bartlett commit fe02752ed6493efb7af28faa3d64d9fd7895d6f1 Author: Andrew Bartlettabart...@samba.org Date: Fri Oct 7 17:24:12 2011 +1100 auth: move gensec_start.c to the top level This does not change who uses gensec for now, but makes it possible to write new gensec modules outside source4/ Andrew Bartlett commit 561d834123a2a8a96954f7cca556f8838ab38b72 Author: Andrew Bartlettabart...@samba.org Date: Fri Oct 7 17:20:33 2011 +1100 auth: move credentials layer to the top level This will allow gensec_start.c to move to the top level. This does not change what code uses the cli_credentials code, but allows the gensec code to be more broadly. Andrew Bartlett commit 1255383140a9b3fbd957c1f7ce47e89c17cc4eda Author: Andrew Bartlettabart...@samba.org Date: Mon Oct 10 13:09:30 2011 +1100 s4-s3-upgrade: Allow import (just without a uid mapping) where getpwnam fails This allows the tests to pass on systems without a jelmer user :-) Andrew Bartlett commit c9bb497f3f7fae8aa6ec4a4a45a2ac4047b640a5 Author: Andrew Bartlettabart...@samba.org Date: Sun Oct 9 23:25:11 2011 +1100 lib/param Avoid dyn_ defines that are not provided by the autoconf build The autoconf build will never use these parameters or load the smb.conf with these defaults, so the defaults are not
Re: [SCM] Samba Shared Repository - branch master updated
Hi folks, I am currently working on adding ctdb and clustered samba build to selftest and autobuild. I have added a ctdb snapshot of current ctdb master into the master branch. I can build ctdb and samba with cluster support. Now I am working on adding a clusteredmember target to Samba3. Eventually, I think we should add ctdb master with full history to samba master and change to developing it here. But for a start/poc, I'll just add a snapshot, see: http://gitweb.samba.org/?p=obnox/samba/samba-obnox.git;a=shortlog;h=refs/heads/master-ctdbtest Cheers - Michael Michael Adam wrote: The branch, master has been updated via 71f5f41 s3:dbwrap_ctdb: fix the build after changing signature of tdb_wrap_open() from e287a72 add lib/param files to .gitignore http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master - Log - commit 71f5f41860e678e6151b969a05729ccf2d738e35 Author: Michael Adam ob...@samba.org Date: Thu Oct 13 16:50:57 2011 +0200 s3:dbwrap_ctdb: fix the build after changing signature of tdb_wrap_open() Autobuild-User: Michael Adam ob...@samba.org Autobuild-Date: Thu Oct 13 18:50:47 CEST 2011 on sn-devel-104 --- Summary of changes: source3/lib/dbwrap/dbwrap_ctdb.c |8 +++- 1 files changed, 7 insertions(+), 1 deletions(-) Changeset truncated at 500 lines: diff --git a/source3/lib/dbwrap/dbwrap_ctdb.c b/source3/lib/dbwrap/dbwrap_ctdb.c index 4c5cd64..e0f98cc 100644 --- a/source3/lib/dbwrap/dbwrap_ctdb.c +++ b/source3/lib/dbwrap/dbwrap_ctdb.c @@ -23,6 +23,7 @@ #include lib/util/tdb_wrap.h #include util_tdb.h #include dbwrap/dbwrap_rbt.h +#include lib/param/param.h #ifdef CLUSTER_SUPPORT @@ -1428,6 +1429,7 @@ struct db_context *db_open_ctdb(TALLOC_CTX *mem_ctx, struct db_ctdb_ctx *db_ctdb; char *db_path; struct ctdbd_connection *conn; + struct loadparm_context *lp_ctx; if (!lp_clustering()) { DEBUG(10, (Clustering disabled -- no ctdb\n)); @@ -1474,7 +1476,11 @@ struct db_context *db_open_ctdb(TALLOC_CTX *mem_ctx, chmod(db_path, mode); } - db_ctdb-wtdb = tdb_wrap_open(db_ctdb, db_path, hash_size, tdb_flags, O_RDWR, 0); + lp_ctx = loadparm_init_s3(db_path, loadparm_s3_context()); + + db_ctdb-wtdb = tdb_wrap_open(db_ctdb, db_path, hash_size, tdb_flags, + O_RDWR, 0, lp_ctx); + talloc_unlink(db_path, lp_ctx); if (db_ctdb-wtdb == NULL) { DEBUG(0, (Could not open tdb %s: %s\n, db_path, strerror(errno))); TALLOC_FREE(result); -- Samba Shared Repository pgpdYZ8iYKSUo.pgp Description: PGP signature
Re: [SCM] Samba Shared Repository - branch master updated
Hi David, On 09/29/2011 02:48 PM, David Disseldorp wrote: The branch, master has been updated via 08573c2 s4: add SMB2_FSCTL opcodes via dbcd59f s3-smb2_server: fix ioctl InputOffset checking via 1848295 s3-smb2_server: SMB2_OP_IOCTL doesn't require at least 1 dyn byte via 392fd0d s4-torture: add smb2 ioctl test suite via bd5e975 s4-torture: remove unchecked read from smb2 create from 27195b3 socket_wrapper: correctly handle dup()/dup2() ref counting http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master Congratulations to your first commit (not your first patch of course) ! And welcome on board! Cheers, Guenther -- Günther DeschnerGPG-ID: 8EE11688 Red Hat gdesch...@redhat.com Samba Team g...@samba.org
Re: [SCM] Samba Shared Repository - branch master updated
Hi Matthieu, diff --git a/source4/ntvfs/posix/python/pyxattr_tdb.c b/source4/ntvfs/posix/python/pyxattr_tdb.c index 2b28aca..a08557e 100644 --- a/source4/ntvfs/posix/python/pyxattr_tdb.c +++ b/source4/ntvfs/posix/python/pyxattr_tdb.c @@ -60,9 +60,8 @@ static PyObject *py_wrap_setxattr(PyObject *self, PyObject *args) status = push_xattr_blob_tdb_raw(eadb, mem_ctx, attribute, filename, -1, blob); if (!NT_STATUS_IS_OK(status)) { - PyErr_FromNTSTATUS(status); talloc_free(mem_ctx); - return NULL; + PyErr_NTSTATUS_IS_ERR_RAISE(status); } talloc_free(mem_ctx); Py_RETURN_NONE; @@ -91,9 +90,8 @@ static PyObject *py_wrap_getxattr(PyObject *self, PyObject *args) status = pull_xattr_blob_tdb_raw(eadb, mem_ctx, attribute, filename, -1, 100, blob); if (!NT_STATUS_IS_OK(status) || blob.length 0) { - PyErr_FromNTSTATUS(status); talloc_free(mem_ctx); - return NULL; + PyErr_NTSTATUS_IS_ERR_RAISE(status); Why this changes? What was the problem? NT_STATUS_IS_ERR() is not the same as !NT_STATUS_IS_OK() metze signature.asc Description: OpenPGP digital signature
Re: [SCM] Samba Shared Repository - branch master updated
On 22/09/2011 21:14, Stefan (metze) Metzmacher wrote: Hi Matthieu, diff --git a/source4/ntvfs/posix/python/pyxattr_tdb.c b/source4/ntvfs/posix/python/pyxattr_tdb.c index 2b28aca..a08557e 100644 --- a/source4/ntvfs/posix/python/pyxattr_tdb.c +++ b/source4/ntvfs/posix/python/pyxattr_tdb.c @@ -60,9 +60,8 @@ static PyObject *py_wrap_setxattr(PyObject *self, PyObject *args) status = push_xattr_blob_tdb_raw(eadb, mem_ctx, attribute, filename, -1, blob); if (!NT_STATUS_IS_OK(status)) { - PyErr_FromNTSTATUS(status); talloc_free(mem_ctx); - return NULL; + PyErr_NTSTATUS_IS_ERR_RAISE(status); } talloc_free(mem_ctx); Py_RETURN_NONE; @@ -91,9 +90,8 @@ static PyObject *py_wrap_getxattr(PyObject *self, PyObject *args) status = pull_xattr_blob_tdb_raw(eadb, mem_ctx, attribute, filename, -1, 100,blob); if (!NT_STATUS_IS_OK(status) || blob.length 0) { - PyErr_FromNTSTATUS(status); talloc_free(mem_ctx); - return NULL; + PyErr_NTSTATUS_IS_ERR_RAISE(status); Why this changes? What was the problem? NT_STATUS_IS_ERR() is not the same as !NT_STATUS_IS_OK() Well the problem is that without this if your status was != OK you'll get error like instead of getting the right exception. ERROR(type 'exceptions.SystemError'): uncaught exception - error return without exception set As PyErr_FromNTSTATUS(status) is not the same as PyErr_SetNTSTATUS(status) If _IS_ERR is not the same then we need to have this code: if (!NT_STATUS_IS_OK(status)) { PyErr_SetNTSTATUS(status); talloc_free(mem_ctx); return NULL; } If you have no objection I'll push a fix like that. -- Matthieu Patou Samba Team http://samba.org
Re: [SCM] Samba Shared Repository - branch master updated
Am 23.09.2011 07:01, schrieb Matthieu Patou: On 22/09/2011 21:14, Stefan (metze) Metzmacher wrote: Hi Matthieu, diff --git a/source4/ntvfs/posix/python/pyxattr_tdb.c b/source4/ntvfs/posix/python/pyxattr_tdb.c index 2b28aca..a08557e 100644 --- a/source4/ntvfs/posix/python/pyxattr_tdb.c +++ b/source4/ntvfs/posix/python/pyxattr_tdb.c @@ -60,9 +60,8 @@ static PyObject *py_wrap_setxattr(PyObject *self, PyObject *args) status = push_xattr_blob_tdb_raw(eadb, mem_ctx, attribute, filename, -1, blob); if (!NT_STATUS_IS_OK(status)) { -PyErr_FromNTSTATUS(status); talloc_free(mem_ctx); -return NULL; +PyErr_NTSTATUS_IS_ERR_RAISE(status); } talloc_free(mem_ctx); Py_RETURN_NONE; @@ -91,9 +90,8 @@ static PyObject *py_wrap_getxattr(PyObject *self, PyObject *args) status = pull_xattr_blob_tdb_raw(eadb, mem_ctx, attribute, filename, -1, 100,blob); if (!NT_STATUS_IS_OK(status) || blob.length 0) { -PyErr_FromNTSTATUS(status); talloc_free(mem_ctx); -return NULL; +PyErr_NTSTATUS_IS_ERR_RAISE(status); Why this changes? What was the problem? NT_STATUS_IS_ERR() is not the same as !NT_STATUS_IS_OK() Well the problem is that without this if your status was != OK you'll get error like instead of getting the right exception. ERROR(type 'exceptions.SystemError'): uncaught exception - error return without exception set As PyErr_FromNTSTATUS(status) is not the same as PyErr_SetNTSTATUS(status) If _IS_ERR is not the same then we need to have this code: if (!NT_STATUS_IS_OK(status)) { PyErr_SetNTSTATUS(status); talloc_free(mem_ctx); return NULL; } yes, that looks more sane to me. metze signature.asc Description: OpenPGP digital signature
Re: [SCM] Samba Shared Repository - branch master updated
On Wed, Jul 06, 2011 at 03:41:02PM +0200, Stefan Metzmacher wrote: The branch, master has been updated via f9601a9 s3:sharesec: return an error if get_share_security() returns NULL from 93ad34c s3:test: catch more errors in test_access_check() in the smbclient_s3 test http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master - Log - commit f9601a91805651963834e37e27800a74931468b7 Author: Stefan Metzmacher me...@samba.org Date: Wed Jul 6 12:31:41 2011 +0200 s3:sharesec: return an error if get_share_security() returns NULL This indicates an error, as it uses get_share_security_default() if no security descriptor is configured. Jeremy, please check. +1 from me - this fixes an error condition. Thanks !
Re: [SCM] Samba Shared Repository - branch master updated
On 2010-05-25 at 07:41 +0200 Stefan (metze) Metzmacher sent off: I think we also need -qhalt=e here, to match -Werror. I'm wondering if --errwarn is the correct thing for the sun case... -qhalt=w and -errwarn=%all are the equivalent options for -Werror. Some configure checks need to bail out errors even on warnings, that's all what this is needed for. Cheers Björn
Re: [SCM] Samba Shared Repository - branch master updated
Hi Metze, On 2010-06-10 at 20:50 +0200 Stefan (metze) Metzmacher sent off: + * h_errno is defined as function call macro for multithreaded applications + * in HP-UX. *this* h_errno is not used in the HP-UX codepath of our nss + * modules, so let's simply comment it out here: + * int h_errno; + */ nss_status_t status; } nss_XbyY_args_t; Aren't this structures of fixed predefined size? Don't we use this to cast a given buffer into something we use? you may very well be right. I didn't see this - the solaris/hpux shared nss code is quite a pain to read If so args-status is wrong now and we better keep an 'int _not_used'. alright, I'll change it this way, thanks! Cheers Björn -- SerNet GmbH, Bahnhofsallee 1b, 37081 Göttingen phone: +49-551-37-0, fax: +49-551-37-9 AG Göttingen, HRB 2816, GF: Dr. Johannes Loxen pgpDGaPi6RKhX.pgp Description: PGP signature
Re: [SCM] Samba Shared Repository - branch master updated
Hi Matthias, On Sat, 2010-04-17 at 13:12 -0500, Matthias Dieter Wallnöfer wrote: - Log - commit c678563efec82e84c9c86309cbf9495950a02b8c Author: Matthias Dieter Wallnöfer mwallnoe...@yahoo.de Date: Sat Apr 17 20:08:15 2010 +0200 s4:kdc/db-glue.c - use TALLOC_FREE insteal of talloc_free for the priv context Also after a free priv could be != NULL and may be freed again. This should fix bug #7365. We tend to avoid TALLOC_FREE in Samba 4, opting for explicitly setting variable to NULL explicitly. This way valgrind will tell us about double frees. If you think we should consider using TALLOC_FREE, I think we should discuss that first - for now, please use talloc_free() and assignments to NULL. Cheers, Jelmer
Re: [SCM] Samba Shared Repository - branch master updated
On Fri, 2010-12-10 at 22:35 +0100, Matthieu Patou wrote: commit 6875a611a71bf6d1623fc62361fc8b145b1f6559 Author: Matthieu Patou m...@matws.net Date: Fri Dec 10 23:16:28 2010 +0300 build: add a function to test if -lc is needed This is needed on openbsd as some linking flags makes mandatory to specify the libc for the linking Urgh, there is something seriously broken if a C compiler doesn't link to libc by default (I can understand a -nostdlib option). Is this default OpenBSD behaviour or something specific to this host? Cheers, Jelmer
Re: [SCM] Samba Shared Repository - branch master updated
On Tue, 2011-01-04 at 05:40 +0100, Andrew Bartlett wrote: commit 640028d158583825ea5ffd1266f099cf8776db5d Author: Andrew Bartlett abart...@samba.org Date: Tue Jan 4 14:41:38 2011 +1100 nsswitch Add talloc depencency for nsstest nsstest does not use talloc, but it includes talloc.h via includes.h, and so without this it cannot find the right internal header. This wasn't noticed before, as most linux developer boxes have a talloc.h. The issue was reported by Adam Tauno Williams awilliam at whitemice.org I'm a bit puzzled by this, as e.g. sn doesn't have talloc.h installed either and neither have most of the build farm hosts. Perhaps it happens if talloc.h is installed but not in /usr/include ? Cheers, Jelmer
Re: [SCM] Samba Shared Repository - branch master updated
Hi Matthias, On Fri, 2010-05-28 at 12:08 -0500, Matthias Dieter Wallnöfer wrote: Summary of changes: source4/web_server/wsgi.c |3 ++- 1 files changed, 2 insertions(+), 1 deletions(-) Changeset truncated at 500 lines: diff --git a/source4/web_server/wsgi.c b/source4/web_server/wsgi.c index 7a23ae4..db1a5f2 100644 --- a/source4/web_server/wsgi.c +++ b/source4/web_server/wsgi.c @@ -301,7 +301,8 @@ static PyObject *create_environ(bool tls, int content_length, struct http_header PyDict_SetItemString(env, CONTENT_TYPE, PyString_FromString(hdr-value)); } else { if (asprintf(name, HTTP_%s, hdr-name) 0) { - continue; + PyErr_NoMemory(); + return NULL; } This will leak the 'env' python object. You probably also want to call Py_DECREF here. Cheers, Jelmer signature.asc Description: This is a digitally signed message part
Re: [SCM] Samba Shared Repository - branch master updated
On Mon, 2010-05-31 at 09:46 +0200, Stefan (metze) Metzmacher wrote: diff --git a/lib/talloc/wscript b/lib/talloc/wscript index 3a87506..61930f6 100644 --- a/lib/talloc/wscript +++ b/lib/talloc/wscript @@ -60,14 +60,14 @@ def build(bld): abi_file='ABI/talloc-%s.sigs' % VERSION, abi_match='talloc* _talloc*', hide_symbols=True, - vnum=VERSION) + vnum=VERSION, is_bundled=True) # should we also install the symlink to libtalloc1.so here? bld.SAMBA_LIBRARY('talloc-compat1', 'compat/talloc_compat1.c', deps='talloc', enabled = bld.env.TALLOC_COMPAT1, - vnum=VERSION) + vnum=VERSION, is_bundled=True) shouldn't we use 'not bld.env.standalone_talloc' instead of 'True' here too? That's a good point, thanks. Fixed in master. Cheers, Jelmer signature.asc Description: This is a digitally signed message part
Re: [SCM] Samba Shared Repository - branch master updated
On Mon, 2010-05-31 at 19:49 +0200, Stefan (metze) Metzmacher wrote: Am 31.05.2010 19:23, schrieb Jelmer Vernooij: The branch, master has been updated via 7f75ee0... ldb: Install ldb_handlers.h header. via fe8302b... ldb: Remove Samba-specific symbols. via 1bc53f0... ldb: Move utility functions to separate file. via 82d56b9... ldb: Fix dependencies when building with system ldb. from 471ed70... s3:smbd map_username() doesn't need sconn anymore http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master - Log - commit 7f75ee025ff9c02763fb9201d94af4b2739c8e03 Author: Jelmer Vernooij jel...@samba.org Date: Mon May 31 18:29:11 2010 +0200 ldb: Install ldb_handlers.h header. commit fe8302b235197e359a20ba0489eb72b54793963e Author: Jelmer Vernooij jel...@samba.org Date: Mon May 31 18:28:09 2010 +0200 ldb: Remove Samba-specific symbols. commit 1bc53f0213b04e30ebf2ac60695c84c5781e4e72 Author: Jelmer Vernooij jel...@samba.org Date: Mon May 31 18:12:05 2010 +0200 ldb: Move utility functions to separate file. commit 82d56b937445585b9e60451daf660aee5ace0523 Author: Jelmer Vernooij jel...@samba.org Date: Mon May 31 16:05:41 2010 +0200 ldb: Fix dependencies when building with system ldb. I'm getting this again: [1862/1965] abi_check: source4/bin/default/source4/lib/ldb/libldb-samba4.so libldb-samba4.so: symbol ldb_wrap_connect has been added - please mark it _PRIVATE_ or update minor version signature: struct ldb_context *(TALLOC_CTX *, struct tevent_context *, struct loadparm_context *, const char *, struct auth_session_info *, struct cli_credentials *, unsigned int) libldb-samba4.so: symbol ldb_register_samba_handlers has been added - please mark it _PRIVATE_ or update minor version signature: int (struct ldb_context *) libldb-samba4.so: symbol ldb_wrap_fork_hook has been added - please mark it _PRIVATE_ or update minor version signature: void (void) libldb-samba4.so: symbol ldb_samba_syntax_by_lDAPDisplayName has been added - please mark it _PRIVATE_ or update minor version signature: const struct ldb_schema_syntax *(struct ldb_context *, const char *) libldb-samba4.so: symbol ldb_samba_syntax_by_name has been added - please mark it _PRIVATE_ or update minor version signature: const struct ldb_schema_syntax *(struct ldb_context *, const char *) Waf: Leaving directory `/home/metze/master4/source4/bin' Build failed: ABI for libldb-samba4.so has changed - please fix library version then build with --abi-update See http://wiki.samba.org/index.php/Waf#ABI_Checking for more information make: *** [all] Fehler 1 Have you tried a git clean -f -x -d ? I can't reproduce the issue with that, and as far as I can tell the buildfarm is happy too. Cheers, Jelmer signature.asc Description: This is a digitally signed message part
Re: [SCM] Samba Shared Repository - branch master updated
On Sat, 2011-05-07 at 02:21 +0200, Jeremy Allison wrote: The branch, master has been updated via ac25835 Fix Samba3 on OpenIndiana. from f7b3909 A couple more off-by-one calculations with strlcpy. http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master - Log - commit ac25835ab7b76226bd59fec9ffef46d5c5817d54 Author: Gordon Ross gordon.w.r...@gmail.com Date: Fri May 6 16:00:08 2011 -0700 Fix Samba3 on OpenIndiana. I'd like Samba to use the native OpenLDAP and MIT Kerberos libs. Attached are some patches to do that. (relative to git master) It does not build for me without these. (OpenIndiana is an off-shoot of OpenSolaris See http://www.openindiana.org) Autobuild-User: Jeremy Allison j...@samba.org Autobuild-Date: Sat May 7 02:20:14 CEST 2011 on sn-devel-104 --- Summary of changes: auth/kerberos/gssapi_pac.c| 24 lib/replace/system/kerberos.h |4 +++- source3/configure.in | 26 +- source3/wscript |2 +- 4 files changed, 49 insertions(+), 7 deletions(-) Changeset truncated at 500 lines: diff --git a/auth/kerberos/gssapi_pac.c b/auth/kerberos/gssapi_pac.c index d89a649..e115cfe 100644 --- a/auth/kerberos/gssapi_pac.c +++ b/auth/kerberos/gssapi_pac.c @@ -23,6 +23,30 @@ #include libcli/auth/krb5_wrap.h +#if 0 +/* FIXME - need proper configure/waf test + * to determine if gss_mech_krb5 and friends + * exist. JRA. + */ +/* + * These are not exported by Solaris -lkrb5 + * Maybe move to libreplace somewhere? + */ +static const gss_OID_desc krb5_gss_oid_array[] = { + /* this is the official, rfc-specified OID */ + { 9, \052\206\110\206\367\022\001\002\002 }, + /* this is the pre-RFC mech OID */ + { 5, \053\005\001\005\002 }, + /* this is the unofficial, incorrect mech OID emitted by MS */ + { 9, \052\206\110\202\367\022\001\002\002 }, + { 0, 0 } +}; Jeremy something looks wrong here. The comments seem to imply the MS and the official OID should be different, (the MS being wrong), yet I see non difference between them. Simo. -- Simo Sorce Samba Team GPL Compliance Officer s...@samba.org Principal Software Engineer at Red Hat, Inc. s...@redhat.com
Re: [SCM] Samba Shared Repository - branch master updated
On Thu, Apr 21, 2011 at 07:14:02PM +0200, Günther Deschner wrote: The branch, master has been updated via 1b7cd33 s3-vfs: try to fix the Tru64 build. via bb65187 s3-proto: move more rpc client prototypes to cli_pipe.h. from 2b608f3 script: add hardcoded library release key id http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master - Log - commit 1b7cd33291ec69f58a68396cf7406a6da3083df3 Author: Günther Deschner g...@samba.org Date: Thu Apr 21 18:26:40 2011 +0200 s3-vfs: try to fix the Tru64 build. Volker, Jeremy please check. Shouldn't this be bracketed with an : #ifdef vfsops +change #endif
Re: [SCM] Samba Shared Repository - branch master updated
Hi Andrew, commit b32f155bae384bd3ae1d9ba9f246a0eca18a2511 Author: Andrew Tridgell tri...@samba.org Date: Mon Mar 28 16:44:32 2011 +1100 s4-dsdb: perform FSMO transfers asynchronously this gives the administrator a proper error message on the command line Pair-Programmed-With: Andrew Bartlett abart...@samba.org commit 358892e8365d052d7b9300737a4c1dc92b26cc0c Author: Andrew Tridgell tri...@samba.org Date: Mon Mar 28 16:20:47 2011 +1100 s4-fsmo: make rootDSE modify for FSMO transfer async this gives the ldap client the error code from the transfer Pair-Programmed-With: Andrew Bartlett abart...@samba.org Not really related to your change, but I think we have a real problem in this code, as rootdse_become_master() is called within a transaction. Which may talks to another local process and then to a remove server. metze signature.asc Description: OpenPGP digital signature
Re: [SCM] Samba Shared Repository - branch master updated
On Wed, 2011-03-09 at 09:04 +0100, Andrew Bartlett wrote: The branch, master has been updated via 9728b5a librpc/ndr use hyper for uid_t/gid_t rather than udlong from c4b52fb s3: Use talloc_tos() in idmap_nss_sids_to_unixids http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master - Log - commit 9728b5a0d5a5c09615a09a97bf217bcacb773579 Author: Andrew Bartlett abart...@samba.org Date: Wed Mar 2 19:06:15 2011 +1100 librpc/ndr use hyper for uid_t/gid_t rather than udlong This has 8 byte alignment, which is what was specified in pidl for these types. I am puzzled, uid_t and gid_t are 32 bit quantitiers on all platforms I know, do you know of platforms that use 64 bit uid/gid types ? If not, why do you use a 64 bit type for them ? Also uid_t and gid_t are signed in many platforms while here you seem to be casting to uint64_t, and that doesn't look right to me. Care to comment ? Simo. -- Simo Sorce Samba Team GPL Compliance Officer s...@samba.org Principal Software Engineer at Red Hat, Inc. s...@redhat.com
Re: [SCM] Samba Shared Repository - branch master updated
On Tue, 2011-03-01 at 07:35 +0100, Volker Lendecke wrote: On Tue, Mar 01, 2011 at 04:33:01AM +0100, Andrew Bartlett wrote: The branch, master has been updated via 5f5ca91 lib/util: new merged debug system via 4acef31 lib/util move debug.[ch] out of the way from b1f68b6 s4-libnet_vampire: Ignore some attributes when building working schema cache http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master - Log - commit 5f5ca913b7abfcf95782339fac2dc8c1541b1126 Author: Andrew Bartlett abart...@samba.org Date: Thu Feb 24 16:14:03 2011 +1100 lib/util: new merged debug system This is the s3 debug system, with a number of changes to tidy it up for common use. The debug class system is simplified by the removal of the ISSET table, the system no longer attempts to cope with assignment of DEBUGLEVEL, and the full class table is always available (rather than just DEBUGLEVEL_CLASS[DBCG_ALL]) from startup. It is also no longer confusingly described as a hack, but as the initial table. The way git diff / git whatchanged works makes it more difficult than necessary to compare the actual code changes if you make changes and move the files at the same time. Next time, can you please split those two changes into two commits? Volker, If you want to see this patch in a slightly better form you can use a command like this: git showh -M -C --patience 5f5ca913b7abfcf95782339fac2dc8c1541b1126 With this one git is told to detect the rename and does show only the differences. you can use the same arguments for generating better patches when you want to send them around for review, like this: git format-patch -M -C --patience --full-index -1 5f5ca913b7abfcf95782339fac2dc8c1541b1126 Simo. -- Simo Sorce Samba Team GPL Compliance Officer s...@samba.org Principal Software Engineer at Red Hat, Inc. s...@redhat.com
Re: [SCM] Samba Shared Repository - branch master updated
On Tue, Mar 01, 2011 at 07:55:28AM -0500, simo wrote: If you want to see this patch in a slightly better form you can use a command like this: git showh -M -C --patience 5f5ca913b7abfcf95782339fac2dc8c1541b1126 With this one git is told to detect the rename and does show only the differences. Ok, thanks. Then I rephrase my comment a bit. It would be highly desirable to make small patches to both implementations until there is no difference anymore or the diff is really trivial. That patch is almost 2000 lines, which is for my taste a LOT too much for a single patch unless there is a very good reason for this. Please in the future divide that kind of patches into much smaller chunks if you do not want to actively feed resistance against any kind of changes to Samba3. With best regards, Volker Lendecke -- SerNet GmbH, Bahnhofsallee 1b, 37081 Göttingen phone: +49-551-37-0, fax: +49-551-37-9 AG Göttingen, HRB 2816, GF: Dr. Johannes Loxen
Re: [SCM] Samba Shared Repository - branch master updated
On 02/28/2011 08:18 PM, Günther Deschner wrote: The branch, master has been updated via 26321c6 s3-torture: fix the build of rpc_open_tcp. from eece80e s3-smbd: Pass tevent context to messaging functions. http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master - Log - commit 26321c6010dbd461e85111daf358e6c23a0b47b2 Author: Günther Deschnerg...@samba.org Date: Mon Feb 28 20:18:33 2011 +0100 s3-torture: fix the build of rpc_open_tcp. Guenther Argl, this slipped to master w/o going through autobuild, sorry, was not meant to be like that :) Guenther -- Günther DeschnerGPG-ID: 8EE11688 Red Hat gdesch...@redhat.com Samba Team g...@samba.org
Re: [SCM] Samba Shared Repository - branch master updated
On Tue, Mar 01, 2011 at 04:33:01AM +0100, Andrew Bartlett wrote: The branch, master has been updated via 5f5ca91 lib/util: new merged debug system via 4acef31 lib/util move debug.[ch] out of the way from b1f68b6 s4-libnet_vampire: Ignore some attributes when building working schema cache http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master - Log - commit 5f5ca913b7abfcf95782339fac2dc8c1541b1126 Author: Andrew Bartlett abart...@samba.org Date: Thu Feb 24 16:14:03 2011 +1100 lib/util: new merged debug system This is the s3 debug system, with a number of changes to tidy it up for common use. The debug class system is simplified by the removal of the ISSET table, the system no longer attempts to cope with assignment of DEBUGLEVEL, and the full class table is always available (rather than just DEBUGLEVEL_CLASS[DBCG_ALL]) from startup. It is also no longer confusingly described as a hack, but as the initial table. The way git diff / git whatchanged works makes it more difficult than necessary to compare the actual code changes if you make changes and move the files at the same time. Next time, can you please split those two changes into two commits? Thanks, Volker -- SerNet GmbH, Bahnhofsallee 1b, 37081 Göttingen phone: +49-551-37-0, fax: +49-551-37-9 AG Göttingen, HRB 2816, GF: Dr. Johannes Loxen
Re: [SCM] Samba Shared Repository - branch master updated
On Fri, 2011-02-18 at 09:30 +0100, Andrew Bartlett wrote: - Log - commit 1ad8e5229f618fc04af371ba52b81f2e7e1f88f5 Author: Andrew Bartlett abart...@samba.org Date: Fri Feb 18 13:59:05 2011 +1100 lib/util/charset Add back setlocale(), but only when called from binaries When called from a library, we don't want to call this, as we may overwrite some of our calling program's context. What in particular re-added the need for this? I'm worried about this change as it means we won't be testing the library code that e.g. gets used as part of OpenChange and Evolution. It would be really nice if we could get rid of our reliance on the system locale for once and for all. :-/ Cheers, Jelmer commit 2a3a86a86f3d1ab97adda563beda7ee35f6a2414 Author: Andrew Bartlett abart...@samba.org Date: Fri Feb 18 13:47:28 2011 +1100 lib/util/charcnv Move iconv handle setup in common We now use the struct smb_iconv_convenience at the core of all our iconv code, and use global_iconv_convenience for the callers that don't specify one. Andrew Bartlett commit 5155a5f5c130ff5b71ce4e37877378a6967046b4 Author: Andrew Bartlett abart...@samba.org Date: Fri Feb 18 11:51:33 2011 +1100 s3-charcnv Don't genreate valid_table on the fly, rely on valid.dat This file is always installed, and is only even required for the old, depricated mangle hash method. Andrew Bartlett commit e4c20230825dd78a2f06839dd7f6fe72f486a910 Author: Andrew Bartlett abart...@samba.org Date: Fri Feb 18 10:52:12 2011 +1100 lib/util Remove #if _SAMBA_BUILD_ == 4 that isn't required any more commit 8afc271e2a9950fda68e99928746623093cbd6ac Author: Andrew Bartlett abart...@samba.org Date: Fri Feb 18 10:24:58 2011 +1100 lib/util/charset Use top level iconv.c in source3 The two files were very similar already, the only change required was to adopt the s3 module registration fucntion name. (NTSTATUS wasn't used as the charset code does not otherwise use that type). Andrew Bartlett Signed-off-by: Andrew Tridgell tri...@samba.org commit bed374215ff040cc995659d42938be35412a794a Author: Andrew Tridgell tri...@samba.org Date: Fri Feb 18 10:10:30 2011 +1100 s4-idl: rename s4 server_id.idl to server_id4.idl this avoids a conflict with the new s3 server_id.idl Pair-Programmed-With: Andrew Bartlett abart...@samba.org commit cfeceedfab64fe8b7f5823d1ada3f6d9b1799adf Author: Andrew Bartlett abart...@samba.org Date: Thu Feb 17 16:29:28 2011 +1100 librpc make ndr-standard a common library commit 177ddcb084e8f8608bf2012a7c58f7b1e56d2acd Author: Andrew Bartlett abart...@samba.org Date: Thu Feb 17 16:19:45 2011 +1100 librpc push NDR_SECURITY and NDR_DCERPC to the top level wscript files There isn't any reson to keep these private to source3 and source4 and more, and doing so creates problems for the top level build. Andrew Bartlett commit a493bf9103e5a36e0aef0d1bf1a055f602955d0f Author: Andrew Tridgell tri...@samba.org Date: Thu Feb 17 14:13:42 2011 +1100 nsswitch: depend on dl The previous merge in this area didn't preserve the 'dl' depedency, which may be required on some platforms. Pair-Programmed-With: Andrew Bartlett abart...@samba.org commit 92faeaeea841dd0bce9460d25429846ae2fdc0af Author: Andrew Bartlett abart...@samba.org Date: Wed Feb 16 16:30:56 2011 +1100 charset Remove use of {isupper,islower,toupper,tolower}_w functions These now call the common _m functions that consider UTF16 code points. This removes the code which will make up a 'lame' table in memory, as this can just as correctly be handled by running the algorithm at runtime (which is to call toupper() and tolower() on characters 128). When used, a top level waf build will always locate the correct table - in the build tree or outside - due to relinking the installed binary. Andrew Bartlett commit 0bad0e3ff2063f009557ab6ad7a442ceaed593ee Author: Andrew Bartlett abart...@samba.org Date: Tue Feb 15 16:34:02 2011 +1100 s3-libads Remove MIT-specific krb5_princ_realm macro calls. When compiled against heimdal, we need to use a more elegant API. Andrew Bartlett commit e4c2f5c474b2c294341edcce8044319a111e1e45 Author: Andrew Bartlett abart...@samba.org Date: Fri Feb 18 16:04:30 2011 +1100 s3-waf use lib/util/wscript_build, and avoid duplicate subsystems commit 83a63baecc0b49dac5547551436f9fca2595f2f7 Author: Andrew Bartlett abart...@samba.org Date: Thu Feb 17 16:39:07 2011 +1100 lib/util Make UTIL_LDB conditional on an s4 build This allows this file to be used in common. commit
Re: [SCM] Samba Shared Repository - branch master updated
On Wed, Jan 19, 2011 at 11:11:01PM +0100, Andreas Schneider wrote: commit 7a97518e12054e1f84236501322b756c5c09d7dd Author: Andreas Schneider a...@samba.org Date: Wed Jan 19 16:22:28 2011 +0100 s3-smbd: Fixed a possible null pointer dereference. index dd6af6d..b573a6c 100644 --- a/source3/smbd/uid.c +++ b/source3/smbd/uid.c @@ -229,7 +229,7 @@ void conn_clear_vuid_cache(connection_struct *conn, uint16_t vuid) bool change_to_user(connection_struct *conn, uint16 vuid) { const struct auth_serversupplied_info *server_info = NULL; - user_struct *vuser = get_valid_user_struct(conn-sconn, vuid); + user_struct *vuser; int snum; gid_t gid; uid_t uid; @@ -242,6 +242,8 @@ bool change_to_user(connection_struct *conn, uint16 vuid) return(False); } + vuser = get_valid_user_struct(conn-sconn, vuid); + /* * We need a separate check in security=share mode due to vuid * always being UID_FIELD_INVALID. If we don't do this then Andreas, can you please explain this commit *exactly*. How does leaving vuser undefined fix this ? Jeremy.
Re: mac mini build broken was Re: [SCM] Samba Shared Repository - branch master updated
Hi Love, After some digging I found the reason of the problem that affected mac minis but also openbsd, and netbsd. On this platform sys/queue.h is included by net/if.h and this include happen quite early in the include tree. But heimdal/lib/roken/roken-common.h do an #undef of SLIST_ENTRY (it's written that it's needed for WinNT) but this undef apply to all the platforms. What I did is to change the scope of this undef, so that it apply in the case where _WIN32 is defined, see changeset 533ba5a Cheers. Matthieu On 02/12/2010 23:28, Love Hörnquist Åstrand wrote: Matthieu, SLIST_ENTRY should be defined by mechqueue.h orsys/queue.h, without cpp output it hard to determine what went wrong. Love 2 dec 2010 kl. 03.45 skrev Matthieu Patou: Hi Andrews, I highly suspect the new import of heimdal to be the root cause of the breakage on mac mini can one of you look on it ? Thanks. Matthieu. On 01/12/2010 09:49, Andrew Tridgell wrote: The branch, master has been updated via b7172e7 s4-drs: cope with invalid NTDS DNs from DsReplicaInfo() via 00ecbdb wintest: cope with w2k3 form of dcdiag output via 06fd5b7 wintest Move stopping of BIND into a new step via c5bea98 s4:heimdal: import lorikeet-heimdal-201012010201 (commit 81fe27bcc0148d410ca4617f8759b9df1a5e935c) from 9c84f98 wintest: make command matching case insensitive by default http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master - Log - commit b7172e7a71b152be687fe5045565c8cd99a73a18 Author: Andrew Tridgelltri...@samba.org Date: Wed Dec 1 16:40:17 2010 +1100 s4-drs: cope with invalid NTDS DNs from DsReplicaInfo() w2k3 sometimes returns a deleted DN Autobuild-User: Andrew Tridgelltri...@samba.org Autobuild-Date: Wed Dec 1 07:48:19 CET 2010 on sn-devel-104 commit 00ecbdbbd13ba191400c6f4185df2dd8e72d1459 Author: Andrew Tridgelltri...@samba.org Date: Wed Dec 1 16:34:16 2010 +1100 wintest: cope with w2k3 form of dcdiag output commit 06fd5b70037728800cfeb2d1989ad8e851f604df Author: Andrew Bartlettabart...@samba.org Date: Wed Dec 1 15:34:19 2010 +1100 wintest Move stopping of BIND into a new step We must run this early, to ensure that BIND isn't alive to write to the zone file after provision has cleaned it up. Andrew Bartlett commit c5bea98ddb2f7967df572160f639da3cba381a87 Author: Andrew Bartlettabart...@samba.org Date: Mon Nov 29 11:24:08 2010 +1100 s4:heimdal: import lorikeet-heimdal-201012010201 (commit 81fe27bcc0148d410ca4617f8759b9df1a5e935c) --- Summary of changes: source4/heimdal/base/baselocl.h | 52 +- source4/heimdal/base/heimbase.c | 15 +- source4/heimdal/cf/make-proto.pl |1 + source4/heimdal/kdc/default_config.c |3 +- source4/heimdal/kdc/kaserver.c| 955 source4/heimdal/kdc/misc.c| 21 +- source4/heimdal/kdc/process.c | 77 -- source4/heimdal/kuser/kinit.c | 14 +- source4/heimdal/kuser/kuser_locl.h|3 + source4/heimdal/lib/asn1/asn1parse.c | 724 + source4/heimdal/lib/asn1/asn1parse.y |4 + source4/heimdal/lib/asn1/gen_template.c |4 +- source4/heimdal/lib/com_err/com_err.h | 20 - source4/heimdal/lib/com_err/com_right.h | 35 +- source4/heimdal/lib/com_err/error.c |8 +- source4/heimdal/lib/com_err/lex.c | 83 +-- source4/heimdal/lib/com_err/lex.h |2 +- source4/heimdal/lib/com_err/lex.l |5 +- source4/heimdal/lib/com_err/parse.c | 265 --- source4/heimdal/lib/com_err/parse.y |5 +- source4/heimdal/lib/gssapi/gssapi/gssapi.h| 85 ++- source4/heimdal/lib/gssapi/gssapi/gssapi_krb5.h | 73 -- source4/heimdal/lib/gssapi/gssapi/gssapi_oid.h| 231 + source4/heimdal/lib/gssapi/gssapi_mech.h | 80 ++- source4/heimdal/lib/gssapi/krb5/acquire_cred.c| 14 +- source4/heimdal/lib/gssapi/krb5/external.c| 238 +++--- source4/heimdal/lib/gssapi/krb5/set_cred_option.c |9 - source4/heimdal/lib/gssapi/mech/gss_mech_switch.c |2 +- source4/heimdal/lib/gssapi/mech/gss_mo.c | 464 ++ source4/heimdal/lib/gssapi/mech/gss_oid.c | 253 ++ source4/heimdal/lib/gssapi/mech/gss_oid_equal.c |2 +- source4/heimdal/lib/gssapi/mech/gss_oid_to_str.c | 31 + source4/heimdal/lib/gssapi/mech/gss_wrap.c| 15 + source4/heimdal/lib/gssapi/mech/mech_locl.h | 15 +- source4/heimdal/lib/gssapi/spnego/external.c | 49 +-
Re: [SCM] Samba Shared Repository - branch master updated
Hi Matthias, As discussed earlier, can you please send this kind of patch to the list for review first? I have my doubts about the usefulness of these changes in general compared to the benefit. Eliminating talloc contexts doesn't really gain us much. The overhead of creating an extra talloc context compared to the overhead of using Python in the first place is negligible. On Sun, 2010-12-12 at 20:51 +0100, Matthias Dieter Wallnöfer wrote: @@ -163,6 +165,8 @@ static PyObject *py_samdb_set_ntds_settings_dn(PyLdbObject *self, PyObject *args } if (!PyObject_AsDn(tmp_ctx, py_ntds_settings_dn, ldb, ntds_settings_dn)) { + PyErr_NoMemory(); + talloc_free(tmp_ctx); return NULL; } This is incorrect, PyObject_AsDn will already set an exception itself. The fact that it fails is not necessarily an indication of an out of memory error. @@ -248,27 +253,19 @@ static PyObject *py_dsdb_get_oid_from_attid(PyObject *self, PyObject *args) PyErr_LDB_OR_RAISE(py_ldb, ldb); - mem_ctx = talloc_new(NULL); - if (mem_ctx == NULL) { -PyErr_NoMemory(); -return NULL; - } - schema = dsdb_get_schema(ldb, NULL); if (!schema) { PyErr_SetString(PyExc_RuntimeError, Failed to find a schema from ldb \n); - talloc_free(mem_ctx); return NULL; } status = dsdb_schema_pfm_oid_from_attid(schema-prefixmap, attid, - mem_ctx, oid); + NULL, oid); PyErr_WERROR_IS_ERR_RAISE(status); ret = PyString_FromString(oid); - - talloc_free(mem_ctx); + talloc_free(discard_const_p(char, oid)); ^^ Is this really necessary? I'd rather have the extra memory context than add an extra discard_const_p. diff --git a/source4/lib/ldb-samba/pyldb.c b/source4/lib/ldb-samba/pyldb.c index e8cdb90..f198d74 100644 --- a/source4/lib/ldb-samba/pyldb.c +++ b/source4/lib/ldb-samba/pyldb.c @@ -19,10 +19,8 @@ License along with this library; if not, see http://www.gnu.org/licenses/. */ -#include Python.h -#include includes.h -#include ldb.h #include lib/ldb/pyldb.h +#include includes.h #include param/pyparam.h #include auth/credentials/pycredentials.h #include ldb_wrap.h Can you please stop reordering include files? There's a good reason Python.h is included first, it prevents warnings on some systems. What is the benefit of this sort of reordering? diff --git a/source4/lib/ldb/pyldb.c b/source4/lib/ldb/pyldb.c index 3bee9ab..44a006f 100644 --- a/source4/lib/ldb/pyldb.c +++ b/source4/lib/ldb/pyldb.c @@ -26,9 +26,6 @@ License along with this library; if not, see http://www.gnu.org/licenses/. */ -#include Python.h -#include replace.h -#include ldb_private.h #include pyldb.h /* There's no Py_ssize_t in 2.4, apparently */ Same here. diff --git a/source4/lib/ldb/pyldb.h b/source4/lib/ldb/pyldb.h index 1f4bdf7..afc8c51 100644 --- a/source4/lib/ldb/pyldb.h +++ b/source4/lib/ldb/pyldb.h @@ -28,6 +28,7 @@ #include Python.h #include talloc.h +#include ldb_private.h typedef struct { PyObject_HEAD ^^^ We can't include ldb_private.h here, it's not installed so this will break system installs of pyldb. diff --git a/source4/lib/ldb/pyldb_util.c b/source4/lib/ldb/pyldb_util.c index 3e015d0..35071f3 100644 --- a/source4/lib/ldb/pyldb_util.c +++ b/source4/lib/ldb/pyldb_util.c @@ -23,10 +23,7 @@ License along with this library; if not, see http://www.gnu.org/licenses/. */ -#include Python.h -#include replace.h #include pyldb.h -#include ldb.h static PyObject *ldb_module = NULL; See above. Python.h is included for a reason. Also, replace.h might not be necessary on your system but necessary on others (as some functionality is not provided by the OS). diff --git a/source4/libnet/py_net.c b/source4/libnet/py_net.c index 9775e24..28dee59 100644 --- a/source4/libnet/py_net.c +++ b/source4/libnet/py_net.c @@ -18,19 +18,15 @@ along with this program. If not, see http://www.gnu.org/licenses/. */ -#include Python.h +#include lib/ldb/pyldb.h ^^ We shouldn't include lib/ldb/pyldb.h directly, but always pyldb.h in case the system pyldb is being used. Cheers, Jelmer signature.asc Description: This is a digitally signed message part
Re: [SCM] Samba Shared Repository - branch master updated
Hi Jelmer, Jelmer Vernooij wrote: On Sun, 2010-12-12 at 20:51 +0100, Matthias Dieter Wallnöfer wrote: @@ -163,6 +165,8 @@ static PyObject *py_samdb_set_ntds_settings_dn(PyLdbObject *self, PyObject *args } if (!PyObject_AsDn(tmp_ctx, py_ntds_settings_dn, ldb,ntds_settings_dn)) { + PyErr_NoMemory(); + talloc_free(tmp_ctx); return NULL; } This is incorrect, PyObject_AsDn will already set an exception itself. The fact that it fails is not necessarily an indication of an out of memory error. ah - sorry. @@ -248,27 +253,19 @@ static PyObject *py_dsdb_get_oid_from_attid(PyObject *self, PyObject *args) PyErr_LDB_OR_RAISE(py_ldb, ldb); - mem_ctx = talloc_new(NULL); - if (mem_ctx == NULL) { - PyErr_NoMemory(); - return NULL; - } - schema = dsdb_get_schema(ldb, NULL); if (!schema) { PyErr_SetString(PyExc_RuntimeError, Failed to find a schema from ldb \n); - talloc_free(mem_ctx); return NULL; } status = dsdb_schema_pfm_oid_from_attid(schema-prefixmap, attid, - mem_ctx,oid); + NULL,oid); PyErr_WERROR_IS_ERR_RAISE(status); ret = PyString_FromString(oid); - - talloc_free(mem_ctx); + talloc_free(discard_const_p(char, oid)); ^^ Is this really necessary? I'd rather have the extra memory context than add an extra discard_const_p. I've really thought hard about this change - but it seems more correct to me. The problem is that the memory context isn't freed when PyERR_WERROR_IS_ERR_RAISE raises an exception. diff --git a/source4/lib/ldb-samba/pyldb.c b/source4/lib/ldb-samba/pyldb.c index e8cdb90..f198d74 100644 --- a/source4/lib/ldb-samba/pyldb.c +++ b/source4/lib/ldb-samba/pyldb.c @@ -19,10 +19,8 @@ License along with this library; if not, seehttp://www.gnu.org/licenses/. */ -#includePython.h -#include includes.h -#includeldb.h #include lib/ldb/pyldb.h +#include includes.h #include param/pyparam.h #include auth/credentials/pycredentials.h #include ldb_wrap.h Can you please stop reordering include files? There's a good reason Python.h is included first, it prevents warnings on some systems. What is the benefit of this sort of reordering? pyldb.h includes Python.h - therefore I had to do the reordering. diff --git a/source4/lib/ldb/pyldb.c b/source4/lib/ldb/pyldb.c index 3bee9ab..44a006f 100644 --- a/source4/lib/ldb/pyldb.c +++ b/source4/lib/ldb/pyldb.c @@ -26,9 +26,6 @@ License along with this library; if not, seehttp://www.gnu.org/licenses/. */ -#includePython.h -#include replace.h -#include ldb_private.h #include pyldb.h /* There's no Py_ssize_t in 2.4, apparently */ Same here. diff --git a/source4/lib/ldb/pyldb.h b/source4/lib/ldb/pyldb.h index 1f4bdf7..afc8c51 100644 --- a/source4/lib/ldb/pyldb.h +++ b/source4/lib/ldb/pyldb.h @@ -28,6 +28,7 @@ #includePython.h #includetalloc.h +#include ldb_private.h typedef struct { PyObject_HEAD ^^^ We can't include ldb_private.h here, it's not installed so this will break system installs of pyldb. Oh sorry - that wasn't clear. I will change this. diff --git a/source4/lib/ldb/pyldb_util.c b/source4/lib/ldb/pyldb_util.c index 3e015d0..35071f3 100644 --- a/source4/lib/ldb/pyldb_util.c +++ b/source4/lib/ldb/pyldb_util.c @@ -23,10 +23,7 @@ License along with this library; if not, seehttp://www.gnu.org/licenses/. */ -#includePython.h -#include replace.h #include pyldb.h -#includeldb.h static PyObject *ldb_module = NULL; See above. Python.h is included for a reason. Also, replace.h might not be necessary on your system but necessary on others (as some functionality is not provided by the OS). But let me think - we include system headers only by libreplace. So at the end it doesn't matter if replace or a system header defines it. Obviously there we have no need for a system call - otherwise the build would have broken. diff --git a/source4/libnet/py_net.c b/source4/libnet/py_net.c index 9775e24..28dee59 100644 --- a/source4/libnet/py_net.c +++ b/source4/libnet/py_net.c @@ -18,19 +18,15 @@ along with this program. If not, seehttp://www.gnu.org/licenses/. */ -#includePython.h +#include lib/ldb/pyldb.h ^^ We shouldn't include lib/ldb/pyldb.h directly, but alwayspyldb.h in case the system pyldb is being used. Could you point out a rule when to use lib/ldb/pyldb.h or pyldb.h? This isn't clear. Cheers, Matthias
Re: [SCM] Samba Shared Repository - branch master updated
On Sun, 2010-12-12 at 21:29 +0100, Matthias Dieter Wallnöfer wrote: Jelmer Vernooij wrote: On Sun, 2010-12-12 at 20:51 +0100, Matthias Dieter Wallnöfer wrote: @@ -248,27 +253,19 @@ static PyObject *py_dsdb_get_oid_from_attid(PyObject *self, PyObject *args) PyErr_LDB_OR_RAISE(py_ldb, ldb); - mem_ctx = talloc_new(NULL); - if (mem_ctx == NULL) { - PyErr_NoMemory(); - return NULL; - } - schema = dsdb_get_schema(ldb, NULL); if (!schema) { PyErr_SetString(PyExc_RuntimeError, Failed to find a schema from ldb \n); - talloc_free(mem_ctx); return NULL; } status = dsdb_schema_pfm_oid_from_attid(schema-prefixmap, attid, - mem_ctx,oid); + NULL,oid); PyErr_WERROR_IS_ERR_RAISE(status); ret = PyString_FromString(oid); - - talloc_free(mem_ctx); + talloc_free(discard_const_p(char, oid)); ^^ Is this really necessary? I'd rather have the extra memory context than add an extra discard_const_p. I've really thought hard about this change - but it seems more correct to me. The problem is that the memory context isn't freed when PyERR_WERROR_IS_ERR_RAISE raises an exception. I suspect you mean PyErr_LDB_OR_RAISE? That macro never does a return at the moment, it's just a stub for ldb = PyLdb_AsLdbContext. Even if it did return, I think we should just avoid using it, and manually check whether py_ldb is a ldb handle and talloc_free and return if it isn't. diff --git a/source4/lib/ldb-samba/pyldb.c b/source4/lib/ldb-samba/pyldb.c index e8cdb90..f198d74 100644 --- a/source4/lib/ldb-samba/pyldb.c +++ b/source4/lib/ldb-samba/pyldb.c @@ -19,10 +19,8 @@ License along with this library; if not, seehttp://www.gnu.org/licenses/. */ -#includePython.h -#include includes.h -#includeldb.h #include lib/ldb/pyldb.h +#include includes.h #include param/pyparam.h #include auth/credentials/pycredentials.h #include ldb_wrap.h Can you please stop reordering include files? There's a good reason Python.h is included first, it prevents warnings on some systems. What is the benefit of this sort of reordering? pyldb.h includes Python.h - therefore I had to do the reordering. But why did you have to change these include lines at all? I can understand removing #include lines that are not necessary but there's no need to use the smallest subset of include lines. If an include file was already processed earlier then the overhead of including it again is minimal. diff --git a/source4/lib/ldb/pyldb_util.c b/source4/lib/ldb/pyldb_util.c index 3e015d0..35071f3 100644 --- a/source4/lib/ldb/pyldb_util.c +++ b/source4/lib/ldb/pyldb_util.c @@ -23,10 +23,7 @@ License along with this library; if not, seehttp://www.gnu.org/licenses/. */ -#includePython.h -#include replace.h #include pyldb.h -#includeldb.h static PyObject *ldb_module = NULL; See above. Python.h is included for a reason. Also, replace.h might not be necessary on your system but necessary on others (as some functionality is not provided by the OS). But let me think - we include system headers only by libreplace. So at the end it doesn't matter if replace or a system header defines it. Obviously there we have no need for a system call - otherwise the build would have broken. replace.h isn't the only way in which we get system headers, e.g. Python.h also includes a bunch - at least stdlib.h, unistd.h, stddef.h, string.h, stdio.h, limits.h and assert.h. diff --git a/source4/libnet/py_net.c b/source4/libnet/py_net.c index 9775e24..28dee59 100644 --- a/source4/libnet/py_net.c +++ b/source4/libnet/py_net.c @@ -18,19 +18,15 @@ along with this program. If not, seehttp://www.gnu.org/licenses/. */ -#includePython.h +#include lib/ldb/pyldb.h ^^ We shouldn't include lib/ldb/pyldb.h directly, but alwayspyldb.h in case the system pyldb is being used. Could you point out a rule when to use lib/ldb/pyldb.h or pyldb.h? This isn't clear. Always use pyldb.h, except in lib/ldb/ itself. Cheers, Jelmer signature.asc Description: This is a digitally signed message part
Re: [SCM] Samba Shared Repository - branch master updated
Hi Jelmer, Jelmer Vernooij wrote: On Sun, 2010-12-12 at 21:29 +0100, Matthias Dieter Wallnöfer wrote: Jelmer Vernooij wrote: On Sun, 2010-12-12 at 20:51 +0100, Matthias Dieter Wallnöfer wrote: @@ -248,27 +253,19 @@ static PyObject *py_dsdb_get_oid_from_attid(PyObject *self, PyObject *args) PyErr_LDB_OR_RAISE(py_ldb, ldb); - mem_ctx = talloc_new(NULL); - if (mem_ctx == NULL) { - PyErr_NoMemory(); - return NULL; - } - schema = dsdb_get_schema(ldb, NULL); if (!schema) { PyErr_SetString(PyExc_RuntimeError, Failed to find a schema from ldb \n); - talloc_free(mem_ctx); return NULL; } status = dsdb_schema_pfm_oid_from_attid(schema-prefixmap, attid, - mem_ctx,oid); + NULL,oid); PyErr_WERROR_IS_ERR_RAISE(status); ret = PyString_FromString(oid); - - talloc_free(mem_ctx); + talloc_free(discard_const_p(char, oid)); ^^ Is this really necessary? I'd rather have the extra memory context than add an extra discard_const_p. I've really thought hard about this change - but it seems more correct to me. The problem is that the memory context isn't freed when PyERR_WERROR_IS_ERR_RAISE raises an exception. I suspect you mean PyErr_LDB_OR_RAISE? That macro never does a return at the moment, it's just a stub for ldb = PyLdb_AsLdbContext. Even if it did return, I think we should just avoid using it, and manually check whether py_ldb is a ldb handle and talloc_free and return if it isn't. No Jelmer, I mean the macro below dsdb_schema_pfm_oid_from_attid. When that one raises an exception, what does succeed to tmp_ctx? I imagine that it never will be freed - therefore I've provided this fix. diff --git a/source4/lib/ldb-samba/pyldb.c b/source4/lib/ldb-samba/pyldb.c index e8cdb90..f198d74 100644 --- a/source4/lib/ldb-samba/pyldb.c +++ b/source4/lib/ldb-samba/pyldb.c @@ -19,10 +19,8 @@ License along with this library; if not, seehttp://www.gnu.org/licenses/. */ -#includePython.h -#include includes.h -#includeldb.h #include lib/ldb/pyldb.h +#include includes.h #include param/pyparam.h #include auth/credentials/pycredentials.h #include ldb_wrap.h Can you please stop reordering include files? There's a good reason Python.h is included first, it prevents warnings on some systems. What is the benefit of this sort of reordering? pyldb.h includesPython.h - therefore I had to do the reordering. But why did you have to change these include lines at all? I can understand removing #include lines that are not necessary but there's no need to use the smallest subset of include lines. If an include file was already processed earlier then the overhead of including it again is minimal. Sorry, this was a mistake - the previous behaviour was better. I will push a revert fix. diff --git a/source4/lib/ldb/pyldb_util.c b/source4/lib/ldb/pyldb_util.c index 3e015d0..35071f3 100644 --- a/source4/lib/ldb/pyldb_util.c +++ b/source4/lib/ldb/pyldb_util.c @@ -23,10 +23,7 @@ License along with this library; if not, seehttp://www.gnu.org/licenses/. */ -#includePython.h -#include replace.h #include pyldb.h -#includeldb.h static PyObject *ldb_module = NULL; See above. Python.h is included for a reason. Also, replace.h might not be necessary on your system but necessary on others (as some functionality is not provided by the OS). But let me think - we include system headers only by libreplace. So at the end it doesn't matter if replace or a system header defines it. Obviously there we have no need for a system call - otherwise the build would have broken. replace.h isn't the only way in which we get system headers, e.g. Python.h also includes a bunch - at least stdlib.h, unistd.h, stddef.h, string.h, stdio.h, limits.h and assert.h. In this special case we don't need the replace.h anymore since I do now include ldb_private.h (which itself includes it). diff --git a/source4/libnet/py_net.c b/source4/libnet/py_net.c index 9775e24..28dee59 100644 --- a/source4/libnet/py_net.c +++ b/source4/libnet/py_net.c @@ -18,19 +18,15 @@ along with this program. If not, seehttp://www.gnu.org/licenses/. */ -#includePython.h +#include lib/ldb/pyldb.h ^^ We shouldn't include lib/ldb/pyldb.h directly, but alwayspyldb.h in case the system pyldb is being used. Could you point out a rule when to use lib/ldb/pyldb.h orpyldb.h? This isn't clear. Always usepyldb.h, except in lib/ldb/ itself. Okay, I will fix this. PS: Could you please start the implementation of the other message attribute-mapping function in pyldb? Cheers, Matthias
Re: [SCM] Samba Shared Repository - branch master updated
Hi Matthias, On Sun, 2010-12-12 at 22:22 +0100, Matthias Dieter Wallnöfer wrote: Jelmer Vernooij wrote: On Sun, 2010-12-12 at 21:29 +0100, Matthias Dieter Wallnöfer wrote: Jelmer Vernooij wrote: On Sun, 2010-12-12 at 20:51 +0100, Matthias Dieter Wallnöfer wrote: @@ -248,27 +253,19 @@ static PyObject *py_dsdb_get_oid_from_attid(PyObject *self, PyObject *args) PyErr_LDB_OR_RAISE(py_ldb, ldb); -mem_ctx = talloc_new(NULL); -if (mem_ctx == NULL) { - PyErr_NoMemory(); - return NULL; -} - schema = dsdb_get_schema(ldb, NULL); if (!schema) { PyErr_SetString(PyExc_RuntimeError, Failed to find a schema from ldb \n); -talloc_free(mem_ctx); return NULL; } status = dsdb_schema_pfm_oid_from_attid(schema-prefixmap, attid, -mem_ctx,oid); +NULL,oid); PyErr_WERROR_IS_ERR_RAISE(status); ret = PyString_FromString(oid); - -talloc_free(mem_ctx); +talloc_free(discard_const_p(char, oid)); ^^ Is this really necessary? I'd rather have the extra memory context than add an extra discard_const_p. I've really thought hard about this change - but it seems more correct to me. The problem is that the memory context isn't freed when PyERR_WERROR_IS_ERR_RAISE raises an exception. I suspect you mean PyErr_LDB_OR_RAISE? That macro never does a return at the moment, it's just a stub for ldb = PyLdb_AsLdbContext. Even if it did return, I think we should just avoid using it, and manually check whether py_ldb is a ldb handle and talloc_free and return if it isn't. No Jelmer, I mean the macro below dsdb_schema_pfm_oid_from_attid. When that one raises an exception, what does succeed to tmp_ctx? I imagine that it never will be freed - therefore I've provided this fix. The same goes for that macro - if it doesn't deal with proper free'ing, then why not avoid it rather than rewrite the rest of the function that uses it? discard_const_p is bad, and we should avoid it unless we really can. diff --git a/source4/lib/ldb/pyldb_util.c b/source4/lib/ldb/pyldb_util.c index 3e015d0..35071f3 100644 --- a/source4/lib/ldb/pyldb_util.c +++ b/source4/lib/ldb/pyldb_util.c @@ -23,10 +23,7 @@ License along with this library; if not, seehttp://www.gnu.org/licenses/. */ -#includePython.h -#include replace.h #include pyldb.h -#includeldb.h static PyObject *ldb_module = NULL; See above. Python.h is included for a reason. Also, replace.h might not be necessary on your system but necessary on others (as some functionality is not provided by the OS). But let me think - we include system headers only by libreplace. So at the end it doesn't matter if replace or a system header defines it. Obviously there we have no need for a system call - otherwise the build would have broken. replace.h isn't the only way in which we get system headers, e.g. Python.h also includes a bunch - at least stdlib.h, unistd.h, stddef.h, string.h, stdio.h, limits.h and assert.h. In this special case we don't need the replace.h anymore since I do now include ldb_private.h (which itself includes it). Is there any particular reason why pyldb_util.c requires ldb_private.h ? We should avoid including it if we can (and thus avoid tying pyldb_util to a specific version of ldb). PS: Could you please start the implementation of the other message attribute-mapping function in pyldb? Sorry, that's on my todo list as are several other things. I'd be happy to review a patch that adds it though. Cheers, Jelmer signature.asc Description: This is a digitally signed message part
Re: [SCM] Samba Shared Repository - branch master updated
Hi Jelmer, Jelmer Vernooij wrote: The same goes for that macro - if it doesn't deal with proper free'ing, then why not avoid it rather than rewrite the rest of the function that uses it? discard_const_p is bad, and we should avoid it unless we really can. well, I agree that discard_const_p isn't so nice. But I think the right solution (if we make use of memory contexts) would be to never derive them from NULL, but from a self instance or kind of this or if not possible (static) the LDB context. diff --git a/source4/lib/ldb/pyldb_util.c b/source4/lib/ldb/pyldb_util.c index 3e015d0..35071f3 100644 --- a/source4/lib/ldb/pyldb_util.c +++ b/source4/lib/ldb/pyldb_util.c @@ -23,10 +23,7 @@ License along with this library; if not, seehttp://www.gnu.org/licenses/. */ -#includePython.h -#include replace.h #include pyldb.h -#includeldb.h static PyObject *ldb_module = NULL; See above. Python.h is included for a reason. Also, replace.h might not be necessary on your system but necessary on others (as some functionality is not provided by the OS). But let me think - we include system headers only by libreplace. So at the end it doesn't matter if replace or a system header defines it. Obviously there we have no need for a system call - otherwise the build would have broken. replace.h isn't the only way in which we get system headers, e.g. Python.h also includes a bunch - at least stdlib.h, unistd.h, stddef.h, string.h, stdio.h, limits.h and assert.h. In this special case we don't need the replace.h anymore since I do now include ldb_private.h (which itself includes it). Is there any particular reason why pyldb_util.c requires ldb_private.h ? We should avoid including it if we can (and thus avoid tying pyldb_util to a specific version of ldb). Okay. PS: Could you please start the implementation of the other message attribute-mapping function in pyldb? Sorry, that's on my todo list as are several other things. I'd be happy to review a patch that adds it though. Okay. Cheers, Matthias
Re: [SCM] Samba Shared Repository - branch master updated
On Sun, 2010-12-12 at 22:50 +0100, Matthias Dieter Wallnöfer wrote: Hi Jelmer, Jelmer Vernooij wrote: The same goes for that macro - if it doesn't deal with proper free'ing, then why not avoid it rather than rewrite the rest of the function that uses it? discard_const_p is bad, and we should avoid it unless we really can. well, I agree that discard_const_p isn't so nice. But I think the right solution (if we make use of memory contexts) would be to never derive them from NULL, but from a self instance or kind of this or if not possible (static) the LDB context. Even if we do that then it should be as a safeguard, not as an excuse not to call talloc_free(). The advantage of hanging them off NULL is that they can more easily be found using valgrind. If we don't free these kinds of contexts explicitly and they are allocated as children of e.g. the ldb context then that means they will stay around for as long as the ldb context stays around, which can be a very long time. So in some cases this will help contain the effect of the leak but it's by no means an alternative. Cheers, Jelmer signature.asc Description: This is a digitally signed message part
Re: mac mini build broken was Re: [SCM] Samba Shared Repository - branch master updated
Hi Love, Thanks to help me to fix this So the best that I can do so far is to provide you with make V=1 output: [1686/2283] Compiling heimdal/lib/gssapi/mech/context.c 10:03:45 runner ccache gcc -fPIC -fno-common -DSTATIC_gssapi_MODULES=NULL -DSTATIC_gssapi_MODULES_PROTO -MD -Idefault/source4/heimdal_build -I../heimdal_build -Idefault/source4/heimdal/lib/gssapi -I../heimdal/lib/gssapi -Idefault/source4/heimdal/lib/gssapi/gssapi -I../heimdal/lib/gssapi/gssapi -Idefault/source4/heimdal/lib/gssapi/spnego -I../heimdal/lib/gssapi/spnego -Idefault/source4/heimdal/lib/gssapi/krb5 -I../heimdal/lib/gssapi/krb5 -Idefault/source4/heimdal/lib/gssapi/mech -I../heimdal/lib/gssapi/mech -Idefault/source4 -I.. -Idefault/lib -I../../lib -Idefault/source4/lib -I../lib -Idefault/source4/include -I../include -Idefault -I../.. -Idefault/lib/socket_wrapper -I../../lib/socket_wrapper -Idefault/source4/heimdal/lib/krb5 -I../heimdal/lib/krb5 -Idefault/source4/heimdal/lib/asn1 -I../heimdal/lib/asn1 -Idefault/source4/heimdal/include -I../heimdal/include -Idefault/lib/replace -I../../lib/replace -Idefault/source4/heimdal/lib/roken -I../heimdal/lib/roken -Idefault/source4/heimdal/lib/wind -I../heimdal/lib/wind -Idefault/source4/heimdal/lib/hcrypto/libtommath -I../heimdal/lib/hcrypto/libtommath -Idefault/lib/util/charset -I../../lib/util/charset -Idefault/lib/nss_wrapper -I../../lib/nss_wrapper -Idefault/lib/talloc -I../../lib/talloc -Idefault/source4/heimdal/lib/hcrypto -I../heimdal/lib/hcrypto -Idefault/source4/heimdal/lib -I../heimdal/lib -Idefault/source4/heimdal/lib/com_err -I../heimdal/lib/com_err -Idefault/source4/heimdal/lib/hx509 -I../heimdal/lib/hx509 -Idefault/source4/param -I../param -Idefault/lib/uid_wrapper -I../../lib/uid_wrapper -Idefault/lib/crypto -I../../lib/crypto -Idefault/source4/dynconfig -I../dynconfig -Idefault/source4/heimdal/base -I../heimdal/base -D_SAMBA_BUILD_=4 -DHAVE_CONFIG_H=1 -D_GNU_SOURCE=1 -D_XOPEN_SOURCE_EXTENDED=1 ../heimdal/lib/gssapi/mech/context.c -c -o default/source4/heimdal/lib/gssapi/mech/context_66.o In file included from ../heimdal/lib/gssapi/mech/mech_locl.h:62, from ../heimdal/lib/gssapi/mech/context.c:2: ../heimdal/lib/gssapi/mech/cred.h:31: error: parse error before 'SLIST_ENTRY' ../heimdal/lib/gssapi/mech/cred.h:31: warning: no semicolon at end of struct or union ../heimdal/lib/gssapi/mech/cred.h:35: error: parse error before '}' token In file included from ../heimdal/lib/gssapi/mech/mech_locl.h:63, from ../heimdal/lib/gssapi/mech/context.c:2: ../heimdal/lib/gssapi/mech/mech_switch.h:33: error: parse error before 'SLIST_ENTRY' ../heimdal/lib/gssapi/mech/mech_switch.h:33: warning: no semicolon at end of struct or union ../heimdal/lib/gssapi/mech/mech_switch.h:38: error: parse error before '}' token In file included from ../heimdal/lib/gssapi/mech/mech_locl.h:64, from ../heimdal/lib/gssapi/mech/context.c:2: ../heimdal/lib/gssapi/mech/name.h:31: error: parse error before 'SLIST_ENTRY' ../heimdal/lib/gssapi/mech/name.h:31: warning: no semicolon at end of struct or union ../heimdal/lib/gssapi/mech/name.h:35: error: parse error before '}' token Waf: Leaving directory `/Users/build/build_farm/samba_4_0_test/source4/bin' Build failed: - task failed (err #1): {task: cc context.c - context_66.o} gmake: *** [all] Error 1 Hope this will be helpfull. Matthieu. On 02/12/2010 23:28, Love Hörnquist Åstrand wrote: Matthieu, SLIST_ENTRY should be defined by mechqueue.h orsys/queue.h, without cpp output it hard to determine what went wrong. Love 2 dec 2010 kl. 03.45 skrev Matthieu Patou: Hi Andrews, I highly suspect the new import of heimdal to be the root cause of the breakage on mac mini can one of you look on it ? Thanks. Matthieu. On 01/12/2010 09:49, Andrew Tridgell wrote: The branch, master has been updated via b7172e7 s4-drs: cope with invalid NTDS DNs from DsReplicaInfo() via 00ecbdb wintest: cope with w2k3 form of dcdiag output via 06fd5b7 wintest Move stopping of BIND into a new step via c5bea98 s4:heimdal: import lorikeet-heimdal-201012010201 (commit 81fe27bcc0148d410ca4617f8759b9df1a5e935c) from 9c84f98 wintest: make command matching case insensitive by default http://gitweb.samba.org/?p=samba.git;a=shortlog;h=master - Log - commit b7172e7a71b152be687fe5045565c8cd99a73a18 Author: Andrew Tridgelltri...@samba.org Date: Wed Dec 1 16:40:17 2010 +1100 s4-drs: cope with invalid NTDS DNs from DsReplicaInfo() w2k3 sometimes returns a deleted DN Autobuild-User: Andrew Tridgelltri...@samba.org Autobuild-Date: Wed Dec 1 07:48:19 CET 2010 on sn-devel-104 commit 00ecbdbbd13ba191400c6f4185df2dd8e72d1459 Author: Andrew Tridgelltri...@samba.org Date: Wed Dec 1 16:34:16 2010 +1100 wintest: cope with w2k3 form of dcdiag output commit