summaryrefslogtreecommitdiff
path: root/source4/torture/raw
AgeCommit message (Collapse)AuthorFilesLines
2007-10-10r22670: changed the RAW-NOTIFY test to support clustered testing (two nodes)Andrew Tridgell1-14/+25
(This used to be commit 12ce85aea16379b9946851806c1782272ee1c48f)
2007-10-10r22579: disable progress printing in the build-farmStefan Metzmacher4-14/+30
metze (This used to be commit 93089ad5e8b6e20c4fa92bf13b0137765aeac689)
2007-10-10r22189: fix compiler warningStefan Metzmacher1-1/+1
metze (This used to be commit 62faa4bce44c438a04e44a9189290333ebee9aa5)
2007-10-10r22087: added a RAW-BENCH-OPEN test that can be used to stress out a clusteredAndrew Tridgell2-0/+249
implementation of a opendb backend. Use it in combination with --unclist to tell smbtorture to connect to multiple nodes in the cluster (This used to be commit b4e38b1b8dbc3ce817f858f39c3f0521c9832fce)
2007-10-10r21710: Add client support for the UNIX_INFO2 info level in the QueryFile,James Peach1-1/+1
QueryPath and FindFirst calls. Add a new torture test to verify the server side. (This used to be commit 7f56da2d1fa0718e5282bb4aea7d9a63a62f0bc7)
2007-10-10r21707: Finally merge my (long-living) perlselftest branch.Jelmer Vernooij1-79/+80
This changes the main selftest code to be in perl rather than in shell script. The selftest script is now no longer a black box but a regular executable that takes --help. This adds the following features: * "make test TESTS=foo" will run only the tests that match the regex "foo" * ability to deal with expected failures. the suite will not warn about tests that fail and are known to fail, but will warn about other failing tests and tests that are succeeding tests but incorrectly marked as failing. * ability to print a summary with all failures at the end of the run It also opens up the way to the following features, which I hope to implement later: * "environments", for example having a complete domains with DCs and domain members in a testenvironment * only set up smbd if necessary (not when running LOCAL tests, for example) * different mktestsetup scripts per target. except for the mktestsetup script, we can use the same infrastructure for samba 3 or windows. (This used to be commit 38f867880beb40c691e9713f854426031310629c)
2007-10-10r21216: fail the RAW-BENCH-LOCK test if the locking doesn't happen evenlyAndrew Tridgell1-1/+10
across the connections (This used to be commit 01ef699981f3bcbbbdd4a47d9b85324aec783451)
2007-10-10r21215: extend the RAW-MUX test to test cancelling a async lock request usingAndrew Tridgell1-1/+29
a SMBexit (This used to be commit 753a85eff6a19af2c7b0884f705e9f66b5f605fe)
2007-10-10r21139: make the RAW-MUX test fail if the locking subsystem doesn't triggerAndrew Tridgell1-0/+8
early byte range lock retries (ie. PENDING locks in brlock) (This used to be commit 09237e3acf524f7e772acce2e652ef30b8ba4d65)
2007-10-10r21125: Tridge, please check this closely if this relaxed notify test isVolker Lendecke1-21/+55
acceptable. In Samba3 with inotify we have the problem that the MODIFIED action is sent second. I think this could also happen in Samba4, depending on the order when the different sockets (unix domain dgram / inotify) are looked at by the events system. This patch accepts the NEW/OLD/MODIFIED entries in any order. Comments welcome! Volker (This used to be commit 72a7771b757c3aca637eb632abebab60bdd2da9f)
2007-10-10r21124: Samba3 does not support create times correctly.Volker Lendecke1-7/+15
Volker (This used to be commit c9886ce63b8cc085e0032ad2f6193c547336856d)
2007-10-10r20886: fixed a valgrind errorAndrew Tridgell1-0/+1
(This used to be commit acc96e42e74005f1c381b0ab3f6c1b2e051ffaec)
2007-10-10r20758: Adapt RAW-UNLINK error returns for wildcard unlinks to Samba3.Volker Lendecke1-4/+33
If someone is feeling like solving that puzzle, contact Jeremy or me for hints :-) Volker (This used to be commit 93e6dbaf5b81661d85539877fd0874480cfa5673)
2007-10-10r20650: revert a bunch of code I didn't mean to commit yetAndrew Tridgell1-1/+1
(This used to be commit b3e2d4908781781a487eaeb683d22eb967e5597d)
2007-10-10r20647: add cluster codeAndrew Tridgell1-1/+1
(This used to be commit 5870830b99a8d76bda1ff5af3fcf8dda9aba50ec)
2007-10-10r20632: The extended RAW-SAMBA3BADPATH test led me to some wrong assumptions, inVolker Lendecke1-261/+2
particular the NT_STATUS_INVALID_PARAMETER thing was badly wrong. Remove the changes based on it. Using gentest is much more effective in this respect, but it will take a while to figure out the wildcard error handling of W2k3. Volker (This used to be commit 33d842e27d7611dd1d3cbfa71d44d63c273ba785)
2007-10-10r20618: Fix a bug in bad_path handling that also exists in 3.0.23: For ↵Volker Lendecke1-5/+16
reply_unlink under Linux we returned NT_STATUS_NOT_A_DIRECTORY. This is because in the bad_path==True condition lstat(2) returns ENOTDIR and not ENOENT. Not sure if we want to necessarily replicate the INVALID_PARAMETER here, but this is what W2k3 does. Jeremy, I tried to call you, but you were not around. So I'll leave it up to you to merge this. Volker (This used to be commit 838afa8f7d2b10460725c1f0b519ce54fb84de88)
2007-10-10r20613: The individual tests are fine, I did run samba3badpath, but ↵Volker Lendecke1-0/+3
raw-samba3hide left test.txt with permissions 000 behind... Volker (This used to be commit fdd501f8032dceb05989c582132b036bb22e6fb1)
2007-10-10r20609: Additional torture test for our bad_path handling. Most of it is ↵Volker Lendecke1-2/+210
disabled for now, there are calls (for example trans2ffirst) which are quite dubious. Volker (This used to be commit 79846f09b98965af0817b362f45d11e69e23c3e4)
2007-10-10r20595: W2k and XP don't always show the top directory as hidden, w2k3 does. ↵Volker Lendecke1-17/+23
I think Samba should follow w2k3 in this respect. Volker (This used to be commit 6094ee9df5820cd26858e9c4eff5dfc7e60181ab)
2007-10-10r20552: Little Samba3 test to force smb_close to return an error. Set delete ↵Volker Lendecke2-1/+72
on close, and then remove perms from the parent dir.... Volker (This used to be commit f24c5052576d4951738f83c3b238d2c251d4553b)
2007-10-10r20549: obvious bug fix..Volker Lendecke1-1/+1
(This used to be commit f0aa63e39050c7ac10b9d51c5cecc753b257be9a)
2007-10-10r20508: Confirm a special case in samba3 reply_getatr that getatr("") always ↵Volker Lendecke1-0/+34
returns HIDDEN. Volker (This used to be commit 83a20912061c663ff422b038b41f01a9c7583afe)
2007-10-10r20497: Two changes to RAW-OPEN that Samba does not fulfil:Volker Lendecke1-5/+25
Samba3 returns the SPARSE flag for files with size > allocation, and on many configs Samba3 does not support EAs. Cope with both. Tridge, please check this is acceptable for this test. Volker (This used to be commit 4df4d37c3d4d9e6c430afd36e88684f02c73700e)
2007-10-10r20480: In Samba3 we don't yet have support for all 4 file times on all ↵Volker Lendecke1-1/+9
platforms (do we have any?? ;-), so don't mark failure to correctly cope with the close write time as critical. Volker (This used to be commit 42070b09cba8adda1e7606e08764b1f202b33321)
2007-10-10r20474: Skip the delete-on-close test on a non-empty directory in raw-unlink forVolker Lendecke1-21/+28
Samba3. This is a known deficiency also taken care of in the base-delete test. Volker (This used to be commit 6b91bb93fad84e054c19f8c9bc4145234f171a3c)
2007-10-10r20311: Check one more samba3 error path for t2mkdirVolker Lendecke1-0/+6
(This used to be commit 0f018cca6f70cb44ab436b3a66edb01e57351eaf)
2007-10-10r20309: Don't treat non-supported EAs as fatal against Samba3.Volker Lendecke1-10/+20
Volker (This used to be commit 52d9c7ed0acbc286151117bb876559e13981fd44)
2007-10-10r20092: added a locking benchmark that should be good for benchmarkingAndrew Tridgell2-2/+199
communitcation in a CIFS clustered server. It tries to keep the connections full by setting up the next lock as each lock is done. The locking pattern is similar to the local filesystem ping pong test in junkcode, forcing a communication between nodes on each request (This used to be commit d57b9fb29860bd03cfa970bcc52ef45d17775638)
2007-10-10r20078: support the large samba3 reads in RAW-READ testAndrew Tridgell1-4/+14
(This used to be commit a4c2a575faf6a20174730524a8b2878eb9fdb50d)
2007-10-10r20059: james was asking about CAP_LARGE_READX and reads close to the 64kAndrew Tridgell1-0/+20
boundary. This test shows that both for Samba4 and w2k3, reads larger than or equal to 64k don't work (return zero bytes). Not that I claim this is very useful behaviour :-) (This used to be commit a10fa12ff5657a96e41b309558786780f294a75f)
2007-10-10r19510: Make smbtorture's `nprocs' parameter a parametric option.Jelmer Vernooij1-1/+1
(This used to be commit f5d6b3ca29788d171d53822a26295ae895a5ed8f)
2007-10-10r19507: Merge my DSO fixes branch. Building Samba's libraries as shared ↵Jelmer Vernooij1-1/+1
libraries works again now, by specifying --enable-dso to configure. (This used to be commit 7a01235067a4800b07b8919a6a475954bfb0b04c)
2007-10-10r19506: fixed remaining parts of RAW-READ and RAW-WRITE tests to check forAndrew Tridgell2-37/+67
CAP_LARGE_FILES (This used to be commit 66bebb210550793d0e7fe1b6a03a5140455ae72f)
2007-10-10r19504: - fixed a free error in file_lines_parseAndrew Tridgell2-0/+10
- added a function to test for large file support - enable CAP_LARGE_FILES only if the test passes - don't test at large offsets if the server doesn't set CAP_LARGE_FILES (This used to be commit c5423ea22bdaa055807d4e6f7ac4be934194fe45)
2007-10-10r19392: Use torture_setting_* rather than lp_parm_* where possible.Jelmer Vernooij5-8/+8
(This used to be commit b28860978fe29c5b10abfb8c59d7182864e21dd6)
2007-10-10r19339: Merge my 4.0-unittest branch. This adds an API for more fine-grainedJelmer Vernooij1-29/+36
output in the testsuite rather than just True or False for a set of tests. The aim is to use this for: * known failure lists (run all tests and detect tests that started working or started failing). This would allow us to get rid of the RPC-SAMBA3-* tests * nicer torture output * simplification of the testsuite system * compatibility with other unit testing systems * easier usage of smbtorture (being able to run one test and automatically set up the environment for that) This is still a work-in-progress; expect more updates over the next couple of days. (This used to be commit 0eb6097305776325c75081356309115f445a7218)
2007-10-10r18585: 50 open searches is still too many for some of the build farm hosts,Andrew Tridgell1-1/+1
which have a ulimit -n of 100. Drop the number of searches to 20. That should limit the maximum total fd count to about 76, most of which is all the listening sockets on the various interfaces (This used to be commit fb5b8c4f8c1ed69e3a813befeeb2a9627e84c454)
2007-10-10r18541: with 100 old style searches we can run out of file descriptors on someAndrew Tridgell1-1/+1
systems. drop to 50 (This used to be commit ce5d5a187264af2b87e2919bdcc5f478cce02df9)
2007-10-10r18371: made the directory depth notify test independent of timingAndrew Tridgell1-11/+29
(This used to be commit 39458e62f576c551df016a7be4d35175dc21696e)
2007-10-10r18355: increase the delays waiting for all events to propogate in theAndrew Tridgell1-4/+4
RAW-NOTIFY test. (This used to be commit 81fd19ba3f38ac19d8bcac56210a83fdfc8237de)
2007-10-10r18320: the raw write size limit test hurts too many hosts. Limit it to 2^33Andrew Tridgell1-1/+7
bytes if --dangerous is not set. (This used to be commit 90e5136ea4a47ecb59426fe547ae92db1fd8657a)
2007-10-10r18319: fixed the directory search resume code on IRIXAndrew Tridgell1-5/+5
The problem was twofold: 1) irix returns 64 bit numbers in telldir(). The protocol uses a 32 bit resume key. We now cope with this properly using the code in pvfs_list_seek_ofs(). 2) irix returns 0xFFFFFFFF from telldir() for the last entry in the directory. When added to DIR_OFFSET_BASE this became DIR_OFFSET_DOTDOT which meant an infinite loop! (This used to be commit 8cce9740ed0da9f08d6821beb4acaa9d28d149c2)
2007-10-10r18301: I discovered how to load the warnings from a build farm build intoAndrew Tridgell3-8/+8
emacs compile mode (hint, paste to a file, and compile as "cat filename"). This allowed me to fix nearly all the warnings for a IA_64 SuSE build very quickly. (This used to be commit eba6c84efff735bb0ca941ac4b755ce2b0591667)
2007-10-10r18103: add torture test for 'RAW_OPEN_OPENX without a leading slash on the ↵Derrell Lipman1-0/+46
path', known to fail on NetApp filers (This used to be commit 9a8c7b0a7673c36a8a48a28ea7ea21caf0c521fa)
2007-10-10r17785: fix compiler warningsStefan Metzmacher2-9/+6
metze (This used to be commit 256fc452a46dfe7449c6ebddb75e0801ceae5dc4)
2007-10-10r17141: Prove that create dispositions with truncate andJeremy Allison1-0/+71
overwrite break oplocks. Jeremy. (This used to be commit 94886f8c4029a9661a0bce2f7d9c41a3a5eaa366)
2007-10-10r17139: Checking in Volker's extended oplock test.Jeremy Allison1-86/+96
Jeremy (This used to be commit 720c425af167373aaedf739d1d7794ada3fcddf6)
2007-10-10r17099: Samba3 now passes RAW-LOCK completely, no need forJeremy Allison1-10/+0
parameters to protect it. Jeremy. (This used to be commit 382c37e3d70c4d259c400dba3d7e4c3846f66c34)
2007-10-10r17020: pass the real error to the failing requestsStefan Metzmacher1-2/+2
metze (This used to be commit 49b96ac44a883c020c69df7a12df154dc4faa4d5)