summaryrefslogtreecommitdiff
path: root/source3/smbd
AgeCommit message (Collapse)AuthorFilesLines
2007-10-10r21279: Get rid of 'aio write behind', this is broken.Volker Lendecke6-175/+2
It should probably better be integrated with our write cache. Volker (This used to be commit 58bfd168b046a97a895aaa3384fd7af8d077a1d5)
2007-10-10r21278: The main goal of this was to get rid of the NetInBuffer / ↵Volker Lendecke2-79/+15
set_InBuffer. But it turns out that this patch actually speeds up the async writes considerably. I tested writing 100.000 times 65535 bytes with the allowed 10 ops in parallel. Without this patch it took about 32 seconds on my dual-core 1.6GHz laptop. With this patch it dropped to about 26 seconds. I can only explain it by better cache locality, NewInBuffer allocates more than 128k, so we jump around in memory more. Jeremy, please check! Volker (This used to be commit 452d51bc6fd41771b9c41ba6391664513d7cf2cd)
2007-10-10r21277: Fix an off by one error in the signal handler for aio: We can ↵Volker Lendecke1-1/+1
actually receive AIO_PENDING_SIZE signals, not one less. Jeremy I'm not merging this to 3_0_25, I want you to look at it first! Volker (This used to be commit 8fd0ec58360a349826508fc361a943650fdd6694)
2007-10-10r21274: As we now have POSIX opens which can unlinkJeremy Allison1-0/+9
with other open files we may have taken the delete code path with more than one share mode entry - ensure we only delete once by resetting the delete on close flag. Jeremy. (This used to be commit 429dff31663e71556b5193b25d5b8e7e09fa5b9c)
2007-10-10r21269: ReformattingVolker Lendecke1-89/+151
(This used to be commit b2eaa733c52113f53b6ff8aea9fce20ede1e3b1f)
2007-10-10r21257: Better fix for bug #4188 :Jeremy Allison1-0/+39
Windows Vista RC1 and RC2 can't delete directory on Samba share based on work by Joe Meadows <jmeadows@webopolis.com>. Jeremy. (This used to be commit 2dab8928769938ab79da7b7ce2d165fc388f9b00)
2007-10-10r21251: Okay, after Jeremy has kindly tested this, check it in :-)Volker Lendecke1-1/+5
Attached find a workaround that works for me. This is not the "correct" fix, to me it seems our DFS referral marshalling is broken. Vista requests level 4, we reply with level 2, and Vista seems not to like that. If we reply with level 3 it seems more happy. Needs more work! Volker (This used to be commit 4c3a9135629415086698da28596f68f7a9db9133)
2007-10-10r21250: Partial fix for BUG 4093: Make %a expand to "Vista"Gerald Carter1-8/+37
based on the flags2 values in the negprot request. This also includes some code for testing the dialect strings for "SMB 2.001" but this is unreliable as Vista only sends that in the 1st negprot and caches the fact that we don't support it. Restartnig the WOrkstation service on the client clears the cache. (This used to be commit d781eeb0e4362b7af1497634d26315498a5257d4)
2007-10-10r21247: In the refactoring I dropped a RESOLVE_DFSPATH callJeremy Allison1-2/+9
from setpathinfo. Return it and make sure all paths go through a dfs resolve. Jeremy. (This used to be commit 7fd3a146c11b70428ab5fbd7dc0272706abf8280)
2007-10-10r21246: Remove incorrect control on unix set info.Jeremy Allison1-6/+2
We can take path or handle based calls. Jeremy. (This used to be commit a19fe0f0c4a809d9b086e904586243949f57ec96)
2007-10-10r21229: Don't indirect through a potentially null pointer :-).Jeremy Allison1-5/+6
Jeremy. (This used to be commit 6723f9d85ccb4883f15658aceb021b4c4a752edf)
2007-10-10r21227: Quick fix for Stevef - ensure the returned data on openJeremy Allison1-6/+10
is 8 byte aligned. Jeremy. (This used to be commit 7645044fcf77644839bb0d74163d93d03284bd67)
2007-10-10r21226: Fix bug #4377 (rename of "foo" -> "Foo" fails).Jeremy Allison1-11/+29
This is actually an interesting case as it exposed bad code in our DFS redirect module (that was where the bug was introduced). Caused by our turning on dfsroot be default. Jeremy. (This used to be commit 5fe25588f57ee59d70a66c755dfe70312e7afd08)
2007-10-10r21225: Couple of fixes from Martin Zielinski mz@seh.de,Jeremy Allison1-3/+2
one typo, one to make sure that time initialization is done before modules that depend on it (printer initialization). Jeremy. (This used to be commit 6df32b4a6a9abd40e963c188b4c8769f15a0ab9b)
2007-10-10r21223: Try and fix bug #4361 - Vista backup fails.Jeremy Allison1-2/+24
Based on work from Joe Meadows <jameadows@webopolis.com>. One for the Vista patchset. Jeremy. (This used to be commit 470a4dc30874b00c599806cb6dbc06fe2775144c)
2007-10-10r21210: Fix POSIX open to return an info level.Jeremy Allison1-54/+96
Jeremy. (This used to be commit 5fa720f04f5d3b736145f9323aa8f4e623c14a41)
2007-10-10r21207: Make the code match the spec :-).Jeremy Allison1-8/+25
Jeremy. (This used to be commit 36afae2d48a3c3e97d1c2c35275f99d51ba2b64a)
2007-10-10r21191: Add in the POSIX open/mkdir/unlink calls.Jeremy Allison8-137/+460
Move more error code returns to NTSTATUS. Client test code to follow... See if this passes the build-farm before I add it into 3.0.25. Jeremy. (This used to be commit 83dbbdff345fa9e427c9579183f4380004bf3dd7)
2007-10-10r21184: Dummy checkin to let the build farm pick up r21183 of Samba4Volker Lendecke1-2/+2
(This used to be commit 22c9511314d563e1f4480099d2990cb5fbb8546a)
2007-10-10r21181: Add some debug, fix the NT_STATUS_IO_TIMEOUT problems in the ↵Volker Lendecke1-1/+8
RAW-NOTIFY test in the build farm. Volker (This used to be commit 3dd1bf667f42bd03a7b7159b9c2ad9ccbd269130)
2007-10-10r21166: Ensure we return the correct "EROFS" error on a non-writableJeremy Allison1-1/+1
filesystem. Jeremy. (This used to be commit 1e25501e8e6c568ab8892e05e0a60197e267d2b5)
2007-10-10r21131: Some notify fixesVolker Lendecke2-1/+16
(This used to be commit b9300e1d5acba4e2537fcc6631d1ee4d53685f64)
2007-10-10r21128: Fix Vista connecting to Samba in share level security.Jeremy Allison2-0/+26
Vista sends the NTLMv2 blob by default in the tconX packet. Make sure we save off the workgroup the user was logged into on the client in the sessionsetupX and re-use it for the NTLMv2 calc. Jeremy. (This used to be commit 45dcf62960c2815c4d8e0c5f4a2d0af24df83290)
2007-10-10r21123: Make notify_fsp() static to notify.c.Volker Lendecke2-50/+71
Volker (This used to be commit d663b729c0b3f4265fa90a5df8534ad52b8f04eb)
2007-10-10r21121: Fix error message, based on RAW-NOTIFYVolker Lendecke1-1/+1
(This used to be commit 2d0a0a0b34fe4b8778d453ccfb5df6bc4534ef62)
2007-10-10r21120: Make notify a bit more robust: Delete the notify records if we ↵Volker Lendecke1-14/+34
figured out that the process holding it is not around anymore. Tridge, please review this and possibly also include it in 4. We can discuss the error message for "unknown target" from messaging_send, INVALID_HANDLE seemed to be most appropriate to me. Volker (This used to be commit 3b3ed7429e54b571375914e473eafc9888b5b204)
2007-10-10r21115: notify_internal.c needs to remove the table entry if a process has ↵Volker Lendecke1-5/+6
crashed. So it needs the specific error message. Make messages.c return NTSTATUS and specificially NT_STATUS_INVALID_HANDLE if sending to a non-existent process. Volker (This used to be commit 3f620d181da0c356c8ffbdb5b380ccab3645a972)
2007-10-10r21111: Reorganize the change notify params a bit. We now have the per-shareVolker Lendecke1-2/+1
parameters change notify = [yes]/no # do we do it at all kernel change notify = [yes]/no # enable/disable inotify Those who want FAM need to say change notify = yes vfs objects = notify_fam Volker (This used to be commit c3a44d8b9606fc516faceb69b8e87bfc8be312f3)
2007-10-10r21109: Send notify_fam.c through the VFS. Lets see how the build farm likes ↵Volker Lendecke1-283/+0
this. Volker (This used to be commit e57a6d4d6341a52de55860e7f97031a63810dbfa)
2007-10-10r21108: Send sys_notify_watch through the VFS, FAM is nextVolker Lendecke3-10/+9
(This used to be commit 603a96761391f36ae9a1c8777d3333ab5c02eb34)
2007-10-10r21107: Adapt notify_fam.c to the new infrastructure. This works in my test, ↵Volker Lendecke1-76/+90
but does not survive the raw-notify test just because famd is too slow. Events are coming in too late and are thus unsyncronized. Now to get that through the VFS. Volker (This used to be commit c84bf6c4f92ffe3728acef7de7cefbc08307ea39)
2007-10-10r21104: Rename struct watch_context to struct inotify_watch_contextVolker Lendecke1-8/+9
(This used to be commit 7ac79203f247ed4e41d1612ea82211db8c0cd644)
2007-10-10r21097: Ensure that when manipulating UNIX permissions viaJeremy Allison1-14/+57
the UNIX extensions we use the correct mask/force mask settings for new and existing files/directories. Bug reported by Juran David <ext-david.juran@nokia.com>. Jeremy. (This used to be commit d84af53666369c314f02faa09b708ec6f971fa28)
2007-10-10r21093: Remove the hash and dnotify backends. Disabling FAM for this ↵Volker Lendecke7-779/+6
checkin, I'm working on that right now. Volker (This used to be commit 01c9fb17281e99300b339d2cfc0acac7ca94843c)
2007-10-10r21092: Ok, that's the one that activates the Samba4 notify backend.Volker Lendecke4-46/+117
Now to clean up / fix lots of stuff. Volker (This used to be commit 9e7443fa1417c01be903b15073825dc4def78d99)
2007-10-10r21091: Fix the build without inotifyVolker Lendecke1-0/+4
(This used to be commit 686345b7ed727db7b1d9bc164208ecfb612837b6)
2007-10-10r21090: Add some helper routinesVolker Lendecke2-0/+49
(This used to be commit 985dde8ed0a7e6da5adf9c55654a4fd362b22acb)
2007-10-10r21089: Do notifies on renameVolker Lendecke1-0/+44
(This used to be commit 2aadb95a7e58777ee7ff024cc9e14f4334970d50)
2007-10-10r21087: Make the param list of notify_fname match notify_triggerVolker Lendecke4-16/+18
(This used to be commit defa28f9c3eda85a072b972fffd2d5de8bcf01f7)
2007-10-10r21084: Make the Samba4 files compile in Samba3, not activated yet.Volker Lendecke3-43/+74
Volker (This used to be commit c24854433a28cc066072a7107e29aa7fe2bec3c8)
2007-10-10r21083: Actually make it static... :-)Volker Lendecke1-1/+1
(This used to be commit 65f473d22bd8f3b04becfc0be369758898fdadb6)
2007-10-10r21082: Make canonicalize_path static to service.c -- we do have ↵Volker Lendecke4-59/+31
conn->connectpath (This used to be commit 06f58096e3785d0e2e0b9f2053d4975e44568e15)
2007-10-10r21080: ReformattingVolker Lendecke1-37/+68
(This used to be commit 705f866a78ec3f1169543bf82fb48ea6580160a9)
2007-10-10r21079: Minimizing diff: Adopt the Samba4 style ChangeNotify flags.Volker Lendecke4-9/+8
Volker (This used to be commit a3c1069b0c3da914e2ac7337fd9e924b1c811d39)
2007-10-10r21077: A step to minimize the diff later: This pulls in unmodified files ↵Volker Lendecke3-176/+987
from Samba4, not compiled yet. This way the modifications become visible later. ntvfs/common/notify.c -> smbd/notify_internal.c ntvfs/sysdep/inotify.c -> smbd/notify_inotify.c Naturally I had to disable notify again :-) Volker (This used to be commit cdb7d582b7397faa5926bff5783da7fef4209948)
2007-10-10r21073: Fix compiler warningVolker Lendecke1-2/+2
(This used to be commit 3977ee027f500c76bcfd3de34301d3959e9b6d2b)
2007-10-10r21067: Finished refactoring setqpathinfo to be readableJeremy Allison1-385/+430
and modifiable by other people. Now to look over the build farm and fix any breakage. Jeremy. (This used to be commit 997259f2479842ee888621ae2d5446797b3a7cb6)
2007-10-10r21064: The core of this patch isVolker Lendecke6-34/+55
void message_register(int msg_type, void (*fn)(int msg_type, struct process_id pid, - void *buf, size_t len)) + void *buf, size_t len, + void *private_data), + void *private_data) { struct dispatch_fns *dfn; So this adds a (so far unused) private pointer that is passed from message_register to the message handler. A prerequisite to implement a tiny samba4-API compatible wrapper around our messaging system. That itself is necessary for the Samba4 notify system. Yes, I know, I could import the whole Samba4 messaging system, but I want to do it step by step and I think getting notify in is more important in this step. Volker (This used to be commit c8ae60ed65dcce9660ee39c75488f2838cf9a28b)
2007-10-10r21063: All case statements are now NTSTATUS returningJeremy Allison2-175/+219
functions. Now to factor out the post processing and make all cases behave the same (no mixture of "early returns" and "break"s. Jeremy (This used to be commit 7e17e54cb729e34c935927fe69a43690c7f446ae)
2007-10-10r21062: Should fix the build plus make all refactoredJeremy Allison2-212/+191
functions return NTSTATUS. Jeremy. (This used to be commit 1f3c2b2abca6f1b3b59b73df038832e14106ff76)
generated by cgit (git 2.34.1) at 2024-08-14 05:24:11 +0000