| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
tactics :-). Protected by #ifdef until used.
Fixed bug in fd_attempt_close() where a pointer to potentially
free'd memory was returned. I hate that.
Added "blocking locks" as a per-share option for performance testing.
Changed is_mangled() so it will return true if called with a pathname
and any component of the pathname was mangled (it was already attempting
to do this, but not checking for a '/' as end-of-mangle).
This should be a better fix for the wierd stat cache bug Andrew identified.
Jeremy.
(This used to be commit 0de01f45980c7bc261248a9cead972a8d8cbd594)
|
|
mangling a name can't increase it's size which isn't true. (imagine a
file called "L B" which mangles to "LB~XX")
The symptoms were that users couldn't run batch files from short
directory names that contained non 8.3 characters (such as spaces).
(This used to be commit c319d8ea3f8b42bb3a8e501642971ed0bdb21583)
|
|
Connections[] is now a local array in server.c
I might have broken something with this change. In particular the
oplock code is suspect and some .dll files aren't being oplocked when
I expected them to be. I'll look at it after I've got some sleep.
(This used to be commit c7ee025ead4a85b6fa44a832047b878451845fb6)
|
|
before testing that the strdup() worked. Fixed.
(This used to be commit 899d0d5de5dd9d080d5c4cb94874d4f939427d1b)
|
|
dir.c & mangle.c. This was possible after I checked the tree code to make
sure that the tree code did not reference functions by macro name. Also,
note that the AVL module has been removed to prevent conflict with the
SplayTree macro defines.
Chris -)-----
(This used to be commit a1f4c8567c3c5264cbfa99867fa894cafab7da81)
|
|
It removed all ocurrences of the following functions :
sprintf
strcpy
strcat
The replacements are slprintf, safe_strcpy and safe_strcat.
It should not be possible to use code in Samba that uses
sprintf, strcpy or strcat, only the safe_equivalents.
Once Andrew has fixed the slprintf implementation then
this code will be moved back to the 1.9.18 code stream.
Jeremy.
(This used to be commit 2d774454005f0b54e5684cf618da7060594dfcbb)
|
|
He also suggested that some systems may implement free() as a macro (but
I think he was looking for an excuse ;). Anyway, I've added a function
to mangle.c that calls free().
Chris -)-----
(This used to be commit 95f7b03285c42e8f5573690939b79afc7e686908)
|
|
Jeremy.
(This used to be commit e74428e4287cbc1557cc79d4930371cfaf3256d4)
|
|
proto.h: The unusual. ;)
reply.c: I changes some function names, and updated reply.c to match.
See mangle.c below for more.
server.c: Changed function names and parameters in file mangle.c, so
changed server.c calls to match. See mangle.c below for more.
mangle.c:
I replaced the caching mechanism used for caching reverse mangled name
maps. The old method was a large array of 256-byte strings. Movement
in the stack (including push and pop) was done by memcpy()ing whole
chunks of memory around.
The new system uses the ubi_Cache module which, in turn, uses a splay
tree. Entries are dynamically allocated using a minimum amount of
memory. Searches are non-linear, which should speed things up a bit,
too. Overall, this should save memory and be faster.
Other changes: I streamlined the is_mangled() test and made other
speed enhancements including replacing some static functions with
macros. Added comments, etc.
Note: Per an E'mail conversation with Andrew, the 'mangled stack'
parameter in smb.conf doesn't do anything anymore. The cache is
now set for 16K bytes maximum memory usage. The mangle stack
parameter is silently ignored. This can easily be changed, but
I'd rather introduce a 'mangled cache memory' parameter and
remove 'mangled stack'.
Remaining problems:
While testing the module, I noticed that something is calling
name_map_mangle() twice. The result is that names which contain
illegal characters are getting mangled twice.
Also, the entire module works by overwriting the input string.
This has a variety of nasty side effects.
Summary:
There's a lot still to be done, but the changes I have in place *should*
work in exactly the same way (except for the mangle stack parameter).
The rest of the bugs and other issues are separate.
Chris -)-----
(This used to be commit 8759bec11ba483b2292b0e513b85c98ed5e3e2d4)
|
|
algorithm was correct).
Finally (I think) fixed the mangled directory stack issue in
scan_directory() correctly. Mangled & non-mangled names are
now being checked correctly. Hurrah to Ulrik Dickow <ukd@kampsax.dk>
who helped isolate this one.
Jeremy.
(This used to be commit 37f5f7b557aa59c7eba4a9f2f2f323d7cec62fc7)
|
|
Kanji support is one case of multibyte character
support, rather than being a specific case in
single byte character support.
This allows us to add Big5 Chinese support (code page 950)
and Korean Hangul support (code page 949) at very little
cost. Also allows us to easily add future multibyte
code pages.
Makefile: Added codepages 949, 950 as we now support more multibyte
codepages.
asyncdns.c: Fixed problem with child being re-spawned when parent killed.
charcnv.c
charset.c
client.c
clitar.c
kanji.c
kanji.h
smb.h
util.c
loadparm.c: Generic multibyte codepage support (adding Big5 Chinese
and Korean Hangul).
nmbd.c: Fixed problem with child being re-spawned when parent killed.
mangle.c: Modified str_checksum so that first 15 characters have more
effect on outcome. This helps with short name mangling as
most 'long' names are still shorter than 15 chars (bug was
foobar_mng and foobar_sum would hash to the same value, with
the modified code they hash differently.
Jeremy.
(This used to be commit 299016338cfb47f0c585875ef9b468121fcee97d)
|
|
This is merely updating the Copyright statements from 1997 to 1998.
It's a once a year thing :-).
NO OTHER CHANGES WERE MADE.
Jeremy.
(This used to be commit b9c16977231efb274e08856f7f3f4408dad6d96c)
|
|
mangle.c server.c proto.h
mangle.c
I am planning to replace the mangled_stack array with a proper stack,
but found many style inconsistencies (no, really). As you might expect,
I have standardized on my own preferences. ;)
I also found a potential problem in create_mangled_stack (which I've
renamed as reset_mangled_stack). If the stack size were passed into
the function as 0 or less, there was the possibility that the array
would have been freed twice. I doubt that this ever happens, but I
don't like to leave holes.
Of course, the fix will be irrelevent once I replace the array with
a linked-list-based stack.
server.c
Changed the call to create_mangled_stack() to a call to reset_mangled_stack().
proto.h
Regenerated to match the above changes. (A real comment! How unusual!)
(This used to be commit 34d1b3e4fa7a1158f3a3c5c47adf0417c7144095)
|
|
CIFS3 spec does not list them as illegal.
This allows things like the control panel icon to be placed on a Samba
drive.
(This used to be commit d2ac9e6fb6e76778154c58217237251d7bb0e98c)
|
|
Jeremy (jallison@whistle.com)
(This used to be commit e7eb1f044d3101679dc7a118820ea5efe0cd837c)
|
|
mangle.c smb.h util.c:
Big merge to allow KANJI support to be in the main
binary without explicitly compiling with it.
locking.c: Fix for smbstatus not being able to read files.
namepacket.c: Removed unneccesary debug statement.
trans2.c: Added Luke's proposed fix (ifdefed out until further testing).
nmblookup.c: Fixed bug where query fails and status is done on bogus IP.
Jeremy (jallison@whistle.com)
(This used to be commit 9196255022ae8c51b527412747b324819bea2c13)
|
|
server.c: Changed comparison from dname to name2 - seems correct - don't know why
I changed it originally. hmmmmm.
trans2.c: Made lanman2 code more similar to scan_directory(). I should
make the trans2 code call scan_dir as they are so similar.
Jeremy(jallison@whistle.com).
(This used to be commit 3909576849025ed21e5282abc892ae7582ee0275)
|
|
Wed May 7 1997: Update for 1.9.17alpha1 release - 'browsefix release'
designed to make browsing across subnets work.
byteorder.h: Updated copyright to 1997.
charcnv.c: Updated copyright to 1997.
charset.c Updated copyright to 1997.
charset.h Updated copyright to 1997.
client.c Updated copyright to 1997.
clientutil.c Updated copyright to 1997.
dir.c Updated copyright to 1997.
fault.c Updated copyright to 1997.
includes.h Updated copyright to 1997.
interface.c Updated copyright to 1997.
ipc.c Updated copyright to 1997.
kanji.c Updated copyright to 1997.
kanji.h Updated copyright to 1997.
loadparm.c Updated copyright to 1997.
locking.c Updated copyright to 1997.
mangle.c Updated copyright to 1997.
message.c Updated copyright to 1997.
nameannounce.c
Made use of WINS subnet explicit.
Added reset_announce_timer() so announcement
can be made immediately when we become a master.
Expanded code to do sync with dmb.
namebrowse.c
Removed redundent checks for AM_MASTER in
sync code. Made use of WINS subnet explicit.
namedbname.c Made use of WINS subnet explicit.
namedbresp.c Made use of WINS subnet explicit.
namedbserver.c Made use of WINS subnet explicit.
namedbsubnet.c
Explicitly add workgroup to WINS subnet
when we become a dmb. Made use of WINS subnet explicit.
namedbwork.c
Made use of WINS subnet explicit. Removed
redundent check_work_servertype() function.
nameelect.c
Explicitly add workgroup to WINS subnet
when we become a master browser. Made use of WINS subnet explicit.
namelogon.c Updated copyright to 1997.
namepacket.c Updated copyright to 1997.
namequery.c Updated copyright to 1997.
nameresp.c
Made use of WINS subnet explicit. Made nmbd fail if
configured as master browser and one exists already.
nameserv.c
Made use of WINS subnet explicit. Remove redundent
logon server and domain master code.
nameserv.h Add emumerate subnet macros.
nameservreply.c Made use of WINS subnet explicit.
nameservresp.c Updated copyright to 1997.
namework.c
Made use of WINS subnet explicit. Updated code to
add sync browser entries to add subnet parameter.
nmbd.c
Added sanity check for misconfigured nmbd.
nmblib.c Updated copyright to 1997.
nmblookup.c Updated copyright to 1997.
nmbsync.c
Removed redundent AM_ANY_MASTER check.
params.c Updated copyright to 1997.
password.c Updated copyright to 1997.
pipes.c Updated copyright to 1997.
predict.c Updated copyright to 1997.
printing.c Updated copyright to 1997.
proto.h
Changed protos for new nmbd code.
quotas.c Updated copyright to 1997.
replace.c Updated copyright to 1997.
reply.c Updated copyright to 1997.
server.c Updated copyright to 1997.
shmem.c Updated copyright to 1997.
smb.h Updated copyright to 1997.
smbencrypt.c Updated copyright to 1997.
smbpasswd.c Updated copyright to 1997.
smbrun.c Updated copyright to 1997.
status.c Updated copyright to 1997.
system.c Updated copyright to 1997.
testparm.c Updated copyright to 1997.
testprns.c Updated copyright to 1997.
time.c Updated copyright to 1997.
trans2.c Updated copyright to 1997.
trans2.h Updated copyright to 1997.
uid.c Updated copyright to 1997.
username.c Updated copyright to 1997.
util.c Updated copyright to 1997.
version.h
Changed to 1.9.17alpha1.
(This used to be commit cf23a155a1315f50d488794a2caf88402bf3e3e6)
|
|
changes a bit less. Also added fix to tricky reply_mv case.
jra@cygnus.com
(This used to be commit f22ac13b1423186d463c9fc5b3588a3dcbd7bc1c)
|
|
filename problem.
Problem conditions:
case sensitive = no
default case = lower
mangle case = yes
preserve case = yes
short preserve case = yes
Problem: Win95 still writes upper case file names.
Patches supplied have been incorporated into mangle.c and server.c
but needs YOSTW to be defined for it to take effect.
Action:
=======
If we are happy with the patch, delete old code in mangle.c, then
remove all YOSTW #ifdefs.
(This used to be commit 84fc674c89e8999f284b59888dfa2cdc47438e88)
|
|
(compiler bugs were the problem)
- minor password cleanups (catch WfWG bug where it sets the password
to a space instead of a NULL)
- fix printing problem for kanji users
- minor cleanups
(This used to be commit 92566ecc315c29da6e9aaa67ddae33e64f5bcc67)
|
|
Bressler)
- machten support from Trevor Strohman (trev@figment.tenon.com)
- added qinfo command to client as part of drag-and-drop printer
support for win95 from David Chappell <chappell@mouse.cc.trincoll.edu>
He also added the "printer driver" option
- use sigblock() on more systems and use sigsetmask(0) instead of
sigunblock() as its more portable. This beats a problem with zombies
on heavilily loaded systems.
- added internals.doc written by David Chappell into the source tree
- get rid of PRINT_COMMAND options from local.h as they are no longer
relevent
- new kanji code from Fujita
- don't set the recursion_available flag on queries in nmbd
- fix a potential bug with pointer subtraction in printing.c
- got rid of error_count code as the real fix (the EOF problem) is now
in
(This used to be commit aa6f8b04d125b5bc00f267abf72b800228aabf7d)
|
|
prototypes automatically using "make proto". This is much less prone
to error than the old method of manually adding prototypes
(This used to be commit b551dc98f7cc194a5fc2e67a4ebae7fd67a01bbc)
|
|
(This used to be commit 45e2ffdc9c4334565e51c550b3a19a9bbb7037ef)
|
|
(This used to be commit 291551d80711daab7b7581720bcd9a08d6096517)
|
