Age | Commit message (Collapse) | Author | Files | Lines |
|
version.h changes rather frequently. Since it is included via includes.h,
this means each C file will be a cache miss. This applies to the following
situations:
* When building a new package with a new Samba version
* building in a git branch after calling mkversion.sh
after a new commit (i.e. virtually always)
This patch improves the situation in the following way:
* remove inlude "version.h" from includes.h
* Use samba_version_string() instead of SAMBA_VERSION_STRING
in files that use no other macro from version.h instead of
SAMBA_VERSION_STRING.
* explicitly include "version.h" in those files that use more
macros from "version.h" than just SAMBA_VERSION_STRING.
Michael
|
|
(This used to be commit fe36d83508a6b1a8de5ad58ec1d0c80cda7b4a21)
|
|
This adds a new macro for a vender version function which will be returned
by the samba_version_string() function if defined.
Signed-off-by: Andreas Schneider <anschneider@suse.de>
Signed-off-by: Stefan Metzmacher <metze@samba.org>
(This used to be commit 8b9301fbc73f28c657ec81c605760a0ff7a1973a)
|
|
This is right now only used there, and in version.c it gave linker errors
because some binaries (e.g. smbmnt) don't link in time.o
(This used to be commit 1f0eaaa5911f893c822465a26fe49ab65afb0730)
|
|
(This used to be commit 85123aacdb13e97c3f44aeded1c80e13af53d83d)
|
|
On Jan 21 16:18, Danilo Almeida wrote:
> Corina wrote:
>
> > + time_t samba_gitcommitdate;
>
> And:
>
> > + SIVAL(pdata,28,extended_info.samba_gitcommitdate);
> > + memcpy(pdata+32,extended_info.samba_version_string,32);
>
> Note that you are dropping bits on a system w/64-bit time_t, and that this has the 2038 problem.
Right. I changed samba_gitcommitdate from time_t to NTTIME and shortened
samba_version_string to 28 bytes. New patch below.
Thanks,
Corinna
(This used to be commit 28aa1c199d3a22cda34afcaab49c0561eeb0abcb)
|
|
(This used to be commit 999647329028147d7c29a3348202641b3e03430e)
|
|
Hi,
while implementing the extra_info version stuff, it occured to me that
samba_version_string() potentially allocates memory which is unused but
never free'd.
If SAMBA_VERSION_VENDOR_PATCH is defined, a second call to asprintf
takes place. The result is stored in tmp_version. Afterwards,
samba_version is set to tmp_version without free'ing samba_version
first. Looks like a simple free(samba_version) is missing. Patch
against 3.2-test below.
Ok, this only happens once over the lifetime of the application, so it's
no big deal, but I though it doesn't hurt to mention it.
Corinna
* lib/version.c (samba_version_string): Free samba_version
before setting to tmp_version.
(This used to be commit 373a23d48f2dd24e65dbf814ea58b4add2322128)
|
|
(This used to be commit a9c62c57db9e580640d0265b08b3178496de76a8)
|
|
which is used by mkversion.sh.
(This used to be commit a62c7f9350322ab435244eafe6053fd5ee618973)
|
|
bugs in various places whilst doing this (places that assumed
BOOL == int). I also need to fix the Samba4 pidl generation
(next checkin).
Jeremy.
(This used to be commit f35a266b3cbb3e5fa6a86be60f34fe340a3ca71f)
|
|
Jeremy.
(This used to be commit 406737cd9a34ac498e5823b380e1768d41e16c1b)
|
|
(This used to be commit b0132e94fc5fef936aa766fb99a306b3628e9f07)
|
|
Jeremy.
(This used to be commit 407e6e695b8366369b7c76af1ff76869b45347b3)
|
|
systems. Fix the build on those boxes
(This used to be commit 0b8ed8136b5c5b9b7bb5615dc8dc27eec7c49506)
|
|
(This used to be commit d40f06018b9197c35a10cc9ab284dd596244f47b)
|
|
(This used to be commit ae452e51b02672a56adf18aa7a7e365eeaba9272)
|