IF YOU WOULD LIKE TO GET AN ACCOUNT, please write an
email to Administrator. User accounts are meant only to access repo
and report issues and/or generate pull requests.
This is a purpose-specific Git hosting for
BaseALT
projects. Thank you for your understanding!
Только зарегистрированные пользователи имеют доступ к сервису!
Для получения аккаунта, обратитесь к администратору.
This patch introduces
struct stat_ex {
dev_t st_ex_dev;
ino_t st_ex_ino;
mode_t st_ex_mode;
nlink_t st_ex_nlink;
uid_t st_ex_uid;
gid_t st_ex_gid;
dev_t st_ex_rdev;
off_t st_ex_size;
struct timespec st_ex_atime;
struct timespec st_ex_mtime;
struct timespec st_ex_ctime;
struct timespec st_ex_btime; /* birthtime */
blksize_t st_ex_blksize;
blkcnt_t st_ex_blocks;
};
typedef struct stat_ex SMB_STRUCT_STAT;
It is really large because due to the friendly libc headers playing macro
tricks with fields like st_ino, so I renamed them to st_ex_xxx.
Why this change? To support birthtime, we already have quite a few #ifdef's at
places where it does not really belong. With a stat struct that we control, we
can consolidate the nanosecond timestamps and the birthtime deep in the VFS
stat calls.
At this moment it is triggered by a request to support the birthtime field for
GPFS. GPFS does not extend the system level struct stat, but instead has a
separate call that gets us the additional information beyond posix. Without
being able to do that within the VFS stat calls, that support would have to be
scattered around the main smbd code.
It will very likely break all the onefs modules, but I think the changes will
be reasonably easy to do.
Fixed.
It turns out there were a number of places where cli_resolve_path() was called
and the error path upon that function failing did not set errno. There were a
couple of places the failure handling code did set errno to ENOENT, so I made
them all consistent, although I think better errno choices for this condition
exist, e.g. EHOSTUNREACH.
Derrell
callers pass in a struct user_auth_info * instead. This commit causes
smbc_set_credentials() to print out a message telling callers to use
smbc_set_credentials_with_fallback() instead, as smbc_set_credentials()
has a broken API (no SMBCCTX * pointer). No more global variables used
in the connection manager API for client dfs calls.
Jeremy.
The statvfs struct isn't guaranteed to be portable across operating
systems. Since libsmbclient isn't actually calling statvfs and just
using the statvfs struct to store similar information, this patch adds
a new portable smbc_statvfs struct. This fixes a few of the failures
in the build farm introduced by:
ae259575c4
Derrell, please check.
- James suggested using gcc's "deprecated" attribute to mark the context
structure fields to generate warnings. This creates a scenario with the
best of all worlds. I'm able to move to an organization that more easily
allows future enhancements, while avoiding any mandatory changes by
applications. Thanks, James!
- Updated WHATSNEW.txt so that it accurately reflects the current state of
affairs.
Derrell
(This used to be commit a67f96fbe9)
- Ensured that all public functions have documentation in libsmbclient.h
- Reformatted for "proper" indentation
- Re-added temporarily-disabled alternate authentication function capability
Derrell
(This used to be commit 64b7150d92)
Given the tacit (if that) approval by some people, and clear disapproval by
others for my proposed clean-up and reorganization of libsmbclient, I've come
up with a slightly different approach. This commit changes back to the
original libsmbclient.h SMBCCTX structure which will maintain ABI
compatibility. I retain, here, the setter and getter functions which all new
code should use. Older programs already compiled should continue to work
fine. Older programs being recompiled will encounter compile-time errors
(intentionally!) so that the code can be corrected to use the setter/getter
interfaces.
Although this doesn't clean up the interface in the way I had wanted, the code
reorganization and requirement for new programs to use the setters and getters
allows future progress to be made on libsmbclient without further muddying up
the interface, while retaining the ABI compatibility that was the big issue
causing disapproval. I hope that this compromise is adequate.
Derrell
(This used to be commit 56429a3d60)
The libsmbclient interface has suffered from difficulty of improvement and
feature enrichment without causing ABI breakage. Although there were a number
of issues, the primary ones were:
(a) the user of the library would manually manipulate the context structure
members, meaning that nothing in the context structure could change other
than adding stuff at the end;
(b) there were three methods of setting options: setting bits in a flags field
within the context structure, setting explicit options variables within an
options structure in the context structure, and by calling the
smbc_option_set() function;
(c) the authentication callback did not traditionally provide enough
information to the callee which required adding an option for a callback
with a different signature, and now there are requests for even more
information at the callback, requiring yet a third signature and option to
set it (if we implement that feature).
This commit provides a reorganization of the code which fixes (a) and (b).
The context structure is now entirely opaque, and there are setter and getter
functions for manipulating it. This makes maintaining ABI consistency much,
much easier.
Additionally, the options setting/getting has been unified into a single
mechanism using smbc_option_set() and smbc_option_get().
Yet to be completed is a refactoring of the authentication callback (c).
The test programs in examples/libsmbclient have been modified (if necessary;
some applications require no changes at all) for the new API and a few have
been minimally tested.
Derrell
(This used to be commit d4b4bae8de)
