mirror of
https://github.com/FirebirdSQL/firebird.git
synced 2025-01-23 04:43:03 +01:00
nightly update
This commit is contained in:
fsg
parent
e602e36bba
commit
7d67262810
161
ChangeLog
161
ChangeLog
@ -1,3 +1,164 @@
|
||||
2004-10-07 22:44 skidder
|
||||
|
||||
* firebird2/builds/posix/Makefile.in.client.fbudf (1.3):
|
||||
|
||||
Build threaded version of fbudf library in SuperServer builds
|
||||
|
||||
2004-10-07 22:25 skidder
|
||||
|
||||
* firebird2/builds/posix/: Makefile.in.client.gpre (1.16),
|
||||
Makefile.in.embed.gpre (1.16):
|
||||
|
||||
Fix POSIX build after recent Claudio fix
|
||||
|
||||
2004-10-07 11:59 robocop
|
||||
|
||||
* firebird2/src/gpre/: gpre.cpp (1.64), gpre.h (1.71), par.cpp
|
||||
(1.53):
|
||||
|
||||
Looking at gpre.c in FB1 and the usage of the variable fatal_env,
|
||||
it's evident that the conversion SEH->C++ try/except screwed the
|
||||
logic. The original idea was that CPR_abort will cause the program
|
||||
to exit gracefully, not to continue parsing, because it's useless
|
||||
and clutters the console with hundreds of messages. As far as I can
|
||||
test, when gpre cannot attach to the db, run-time failures can be
|
||||
detected by gpre_static or a false result is converted in
|
||||
exception. The new gpre_exception propagates up to the highest
|
||||
level where it allows the program to exit gracefully as it was in
|
||||
FB1. Until now, the behavior in FB1.5 & FB2 has been annoying. When
|
||||
we fail to connect to the db, anything more than we do is useless;
|
||||
we now will print a message, clean the house and exit. Hope
|
||||
everyone can compile my change.
|
||||
|
||||
2004-10-07 11:46 robocop
|
||||
|
||||
* firebird2/src/jrd/gds.cpp (1.154):
|
||||
|
||||
VMS-related changes. Just in case.
|
||||
|
||||
2004-10-07 11:42 robocop
|
||||
|
||||
* firebird2/src/: dudley/extract.epp (1.39), jrd/common.h (1.126):
|
||||
|
||||
Treat literal strings as constants, please.
|
||||
|
||||
2004-10-07 11:27 robocop
|
||||
|
||||
* firebird2/src/: common/fb_exception.cpp (1.14),
|
||||
include/fb_exception.h (1.12):
|
||||
|
||||
Fixed a bug in fatal_exception's constructor that causes
|
||||
fill_status to write garbage in the status vector. Overload what()
|
||||
for the same class. It helps gpre to quit on failures (gpre changes
|
||||
still not posted).
|
||||
|
||||
2004-10-07 11:24 robocop
|
||||
|
||||
* firebird2/src/jrd/dyn_def.epp (1.80):
|
||||
|
||||
Ugh, a change that belong to another, still not posted
|
||||
"configuration" slipped in. Go back.
|
||||
|
||||
2004-10-07 11:19 robocop
|
||||
|
||||
* firebird2/src/: burp/backup.epp (1.66), qli/meta.epp (1.50):
|
||||
|
||||
Misc.
|
||||
|
||||
2004-10-07 11:15 dimitr
|
||||
|
||||
* firebird2/src/jrd/: btr.cpp (1.122), cch.cpp (1.112), cmp.cpp
|
||||
(1.152), evl.cpp (1.109), exe.cpp (1.125), idx.cpp (1.75), inf.cpp
|
||||
(1.49), met.epp (1.143), opt.cpp (1.196), pag.cpp (1.83), pcmet.epp
|
||||
(1.30), rlck.cpp (1.31), rng.cpp (1.32), sdw.cpp (1.53), tpc.cpp
|
||||
(1.19), tra.cpp (1.91), tra.h (1.41), tra_proto.h (1.16), vio.cpp
|
||||
(1.114):
|
||||
|
||||
1) Locks related cleanup 2) Got rid of TRA_nowait + transaction
|
||||
waits related cleanup 3) Return isc_lock_timeout when it's
|
||||
necessary TODO: Enhance SET TRANSACTION
|
||||
|
||||
2004-10-07 11:02 robocop
|
||||
|
||||
* firebird2/src/jrd/: dfw.epp (1.131), dyn_def.epp (1.79):
|
||||
|
||||
Possible buffer overrun.
|
||||
|
||||
2004-10-07 11:01 robocop
|
||||
|
||||
* firebird2/src/jrd/: isc.cpp (1.52), isc_proto.h (1.12):
|
||||
|
||||
Possible buffer overrun (SOLARIS & VMS) strncpy does not guarantee
|
||||
to write the null terminator!
|
||||
|
||||
2004-10-07 10:43 robocop
|
||||
|
||||
* firebird2/src/jrd/utl.cpp (1.74):
|
||||
|
||||
Misc
|
||||
|
||||
2004-10-07 10:35 robocop
|
||||
|
||||
* firebird2/src/: common/classes/fb_string.cpp (1.23),
|
||||
include/fb_string.h (1.30):
|
||||
|
||||
recalculate_length(), copy_to(), safe copy to a char array,
|
||||
constness, style.
|
||||
|
||||
2004-10-07 10:33 robocop
|
||||
|
||||
* firebird2/src/jrd/idx.cpp (1.74):
|
||||
|
||||
Little simplification.
|
||||
|
||||
2004-10-07 10:29 robocop
|
||||
|
||||
* firebird2/src/dudley/extract.epp (1.38):
|
||||
|
||||
Clarification on badly written old comment (problem with embedded
|
||||
blanks in identifiers)
|
||||
|
||||
2004-10-07 10:28 robocop
|
||||
|
||||
* firebird2/src/dsql/gen.cpp (1.71):
|
||||
|
||||
1.- Style. 2.- The const_cast is no longer needed.
|
||||
|
||||
2004-10-07 10:27 robocop
|
||||
|
||||
* firebird2/src/dsql/dsql.cpp (1.127):
|
||||
|
||||
Style
|
||||
|
||||
2004-10-07 10:25 robocop
|
||||
|
||||
* firebird2/src/burp/burp.h (1.59):
|
||||
|
||||
Fix a problem caused by cleanup changing #define to pointer to
|
||||
literal string. Due to sizeof being involved, gbak could
|
||||
misinterpret a file as been a valid part of a multi-file backup.
|
||||
|
||||
2004-10-07 10:23 robocop
|
||||
|
||||
* firebird2/src/burp/spit.h (1.8):
|
||||
|
||||
Unused crud
|
||||
|
||||
2004-10-07 10:21 robocop
|
||||
|
||||
* firebird2/src/alice/: alice_meta.epp (1.43), tdr.cpp (1.42):
|
||||
|
||||
Misc and treat string literals as constants.
|
||||
|
||||
2004-10-07 08:15 skidder
|
||||
|
||||
* firebird2/src/jrd/vio.cpp (1.113):
|
||||
|
||||
Report lock manager error code to the user when pessimistic lock
|
||||
conflict is encountered. This change fixes a problem with deadlock
|
||||
reporting for pessimistic locks so it should be backported to
|
||||
B1_5_Release
|
||||
|
||||
2004-10-06 13:44 arnobrinkman
|
||||
|
||||
* firebird2/src/jrd/idx.cpp (1.73):
|
||||
|
||||
Loading…
Reference in New Issue
Block a user