Skip to content

Comments

V2.27#8

Open
rollingxiaoxiao wants to merge 40 commits intoquixadhal:masterfrom
fluffos:v2.27
Open

V2.27#8
rollingxiaoxiao wants to merge 40 commits intoquixadhal:masterfrom
fluffos:v2.27

Conversation

@rollingxiaoxiao
Copy link

No description provided.

thefallentree and others added 30 commits March 3, 2013 16:33
Compile warning fixes (sunyc@lpmuds.net)
add_action moved objects fix (vorpal@lpmuds.net)
Properly report 64 bit linux when starting (sunyc@lpmuds.net)
heart_beat() now sets current_interactive
Changes for C99 (sunyc@lpmuds.net)
GMCP telnet negotiation now following the correct protocol, probably
Websocket support, add a port in the config file with type websocket, when it
        gets a connection on that it will ask the master object for a
        connection as usual, this object will need to do the websocket
        handshake, incoming data will go to process_input(), after finishing
        the handshake call websocket_handshake_done(), after calling that it
        will act like a normal user connection on the mud side, the other side
        is a binary websocket. I suggest you change to the normal login object
        after the handshake.
Added a break at the end of mssp activation to stop it from turning on mccp as well
Stop skipping next user input when we get user input (no idea why that was there)
        (vorpal@lpmuds.net)
Several LPC stack fixes (sunyc@lpmuds.net)
Added several missing type checks (sunyc@lpmuds.net)
Fixed wrong int/long variable types, this includes fixing sha1, so if you used
        that for passwords on 64bit, you'll need to figure out how to get
        around sha1 results changing.(sunyc@lpmuds.net, me (with suggestions from
        (vorpal@lpmuds.net)))
sizeof now returns an actual number (reported by vorpal@lpmuds.net)
Fixed crasher in local_time() (sunyc@lpmuds.net)
Overflow improvement on eval_cost calculation (sunyc@lpmuds.net)
Stop reading past the end of the read data in read_file()
Fixed possible crasher in using floats
Fixed crasher in call_other type warnings (sunyc@lpmuds.net)
Several testsuite improvements (sunyc@lpmuds.net)
Added GCmalloc (sunyc@lpmuds.net)
Mapping fix for 32bit
Fixed crasher in f_unique_mapping (reported by sunyc@lpmuds.net)
options.h fixes (sunyc@lpmuds.net, me)
Added placeholder for now missing old binary config (sunyc@lpmuds.net)
File closing fix for compressed files (sunyc@lpmuds.net)
Fixed zonetime crasher (sunyc@lpmuds.net)
Fixed is_daylight_savings_time crasher (sunyc@lpmuds.net)
Fixed remove_get_char crasher (reported by sunyc@lpmuds.net)

FluffOS 2.25
Fixed the defer() fix
Added new test_load() efun which will try to load a file, it will return 0 when
      it failed, 1 when the file can load, and error if there was a compilation
      error
testlib stuff is actually in there
tiny fix to error messages (missing breaks)
actually contains Properly report 64 bit linux when starting (sunyc@lpmuds.net)
Added timer_settime/timer_gettime alternative implementation
for eval_cost.  This fixes the issue where get_eval returns
negative numbers when the elapsed wall time is greater than
the CPU time spent.

The reason for using POSIX timers rather than merely adding
a getitimer() call to the default implementation is better
precision.  On Linux, the values returned by getitimer()
are rounded off to the nearest jiffy (1/250 of a second),
whereas the POSIX timers have nanosecond precision.
It will pass as long as it is stored as C 64-bit int & double,
as well as parser correctly parse constant literals.
It correctly launch and checks testsuite now.
Change all reference in the source code to use two new types.
remove INT_32 macro, no one will be using.
* Fixed build issue with newer openssl where support for hash
  functions such as MD2 can be conditionally disabled
* Added -lcrypto to edit_source.c when PACKAGE_CRYPTO is enabled
* Efficiency improvements to the hexdump function
* Fixed improper use of strlen on byte-counted strings
* Fixed memory leak of error string on "unknown hash" error
* Fixed memory corruption from storing the wrong type of malloced
  memory on the stack
* Added support for sha512, sha384, sha256, and sha224
Conflicts:
	src/testsuite/single/inh.c
Conflicts:
	src/testsuite/etc/config.test
	src/testsuite/single/tests/efuns/call_out.c
Compile with --std=c99 no longer defines linux, this caused some
old macro to fail.
Also improve the existing read_file tests.

Conflicts:
	src/testsuite/single/tests/efuns/read_file.c
Conflicts:
	src/local_options.test
… (driver timezone would be changed after efun/error in efun)
dwwodan and others added 10 commits April 14, 2013 13:44
Revert "Rewrite apply_low(), fix cache poison issue."

This reverts commit 4c13478.
sha1() now works correctly on strings 8192 bytes and longer
sha1() now works correctly on strings 55 bytes long
Enabled PACKAGE_SHA1 in local_options.test and added unit tests
Also fixed various compile problem on CYGWIN.

Conflicts:
	src/efuns_port.c
	src/eval.h
	src/posix_timers.c
	src/testsuite/single/tests/compiler/64bit.c
Conflicts:
	src/Makefile.in
	src/interpret.c
This reverts commit 9732b60.

Conflicts:
	src/Makefile.in
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants