===================================================================
RCS file: /cvsrepo/anoncvs/cvs/www/plus.html,v
retrieving revision 1.1423
retrieving revision 1.1424
diff -u -r1.1423 -r1.1424
--- www/plus.html 2019/03/21 16:03:07 1.1423
+++ www/plus.html 2019/03/25 02:02:13 1.1424
@@ -82,12 +82,49 @@
+
+
+- Raised the maximum columns and rows to 160 for efifb(4) now that it supports remapping the framebuffer in write combining mode.
+
- Fixed multiple buffer over-reads in ctfdump(1).
+
+
- Changed dwiic(4) to fetch timing parameters from ACPI and have it attach on Intel 300 series devices.
+
- Improved and enabled isochronous transfers in xhci(4). This allows laptop cameras on many newer laptops to work properly.
+
- Fixed rad(8) to only accept one ICMPv6 and one route socket. This fixed a scenario where the ICMPv6 socket was also used as the route socket for a short while.
+
- Changed slaacd(8) and rad(8) to refuse a control socket if it already has one.
+
- Fixed ieee80211(9) autojoin for WEP networks by actually saving the WEP auth key.
+
- Fixed display glitches in tmux(1) to aid with debugging formats such as the status line format.
+
- Changed printing of bgpd(8) configuration to show options set to non-default values.
+
- Added a fallback of direct serial hardware access to uefiboot(8) for those UEFI machines that have serial ports but whose firmware does not have the serial io protocol.
+
+
- Added acpicmos(4) to RAMDISK_CD for amd64.
+
- Made searching for uppercase characters in less(1) use mbtowc(3) rather than the buggy and outdated step_char() and correctly use iswupper(3) rather than isupper(3), further improving UTF-8 support.
+
- Fixed the balancing of octciu(4) work queue interrupts.
+
+
- Backported support for probing ARMv8 hardware acceleration capabilities on armv7 in preparation for adding support for the probing code for arm64.
+
- Improved speed of C++ unwinding by implementing a red-black tree based cache in libunwind.
+
+
- Made rad(8) imsg(3) handling more paranoid.
+
- Fixed resizing of tmux(1) control clients.
+
- Cleaned up less(1) UTF-8 handling in backc() and do_append().
+
- Modified tmux(1) to allow multiple modes to be open in a pane.
+
- Changed tset(1) and ttys(5) to use hardware tabs as the default since almost all terminals now support them.
+
- Added a new driver for USB Audio Class v2.0 devices to uaudio(4). The new driver maps audio blocks to USB transfers allowing precise synchronization and better reliability. This replaces the current one for USB Audio Class v1.0 devices.
+
- Added copy_output() and underrun() methods to audio(4) to support drivers using bounce buffers (e.g. uaudio(4)).
+
- Fixed delayed completions for isochronous transfers using uhci(4).
+
- Add support for origin mode to tmux(1).
+
+
- Separated fields output by pcidump(8) with commas since some fields are strings now.
+
- Made slaacd(8) imsg(3) more paranoid for everything but the control socket.
+
- Made vmd(8) to wait for the controlling end of a pty to become writeable before waiting for read events. This fixed an issue where a recent kqueue pty change caused vmd to hammer the log.
-
- Deleted an unnecessary call to setlocale(3) from touch(1).
+
- Added amdgpu to the list of possible driver firmware for fw_update(1).
+
- Taught pcidump(8) about pci class and subclass names.
+
- Moved adjtimedelta from kern_time.c to kern_tc.c in preparation for upcoming MP-safety changes to the timecounting layer.
+
- Deleted unnecessary calls to setlocale(3) from touch(1), uuencode(1), uudecode(1), and spell(1). Also made usage() static and return from main() rather than exit(3).
- Switched to the use of safe memory reclamation (SMR) in octcrypto(4/octeon) to reduce the need for atomic operations.
- Made it possible to build xterm(1) with OPT_TRACE.
-
- Removed a exit(3) in main(), in ongoing work to make things more readable.
- Changed acme-client(1) to use temporary strings for calls to basename(3) to make the code more portable, as basename can change the contents of its argument on other platforms.
- Fixed 'announce inet none' for bgpd(8). It had cleared more than just the AFI/SAFI pairs where the AFI is inet.