diff options
author | Alan Cox <alan@lxorguk.ukuu.org.uk> | 2007-02-10 01:45:57 -0800 |
---|---|---|
committer | Linus Torvalds <torvalds@woody.linux-foundation.org> | 2007-02-11 10:51:32 -0800 |
commit | 78137e3b34e122949e6de36a894fb5843664b8f9 (patch) | |
tree | cca5f5409643bf6adbe8387203ed0fa076d36cb1 /lib/kobject.c | |
parent | 996a07bcb62c5935248e238a1150089f3d99a6fb (diff) | |
download | lwn-78137e3b34e122949e6de36a894fb5843664b8f9.tar.gz lwn-78137e3b34e122949e6de36a894fb5843664b8f9.zip |
[PATCH] tty: improve encode_baud_rate logic
Mostly so people can see the work in progress. This enhances the encode
function which isn't currently used in the base tree but is when using some of
the testing tty patches.
This resolves a problem with some hardware where applications got confusing
information from the tty ioctls. Correct but confusing.
In some situations asking for, say, 9600 baud actually gets you 9595 baud or
similar near-miss values. With the old code this meant that a request for
B9600 got a return of BOTHER, 9595 which programs interpreted as a failure.
The new code now works on the following basis
- If you ask for specific rate via BOTHER, you get a precise return
- If you ask for a standard Bfoo rate and the result is close you get a Bfoo
return
- If you ask for a standard Bfoo rate and get something way off you get a
BOTHER/rate return
This seems to fix up the cases I've found where this broke compatibility.
Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Diffstat (limited to 'lib/kobject.c')
0 files changed, 0 insertions, 0 deletions