| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
|
| |
callbacks
|
| |
|
|
| |
device timers are instantiated
|
| | |
|
| |
|
|
|
| |
Saves a lot of typing { *this }, { *this }... Could be applied in more places,
I just did a few devices to demonstrate it.
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
There are a couple of issues here:
* Firstly, the determining factor for whether I64 or ll is needed as the
size qualifier doesn't depend entirely on whether you're using MSVC or
not, it depends on the the C runtime library. The msys64 C runtime
apes the Microsoft runtime and requires I64 as well, so a lot of files
where LLFORMAT had been copy/pasted wouldn't work right if you
depeneded on it anyway.
* Secondly, one of the major benefits of util::string_format,
util::stream_format and logerror in their current forms is that you
don't need size specifiers for integers at all (unless you're using
them to truncate a value, but in that case you're better off casting
the argument for clarity). If you do use them, logerror and friends
will accept both Microsoft- and SUS-style independently of the C
runtime.
So please everyone, don't go gratuitously throwing these around, and if
you must use them (e.g. for osd_printf_* functions), make sure your
conditionals are actually correct. Also, conisder using
util::string_format to do the formatting into a temporary string and
just feeding that to the output function so you can avoid the mess.
|
| | |
|
| |
|
|
| |
methods for RSA and RSB inputs
|
| |
|
|
| |
Obviously, the cycle time has to be divided by two, not multiplied by two. The divider isn't necessarily an even number, though.
|
| |
|
|
|
|
|
| |
- Streamline divider calculation
- Rewrite comments (less copying from the datasheet)
- Implement device_clock_changed and use clocks_to_attotime
- Add timer_enable method for future use
|
| | |
|
| |
|