[BRLTTY] BRLTTY in GRUB2

Vladimir 'φ-coder/phcoder' Serbinenko phcoder at gmail.com
Thu Mar 29 19:00:06 EDT 2012


On 30.03.2012 00:45, Dave Mielke wrote:
> [quoted lines by Vladimir 'φ-coder/phcoder' Serbinenko on 2012/03/29 at 18:35 +0200]
>
> Is there a way I can know which headers Grub has in addition to <grub/*.h>?
>
> At the moment, I'm specifically curious about <time.h>.
They are in grub-core/lib/posix-wrap/include. They aren't properly
speaking GRUB header, more of compatibility header.
>> I've added a layer to have fopen/fread/fclose (preferred) and
>> open/read/close.
> So it's okay if we just continue to use those.
OK
> I understand that writing isn't supported, and shall probably be introducing 
> HAVE_WRITE shortly.
I've made an fwrite/write which outputs to console if handler is either
stdout or stderr and returns EROFS on any other handler. Also fopen with
mode other than "r" fails with EROFS.
>> int grub_getkey_noblock (void);
>> Returns GRUB_TERM_NO_KEY if no key is available.
> Am I correct that this is specifically for the console?
Yes
>  I think it'd be best if 
> we don't try to do any keyboard monitoring in grub.
What is keyboard monitoring and how it works?
>> grub_serial_port *port);
> I'm curious: In grub_serial_port, why is it "*next" but "**prev"?
All available serial ports are in double-linked list.
> Is there a function which takes a name and returns a grub_serial_port*,
struct grub_serial_port *grub_serial_find (const char *name);

>
> What do serial port names look like?
com<number>, usb<number>, port<I/O address>
> Am I correct that we should maintain the serial attributes in our copy of a 
> grub_serial_config, and then call grub_serial_port_configure to flush them?
Yes
>> Note that both grub_getkey_noblock and grub_serial_port_fetch consume
>> the character/key if one is available but don't wait if none is.
> Does grub_serial_port_fetch need to do the read to discover that no data is 
> pending, or is it possible for that condition to be tested without doing an 
> actual read? If it's possible to just to the test, is there already a way to do 
> it or could one be added?
This makes little sense for NS8250 since it's just one additional port
read. On USB you need to poll the input anyway for this.


-- 
Regards
Vladimir 'φ-coder/phcoder' Serbinenko


-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 294 bytes
Desc: OpenPGP digital signature
URL: <http://mielke.cc/pipermail/brltty/attachments/20120330/1bce0bbc/attachment.sig>


More information about the BRLTTY mailing list