Tomasz Malesinski wrote:
Iriver iFP has small amount of IRAM, so I need to put some things out
of it. I posted a patch (FS#5857) some time ago which controls what
goes to IRAM on different targets.
I simply define {ICODE,IBSS,IDATA}_ATTR_PURPOSE macros, which are
either defined to be empty or to
Peter D'Hoye wrote:
amiconn, preglow:
I'm unable to spend time on this issue, maybe one of you can check it out?
I'm sorry, I've got another busy week and weekend going on.
Thom
David Bleakley wrote:
I agree with XavierGr, although it is really not very difficult to
press the oposite direction once to get to the previous place.
If you keep the button pressed, it will increment faster than 0.1%, so I
don't see
the problem. The problem with pressing the opposite direct
Tomasz Malesinski wrote:
1. Adjusting what is in IRAM and what is not.
I started implementing the idea discussed some time ago on IRC. I
introduced macors like IBSS_ATTR_MPC_SAMPLE_BUF. The macros may be
defined in config-.h to be empty or not. The default values are
defined in config.h or, as
Rocker wrote:
**Any chance of the voice UI on Apple machines? If not, should this be
advised on the user list so we blind folk don't go out and buy IPODS?
Cheers...rocker
According to commit mails, voice UI on iPods was just implemented by
Amiconn.
Thom
Sander Sweers wrote:
Hello All,
As of last Friday my ipod nano produces a battery failure after it gets
disconnected from the pc without eject. After that I need to do a reset
(middle + menu button) for it to come alive again.
This happens to me all the time when developing, and seems to
b
gl wrote:
IDATA_ATTR is for data that is initialised. Then you have IBSS_ATTR
for data which is
uninitialised, and ICONST_ATTR for constant data (lookup tables and
the like).
Thanks, is this type of stuff documented anywhere?
There's not much more to it, and no, no docs AFAIK. This level o
gl wrote:
We don't use malloc and free in Rockbox. Allocate whatever memory you
need with static buffers.
Those two buffer sizes are unknown at compile time: one I can talk to
David about, the other because I was anticipating packing to various
formats, eg. 8bit. I guess I could statically
gl wrote:
I have the Wavpack record code written (for now hard-coded into
pcm_record.c). I can't get it to compile though. For one, malloc
and free (which I use to allocate the block buffers) don't resolve. I
read the basic RB framework primer on the site about the seperation of
firmware
gl wrote:
Just not maintained, and it was for the old Win32 build which we are now
trying to phase out in favour of the SDL one.
Check. Is the old MSDev project still in CVS, or does somebody have a
copy? I googled one link but it's vanished. Would anybody else be
interested in it?
--
gl wrote:
The sim used to be able to be built in MSDev, but not anymore.
Did something break compatibility or was it just not maintained? I'm
happy to update an old project if there is one.
Just not maintained, and it was for the old Win32 build which we are now
trying to phase
out i
gl wrote:
Can the sim be built and debugged on MSDev? Has anybody tried and
wants to share their project?
I've already set up a VC8 project that can do an all-in build via
Cygwin ((re)build, zip, install to device) so you can browse and edit
the code there. If anybody wants that, let me
gl wrote:
Are the H1x0 full-duplex capable?
It would be fun to support basic multitrack recording (record one mono
or stereo track whilst playing back a mix of several others).
I remember the guy originally working on the codec driver in the H1x0
telling
me that it supported full duplex a
Christi Alice Scarborough wrote:
David,
I don't know if I'm alone in this desire, but I think it could be very
valuable for Rockbox to handle correction files, if that's possible
within the constraints of the platform. In this way it would be
possible to carry around lossless versions of your
gl wrote:
I'll check with him. He likes to stay backwards compatible. In the
meantime, those files should be refused. Some seem to be, but others
try to play regardless and cause the problem.
Sure he is, but it might also be that pre-4.0 files require other/more
resources which
would make
gl wrote:
Correction, it seems _older_ (probably 3.98) .wv files don't play
correctly - my newer ones do work. I'll need to check the details...
--
gl
Then small wonder, our decoder (a modified version of the tiny decoder from
wavpack.com) does not support pre-4.0 files. You'll have to as
Christi Alice Scarborough wrote:
Likewise. Although I'm not sure it's worth taking it out before we
have a single image solution.
Christi
I think it should be taken out as soon as possible. It's flawed in the
sense that it can't possibly detect whether the cache is invalid
without destroyin
Bluechip wrote:
Pretty much got a proper core written for an Audio file processor now.
I have another simple question about MPEG Audio files.
Am I ever likely to encounter an MPEG Audio file which will contain
AUDIO frames which differ in VERSION, LAYER, SAMPLERATE and/or
CHANNELS? Eg. Some
Magnus Holmgren wrote:
Jens Arnold wrote:
My patch removes clipping prevention altogether (like we had it
on Archos Recorders & Ondios for years, btw).
And makes fade on stop pretty ugly (sounds get chopped rather than
change smoothly), it seems... :)
Magnus
We should implement fade on
Jens Arnold wrote:
Now I am asking for opinions/ suggestions on this. I would
really like to commit this...
This is more or less exactly what I've wanted all along, so thumbs up
from me.
Thom
Tony Lenox wrote:
At the moment it is: "Prevent clipping: scale volume/scale bass"
If we are to add another setting, I'd also like an option that
doesn't limit anything at all, but just let's the audio potentially
clip.
Thom
Steve Moskovchenko wrote:
Hello.
My friend is looking to upgrade his iH340 to 60GB by buying a 60GB hard
drive of the same size and connector. Is this possible? There is one on
the Toshiba web site but it is only sold to manufacturers. Where can
such a hard drive be purchased?
Thanks
-- Steve
Magnus Holmgren wrote:
Hi,
Should the crossfeed really lower the volume that much? I need to
increased the volume by 10 steps to get back to the "non-crossfeeded"
volume...
Magnus
Should, no. Does, yes. The highpass filter used in the current incarnation
of crossfeed is a one-pole no-zer
Frieder Ferlemann wrote:
Do you know if the input data size assumption is valid?
No I don't. Please excuse me for jumping into the topic without having
any Rockbox context.
I just found out, it's not. The block size will usually be even, but the last
block will sometimes have an odd number o
Frieder Ferlemann wrote:
Hi,
thanks for adding the EMAC routines!)
You're welcome!
a) align labels to the size of the cache line
I guess this is a decent idea, though I was pretty much planning to
put the function in IRAM anyway, since we have no more data to put
there.
b) clears the accumu
Frederic Devernay wrote:
On the processors we use (ColdFire and SH7034), is the following
instruction authorized, even if s and d are not aligned?
((*(long long *)(d)) = (*(long long *)(s)))
Unaligned accesses will be slower than usual accesses on Coldfire, and outright
crash the SH processors
26 matches
Mail list logo