Re: Bumping arm and m68k toolchains to gcc 4.9.4

cessary up front but would make it fairly easy to spot serious perf regressions and to get any benefits from better optimizers. Glad to see people still working on it. BR Nils >

Re: 540e5d1: Forget about fixedpoint.c in any HWCODEC bin.

rbcodec is no longer looking in global_settings. It's been a while since i looked at it but i think warble uses some of the audio output code from sim/sdlapp. Looks like it also uses some common fw code and some rb headers. Nils

Re: 540e5d1: Forget about fixedpoint.c in any HWCODEC bin.

On Fri, Apr 19, 2013 at 2:39 AM, Michael Sevakis wrote: > Changes over here: > http://gerrit.rockbox.org/r/#/**c/456/ > > BTW, ideally the rbcodec lib shouldn't rely on any code elsewhere in the rb tree.

Re: Opus codec developments

looking at, the only way to know is to bench it. Sections 3.7 and 3.8 in [1] have the instruction timings worth noting is that there's a stall when using movclr after mac on the same accumulator [1] cache.freescale.com/files/32bit/doc/ref_manual/MCF5249UM.pdf Nils

Re: Opus codec developments

hem to c vars. Take a look at how this is done in other functions, there are lots of them, an example is MULT16_32_Q15 in the fixed_generic.h file you were editing. Nils

Re: Freeze update/USB issues

it's not solve, postpone the release a little bit but make it clear that > this is a top priority. I agree, if we don't see promising progress on this in a week or so i think we should unfreeze to avoid getting stuck in a perpetual freeze like for the first attempt at 3.0 Nils

Re: Pre-release testing framework

W. Kernighan I would suggest focusing on things that are more likely to be broken for specific targets, like we've seen with playback/radio/recording while more or less purely sw stuff like dsp or the codecs themselves , while of course worthwile to test, would be more likely to be broken for more targets if they break. Nils

Re: Rockbox's Google Summer of Code 2012 Status

roblem that probably could have been solved more easily with the right questions asked to the right people on irc. I think demanding that the students are available on irc for a while every day they work on the project is not unreasonable. I also think that as much as possible of the student mentor communication should go through open channels so others can chime in and follow the project. I'm sure we have suggested this in the past. * Read: "I :)" End of rant Nils

Re: plugins

don't think any of these require a complicated keymap, most of them probably just use lists anyway. Some people might like pictureflow too. Nils

Re: We've moved to git! No more SVN!

Great job guys! On 1/10/12, Torne Wuff wrote: > Hi folks, > > Just to let everyone know, we have now shut off commit access to the > SVN server, and enabled commit access to the Git repository hosted on > our Gerrit instance. This means that your svn checkouts are now > obsolete, and to get newer

Re: having pluginlib-action handling more/all plugin.

back and testing and perhaps moving on from there as that does not have the same potential for wasted effort and systematic errors as a single huge patch for every plugin has. It's also more likely to get testing than a "test every plugin" patch. Nils

Re: 3.10: Performance regression

> > There was a work in that area for the software mixer. Perhaps that's the > reason? The chunk size was about 32K before that work. Don't know the > current value. > > Best regards. > Should be failry easy to bisect. Nils

Re: Ladies and gentlemen, we have sound on HiFiMan HM-601.

On Mon, Oct 17, 2011 at 10:11 PM, Andrew Ryabinin wrote: > I'm announce it with some delay cause I didn't know about this tradition > till now. > The first time HM-601 played was on the 10.10.2011, but playback speed was > little bit accelerated. > The first time DAP played with normal speed was 1

Re: Im out

t; commits). > > > Jonathan > > P.S This isnt about kugels mailing list thread, though it sure made my > descision easier > Sorry to see you go and thanks for all the hard work. Nils

Re: Git/gerrit migration status and next steps

people to review, if it >> turns out it doesn't work, we disable it. >> >> Nils > > I just don't see why if you're see keen on testing this feature, why you > can't just *voluntarily* have your contributions s

Re: Git/gerrit migration status and next steps

On Wed, Sep 7, 2011 at 10:53 AM, Thomas Martitz wrote: > Am 07.09.2011 10:29, schrieb Nils Wallménius: >> >> I think it could be interesting to test it actually. >> >> > > Testing it will probably show that it works in the beginning, but it will > not show tha

Re: Git/gerrit migration status and next steps

I think it could be interesting to test it actually. Nils

Re: Thoughts on FS#12124 - drawing the lists with the skin engine

cted item Last (so the viewport doesn't move again) and by > allowing you to specify a different viewport label to draw for the > selected item *which doesn't need to have the same rectangle size* > (which should make some nice themes). > Does it work with the "Screen Scrolls Out of View" setting which can cause all items in view to scroll? Nils

Re: Sim build failing?

On Sun, Jun 12, 2011 at 1:29 PM, Magnus Holmgren wrote: > On Sun, Jun 12, 2011 at 10:27, Nils Wallménius > wrote: > >> On Sat, Jun 11, 2011 at 12:11 PM, Thomas Jarosch wrote: > >>> If EOVERFLOW would be the only error, we can work around that. >>> If not

Re: Sim build failing?

t; Thomas > None of the build clients use cygwin. Nils

Re: FS#11931 - Do a short rewind when playback is paused

ng renamed and moved > ("Duration to Rewind" in the pause-phones menu to "Rewind on Pause" in > the playback settings). Sounds like a good idea to me Nils

Re: "Rockbox is an open source firmware for mp3 players"

ever I don't know how to say this is a snappy sound-bite, so no big > complaints here. > > Alex > I like "music player" more than "media player" but maybe "software" isn't needed. Nils

Coldfire toolchain updated!

ygwin if the guide in the wiki was used as that did not previously include a c++ compiler. Installing the g++ package solves this. Happly compiling! Nils

Re: Upgrading coldfire gcc

would be great if someone could please test them: http://dl.dropbox.com/u/17484767/m5_fw.bin http://dl.dropbox.com/u/17484767/x5_fw.bin Nils

Re: Ladies and gentlemen, ...

On Mon, Dec 27, 2010 at 10:36 PM, Michael Sparmann wrote: > ... we have sound on the iPod Classic! Congratulations! Nils

Re: Upgrading coldfire gcc

Forgot one thing: The linux-amd64 bit patch is not needed anymore but i don't know about the CYGWIN/Darwin/FreeBSD/Interix/SunOS patch, could someone with one of these platforms look at that?

Upgrading coldfire gcc

7/rockbox-hd300.zip http://dl.dropbox.com/u/17484767/rockbox-m3.zip http://dl.dropbox.com/u/17484767/rockbox-m5.zip http://dl.dropbox.com/u/17484767/rockbox-x5.zip patches are posted here http://www.rockbox.org/tracker/task/7832 Nils

Re: how to avoid a "make install" on simdisk after each change

On Fri, Dec 3, 2010 at 7:26 PM, Nils Wallménius wrote: > On Fri, Dec 3, 2010 at 5:19 PM, Mike Giacomelli wrote: >> >>> unfortunately I am not fit enough with the Makefiles to do this for >>> myself, so I hope somebody can do this for me. >> >> >>

Re: how to avoid a "make install" on simdisk after each change

s/filesystems with lazy file copying. Nils

Re: FS#8806 - MikMod MOD, S3M, IT, XM player

On Thu, Nov 4, 2010 at 6:08 PM, Thomas Martitz wrote: > The codec buffer is bigger than the plugin buffer, doesn't the plugin buffer > limit it even more? Plugins can capture the audio buffer though. Nils

Re: [ARM] gcc 4.6

k those are harmless and can be left alone untill we want to switch to gcc 4.6 or ffmpeg fixes them. Nils

Re: Segfault with Faster MDCT patch and -fPIC

gcc lets you specify the pic register with -mpic-register= so it would be a qucick test to try with r11 or something. Nils

Re: gain settings in recording screen

pport something that hardware can't > handle. > > Another thing is that currently every driver claims 0.1dB steps for > gain in audiohw_settings[] which should be IMO fixed. > > looking forward to hear what do You think > Marcin Bukat (wodz) > I agree, telling the user the truth is a Good Thing (TM). Nils

Re: P#11260 Task closed by nls: lets keybox import its database from an external text file

aving to re-type the passwords, but they are never transferred > across in plaintext anywhere. > > This seems like a reasonable compromise to me. > Sounds reasonable to me too. Nils

Re: P#11260 Task closed by nls: lets keybox import its database from an external text file

nality inside > Rockbox. > > Any idea would be highly appreciated > Gerhard > > On Wednesday 18 August 2010, Rockbox wrote: >> The following task is now closed: >> >> FS#11260 - lets keybox import its database from an external text file >> User who did thi

Re: Proposing a small change to the sansa fuze keymap.

urself agrue very strongly against that keymap change and you say that you don't use stop so why are you agains this? Even though the fuze and e200 button layouts are similar the non-wheel-button on the front is labeled differently and that "home" stops playback makes little sense to me. Nils

Re: Release 3.7 timeline?

On Mon, Aug 16, 2010 at 6:37 PM, David Hall wrote: > If not RaaA, what will ever justify a bump to 4.0? > > And IF this is 4.0 time, I believe a coordinated publiciy campaign > should be part of the release schedule. > > 4.0 for RaaA sounds like a good idea to me but i don't think it's release qu

Re: Proposing a small change to the sansa fuze keymap.

awkward. as for consistency that depends on what you want to be consistent with. I agree that consistency within the targets' keymap is the most important but i don't think this change makes it less consistent. And last, if keeping it on long home is very important to some people, would it be ok with you to add a stop action to long play/pause anyway? thanks for commenting people :) Nils

Proposing a small change to the sansa fuze keymap.

to either one of them is easy. I prefer moving it to long play/pause since the power switch is kind of hard to operate. Comments? Nils

Re: GPLv3

in fact at least one file, firmware/libc/strstr.c which comes from linux which is gplv2-only. Maybe we should replace that with a more liberally licenced one? Nils

Re: warn_unused_result

able by making the build clients do "make veryclean" before each build? Nils

Re: CRC checks in codecs

That the crc is correct is in no way a guarantee that the data is in spec. However as there seems to be several people against this i'll drop the idea and just use it privately. Nils

Re: CRC checks in codecs

enabled as default would be ok or would you rather i'd leave it alone. I think at least the vorbis loop can be optimized a little but it doesn't really change my opinion of this checking. And as lear stated mp3 does it as well as a bunch of others. Nils

CRC checks in codecs

to disable it or should i make a setting for it? Nils

Re: [RaaA] Weekly status report

llation packages for various OS's. Nils

Re: [RaaA] Weekly status report

hat but i think that can come at a later time, especially if we want to integrate something like that into the build system. Nils

Re: Cowon D2 EABI

r flags, short delay loops in asm solved it and looks much cleaner than the use of volatile or a c loop with a "nop" in it to me. Nils

Re: dev blog? was Re: Cowon D2 EABI

On Tue, Jun 15, 2010 at 9:58 AM, Alex Parker wrote: > > Is it perhaps time to have a news/user front page, and stick all the dev > stuff on a dev "front" page (with a big old link from the new front page? > Alex Yes! Nils

Re: Cowon D2 EABI

days. The news section is hardly used anyway. Nils

Re: svn properties

i say we do that. But maybe we should hold the change off for a few days so people get a bit more chance to comment. Nils

Re: svn properties

ve no objection to switching to forced LF. I also don't object to dropping the keywords property, i see little point in it. The mime-type should be possible to set up server side IIUC, just add the correct property to files with image file endings. Nils

Re: [RaaA] Weekly status report

browse the entire fs. 3) I think your CONFIG_PLATFORM idea is good but would not like reuse of CONFIG_CPU for something else. So i would suggest adding a #define such as APPLICATION that only RaaA defines and that both sim and RaaA defines the PLATFORM_HOSTED. Nils

Re: libcook not working on Solaris

same. > > Regards, > Vikas >> >> Any help will be greatly appreciated. >> >> Regards, >> Vikas >> > Since your machine is big endian and wav files are little endian, you need to byteswap the samples before they are written to the file, i am pretty sure this is your problem, since the standalone app doesn't do this. Nils

Re: Proposal for Bounty to get Linux & Rockbox ported to the iPod Nano 4G and 5G

t; consensus... > > Dave. Mr. Someone's TODO list works (worked? i't sbeen pretty dead for a while) pretty well, with mostly things we agreed on being added and not turning into a general wishlist. So i think a similar wiki page would be fine for this, with a little header saying that. Nils

Re: NoDo or not?

> I just want a key I can assign on the fly so I don't have to drill down to > context menus all the time.  I don't want user-switchable keymaps.  I don't > want to have to load an entirely new keymap just to change functions.  I > merely want what I want, when I want it. :) > > And I feel this is

Re: Idea Discussion: making voice more usable throughout Rock box

On Sun, Mar 28, 2010 at 8:03 PM, Scott Berry wrote: > Yup, I meant the .talk files sorry.  I see your point that's the whole > reason for bringing this up is to see what works and doesn't.  Where do I > find what programming languages you are using so I can get working on some > of these maybe? >

Re: NoDo or not?

I would rather see a solution for completely remapping the keys than this scheme with one or a few remappable buttons only. I know this has been argued to death over and over but the only two arguments against it are a) support nightmare b) unwanted complexity a) is easily dealt with by telling us

Re: Proposal to update GCC for ARM

b, so we > aren't completely in the dark. > > Best regards. > Surely you can build only doom with long enums? (that is if it turns out to be too much work or too intrusive to fix) Nils

Re: Proposal to update GCC for ARM

On Tue, Mar 9, 2010 at 2:13 AM, Thomas Martitz wrote: > Am 09.03.2010 01:50, schrieb Nils: >> >> The flac codec is built with the core optimization level (-O1 on arm). >> Earlier testing done on the gigabeast with gcc 4.4.2 >> showed that compiling the flac codec wit

Re: Proposal to update GCC for ARM

on other arms, we need just a small tweak to get a flac that's even faster than what we currently have. Nils

Re: uchida: r24878 - in trunk/apps: . codecs metadata

On Wed, Feb 24, 2010 at 7:26 PM, Alex Parker wrote: > In addition, personally I'd quite like people to be on IRC when they commit > stuff for just this sort of query.  Anyone agree? > Thanks, > Alex > That does sound like a good rule of thumb. Nils

Re: multifont?

to the biggest argument for the unified buffer, even if i understand you nor anyone else is eager to implement it. That if you use a lot of glyphs from one font and only a few from another you either waste a lot of memory or spin the disk a lot. Nils

Re: ffmpeg coordination effort!

On 2/10/10, Daniel Stenberg wrote: > I don't consider that any less reason to talk to ffmpeg, but just a reason > to also involve the Tremor people as well. Don't you? > > To me, as a person not very closely involved in the codecs area, a way > forward could work like: > > - we setup a separate m

Re: FS#10898 - playlist viewer in the WPS

> So, any thoughts? comments? Cool! I always thought the limitation to only show info on the next track was a bit weird.

Re: Help getting frotz plugin to work on lowmem targets

Have you tried building with -Os?

Re: Bug or intentional?

On Sun, Dec 27, 2009 at 4:25 AM, Paul Louden wrote: > Apparently within the WPS context menu some submenus (for example, sound > settings) can be entered and exited without leaving to the WPS. Meanwhile > others (like Playlist) cannot. Attempting to leave them by pressing "left" > will take you to

Re: I suggest we postpone release 14 days

On Wed, Dec 16, 2009 at 3:07 PM, Daniel Stenberg wrote: > On Wed, 16 Dec 2009, Alex Bennee wrote: > >> Well one suggestion would be to have nominate a person responsible for >> each Flyspray Category area so at least there is a first port of call to >> contact. > > That's not a bad idea. I'm not s

Re: Change the erase dynamic playlist warning to a menu with several options including a quick save

On Sun, Dec 6, 2009 at 10:02 PM, Alex Bennee wrote: > This allows a quick save, or a save via the filename editor rather > than forcing the user to exit and select "Save Playlist" via the other > playlist menus. I like it

Re: Bricked my H300 :(

On Fri, Nov 20, 2009 at 8:40 PM, pondlife wrote: > That makes me feel a bit better - as if I bricked so that you wouldn't have > to...:) > Sorry to hear about your h300 :( IIRC there was another trick to boot the OF, besides REC, double or multi tapping the play/power button, did you try this? >

Re: H300 bootloader - another (final?) push

pondlife, great initiative! I volounteer my h300 as a guniea pig since i don't use it at all anymore but i won't have access to it for another few weeks (~dec 20). My favourite annoyance with the h300 bootloader is that it starts the OF when connecting the charger while it's off, is this fixed in

Re: Tuner in an apparently stable state

I think you have done a great job and that the plugin is in a committable state, if no objections come up, I can commit it for you :) Nils

Re: C99 and EOL comments

On Thu, Aug 6, 2009 at 6:10 PM, David Johnston wrote: > So I understand from what I've read in various places that Rockbox > endeavors to comply with a pre-C99 C standard (C89, I guess, since we > don't use KNR-style function definitions).  My question is, why?  I > think someone mentioned that it'

Re: Fw: Sansa C200V2

ates. That said a port may be usable to you but still lacking in some areas we find important enough to not call it supported. The only way to find out is to test it really and see what you think. Nils

Re: Ladies and gentlemen ...

Congratulations!

Re: Replaygain without a setting, and other menu cleaning.

I'm very much in favour of replaygain being on by default especially since it really has no significant side effects for users that don't have the tags. I don't like the timestretch being on by default mainly because it will just wase ram for most users. On 6/19/09, Paul Louden wrote: > gl wrote:

Re: saving compiler warnings to a file

On Sun, May 31, 2009 at 5:49 PM, Nils wrote: > > > On Sun, May 31, 2009 at 3:14 PM, alex wallis > wrote: > >> Hi list. >> I was wondering if someone could please tell me is there a way when I am >> compiling under cygwin to save any compiler warnings I receiv

Re: saving compiler warnings to a file

On Sun, May 31, 2009 at 3:14 PM, alex wallis wrote: > Hi list. > I was wondering if someone could please tell me is there a way when I am > compiling under cygwin to save any compiler warnings I receive to a file? as > I am getting a few with some patches in place that I would like to find out > i

Re: Mpio HD200

t; Marcin Bukat > I'd suggest registering in our wiki and creating a page there containing info about this player and the port and attaching the x-rays there. Nils

Re: Ladies and Gentlemen we have SOUND! (x2)

On Wed, May 13, 2009 at 11:18 PM, Mark Arigo wrote: > Quietly I've had 3 new ports in progress: Samsung YH-820, 920, and 925. All > 3 are based on the Portalplayer 5020, so the porting has been quite easy > once the LCD was working. Today, I finally had some time and worked on the > audio driver

Re: Buffering two files

As a slightly related note, to make midi work as a codec we would need to be able to buffer megabytes of "patches" so making MoB handle this kind of multi file buffering could help a theoretical future midi codec. Nils

Re: Rockbox as an Application and Google summer of code.

I have never done any coding on any of the platforms where RaaA would be used but i assume our existing UI would be a pretty bad fit in any multi app environment so i think it would make more sense to code a separate UI for RaaA and modularize the rest of rockbox to make it easier to hook this ui i

Re: Building Rockbox faster with make -j

and > uses ccache") and building once, I was able to build again from a clean > state with -j 4 in 4:30. > > -- > Boris > Boris, if you are looking for a major speedup, try building on native linux, i can do a build with make -j3 in about one minute on my 3 year old pentium dual core (using ccache) and your cpu is _much_ faster. Nils

Re: Using gcc 4.4.0 to compile Rockbox

On Tue, Feb 24, 2009 at 4:49 PM, Mark Allums wrote: > Nils wrote: > >> Also could you run some benchmarks on the codecs to see how it affects >> performance? >> >> Nils >> > > > Attached is a spreadsheet in OpenOffice.org .ods format. I can supply

Re: Using gcc 4.4.0 to compile Rockbox

intended to test this myself at some point, keep in mind that crashes in rockbox could come from compiler bugs so don't beat yourself up over them if they make no sense ;) Also could you run some benchmarks on the codecs to see how it affects performance? Nils

Re: Call for testers and DSP brains

e incorrectly. > jhMikeS, committed a fix in r19991 that fixed a coldfire specific lockup bug related to pcm, that might have been causing this too. Nils

Re: Summer of Code 2009

roblem of which buttons to use and how but i think these could be interesting problems to solve. So what do people think of this idea? Nils

Re: Default settings for the release

On Wed, Dec 17, 2008 at 1:28 PM, Paul Louden wrote: > Nils wrote: > >> >> I'm not so sure about turning replaygain on by default, sure _I_ use it >> but for people with messy collections who may have some tracks with RG tags >> and some without, this could

Re: Default settings for the release

ndlife > > I'm not so sure about turning replaygain on by default, sure _I_ use it but for people with messy collections who may have some tracks with RG tags and some without, this could lead to pretty big jumps in volume. Nils

Re: Default settings for the release

> > Warn when erasing dynamic playlist > > I think this warning should be ON by default. The dynamic playlists is one > of the most misunderstood features of Rockbox and this warning helps users > understand what is going on. > I'd like to have this On by default too. Nils

Re: Default settings for the release

t's less annoying but doesn't defeat the purpose of the > setting. The point of defaults is to be reasonably nice to most users and especially new users because experienced users will know how to change settings anyway. Nils

Re: No compiler optimization for libmad?

it was lost in the rework. I have committed a fix now, thanks! Nils

Re: patch discussion - FS#9173

On Tue, Nov 25, 2008 at 4:00 PM, Jonathan Gordon <[EMAIL PROTECTED]> wrote: > Nils message about a tracker cleanup week has made me decide to try > and either get this committed or dropped. > http://www.rockbox.org/tracker/task/9173 > > The aim of this patch is to centra

Tracker cleanup week!

rm bugs, commit patches that are deemed good and of course fix bugs. I changed all open tasks that were marked "due in 3.0" to "due in 3.1" mark the tasks you feel are especially important for the next release as "due in 3.1". For the best 3.1 we can create! :) Nils

Re: Backlight fading - FS#6800

gets have the 2 options so it's better to be consistent > (unless this could be simplified for them too). > > pondlife > > If i would use backlight fading (still don't have a target that supports it), i would probably enable only fade out or a very short fade in so i think it is good to have two options here. Nils

Re: give .cfg settings grouping names?

eal" benefit I agree. - could lead to out-of-menu options > I don't understand what you mean by this, how could grouping the settings lead to more out-of-menu options? (btw we already have at least a couple.) Nils

Re: Should PLAY button always go to the WPS?

othing at all in the playlist viewer, and being able to stop in there would be nice :) Nils

Re: Should PLAY button always go to the WPS?

to start by accident. I think it should be made into a menu item, System -> disk -> free space scan maybe? Nils

Re: Release schedule proposal

I like the idea of time based releases too, waiting around for "significant" features will make the time frame very unpredictable. If we have to wait a long time, more issues are likely to have popped up, documentation more out of sync etc. so each release would mean a lot more work IMHO.

Re: [PATCH] Building ARM cross compiler on ubuntu 8.10

> > > Maybe it's time we consider upgrading binutils? That should not bring any > major issues (tm) and has the added bonus of including amiconn's fix for a > gas bug hitting coldfire. It is also a requirement for building gcc =< 4.3 > for coldfire (without patches)

Re: [PATCH] Building ARM cross compiler on ubuntu 8.10

ecause of a buffer overflow. > > -- > Rafaël Carré > Maybe it's time we consider upgrading binutils? That should not bring any major issues (tm) and has the added bonus of including amiconn's fix for a gas bug hitting coldfire. It is also a requirement for building gcc =< 4.3 for coldfire (without patches). nils

  1   2   >