You can subscribe to this list here.
2007 |
Jan
|
Feb
|
Mar
|
Apr
(22) |
May
(36) |
Jun
(27) |
Jul
(55) |
Aug
(86) |
Sep
(8) |
Oct
(78) |
Nov
(58) |
Dec
(40) |
---|---|---|---|---|---|---|---|---|---|---|---|---|
2008 |
Jan
(100) |
Feb
(29) |
Mar
(176) |
Apr
(64) |
May
(199) |
Jun
(84) |
Jul
(31) |
Aug
(57) |
Sep
(61) |
Oct
(148) |
Nov
(125) |
Dec
(106) |
2009 |
Jan
(15) |
Feb
(34) |
Mar
(42) |
Apr
(66) |
May
(101) |
Jun
(41) |
Jul
(12) |
Aug
(14) |
Sep
(41) |
Oct
(82) |
Nov
(52) |
Dec
(76) |
2010 |
Jan
(158) |
Feb
(53) |
Mar
(32) |
Apr
(43) |
May
(44) |
Jun
(40) |
Jul
(49) |
Aug
(43) |
Sep
(123) |
Oct
(78) |
Nov
(41) |
Dec
(135) |
2011 |
Jan
(67) |
Feb
(70) |
Mar
(83) |
Apr
(97) |
May
(47) |
Jun
(65) |
Jul
(48) |
Aug
(8) |
Sep
(84) |
Oct
(28) |
Nov
(8) |
Dec
(9) |
2012 |
Jan
(26) |
Feb
(19) |
Mar
(220) |
Apr
(262) |
May
(107) |
Jun
(145) |
Jul
(49) |
Aug
(38) |
Sep
(111) |
Oct
(32) |
Nov
(34) |
Dec
(45) |
2013 |
Jan
(28) |
Feb
(4) |
Mar
(6) |
Apr
(26) |
May
(38) |
Jun
(79) |
Jul
(49) |
Aug
(70) |
Sep
(76) |
Oct
(85) |
Nov
(80) |
Dec
(70) |
2014 |
Jan
(167) |
Feb
(114) |
Mar
(181) |
Apr
(95) |
May
(211) |
Jun
(41) |
Jul
(41) |
Aug
(14) |
Sep
(52) |
Oct
(60) |
Nov
(59) |
Dec
(72) |
2015 |
Jan
(88) |
Feb
(20) |
Mar
(89) |
Apr
(15) |
May
(40) |
Jun
(25) |
Jul
(50) |
Aug
(97) |
Sep
(41) |
Oct
(90) |
Nov
(88) |
Dec
(152) |
2016 |
Jan
(60) |
Feb
(52) |
Mar
(63) |
Apr
(79) |
May
(41) |
Jun
(66) |
Jul
(31) |
Aug
(28) |
Sep
(4) |
Oct
(48) |
Nov
(45) |
Dec
(3) |
2017 |
Jan
(44) |
Feb
(23) |
Mar
(42) |
Apr
(39) |
May
(24) |
Jun
(42) |
Jul
(1) |
Aug
(14) |
Sep
|
Oct
(4) |
Nov
(7) |
Dec
(14) |
2018 |
Jan
(67) |
Feb
(70) |
Mar
(19) |
Apr
(35) |
May
(25) |
Jun
(9) |
Jul
(1) |
Aug
(6) |
Sep
(13) |
Oct
(29) |
Nov
(15) |
Dec
(20) |
2019 |
Jan
(35) |
Feb
(4) |
Mar
(4) |
Apr
(1) |
May
(6) |
Jun
(2) |
Jul
(7) |
Aug
|
Sep
(6) |
Oct
(5) |
Nov
(3) |
Dec
|
2020 |
Jan
|
Feb
|
Mar
(15) |
Apr
|
May
(13) |
Jun
(2) |
Jul
(4) |
Aug
(48) |
Sep
(15) |
Oct
(2) |
Nov
(9) |
Dec
(5) |
2021 |
Jan
|
Feb
|
Mar
(2) |
Apr
(3) |
May
(34) |
Jun
(8) |
Jul
(4) |
Aug
|
Sep
(6) |
Oct
(28) |
Nov
(2) |
Dec
(4) |
2022 |
Jan
(1) |
Feb
(5) |
Mar
|
Apr
(16) |
May
|
Jun
|
Jul
(1) |
Aug
(6) |
Sep
(6) |
Oct
|
Nov
(1) |
Dec
|
2023 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
(5) |
Aug
|
Sep
(2) |
Oct
|
Nov
|
Dec
|
2024 |
Jan
(2) |
Feb
(3) |
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
From: Alyssa R. <hi...@al...> - 2024-02-10 12:13:40
|
> You mentioned that musl libc required linking against argp-standalone. If > you would like to prepare a follow up patch to the SCons build scripts which > > * detects whether musl libc is in use, and > > * adds argp-standalone to the link libraries if musl libc is in use, > > I would be happy to merge it too. This would avoid the need to know that > argp-standalone must be added to the linker library list when building > against musl libc. This would be great, but I don't know anything about SCons, so I don't know how to do it. For the build systems I am familiar with, the best way to do this is to check whether argp is provided by libc, and then fall back to depending on argp-standalone if not. This way there's no need to hardcode which libcs provide it and which don't. |
From: Jonathan W. <jw...@ju...> - 2024-02-07 11:39:09
|
Hi Alyssa On Mon, Feb 05, 2024 at 12:29:53PM +1030, Jonathan Woithe wrote: > On Tue, Jan 30, 2024 at 10:34:54PM +0100, Alyssa Ross wrote: > > * <stdint.h> is needed for int32_t > > * sigval_t is a non-standard alias for union sigval. > > > > With these two changes, as well as linking with argp-standalone, I was > > able to build ffado for musl libc. > > Thanks for this contribution and apologies for not responding sooner. I > hope to have time this week to verify it and merge it into our repository. > Once that happens it will be included in the next version of FFADO when that > is released. The timing of the next version is not yet known. Your patch has been added to the FFADO repository as commit r2846. Thanks for preparing it. You mentioned that musl libc required linking against argp-standalone. If you would like to prepare a follow up patch to the SCons build scripts which * detects whether musl libc is in use, and * adds argp-standalone to the link libraries if musl libc is in use, I would be happy to merge it too. This would avoid the need to know that argp-standalone must be added to the linker library list when building against musl libc. Regards jonathan |
From: Jonathan W. <jw...@ju...> - 2024-02-05 02:00:24
|
Hi Alyssa On Tue, Jan 30, 2024 at 10:34:54PM +0100, Alyssa Ross wrote: > * <stdint.h> is needed for int32_t > * sigval_t is a non-standard alias for union sigval. > > With these two changes, as well as linking with argp-standalone, I was > able to build ffado for musl libc. Thanks for this contribution and apologies for not responding sooner. I hope to have time this week to verify it and merge it into our repository. Once that happens it will be included in the next version of FFADO when that is released. The timing of the next version is not yet known. Regards jonathan |
From: Alyssa R. <hi...@al...> - 2024-01-30 21:51:18
|
* <stdint.h> is needed for int32_t * sigval_t is a non-standard alias for union sigval. With these two changes, as well as linking with argp-standalone, I was able to build ffado for musl libc. --- libffado/libffado/ffado.h.in.orig 2024-01-30 22:19:13.561711246 +0100 +++ libffado/libffado/ffado.h.in 2024-01-30 22:19:25.936762584 +0100 @@ -30,6 +30,7 @@ #define FFADO_MAX_NAME_LEN 256 +#include <stdint.h> #include <stdlib.h> #define FFADO_STREAMING_MAX_URL_LENGTH 2048 --- libffado/src/libutil/PosixMessageQueue.h.orig 2024-01-30 22:25:02.558221614 +0100 +++ libffado/src/libutil/PosixMessageQueue.h 2024-01-30 22:25:36.462373787 +0100 @@ -115,7 +115,7 @@ private: bool doOpen(enum eDirection t, int, enum eBlocking); - static void notifyCallbackStatic(sigval_t t) { + static void notifyCallbackStatic(union sigval t) { PosixMessageQueue *obj; obj = static_cast<PosixMessageQueue *>(t.sival_ptr); obj->notifyCallback(); |
From: Jonathan W. <jw...@ju...> - 2024-01-06 18:35:02
|
The FFADO project (https://ffado.org) announces the availability of FFADO version 2.4.8. This is a maintenance release to address issues encountered since version 2.4.7. This is a source-only release which can be downloaded from http://ffado.org/files/libffado-2.4.8.tgz Changes since FFADO 2.4.7: * Make various mixer modules work under Python 3.10 and later: Presonus Firebox, Presonus FP10, Presonus Inspire1394 and Yamaha Go. * Deal with the removal of SafeConfigParser in Python 3.12. Thanks to those who have helped with this release, including Andrew Scherkus and Nils Philippsen. Jonathan Woithe (on behalf of ffado.org) |
From: Jonathan W. <jw...@ju...> - 2023-09-30 12:38:15
|
Hi Nils On Sat, Sep 30, 2023 at 02:03:16PM +0200, Nils Philippsen wrote: > Python 3.12 finally removed SafeConfigParser which has been deprecated > since Python 3.2, this prevents ffado-mixer from being used on systems > where this version is the default (like Fedora 39). The attached patch > switches to using ConfigParser instead, I’ve applied it in the Fedora > package. Thanks for the patch. It's been applied to trunk as r2842 and will be in the next version of FFADO. Regards jonathan |
From: Nils P. <ni...@ti...> - 2023-09-30 12:20:38
|
Hi! Python 3.12 finally removed SafeConfigParser which has been deprecated since Python 3.2, this prevents ffado-mixer from being used on systems where this version is the default (like Fedora 39). The attached patch switches to using ConfigParser instead, I’ve applied it in the Fedora package. Cheers, Nils -- Nils Philippsen / Wilhelmstraße 22 / 71229 Leonberg / Germany ni...@ti... / ni...@re... PGP fingerprint: D0C1 1576 CDA6 5B6E BBAE 95B2 7D53 7FCA E9F6 395D |
From: Andrew S. <asc...@gm...> - 2023-07-06 15:35:55
|
On Thu, Jul 6, 2023 at 2:15 AM Jonathan Woithe <jw...@ju...> wrote: > Hi Andrew > > On Wed, Jul 05, 2023 at 11:58:54AM -0400, Andrew Scherkus wrote: > > On Mon, Jul 3, 2023 at 7:22 PM Jonathan Woithe <jw...@ju...> > wrote: > > > > Based on code inspection alone I'm pretty sure presonus_fp10.py and > > > > others have the same bug as it seems to be rooted in calling > > > > setValue() with the output of db2vol() or pan2val() ... but given > I'm > > > > new to this project and don't have those devices on hand I'm a little > > > > wary of making changes without being unable to test! > > > > > > You are quite possibly right. If you'd like to prepare a draft patch > > > containing the changes you think are needed I would be happy to review > it > > > and apply as required. > > > > I did a manual audit of all setValue() calls and fixed a few more spots. > > Thanks for doing that and providing the patch. Your fixes for the FP10, > Inspire1394 and Yamaha Go mixer modules have been committed as r2841. > Thanks! > Based on my understanding, dbus_util.py already returns ints so casts are > > only required in spots where some additional floating-point math is > > performed to convert volume/pan values into UI slider values. > > I haven't investigated the issue too deeply. The common pattern to the > issue we've seen in the past has been > > var = self.hw.getContignuous(...) > [potentially do some math on "var"] > ctl.setValue(var) > > getContignuous() did not always include the int() cast in its return > value. > It was added because the return value of the getValue() call ended up being > a float under python 3.10 if my memory serves me correctly. Putting the > casts in getContignuous() was an attempt to allow for all ctrl.setValue() > calls which passed the value through. However, evidently some intermediate > math operations are performed at least at some call sites, which is why the > casts you've identified are still needed. > Makes sense. I had assumed based on the naming of getContignuous() vs getDiscrete() that the former would return floats and the latter ints, but I understand how that'd seem like an easy spot to add a fix except for those few places that ended up doing additional floating-point math! > > The only other tricky one to audit was matrixmixer.py where it does still > > use floats in some places but they're either rendered as text, used > > with QDoubleSpinBox that accepts floats as values, or uses helper > functions > > like fromDBvalue() that round and cast to int already. It's possible > there > > are still some missing casts lurking in there. > > The MatrixMixer functionality has, I believe, been reasonably well > exercised > under python 3.10 by other users in the past. They provided a number of > type-related fixes to address the issues they uncovered with their devices: > > * r2818, 2021-11-28, from Nils Philippsen > * r2821, 2022-02-15, from Daniel Baeuerlein > * r2825, 2022-04-15, from Nils Philippsen > * r2826, 2022-04-26, from Edmund Raile > * r2827, 2022-04-29, from Edmund Raile > > As a result I think MatrixMixer is probably okay now. However, that > doesn't > mean that ongoing use won't turn up one or two others, as you suggest. > > Regards > jonathan > |
From: Jonathan W. <jw...@ju...> - 2023-07-06 06:15:22
|
Hi Andrew On Wed, Jul 05, 2023 at 11:58:54AM -0400, Andrew Scherkus wrote: > On Mon, Jul 3, 2023 at 7:22 PM Jonathan Woithe <jw...@ju...> wrote: > > > Based on code inspection alone I'm pretty sure presonus_fp10.py and > > > others have the same bug as it seems to be rooted in calling > > > setValue() with the output of db2vol() or pan2val() ... but given I'm > > > new to this project and don't have those devices on hand I'm a little > > > wary of making changes without being unable to test! > > > > You are quite possibly right. If you'd like to prepare a draft patch > > containing the changes you think are needed I would be happy to review it > > and apply as required. > > I did a manual audit of all setValue() calls and fixed a few more spots. Thanks for doing that and providing the patch. Your fixes for the FP10, Inspire1394 and Yamaha Go mixer modules have been committed as r2841. > Based on my understanding, dbus_util.py already returns ints so casts are > only required in spots where some additional floating-point math is > performed to convert volume/pan values into UI slider values. I haven't investigated the issue too deeply. The common pattern to the issue we've seen in the past has been var = self.hw.getContignuous(...) [potentially do some math on "var"] ctl.setValue(var) getContignuous() did not always include the int() cast in its return value. It was added because the return value of the getValue() call ended up being a float under python 3.10 if my memory serves me correctly. Putting the casts in getContignuous() was an attempt to allow for all ctrl.setValue() calls which passed the value through. However, evidently some intermediate math operations are performed at least at some call sites, which is why the casts you've identified are still needed. > The only other tricky one to audit was matrixmixer.py where it does still > use floats in some places but they're either rendered as text, used > with QDoubleSpinBox that accepts floats as values, or uses helper functions > like fromDBvalue() that round and cast to int already. It's possible there > are still some missing casts lurking in there. The MatrixMixer functionality has, I believe, been reasonably well exercised under python 3.10 by other users in the past. They provided a number of type-related fixes to address the issues they uncovered with their devices: * r2818, 2021-11-28, from Nils Philippsen * r2821, 2022-02-15, from Daniel Baeuerlein * r2825, 2022-04-15, from Nils Philippsen * r2826, 2022-04-26, from Edmund Raile * r2827, 2022-04-29, from Edmund Raile As a result I think MatrixMixer is probably okay now. However, that doesn't mean that ongoing use won't turn up one or two others, as you suggest. Regards jonathan |
From: Andrew S. <asc...@gm...> - 2023-07-05 15:59:13
|
On Mon, Jul 3, 2023 at 7:22 PM Jonathan Woithe <jw...@ju...> wrote: > Hi Andrew > > On Mon, Jul 03, 2023 at 05:21:49PM -0400, Andrew Scherkus wrote: > > I spent some time playing around with FFADO and my PreSonus Firebox and > > found there were some missing float to int casts when using ffado-mixer > > when running with Python 3.10.6. Looks like there were some similar fixes > > over the past year. > > Indeed. We're slowly identifying the locations where these casts are > required as more systems shift to python 3.10 or later. Thanks for your > patch: it has been applied as r2840. > Awesome - thanks! > > Based on code inspection alone I'm pretty sure presonus_fp10.py and > others > > have the same bug as it seems to be rooted in calling setValue() with the > > output of db2vol() or pan2val() ... but given I'm new to this project and > > don't have those devices on hand I'm a little wary of making changes > > without being unable to test! > > You are quite possibly right. If you'd like to prepare a draft patch > containing the changes you think are needed I would be happy to review it > and apply as required. > I did a manual audit of all setValue() calls and fixed a few more spots. Based on my understanding, dbus_util.py already returns ints so casts are only required in spots where some additional floating-point math is performed to convert volume/pan values into UI slider values. The only other tricky one to audit was matrixmixer.py where it does still use floats in some places but they're either rendered as text, used with QDoubleSpinBox that accepts floats as values, or uses helper functions like fromDBvalue() that round and cast to int already. It's possible there are still some missing casts lurking in there. Andrew |
From: Jonathan W. <jw...@ju...> - 2023-07-03 23:53:24
|
Hi Andrew On Mon, Jul 03, 2023 at 05:21:49PM -0400, Andrew Scherkus wrote: > I spent some time playing around with FFADO and my PreSonus Firebox and > found there were some missing float to int casts when using ffado-mixer > when running with Python 3.10.6. Looks like there were some similar fixes > over the past year. Indeed. We're slowly identifying the locations where these casts are required as more systems shift to python 3.10 or later. Thanks for your patch: it has been applied as r2840. > Based on code inspection alone I'm pretty sure presonus_fp10.py and others > have the same bug as it seems to be rooted in calling setValue() with the > output of db2vol() or pan2val() ... but given I'm new to this project and > don't have those devices on hand I'm a little wary of making changes > without being unable to test! You are quite possibly right. If you'd like to prepare a draft patch containing the changes you think are needed I would be happy to review it and apply as required. Regards jonathan |
From: Andrew S. <asc...@gm...> - 2023-07-03 21:22:08
|
Hi there, I spent some time playing around with FFADO and my PreSonus Firebox and found there were some missing float to int casts when using ffado-mixer when running with Python 3.10.6. Looks like there were some similar fixes over the past year. Based on code inspection alone I'm pretty sure presonus_fp10.py and others have the same bug as it seems to be rooted in calling setValue() with the output of db2vol() or pan2val() ... but given I'm new to this project and don't have those devices on hand I'm a little wary of making changes without being unable to test! Thanks in advance, Andrew |
From: Jonathan W. <jw...@ju...> - 2022-11-27 11:51:40
|
The FFADO project announces the availability of FFADO version 2.4.7. This is a bug-fix release to address issues encountered since version 2.4.6. This is a source-only release: libffado-2.4.7.tgz. Changes since FFADO 2.4.6: * Address additional type-related issues encounted in ffado-mixer when using python 3.10. Components affected were the crossbar router and the M-Audio device mixer. * Update config.guess to the latest version. This supports newer architectures such as RISCV64. * Address compiler warnings reported by newer versions of gcc. * Add a configuration entry for the Apogee Duet. Thanks to those who have helped with this release, including Salvador O.G., Letu Ren, Nils Philippsen and Greg Dorian. Jonathan Woithe (on behalf of ffado.org) |
From: Pander <pa...@us...> - 2022-09-04 09:44:28
|
El 4 de septiembre de 2022 10:17:04 CEST, Jonathan Woithe <jw...@ju...> escribió: >Hi Pander > >On Sat, Sep 03, 2022 at 04:08:08PM +0200, Pander via FFADO-devel wrote: >> El 3 de septiembre de 2022 15:11:52 CEST, Jonathan Woithe <jw...@ju...> escribió: >> >This python 3.10 issue is slightly different to the ones we've found >> >earlier. Those have been associated with return values from the dbus >> >controls, whereas this one affects widget geometries. >> >> Perhaps use this to prevent more of these issues >> >> https://docs.python.org/3/library/typing.html > >Thanks for the suggestion. It sounds potentially helpful, but not being an >expert in python myself, I can't immediately see how we could use the typing >module to fix (or at least highlight) the type-related issues introduced by >Python 3.10. I'm happy to read up on the details if there's a description >of the procedure somewhere. Of course we're happy to consider contributed >patches which might do this. Note however that these shouldn't break >ffado-mixer on earlier python versions. I haven't used it myself yet, only heard a lot of good things about it. > >Regards > jonathan |
From: Jonathan W. <jw...@ju...> - 2022-09-04 08:17:16
|
Hi Pander On Sat, Sep 03, 2022 at 04:08:08PM +0200, Pander via FFADO-devel wrote: > El 3 de septiembre de 2022 15:11:52 CEST, Jonathan Woithe <jw...@ju...> escribió: > >This python 3.10 issue is slightly different to the ones we've found > >earlier. Those have been associated with return values from the dbus > >controls, whereas this one affects widget geometries. > > Perhaps use this to prevent more of these issues > > https://docs.python.org/3/library/typing.html Thanks for the suggestion. It sounds potentially helpful, but not being an expert in python myself, I can't immediately see how we could use the typing module to fix (or at least highlight) the type-related issues introduced by Python 3.10. I'm happy to read up on the details if there's a description of the procedure somewhere. Of course we're happy to consider contributed patches which might do this. Note however that these shouldn't break ffado-mixer on earlier python versions. Regards jonathan |
From: Jonathan W. <jw...@ju...> - 2022-09-04 08:08:12
|
Hi Nils On Sat, Sep 03, 2022 at 10:41:56PM +0930, Jonathan Woithe wrote: > On Sat, Sep 03, 2022 at 12:43:03AM +0200, Nils Philippsen wrote: > > I tripped another one of these tonight -- this time it crashed when > > switching to the crossbar router tab. I've attached the patch that > > fixes it for me. > > Thanks for sending this through. I'll apply it when I'm next at by > development system, which will be within a day or so. It's now applied as r2835. Thanks again for the patch. Regards jonathan |
From: Pander <pa...@us...> - 2022-09-03 14:33:13
|
El 3 de septiembre de 2022 15:11:52 CEST, Jonathan Woithe <jw...@ju...> escribió: >Hi Nils > >On Sat, Sep 03, 2022 at 12:43:03AM +0200, Nils Philippsen wrote: >> I tripped another one of these tonight -- this time it crashed when >> switching to the crossbar router tab. I've attached the patch that >> fixes it for me. > >Thanks for sending this through. I'll apply it when I'm next at by >development system, which will be within a day or so. > >This python 3.10 issue is slightly different to the ones we've found >earlier. Those have been associated with return values from the dbus >controls, whereas this one affects widget geometries. Perhaps use this to prevent more of these issues https://docs.python.org/3/library/typing.html > >Regards > jonathan > > >_______________________________________________ >FFADO-devel mailing list >FFA...@li... >https://lists.sourceforge.net/lists/listinfo/ffado-devel |
From: Jonathan W. <jw...@ju...> - 2022-09-03 13:12:08
|
Hi Nils On Sat, Sep 03, 2022 at 12:43:03AM +0200, Nils Philippsen wrote: > I tripped another one of these tonight -- this time it crashed when > switching to the crossbar router tab. I've attached the patch that > fixes it for me. Thanks for sending this through. I'll apply it when I'm next at by development system, which will be within a day or so. This python 3.10 issue is slightly different to the ones we've found earlier. Those have been associated with return values from the dbus controls, whereas this one affects widget geometries. Regards jonathan |
From: Nils P. <ni...@ti...> - 2022-09-02 23:01:52
|
Hey Jonathan, I tripped another one of these tonight -- this time it crashed when switching to the crossbar router tab. I've attached the patch that fixes it for me. Cheers, Nils -- Nils Philippsen <ni...@ti...> |
From: Jonathan W. <jw...@ju...> - 2022-08-20 12:25:13
|
Hi Benoît On Fri, Aug 19, 2022 at 02:53:57PM +0200, Benoît Delcour wrote: > Of course the ffado project is welcoming any patch making life easier > for distribution maintainers. Yes, we have upgraded config.guess as Letu suggested. Hopefully this will address that particular issue. We would certainly welcome other patches if they are needed to work with your architecture. > PS2: @jonathan > benoit@poulidor:~$ ffado-mixer > Traceback (most recent call last): > File "/usr/share/ffado-mixer-qt4/ffado/panelmanager.py", line 460, > in updatePanels > self.addPanel(idx) > File "/usr/share/ffado-mixer-qt4/ffado/panelmanager.py", line 339, in addPanel > mixerwidget.initValues() > File "/usr/share/ffado-mixer-qt4/ffado/mixer/maudio_bebob.py", line > 601, in initValues > ctl.setValue(vol) > TypeError: setValue(self, int): argument 1 has unexpected type 'float' Thanks for reporting this bug. It has already been fixed in our repository (see r2832) and will be present in the next version of FFADO. Regards jonathan |
From: Letu R. <fan...@gm...> - 2022-08-19 16:22:04
|
Hi, Benoît > > Hi, I'm building packages for ArchLinux RISCV. > I would be extremely interested if there is at least 1 SBC featuring > a RISCV and PCI-E. > As I read it, your report is about building some software for which > the hardware support is not there. > Of course the ffado project is welcoming any patch making life easier > for distribution maintainers. I'm using SiFive HiFive Unmatched which has x16 PCIe® Gen 3 Expansion Slot (8-lanes Useable). RISC-V with PCI-E support. Really cool! |
From: Benoît D. <ben...@gm...> - 2022-08-19 12:59:31
|
Hi, jonathan and Letu. > Hi, I'm building packages for ArchLinux RISCV. I would be extremely interested if there is at least 1 SBC featuring a RISCV and PCI-E. As I read it, your report is about building some software for which the hardware support is not there. Of course the ffado project is welcoming any patch making life easier for distribution maintainers. Back to guitar, audition next week. PS: RISCV + PCI 64 bit would be even better, because I already own the TI card... PS2: @jonathan benoit@poulidor:~$ ffado-mixer Traceback (most recent call last): File "/usr/share/ffado-mixer-qt4/ffado/panelmanager.py", line 460, in updatePanels self.addPanel(idx) File "/usr/share/ffado-mixer-qt4/ffado/panelmanager.py", line 339, in addPanel mixerwidget.initValues() File "/usr/share/ffado-mixer-qt4/ffado/mixer/maudio_bebob.py", line 601, in initValues ctl.setValue(vol) TypeError: setValue(self, int): argument 1 has unexpected type 'float' Abandon On M-Audio FW410. I didn't even test the motu ultralite, more to come when I'm done w/ stressful real life. Le mer. 17 août 2022 à 05:17, Letu Ren <fan...@gm...> a écrit : > > Hi, jonathan > > > Thanks for bringing this to our attention. I have updated config.guess to > > the latest version as you suggest. This is in r2833 and will be in the next > > FFADO release. > > Thanks for quick fix > > > _______________________________________________ > FFADO-devel mailing list > FFA...@li... > https://lists.sourceforge.net/lists/listinfo/ffado-devel |
From: Letu R. <fan...@gm...> - 2022-08-17 03:17:17
|
Hi, jonathan > Thanks for bringing this to our attention. I have updated config.guess to > the latest version as you suggest. This is in r2833 and will be in the next > FFADO release. Thanks for quick fix |
From: Jonathan W. <jw...@ju...> - 2022-08-16 23:22:48
|
Hi Letu On Tue, Aug 16, 2022 at 04:39:19PM +0800, Letu Ren wrote: > Hi, I'm building packages for ArchLinux RISCV. Building fails because > admin/config.guess is outdated and cannot recognize RISCV64 > architecture. The log is listed below. > : Thanks for bringing this to our attention. I have updated config.guess to the latest version as you suggest. This is in r2833 and will be in the next FFADO release. Regards jonathan |
From: Letu R. <fan...@gm...> - 2022-08-16 08:39:38
|
Hi, I'm building packages for ArchLinux RISCV. Building fails because admin/config.guess is outdated and cannot recognize RISCV64 architecture. The log is listed below. ``` admin/config.guess: unable to guess system type This script, last modified 2013-06-10, has failed to recognize the operating system you are using. It is advised that you download the most up to date version of the config scripts from http://git.savannah.gnu.org/gitweb/?p=config.git;a=blob_plain;f=config.guess;hb=HEAD and http://git.savannah.gnu.org/gitweb/?p=config.git;a=blob_plain;f=config.sub;hb=HEAD If the version you run (admin/config.guess) is already up to date, please send the following data and any information you think might be pertinent to <con...@gn...> in order to provide the needed information to handle your system. config.guess timestamp = 2013-06-10 uname -m = riscv64 uname -r = 5.18.3-arch1-1 uname -s = Linux uname -v = #1 SMP PREEMPT Sun, 12 Jun 2022 18:42:25 +0000 /usr/bin/uname -p = unknown /bin/uname -X = hostinfo = /bin/universe = /usr/bin/arch -k = /bin/arch = /usr/bin/oslevel = /usr/convex/getsysinfo = UNAME_MACHINE = riscv64 uname -r = 5.18.3-arch1-1 uname -s = Linux uname -v = #1 SMP PREEMPT Sun, 12 Jun 2022 18:42:25 +0000 /usr/bin/uname -p = unknown /bin/uname -X = hostinfo = /bin/universe = /usr/bin/arch -k = /bin/arch = /usr/bin/oslevel = /usr/convex/getsysinfo = UNAME_MACHINE = riscv64 UNAME_RELEASE = 5.18.3-arch1-1 UNAME_SYSTEM = Linux UNAME_VERSION = #1 SMP PREEMPT Sun, 12 Jun 2022 18:42:25 +0000 CalledProcessError: Command '('/bin/sh', 'admin/config.guess')' returned non-zero exit status 1.: File "/build/libffado/src/libffado-2.4.6/SConstruct", line 473: config_guess = conf.ConfigGuess() File "/usr/lib/python3.10/site-packages/SCons/SConf.py", line 713: ret = self.test(context, *args, **kw) File "/build/libffado/src/libffado-2.4.6/SConstruct", line 159: ret = check_output(("/bin/sh", "admin/config.guess")).rstrip() File "/usr/lib/python3.10/subprocess.py", line 420: return run(*popenargs, stdout=PIPE, timeout=timeout, check=True, File "/usr/lib/python3.10/subprocess.py", line 524: raise CalledProcessError(retcode, process.args, Trying to find the system trils ``` I've also tested if admin/config.guess is updated, ffado can be built successfully. So, could you please update admin/config.guess? |