How to get BT working on Pi3B. #1314
Comments
|
You really need to post this sort of question on the forum, but in short sudo apt-get install pi-bluetooth should install a load of stuff you need. Not sure how to get the panel icon On 1 March 2016 at 15:42, Clive Messer [email protected] wrote:
|
|
sudo apt-get install blueman gets you the GUI (which is just about functional). |
|
Guys, I'm not using Raspbian. I never will be using Raspbian, I have no interest in Raspbian. I can see that the BT section of chip is connected to UART from DT config. I don't want to install Debian packages. I am asking for the technical info to be able to get it working on any distribution. Don't want blueman, LXDE, any desktop gui..... blah, blah, blah. Just want the device to be recognised. |
|
But Raspbian is the supported way. All the information is out there for smart, motivated people to use, but don't expect to be spoon-fed. |
|
Sorry, Raspbian may be the only supported OS. I can understand that. What I need is the lower level info to make this work with an OS that is not Raspbian. That's the info I am looking for. Sorry, I must be really stupid. Would you please educate a stupid person where he should look, to find the information I seek. I don't require spoon-feeding, just a pointer to the info. I have looked! |
|
@pelwell Phil, at least tell me what I am looking for. A Broadcom technical document? Documentation in the PiF documentation github? |
|
You need the modified BlueZ package found in Raspbian (source available) and the systemd script called hciattach.service from Raspbian. |
|
http://archive.raspberrypi.org/debian/pool/main/b/bluez/bluez_5.23-2+rpi1.debian.tar.xz |
|
@XECDesign Thank you for the help. Where are those params documented?
I have vague recollections of some sort of firmware upload needing to be done before the attach. (Think that was with Wandboard which used a BRCM chipset for wifi/BT). Does something like that need to happen with the Pi3B? |
|
Have you tried typing hciattach into google? |
|
Yes, but it returns rather a lot of results. None of which seem to be telling me where to obtain the firmware file. Is it available from PiF github? Or only buried in a Raspbian package? |
|
I'm going to ask nicely...... Is it possible to make any firmware files needed to support Pi3B wifi/BT operation available directly from the PiF github, rather than them just being available via specific distribution packages? |
|
I am using a BT module on Pi2, first I enable the UART flow control by: You also need to remove the console=ttyAMA0 in /boot/cmdline.txt |
|
I don't have a Pi3 at this moment but I tried my Pi2 with NOOBS version 1.8.0, BT works natively, only need to make some scripts. |
|
Did you manage to get wifi to work on the Pi3B without Raspbian? I'm trying to figure out how to do this but I'm kind of in the dark on this... |
|
@davidpronk Not yet. I got as far as grabbing those patches Phil made to modify the bluez hciattach from the Raspbian package, and rebuilt my Fedora bluez package..... Got the firmware, BCM43430A1.hcd, from Raspbian image..... At which point I was hoping it would just work, but still getting a timeout..... $ sudo /usr/bin/hciattach -n /dev/ttyAMA0 bcm43xx 921600 noflow - I hope to find some time tomorrow to try and figure it out...... The last time I was fighting with Broadcom bluetooth on another SBC, it was still the "two shot" brcm-patchram-plus to initialize firmware, followed by hciattach..... At least now, it's just hciattach responsible for firmware upload and connecting the UART to the Bluetooth subsystem. |
|
@pelwell "All the information is out there for smart, motivated people to use, but don't expect to be spoon-fed". Completely disagree with this one. |
Do you not see the contradiction here? And to reiterate:
|
Not at all, because I don't consider that a proper documentation. |
@pelwell Are you familiar with the concept of a BSP/SDK? That's what is being asked for. (Not being told that one can find out the necessary info from a Raspbian image.) |
|
Do you have any examples of previous RPi BSP/SDKs? |
Did you ever release a board, before now, that provided Bt/wifi, that required proprietary firmware to make it work, (not hosted in linux-firmware), and required modification to userspace software, to make it work? Phil, I'm not trying to wind you up. Eben's blog, named you .....
.... so it would seem that you might be the best person to put the information together in a format that would be suitable, for other people, who do not wish to use Raspbian, to be able to get it working on whatever other OS, they choose to use. |
|
@clivem And then it hangs... By the way, try this with fresh boot, since it appears that device reset isn't properly working... |
|
Okay, didn't notice the "-n" option prevents detaching from controlling terminal. It's working now, but only at 115200... |
|
Are you using the patched hciattach source from the Raspbian distribution? |
|
Phil, where does your patch set fit in with what was committed upstream to bluez, from Loic Poulain, back in 2014. http://git.kernel.org/cgit/bluetooth/bluez.git/commit?id=beb4892e96785ba9ff429fb940126a815efd47fd Have you submitted your patches to upstream? No need to, because it has already been dealt with upstream and you just need to patch the Rasbian version? |
|
@pelwell Nope, but I'm using a patched version of the OpenWrt one: |
|
This is madness..... If we need to modify userspace code, can we please have "reference" patches against current upstream bluez, not some antique version that Debian/Raspbian is using..... Pretty please? |
|
Since the firmware is the most important data for a modem, increasing the speed before uploading the firmware on a modem without flow control is not a great idea. The other patches add h5 support, but this isn't being used yet. |
|
I tested those patches and there are no more tiemouts, however with baud rates different to 115200 the device doesn't show up: 921600: Default (3000000): |
|
I've patched the bluez 5.36 Fedora 23 package.... http://www.squeezecommunity.org/repo/fedora/23/testing/SRPMS/bluez-5.36-1.fc23.2.src.rpm $ sudo hciattach /dev/ttyAMA0 bcm43xx 921600 noflow - $ sudo hciconfig hci0 up $ sudo hciconfig dev $ sudo hcitool scan |
|
@pelwell BTW, the gist patches.... Thank you, and not wishing to appear ungrateful, but the first three don't apply until the spaces are changed to tabs..... ;) |
|
@clivem I get the following when I use your hciattach parameters: Timeouts, timeouts everywhere xD |
|
Gist drag & drop doesn't work for me - why isn't there an upload button? - and the tabs got lost in the cut & paste. It should be fixed now. |
|
FYI there is a new overlay - pi3-miniuart-bt - that enables Bluetooth using the mini-UART. There may be throughput limitations, and it is even more important to clamp cpu_freq to 250, so caveat emptor. You will also need to edit the systemd service See 866cf94. |
|
I did more tests yesterday and I was able to get WiFi working. On the other hand, I did some more tests and I couldn't get Bluetooth working at baud rates > 115200 (timeouts), but it was working perfectly with baud rates <= 115200. BTW thanks for sharing those links @clivem. |
|
Can I just ask, is the internal bluetooth implementation in the standard release for rpi3 supposed to work? I seem to have many compatibility issues connecting devices, but the same devices work on the same pi3 with an external bluetooth dongle. |
|
@pelwell I see a couple of patches you sent for bluez5. Are these patches intended to go upstream in bluez5 or are they rpi3 specific? |
|
They are Pi3 specific, at least in their current form. Any upstream patch would have to be more extensive, adding extra command-line parameters to make the behaviour configurable. |
|
@pelwell Thank you, I applied your patches to BlueZ 5.39, please look here for context: ukBaz/python-bluezero#30 There are a few other Bluetooth/BlueZ/BLE related issues going on but without going into details here I wondered if you happen have any other Pi3 (or Pi2) specific BlueZ patches? Thanks |
|
@clivem has your issue been resolved? If so, please close this issue. Thanks. |
@clivem have you ever manage to wrap this up? What is the status of the firmware.bin? Since the raspberry pi org is redistributing them, they should be redistributable implicitly already? If so, what is the canonical source for these firmware's? They still haven't shown up in linux-firmware.git ... |
it works for me |
Patch series "migrate_pages: fix deadlock in batched synchronous
migration", v2.
Two deadlock bugs were reported for the migrate_pages() batching series.
Thanks Hugh and Pengfei. Analysis shows that if we have locked some other
folios except the one we are migrating, it's not safe in general to wait
synchronously, for example, to wait the writeback to complete or wait to
lock the buffer head.
So 1/3 fixes the deadlock in a simple way, where the batching support for
the synchronous migration is disabled. The change is straightforward and
easy to be understood. While 3/3 re-introduce the batching for
synchronous migration via trying to migrate asynchronously in batch
optimistically, then fall back to migrate synchronously one by one for
fail-to-migrate folios. Test shows that this can restore the TLB flushing
batching performance for synchronous migration effectively.
This patch (of 3):
Two deadlock bugs were reported for the migrate_pages() batching series.
Thanks Hugh and Pengfei! For example, in the following deadlock trace
snippet,
INFO: task kworker/u4:0:9 blocked for more than 147 seconds.
Not tainted 6.2.0-rc4-kvm+ #1314
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
task:kworker/u4:0 state:D stack:0 pid:9 ppid:2 flags:0x00004000
Workqueue: loop4 loop_rootcg_workfn
Call Trace:
<TASK>
__schedule+0x43b/0xd00
schedule+0x6a/0xf0
io_schedule+0x4a/0x80
folio_wait_bit_common+0x1b5/0x4e0
? __pfx_wake_page_function+0x10/0x10
__filemap_get_folio+0x73d/0x770
shmem_get_folio_gfp+0x1fd/0xc80
shmem_write_begin+0x91/0x220
generic_perform_write+0x10e/0x2e0
__generic_file_write_iter+0x17e/0x290
? generic_write_checks+0x12b/0x1a0
generic_file_write_iter+0x97/0x180
? __sanitizer_cov_trace_const_cmp4+0x1a/0x20
do_iter_readv_writev+0x13c/0x210
? __sanitizer_cov_trace_const_cmp4+0x1a/0x20
do_iter_write+0xf6/0x330
vfs_iter_write+0x46/0x70
loop_process_work+0x723/0xfe0
loop_rootcg_workfn+0x28/0x40
process_one_work+0x3cc/0x8d0
worker_thread+0x66/0x630
? __pfx_worker_thread+0x10/0x10
kthread+0x153/0x190
? __pfx_kthread+0x10/0x10
ret_from_fork+0x29/0x50
</TASK>
INFO: task repro:1023 blocked for more than 147 seconds.
Not tainted 6.2.0-rc4-kvm+ #1314
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
task:repro state:D stack:0 pid:1023 ppid:360 flags:0x00004004
Call Trace:
<TASK>
__schedule+0x43b/0xd00
schedule+0x6a/0xf0
io_schedule+0x4a/0x80
folio_wait_bit_common+0x1b5/0x4e0
? compaction_alloc+0x77/0x1150
? __pfx_wake_page_function+0x10/0x10
folio_wait_bit+0x30/0x40
folio_wait_writeback+0x2e/0x1e0
migrate_pages_batch+0x555/0x1ac0
? __pfx_compaction_alloc+0x10/0x10
? __pfx_compaction_free+0x10/0x10
? __this_cpu_preempt_check+0x17/0x20
? lock_is_held_type+0xe6/0x140
migrate_pages+0x100e/0x1180
? __pfx_compaction_free+0x10/0x10
? __pfx_compaction_alloc+0x10/0x10
compact_zone+0xe10/0x1b50
? lock_is_held_type+0xe6/0x140
? check_preemption_disabled+0x80/0xf0
compact_node+0xa3/0x100
? __sanitizer_cov_trace_const_cmp8+0x1c/0x30
? _find_first_bit+0x7b/0x90
sysctl_compaction_handler+0x5d/0xb0
proc_sys_call_handler+0x29d/0x420
proc_sys_write+0x2b/0x40
vfs_write+0x3a3/0x780
ksys_write+0xb7/0x180
__x64_sys_write+0x26/0x30
do_syscall_64+0x3b/0x90
entry_SYSCALL_64_after_hwframe+0x72/0xdc
RIP: 0033:0x7f3a2471f59d
RSP: 002b:00007ffe567f7288 EFLAGS: 00000217 ORIG_RAX: 0000000000000001
RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007f3a2471f59d
RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000005
RBP: 00007ffe567f72a0 R08: 0000000000000010 R09: 0000000000000010
R10: 0000000000000010 R11: 0000000000000217 R12: 00000000004012e0
R13: 00007ffe567f73e0 R14: 0000000000000000 R15: 0000000000000000
</TASK>
The page migration task has held the lock of the shmem folio A, and is
waiting the writeback of the folio B of the file system on the loop block
device to complete. While the loop worker task which writes back the
folio B is waiting to lock the shmem folio A, because the folio A backs
the folio B in the loop device. Thus deadlock is triggered.
In general, if we have locked some other folios except the one we are
migrating, it's not safe to wait synchronously, for example, to wait the
writeback to complete or wait to lock the buffer head.
To fix the deadlock, in this patch, we avoid to batch the page migration
except for MIGRATE_ASYNC mode. In MIGRATE_ASYNC mode, synchronous waiting
is avoided.
The fix can be improved further. We will do that as soon as possible.
Link: https://lkml.kernel.org/r/[email protected]
Link: https://lore.kernel.org/linux-mm/[email protected]/
Link: https://lore.kernel.org/linux-mm/[email protected]/
Link: https://lore.kernel.org/linux-mm/20230227110614.dngdub2j3exr6dfp@quack3/
Link: https://lkml.kernel.org/r/[email protected]
Fixes: 5dfab10 ("migrate_pages: batch _unmap and _move")
Signed-off-by: "Huang, Ying" <[email protected]>
Reported-by: Hugh Dickins <[email protected]>
Reported-by: "Xu, Pengfei" <[email protected]>
Cc: Jan Kara <[email protected]>
Cc: Baolin Wang <[email protected]>
Cc: Christoph Hellwig <[email protected]>
Cc: Stefan Roesch <[email protected]>
Cc: Tejun Heo <[email protected]>
Cc: Xin Hao <[email protected]>
Cc: Zi Yan <[email protected]>
Cc: Yang Shi <[email protected]>
Cc: Matthew Wilcox <[email protected]>
Cc: Mike Kravetz <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
I figured this was going to be really simple, with something like using hciattach to the UART, to make the device visible.....
Any instructions anywhere?
The text was updated successfully, but these errors were encountered: