[meta-freescale] [PATCH 51/52] bluez5: Apply patches for QCA6174 and QCA9377-3

Wed Aug 1 12:03:07 PDT 2018

Thanks, Max. It does look like these patches don't belong on meta-freescale. I will remove this patch. Thanks for the suggestions on the content as well.

Tom

> -----Original Message-----
> From: Max Krummenacher [mailto:max.oss.09 at gmail.com]
> Sent: Wednesday, August 1, 2018 9:24 AM
> To: Tom Hochstein <tom.hochstein at nxp.com>; meta-
> freescale at yoctoproject.org
> Subject: Re: [meta-freescale] [PATCH 51/52] bluez5: Apply patches for
> QCA6174 and QCA9377-3
> 
> Resent, due to wrong sender address. Also cut out unneeded quote lines.
> 
> Hi
> 
> Does this really belong into meta-freescale?
> It looks like it is needed because some NXP boards use these chips.
> However they are not specific to NXP SoCs.
> 
> And see inline comments below, I start them with 'MAX:':
> 
> Regards
> Max Krummenacher
> 
> Am Montag, den 30.07.2018, 14:44 -0500 schrieb Tom Hochstein:
> > Signed-off-by: Tom Hochstein <tom.hochstein at nxp.com>
> ---
>  recipes-
> > connectivity/bluez5/bluez5_5.%.bbappend    |   14 +
>  ...th-Add-bluetooth-support-for-QCA6174-
> > chip.patch | 2404 ++++++++++++++++++++
>  ...tach-set-flag-to-enable-HCI-reset-on-init.patch |   32 +
>  ...nstead-of-strlcpy-with-strncpy-to-avoid-r.patch |   35 +
>  4 files changed, 2485 insertions(+)
>  c
> > reate mode 100644 recipes-connectivity/bluez5/bluez5_5.%.bbappend
>  create mode 100644 recipes-
> > connectivity/bluez5/files/0001-bluetooth-Add-bluetooth-support-for-
> QCA6174-chip.patch
>  create mode
> > 100644 recipes-connectivity/bluez5/files/0002-hciattach-set-flag-to-
> enable-HCI-reset-
> on-init.patch
>  create mode 100644 recipes-connectivity/bluez5/files/0003-hciattach-
> instead-of-strlcpy-with-
> strnc
> > py-to-avoid-r.patch
> 
> diff --git a/recipes-connectivity/bluez5/bluez5_5.%.bbappend b/recipes-
> connectivity/bluez5/bluez5_5.%.bbappend
> new file mode 100644
> index 0000000..00eb248
> --- /dev/null
> +++ b/recipes-connectivity/bluez5/bluez5_5.%.bbappend
> @@ -0,0 +1,14 @@
> +FILESEXTRAPATHS_prepend := "${THISDIR}/files:"
> +
> +# Do not start the service during system boot up
> +INITSCRIPT_PARAMS_${PN} = "stop 20 0 1 6 ."
> 
> Max:
> That one would merit some text in the commit log. This changes behavior for
> all
> users of bluez5. But it looks that the line gets removed in the follow up patch
> so probably it can be deleted for good?
> 
> > +
> +# Add patch for module bcm43xx
> +# Add patches for QCA modules with Qca6174 and Qca9377-3 chips
> +SR
> > C_URI += " \
> +            file://0001-bluetooth-Add-bluetooth-support-for-QCA6174-
> chip.patch \
> +
> >         file://0002-hciattach-set-flag-to-enable-HCI-reset-on-init.patch \
> +            file://0003
> > -hciattach-instead-of-strlcpy-with-strncpy-to-avoid-r.patch \
> +            file://0004-Add-support-
> > for-Tufello-1.1-SOC.patch \
> +            file://0005-bluetooth-Add-support-for-multi-baud-
> > rate.patch \
> 
> Max:
> Patch 0004 and 0005 are only added with a follow up patch. IMHO one could
> squash the QCAxxxx and
> Tufello patches, if that is not desired then only add 0004 and 0005 to SRC_URI
> in the patch which
> actually does also add the corresponding files.
> 
> > +"
> diff --git a/recipes-connectivity/bluez5/files/0001-bluetooth-Add-bluetooth-
> support-for-QCA6174-
> chip.patch b/recipes-connectivity/bluez5/files/0001-bluetooth-Add-
> bluetooth-support-for-QCA6174-
> ch
> > ip.patch
> new file mode 100644
> index 0000000..5ae3c98
> --- /dev/null
> +++ b/recipes-
> > connectivity/bluez5/files/0001-bluetooth-Add-bluetooth-support-for-
> QCA6174-
> chip.patch
> @@ -0,0
> > +1,2404 @@
> +From d509b84f237c11874087d7ea527e5ba2f460ed2c Mon Sep 17 00:00:00
> 2001
> +From: Fugang
> > Duan <fugang.duan at nxp.com>
> +Date: Tue, 29 Aug 2017 10:12:57 +0800
> +Subject: [PATCH 1/5] bluetooth :
> > Add bluetooth support for QCA6174 chip.
> +
> +Register the QCA6174 initialization routine with
> > hciattach for
> +downloading firmware patches to the bluetooth controller.
> +Add optional support 'f'
> > to control installation of line
> +discipline driver. Invoke hciattach from command line and
> +download
> > the firmware patches:
> +        hciattach /dev/ttyHS0 qca 3000000 -t120 flow -f0
> +
> +cherry-pick and
> > merged from:
> +
> 	https://emea01.safelinks.protection.outlook.com/?url=https%3A%2
> F%2Fsource.codeaurora.org%2Fquic%2Fla%2Fplatform%2Fexternal%2Fbluet
> ooth%2Fbluez&data=02%7C01%7Ctom.hochstein%40nxp.com%7C1702c
> 4de42e64b5ecefc08d5f7ba7830%7C686ea1d3bc2b4c6fa92cd99c5c301635%7C
> 0%7C0%7C636687302609941480&sdata=lcDl91WPxuWAYgIiDk9sl7xyM3i
> CEeifi0I8a9EnZpk%3D&reserved=0
> +	branch: LNX.LE.5.3
> +	commit: commit bb96f3b759e0b99db70014302ca12929fb42f554
> Max:
> I think a proper Upstream-Status: (with upstream meaning bluez.org) is
> missing.
> 
> > +
> +Change-Id: I87f2927d7096904071a02d73d3afef0dc34db414
> +Signed-off-by: Rupesh Tatiya <rtatiya at codea
> > urora.org>
> +Signed-off-by: Fugang Duan <fugang.duan at nxp.com>
> +---
> + Makefile.tools         |    3 +-
> +
> > tools/hciattach.c      |   26 +-
> + tools/hciattach.h      |    2 +
> + tools/hciattach_rome.c | 1864
> > ++++++++++++++++++++++++++++++++++++++++++++++++
> + tools/hciattach_rome.h |  371 ++++++++++
> + 5
> > files changed, 2264 insertions(+), 2 deletions(-)
> +
> 
> diff --git a/recipes-connectivity/bluez5/files/0002-hciattach-set-flag-to-
> enable-HCI-reset-on-
> init.patch b/recipes-connectivity/bluez5/files/0002-hciattach-set-flag-to-
> enable-HCI-reset-on-
> init.patch
> new file mode 100644
> index 0000000..a9ce17d
> --- /dev/null
> +++ b/recipes-connectivity/bluez5/files/0002-hciattach-set-flag-to-enable-
> HCI-reset-on-init.patch
> @@ -0,0 +1,32 @@
> +From 3648d3c33b1dd0e11707328d87101d8044d23302 Mon Sep 17 00:00:00
> 2001
> +From: Fugang Duan <fugang.duan at nxp.com>
> +Date: Tue, 29 Aug 2017 10:21:09 +0800
> +Subject: [PATCH 2/5] hciattach: set flag to enable HCI reset on init
> +
> +On some qca devices, correct setup of BT+WLAN co-existance requires
> +HCI reset being set at the start of power on sequence. Sending HCI
> +resets has no side effect.
> +
> 
> Max:
> Upstream-Status is missing.
> 
> > +Change-Id: I71cb367d10d4d19d82b41af6a4a0b8b2f770f691
> +Signed-off-by: Rupesh Tatiya <rtatiya at codeau
> > rora.org>
> +---
> 
> > connectivity/bluez5/files/0003-hciattach-instead-of-strlcpy-with-strncpy-
> to-avoid-
> r.patch
> @@ -0,0
> > +1,35 @@
> +From 5877685b10560c40f1c4af375e30adf2a631fefc Mon Sep 17 00:00:00
> 2001
> +From: Fugang Duan
> > <fugang.duan at nxp.com>
> +Date: Tue, 29 Aug 2017 14:54:44 +0800
> +Subject: [PATCH 3/5] hciattach:
> > instead of strlcpy with strncpy to avoid rome
> + build error
> +
> +Instead of strlcpy with strncpy to
> > avoid rome build error.
> +
> +igned-off-by: Fugang Duan <fugang.duan at nxp.com>
> 
> Max:
> Typo, missing 'S'
> Upstream-Status is missing.
> 
> > +---
> + tools/hciattach_rome.c | 4 ++--
> + 1 file changed, 2 insertions(+), 2 deletions(-)
> +
> +diff --git
> > a/tools/hciattach_rome.c b/tools/hciattach_rome.c
> +index 242a49f..c3e9fde 100644