From: Gary Bisson <gary.bisson@boundarydevices.com>
To: buildroot@busybox.net
Subject: [Buildroot] [PATCH 4/8] imx-vpu-hantro: new package
Date: Thu, 26 Jul 2018 12:02:50 +0200 [thread overview]
Message-ID: <20180726100250.GB5860@t450s.lan> (raw)
In-Reply-To: <3a66b04f-74dc-5296-4511-99cf5069f787@mind.be>
Hi Arnout,
On Thu, Jul 26, 2018 at 11:41:44AM +0200, Arnout Vandecappelle wrote:
>
>
> On 25-07-18 17:01, Gary Bisson wrote:
> > This package provides the user-space libraries needed to use the Hantro
> > VPU present in processors such as the i.MX8MQ.
> >
> > Signed-off-by: Gary Bisson <gary.bisson@boundarydevices.com>
> > ---
> > package/freescale-imx/Config.in | 4 +-
> > ...on.h-header-inclusion-to-be-standard.patch | 44 +++++++++++++++++++
> > .../freescale-imx/imx-vpu-hantro/Config.in | 15 +++++++
> > .../imx-vpu-hantro/imx-vpu-hantro.hash | 2 +
> > .../imx-vpu-hantro/imx-vpu-hantro.mk | 42 ++++++++++++++++++
> > 5 files changed, 106 insertions(+), 1 deletion(-)
> > create mode 100644 package/freescale-imx/imx-vpu-hantro/0001-Fix-ion.h-header-inclusion-to-be-standard.patch
> > create mode 100644 package/freescale-imx/imx-vpu-hantro/Config.in
> > create mode 100644 package/freescale-imx/imx-vpu-hantro/imx-vpu-hantro.hash
> > create mode 100644 package/freescale-imx/imx-vpu-hantro/imx-vpu-hantro.mk
> >
> > diff --git a/package/freescale-imx/Config.in b/package/freescale-imx/Config.in
> > index 5e0c7ff291..9b47958ff3 100644
> > --- a/package/freescale-imx/Config.in
> > +++ b/package/freescale-imx/Config.in
> > @@ -63,7 +63,8 @@ config BR2_PACKAGE_FREESCALE_IMX_HAS_VPU
> > default y if BR2_PACKAGE_FREESCALE_IMX_PLATFORM_IMX27ADS || \
> > BR2_PACKAGE_FREESCALE_IMX_PLATFORM_IMX51 || \
> > BR2_PACKAGE_FREESCALE_IMX_PLATFORM_IMX53 || \
> > - BR2_PACKAGE_FREESCALE_IMX_PLATFORM_IMX6Q
> > + BR2_PACKAGE_FREESCALE_IMX_PLATFORM_IMX6Q || \
> > + BR2_PACKAGE_FREESCALE_IMX_PLATFORM_IMX8M
>
> Wouldn't it make more sense to make a new _HAS_VPU_HANTRO option? You're later
> changing some 'depends' options to depending on the provider instead of
> depending on the VPU, so that makes sense.
Yes we could, I just didn't do it as first as I recall someone not being
fond of the FREESCALE_IMX_HAS_XXX options.
But if we remove the virtual package as Thomas suggested that would be
best indeed.
> > config BR2_PACKAGE_FREESCALE_IMX_HAS_VIV_GPU
> > bool
> > @@ -80,6 +81,7 @@ source "package/freescale-imx/imx-m4fwloader/Config.in"
> > source "package/freescale-imx/imx-parser/Config.in"
> > source "package/freescale-imx/imx-uuc/Config.in"
> > source "package/freescale-imx/imx-vpu-cnm/Config.in"
> > +source "package/freescale-imx/imx-vpu-hantro/Config.in"
> > source "package/freescale-imx/imx-vpuwrap/Config.in"
> > source "package/freescale-imx/firmware-imx/Config.in"
> > if (BR2_PACKAGE_FREESCALE_IMX_PLATFORM_IMX51 || BR2_PACKAGE_FREESCALE_IMX_PLATFORM_IMX53)
> > diff --git a/package/freescale-imx/imx-vpu-hantro/0001-Fix-ion.h-header-inclusion-to-be-standard.patch b/package/freescale-imx/imx-vpu-hantro/0001-Fix-ion.h-header-inclusion-to-be-standard.patch
> > new file mode 100644
> > index 0000000000..951ead9824
> > --- /dev/null
> > +++ b/package/freescale-imx/imx-vpu-hantro/0001-Fix-ion.h-header-inclusion-to-be-standard.patch
> > @@ -0,0 +1,44 @@
> > +From 872c82e7cbb9a0a0e761e8ac70fc28e19a55b4c3 Mon Sep 17 00:00:00 2001
> > +From: Gary Bisson <gary.bisson@boundarydevices.com>
> > +Date: Thu, 12 Jul 2018 11:38:28 +0200
> > +Subject: [PATCH] Fix ion.h header inclusion to be standard
> > +
> > +NXP "solution" was to manually copy the header to include/linux.
> > +Let's point the Makefile to the proper (mainline) location instead:
> > +https://elixir.bootlin.com/linux/v4.17/source/drivers/staging/android/uapi/ion.h
> > +
> > +Signed-off-by: Gary Bisson <gary.bisson@boundarydevices.com>
>
> Is it now possible to upstream things to codeaurora? Did you do that?
Not that I know of. I've tried sending emails to the owner of the repo +
main contributor but never received an answer.
The best approach is to have that patch in the meta-freescale BSP, then
NXP will look at it (it is the case here).
> > +---
> > + Makefile | 2 ++
> > + decoder_sw/software/linux/dwl/dwl_linux.c | 2 +-
> > + 2 files changed, 3 insertions(+), 1 deletion(-)
> > +
> > +diff --git a/Makefile b/Makefile
> > +index b74e23a..a5ce22b 100755
> > +--- a/Makefile
> > ++++ b/Makefile
> > +@@ -11,6 +11,8 @@ INCLUDE_HEADERS = -I./decoder_sw -I$(SOURCE_ROOT)/source/inc -I$(SOURCE_ROOT)/so
> > + INCLUDE_HEADERS += -I$(SOURCE_ROOT)/linux/memalloc
> > + #INCLUDE_HEADERS += -I$(SOURCE_ROOT)/linux/ldriver
> > + INCLUDE_HEADERS += -I$(LINUX_KERNEL_ROOT)/include/uapi -I$(LINUX_KERNEL_ROOT)/include
> > ++# ION header location
> > ++INCLUDE_HEADERS += -I$(LINUX_KERNEL_ROOT)/drivers/staging/android/uapi
> > +
> > + CFLAGS += -DDEC_MODULE_PATH=\"/dev/mxc_hantro\" -DUSE_FAKE_RFC_TABLE -DFIFO_DATATYPE=void* -DNDEBUG -DDOWN_SCALER \
> > + -DUSE_EXTERNAL_BUFFER -DUSE_FAST_EC -DUSE_VP9_EC -DGET_FREE_BUFFER_NON_BLOCK \
> > +diff --git a/decoder_sw/software/linux/dwl/dwl_linux.c b/decoder_sw/software/linux/dwl/dwl_linux.c
> > +index 8183660..ed37d86 100644
> > +--- a/decoder_sw/software/linux/dwl/dwl_linux.c
> > ++++ b/decoder_sw/software/linux/dwl/dwl_linux.c
> > +@@ -41,7 +41,7 @@
> > + #include "dwl.h"
> > + #include <linux/hantrodec.h>
> > + #ifdef USE_ION
> > +-#include <linux/ion.h>
> > ++#include <ion.h>
> > + #ifdef ANDROID
> > + #include <linux/mxc_ion.h>
> > + #endif
> > +--
> > +2.18.0
> > +
> > diff --git a/package/freescale-imx/imx-vpu-hantro/Config.in b/package/freescale-imx/imx-vpu-hantro/Config.in
> > new file mode 100644
> > index 0000000000..5a1255164e
> > --- /dev/null
> > +++ b/package/freescale-imx/imx-vpu-hantro/Config.in
> > @@ -0,0 +1,15 @@
> > +comment "imx-vpu-hantro needs an i.MX platform with VPU support"
> > + depends on BR2_aarch64
> > + depends on !BR2_PACKAGE_FREESCALE_IMX_HAS_VPU
> > +
> > +config BR2_PACKAGE_IMX_VPU_HANTRO
> > + bool "imx-vpu-hantro"
> > + depends on BR2_aarch64 # Only relevant for i.MX8
>
> If you make a new BR2_PACKAGE_FREESCALE_IMX_HAS_VPU_HANTRO, this becomes
> unnecessary. Actually the dependency on BR2_arm in imx-vpu is also redundant now
> IMO.
True.
> > + depends on BR2_PACKAGE_FREESCALE_IMX_HAS_VPU
> > + help
> > + Library of userspace helpers specific for the NXP i.MX CPUs
> > + integrating a Hantro Video Processing Unit (VPU). It requires
> > + a kernel that includes the i.MX specific headers to be built.
>
> You copied this from imx-vpu, but it is wrong there... Look at commit
> 0ac0982fb7e936c48c0925b272eb18e0195e1c47 which removed the Linux dependency but
> failed to update the help text...
Yep, will update help text for imx-vpu. Here it is required.
> So, you need to add a dependency on BR2_LINUX_KERNEL and the corresponding comment.
Yes.
Thanks,
Gary
next prev parent reply other threads:[~2018-07-26 10:02 UTC|newest]
Thread overview: 22+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-07-25 15:01 [Buildroot] [PATCH 0/8] imx: update multimedia packages to 4.9.88_2.0.0_ga Gary Bisson
2018-07-25 15:01 ` [Buildroot] [PATCH 1/8] firmware-imx: bump to version 7.5 Gary Bisson
2018-07-28 21:53 ` Arnout Vandecappelle
2018-07-29 13:08 ` Thomas Petazzoni
2018-07-30 8:59 ` Gary Bisson
2018-07-25 15:01 ` [Buildroot] [PATCH 2/8] imx-vpu: rename package to imx-vpu-cnm Gary Bisson
2018-07-25 16:26 ` Baruch Siach
2018-07-26 6:45 ` Gary Bisson
2018-07-25 15:01 ` [Buildroot] [PATCH 3/8] imx-vpu-cnm: bump version to 5.4.38 Gary Bisson
2018-07-25 15:01 ` [Buildroot] [PATCH 4/8] imx-vpu-hantro: new package Gary Bisson
2018-07-25 16:29 ` Baruch Siach
2018-07-26 6:45 ` Gary Bisson
2018-07-26 9:41 ` Arnout Vandecappelle
2018-07-26 10:02 ` Gary Bisson [this message]
2018-07-25 15:01 ` [Buildroot] [PATCH 5/8] imx-vpu: new virtual package Gary Bisson
2018-07-25 15:01 ` [Buildroot] [PATCH 6/8] imx-vpuwrap: bump version to 4.3.5 Gary Bisson
2018-07-25 15:01 ` [Buildroot] [PATCH 7/8] imx-codec: " Gary Bisson
2018-07-25 15:01 ` [Buildroot] [PATCH 8/8] imx-parser: " Gary Bisson
2018-07-26 9:26 ` [Buildroot] [PATCH 0/8] imx: update multimedia packages to 4.9.88_2.0.0_ga Arnout Vandecappelle
2018-07-26 9:45 ` Thomas Petazzoni
2018-07-26 9:58 ` Gary Bisson
2018-07-28 8:08 ` Arnout Vandecappelle
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20180726100250.GB5860@t450s.lan \
--to=gary.bisson@boundarydevices.com \
--cc=buildroot@busybox.net \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox