All of lore.kernel.org
 help / color / mirror / Atom feed
From: Kilian Zinnecker via buildroot <buildroot@buildroot.org>
To: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>,
	Giulio Benetti <giulio.benetti@benettiengineering.com>,
	Quentin Schulz <quentin.schulz@theobroma-systems.com>,
	Andreas Ziegler <br015@umbiko.net>,
	buildroot@buildroot.org
Subject: Re: [Buildroot] [PATCH v8 1/3] package/rockchip-rkbin: new package
Date: Mon, 07 Aug 2023 13:45:40 +0200	[thread overview]
Message-ID: <8255697.T7Z3S40VBb@kilian-aisec> (raw)
In-Reply-To: <20230726214126.GC3544423@scaer>

Hello Yann, all

thanks for your review, fixes and applying the patches. And sorry for the late 
answer. What I write now is imho not so important anymore, still I want to 
give some reply:

> > This patch adds a package for the Rockchip ATF binary blobs. These
> > binaries are needed to build U-Boot for some Rockchip SoCs (e.g.,
> > RK3588). One can config a custom version and manually define which
> > blobs (for bl31, tpl and optee) to use from the repository.
> 
> So, there was no rationale about why we sould need a custom version [0],
> and the defconfig you provided was actually using the same version as
> the predefined one, so we did not have any usage for it in-tree.
> 
> [0] yes, I read your reply about the question from Thomas, but
> discussing with him we conluded that we indeed did not want to have a
> custom version for now. So I dropped it.
> 
> If and when the need arises, then it will be time to add it back; see
> below.

I personally still favor it, but its not so important to me and I can 
understand, that you don't see much need at the moment and probably prefer to 
keep it simple, so I am totally fine with dropping the custom version stuff. As 
you say: If need arises, we can then revaluate.

[--SNIP--]

> > +config BR2_PACKAGE_ROCKCHIP_RKBIN_TPL_FILENAME
> > +	string "Rockchip rkbin tpl file path"
> > +	help
> > +	  Full path to the tpl file inside the rkbin repository. The
> 
> I've driopped the "Full" qualifier, because that is not a full path, as
> it is relative to the top of the rockchip-rkbin directory.
> 
> Ditto for the other help texts.

Agreed. However, you introduced some typo on one occasion, by snipping "Full" 
and "path" to "Fath ;)

https://gitlab.com/buildroot.org/buildroot/-/blob/
8d1180aa483b624f1be27bd0858c2f6665ebf36c/package/rockchip-rkbin/Config.in#L12

> > diff --git a/package/rockchip-rkbin/rockchip-rkbin.hash
> > b/package/rockchip-rkbin/rockchip-rkbin.hash new file mode 100644
> > index 0000000000..cb71226556
> > --- /dev/null
> > +++ b/package/rockchip-rkbin/rockchip-rkbin.hash
> > @@ -0,0 +1,2 @@
> > +# Locally computed
> > +sha256  bd8d19ace202ff26d1c0b4d7744cd467cd0093801dc674dde57290159eedee2b 
> > rockchip-rkbin-b4558da0860ca48bf1a571dd33ccba580b9abe23-br1.tar.gz
> I've added a hash for the license file. Please check with:
> 
>     $ make legal-info
> or:
>     $ make rockchip-rkbin-legal-info

Thanks, seems to work on my PC.

> > diff --git a/package/rockchip-rkbin/rockchip-rkbin.mk
> > b/package/rockchip-rkbin/rockchip-rkbin.mk new file mode 100644
> > index 0000000000..74787585e8
> > --- /dev/null
> > +++ b/package/rockchip-rkbin/rockchip-rkbin.mk
> > @@ -0,0 +1,46 @@
> > +#########################################################################
> > ####### +#
> > +# rockchip-rkbin
> > +#
> > +#########################################################################
> > ####### +
> > +ROCKCHIP_RKBIN_VERSION = $(call
> > qstrip,$(BR2_PACKAGE_ROCKCHIP_RKBIN_VERSION)) +ROCKCHIP_RKBIN_SITE =
> > https://github.com/rockchip-linux/rkbin.git +ROCKCHIP_RKBIN_SITE_METHOD =
> > git
> > +ROCKCHIP_RKBIN_LICENSE = PROPRIETARY
> > +ROCKCHIP_RKBIN_LICENSE_FILES = LICENSE
> 
> Since the license file was only recently added, it should not be defined
> except for the known version.
> 
> Furthermore, I think hadling the custom version, if we re-introduce it
> later, should be done with:
> 
> Config.in:
> 
>     config BR2_PACKAGE_ROCKCHIP_RKBIN_CUSTOM_VERSION
>         bool "Use a custom version"
>         help
>           Some binaries for older hipsets mey get pruned from the
>           latest revisions of the repository. Say 'y' here if your
>           SoC uses such older binaries.
> 
>     config BR2_PACKAGE_ROCKCHIP_RKBIN_CUSTOM_VERSION_VALUE
>         string "Custom version"
>         depends on BR2_PACKAGE_ROCKCHIP_RKBIN_CUSTOM_VERSION
> 
> rocjchip-rkbin.mk:
> 
>     ifeq ($(BR2_PACKAGE_ROCKCHIP_RKBIN_CUSTOM_VERSION),)
>     ROCKCHIP_RKBIN_VERSION = b4558da0860ca48bf1a571dd33ccba580b9abe23
>     ROCKCHIP_RKBIN_LICENSE_FILES = LICENSE
>     else
>     ROCKCHIP_RKBIN_VERSION = $(call
> qstrip,$(BR2_PACKAGE_ROCKCHIP_RKBIN_CUSTOM_VERSION_VALUE)) ifeq
> ($(ROCKCHIP_RKBIN_VERSION),)
>     $(error blabla check config blabla)
>     endif
>     BR_NO_CHECK_HASH_FOR += $(ROCKCHIP_RKBIN_SOURCE)
>     endif

Great, thanks! I did actually have in mind, that the custom version and the 
(in older versions not existing) LICENSE file may cause problems. But I didn't 
know what to do about it and forgot to ask here. Thanks!

[--SNIP--]

> > +ifneq ($(BR2_PACKAGE_ROCKCHIP_RKBIN_TPL_FILENAME),"")
> > +ROCKCHIP_RKBIN_TPL_FILENAME = $(call
> > qstrip,$(BR2_PACKAGE_ROCKCHIP_RKBIN_TPL_FILENAME)) +endif
> > +
> > +ifneq ($(BR2_PACKAGE_ROCKCHIP_RKBIN_TEE_FILENAME),"")
> > +ROCKCHIP_RKBIN_TEE_FILENAME = $(call
> > qstrip,$(BR2_PACKAGE_ROCKCHIP_RKBIN_TEE_FILENAME)) +endif
> > +
> > +define ROCKCHIP_RKBIN_INSTALL_IMAGES_CMDS
> > +	$(if $(ROCKCHIP_RKBIN_BL31_FILENAME), \
> > +		cp $(@D)/$(ROCKCHIP_RKBIN_BL31_FILENAME)
> > $(BINARIES_DIR)/${ROCKCHIP_RKBIN_BL31_FILENAME##*/}) +	$(if
> > $(ROCKCHIP_RKBIN_TPL_FILENAME), \
> > +		cp $(@D)/$(ROCKCHIP_RKBIN_TPL_FILENAME)
> > $(BINARIES_DIR)/${ROCKCHIP_RKBIN_TPL_FILENAME##*/}) +	$(if
> > $(ROCKCHIP_RKBIN_TEE_FILENAME), \
> > +		cp $(@D)/$(ROCKCHIP_RKBIN_TEE_FILENAME)
> > $(BINARIES_DIR)/${ROCKCHIP_RKBIN_TEE_FILENAME##*/})
> I've simplified this with a $(foreach...)

Thanks, much better of course!

> Applied to master with these changes, thanks.

Thanks for reviewing!

Best regards,
Kilian




_______________________________________________
buildroot mailing list
buildroot@buildroot.org
https://lists.buildroot.org/mailman/listinfo/buildroot

  reply	other threads:[~2023-08-07 11:45 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-07-26 18:04 [Buildroot] [PATCH v8 0/3] Add board support for Radxa Rock 5B Kilian Zinnecker via buildroot
2023-07-26 18:04 ` [Buildroot] [PATCH v8 1/3] package/rockchip-rkbin: new package Kilian Zinnecker via buildroot
2023-07-26 21:41   ` Yann E. MORIN
2023-08-07 11:45     ` Kilian Zinnecker via buildroot [this message]
2023-08-07 17:29       ` Yann E. MORIN
2023-07-26 18:04 ` [Buildroot] [PATCH v8 2/3] boot/uboot: Add support for Rockchip binary blobs Kilian Zinnecker via buildroot
2023-07-26 21:43   ` Yann E. MORIN
2023-08-07 11:46     ` Kilian Zinnecker via buildroot
2023-07-26 18:04 ` [Buildroot] [PATCH v8 3/3] configs/rock5b: new defconfig Kilian Zinnecker via buildroot
2023-07-26 21:45   ` Yann E. MORIN

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=8255697.T7Z3S40VBb@kilian-aisec \
    --to=buildroot@buildroot.org \
    --cc=br015@umbiko.net \
    --cc=giulio.benetti@benettiengineering.com \
    --cc=kilian.zinnecker@mail.de \
    --cc=quentin.schulz@theobroma-systems.com \
    --cc=thomas.petazzoni@bootlin.com \
    --cc=yann.morin.1998@free.fr \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.