All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Shiju Sivasankaran" <shiju@ti.com>
To: "Bajjuri, Praneeth" <praneeth@ti.com>,
	"meta-ti@lists.yoctoproject.org" <meta-ti@lists.yoctoproject.org>
Subject: Re: [dunfell/master PATCH] ti-rtos: update metadata and version of am64x rtos firmware
Date: Fri, 26 Mar 2021 03:49:09 +0000	[thread overview]
Message-ID: <b7b048b378b14dde8ddf2b453307e162@ti.com> (raw)
In-Reply-To: <1a2f7012-0c7b-33fe-5fe4-a0621ebcf1fb@ti.com>

<1>
Yes, All the FW are hosted on same repo processor-sdk/coresdk_rtos_releases. But its only hosting purpose, these FW are builds from PDK (all device except AM64x) and mcu-plus-sdk (AM64x alone)
Similarly metadata.inc will be used to get the version/urls details for all devices except AM64x and mcusdk_metadata.inc is used for AM64x. Same metadata.inc for mcu-plus-sdk is a big risk as any PDK build of AM64x can override this.

<2>
coresdk_rtos_releases  has 2 branches - master & next.  All metadata files and tags are applied on next and during RC period coreSDK team merge next to master.  At the same time next-branch also carry the metadata files from daily build as well.  During this merge they make sure only release builds of all devices on that tag/commit, as only release build points the FW url from external location.  If any daily build of any device happens in between the tagging and as daily build has TI internal url for FW will spoil the purpose.  AM65x, J7 & J7200 builds of PSDK should have happen one after another without any daily builds in between, this can be controlled as all these are built from same codebase of PDK.  As AM64x builds from mcu-plus-sdk, a different codebase, code-freeze and RC timing rarely match.  Rebuild of all devices and recipe updates will arises if builds from both repo doesn't happen at same time. Separate TI_RTOS_METADATA_SRCREV_am64xx is added to avoid this syncing of builds from two different repos.

<3>
I will correct the typo in the commit-comment



-----Original Message-----
From: Bajjuri, Praneeth 
Sent: Friday, March 26, 2021 3:04 AM
To: Sivasankaran, Shiju <shiju@ti.com>; meta-ti@lists.yoctoproject.org
Subject: Re: [dunfell/master PATCH] ti-rtos: update metadata and version of am64x rtos firmware



On 3/25/2021 1:30 AM, Shiju Sivasankaran wrote:
> updated rtos firmware
> am64xx: 07.03.00.13
> 
> rtos fw of am64x build and upload from mcu-plus-sdk repo.
> rtos fw of am65x, j7 and j7200 build and upload from pdk repo.

But all of the fws are coming from the same repo processor-sdk/coresdk_rtos_releases where metadata.inc is updated on master with  right tag and public urls inside the metadata_*.inc isnt it ?

> a separate TI_RTOS_METADATA_SRCREV_am64xx for pining am64x fw from 
> macu-plus-sdk

typo s/macu?

  avoids syncing of builds from two different repos.

TI_RTOS_METADATA_SRCREV is pointing to master branch commitsha where the given tag is present. I dont see why we have to seperate this macro for each device seperately.


Add your signoff by.

> ---
>   recipes-ti/ti-rtos-bin/ti-rtos-firmware.bb | 6 +++---
>   recipes-ti/ti-rtos-bin/ti-rtos-metadata.bb | 1 +
>   2 files changed, 4 insertions(+), 3 deletions(-)
> 
> diff --git a/recipes-ti/ti-rtos-bin/ti-rtos-firmware.bb b/recipes-ti/ti-rtos-bin/ti-rtos-firmware.bb
> index e85948b2..fe020aa3 100644
> --- a/recipes-ti/ti-rtos-bin/ti-rtos-firmware.bb
> +++ b/recipes-ti/ti-rtos-bin/ti-rtos-firmware.bb
> @@ -23,8 +23,8 @@ DEFAULT_RTOS_FAMILY = "jacinto"
>   DEFAULT_RTOS_VERSION = "07_03_00_21"
>   DEFAULT_RTOS_VERSION_DOT = "07.03.00.21"
>   
> -DEFAULT_RTOS_VERSION_am64xx = "07_03_00_03"
> -DEFAULT_RTOS_VERSION_DOT_am64xx = "07.03.00.03"
> +DEFAULT_RTOS_VERSION_am64xx = "07_03_00_13"
> +DEFAULT_RTOS_VERSION_DOT_am64xx = "07.03.00.13"
>   
>   DEFAULT_RTOS_VERSION_am65xx = "07_03_00_22"
>   DEFAULT_RTOS_VERSION_DOT_am65xx = "07.03.00.22"
> @@ -51,7 +51,7 @@ DEFAULT_FIRMWARE_SHA256SUM = "unknown"
>   DEFAULT_FIRMWARE_SHA256SUM_j7 = "fc7805ed3d6c1801efb2b85ac9af99a8657aa3a7a9327bc16a1163d566f11c2c"
>   DEFAULT_FIRMWARE_SHA256SUM_j7200-evm = "71f282ba6768ad3b603a17d30851cc621986c88311c174960a50196bfc79f85c"
>   DEFAULT_FIRMWARE_SHA256SUM_am65xx = "61e0be08bea8ab1055645bd96504a6a29c70318c5b277237dee9981cd94d7f79"
> -DEFAULT_FIRMWARE_SHA256SUM_am64xx = "bc2d3d603ad30006ea3fefa3e777c26c19b95ee6d6f5a630259009f2f5a4996b"
> +DEFAULT_FIRMWARE_SHA256SUM_am64xx = "0c43cea6ccaf45b3d7c48996f1c561e814a6e409a330da5d45877c82ce7a299a"
>   
>   # Use weak assignment for CORESDK_RTOS_* variables to use defaults if not yet set
>   CORESDK_RTOS_FAMILY ?= "${DEFAULT_RTOS_FAMILY}"
> diff --git a/recipes-ti/ti-rtos-bin/ti-rtos-metadata.bb b/recipes-ti/ti-rtos-bin/ti-rtos-metadata.bb
> index 83e0e901..f052c627 100644
> --- a/recipes-ti/ti-rtos-bin/ti-rtos-metadata.bb
> +++ b/recipes-ti/ti-rtos-bin/ti-rtos-metadata.bb
> @@ -19,6 +19,7 @@ PLAT_SFX_am64xx = "/am64x"
>   TI_RTOS_METADATA_URI ?= "git://git.ti.com/processor-sdk/coresdk_rtos_releases.git"
>   TI_RTOS_METADATA_PROTOCOL ?= "git"
>   TI_RTOS_METADATA_SRCREV ?= "3cda650c9e22eb4fd4eb5797ad998991a6124243"
> +TI_RTOS_METADATA_SRCREV_am64xx ?= "051de801b539f4acc6e7630684f9993e34332dde"

Why this?
Why TI_RTOS_METADATA_SRCREV ?= "051de801b539f4acc6e7630684f9993e34332dde"
willnot work for all devices?


>   TI_RTOS_METADATA_BRANCH ?= "master"
>   TI_RTOS_METADATA_DIR ?= "${PLAT_SFX}"
>   TI_RTOS_METADATA_FILE ?= "${S}${TI_RTOS_METADATA_DIR}/metadata.inc"
> 

  reply	other threads:[~2021-03-26  3:49 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-03-25  6:30 [dunfell/master PATCH] ti-rtos: update metadata and version of am64x rtos firmware Shiju Sivasankaran
2021-03-25 21:34 ` praneeth
2021-03-26  3:49   ` Shiju Sivasankaran [this message]
  -- strict thread matches above, loose matches on Subject: below --
2021-03-26  7:17 Shiju Sivasankaran
2021-03-29  4:11 Shiju Sivasankaran
2021-03-29  4:41 ` praneeth
2021-03-31 18:56 Shiju Sivasankaran
2021-04-05  4:08 Shiju Sivasankaran

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=b7b048b378b14dde8ddf2b453307e162@ti.com \
    --to=shiju@ti.com \
    --cc=meta-ti@lists.yoctoproject.org \
    --cc=praneeth@ti.com \
    /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.