buildroot.busybox.net archive mirror
 help / color / mirror / Atom feed
From: Peter Korsgaard <peter@korsgaard.com>
To: Arnout Vandecappelle <arnout@mind.be>
Cc: Romain Naour <romain.naour@smile.fr>,
	buildroot@buildroot.org,
	"Yann E. MORIN" <yann.morin.1998@free.fr>,
	Thomas Petazzoni <thomas.petazzoni@bootlin.com>,
	Andreas Dannenberg <dannenberg@ti.com>
Subject: Re: [Buildroot] [PATCH v9 06/11] board/ti/am62x_sk|am64x_sk: switch to TI SDK v8.6 sources
Date: Sun, 25 Jun 2023 21:14:27 +0200	[thread overview]
Message-ID: <87v8fbnylo.fsf@48ers.dk> (raw)
In-Reply-To: <1383cf3a-fa81-75b7-1e3f-45a82ebb4c7d@mind.be> (Arnout Vandecappelle's message of "Sun, 25 Jun 2023 20:59:01 +0200")

>>>>> "Arnout" == Arnout Vandecappelle <arnout@mind.be> writes:

Hi,

 >> Correct, but given the poor quality of most vendor kernels, using a
 >> board with mainline is typically a lot nicer (if the IP blocks you care
 >> about are supported naturally).

 >  I'm not sure that you can claim most vendor kernels are poor
 >  quality. At least NXP and Microchip have decent vendor
 > kernels. Conversely, when vendor kernels are horrible, mainline is
 > often not super usable either (because, of course, those vendors don't
 > invest in upstreaming either).

Sure, not all vendor kernels are equal, but vendor drivers are often
written by people not very familiar with the subsystem, so things get
handled in "odd ways", and as they have not been submitted upstream they
don't get any review feedback to clean it up. The vendor kernels of high
quality is typically the ones where the vendor developers interact with
mainline (so their drivers get reviewed by the subsystem maintainers).
E.G. for the Microchip and NXP cases, the basics are anyway in mainline,
so it is normal that the identical parts in the vendor kernel also looks
sensible.

I am not aware of any situations where the mainline kernel is of lower
quality than the vendor kernel. It will often have less features, but the
support that is there is often cleaner.


 >> Same about the support, E.G. we had issues with various vendor
 >> kernels/bootloaders breaking with newer toolchain versions.
 >> 
 >> >  That said, I think for each board we should look at what the vendor
 >> >  kernel really brings. If everything, including GPU, is working with
 >> > the upstream kernel, it doesn't make sense to use the vendor kernel. I
 >> > don't know if that's the case in this specific situation.
 >> >  Yann, Peter, Romain, Thomas, what do you think?

 >> Agreed. I would in general say go for mainline, unless there is no
 >> decent support in mainline. I would not necessarily think that GPU
 >> support is a requirement, but at least the things used by the defconfig
 >> (E.G. storage, serial, network).

 >  That's exactly the opposite of what I was saying, so not agreed I guess :-)

I guess it is a question of glass half full vs half empty. You wrote
".. we should look at what the vendor kernel really brings. If
everything, including GPU, is working with the upsteam kernel, it
doesn't make sense to use the vendor kernel", which I agreed with,
EXCEPT that I think the cutoff point should be lower than "everything
supported", but just if mainline is suitable for booting a Buildroot
defconfig, E.G. working serial/storage/network.

But like usual, this is largely up to the person contributing the
defconfig.

-- 
Bye, Peter Korsgaard
_______________________________________________
buildroot mailing list
buildroot@buildroot.org
https://lists.buildroot.org/mailman/listinfo/buildroot

  reply	other threads:[~2023-06-25 19:14 UTC|newest]

Thread overview: 54+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-06-22 16:02 [Buildroot] [PATCH v9 00/11] add support for TI's AM64x and AM62x boards Andreas Dannenberg via buildroot
2023-06-22 16:02 ` [Buildroot] [PATCH v9 01/11] boot/ti-k3-r5-loader: allow for full build source customization Andreas Dannenberg via buildroot
2023-06-24 21:23   ` Yann E. MORIN
2023-06-25 13:21     ` Arnout Vandecappelle via buildroot
2023-06-25 13:35       ` Yann E. MORIN
2023-06-26 19:44   ` Julien Olivain
2023-06-26 19:53   ` Julien Olivain
2023-06-22 16:02 ` [Buildroot] [PATCH v9 02/11] boot/ti-k3-image-gen: new package Andreas Dannenberg via buildroot
2023-06-24 22:28   ` Yann E. MORIN
2023-08-08 23:38   ` Bryce Johnson
2023-08-15  7:15     ` Andreas Dannenberg via buildroot
2023-08-15 22:54       ` Bryce Johnson
2023-06-22 16:02 ` [Buildroot] [PATCH v9 03/11] boot/uboot: add support for building the TI K3 DM into U-Boot Andreas Dannenberg via buildroot
2023-06-25  7:02   ` Yann E. MORIN
2023-06-25  7:08     ` Yann E. MORIN
2023-06-22 16:02 ` [Buildroot] [PATCH v9 04/11] board/ti/am64x_sk: add new board Andreas Dannenberg via buildroot
2023-06-25  5:41   ` François Perrad
2023-06-25 13:43   ` Yann E. MORIN
2023-06-22 16:02 ` [Buildroot] [PATCH v9 05/11] board/ti/am62x_sk: " Andreas Dannenberg via buildroot
2023-06-25  5:42   ` François Perrad
2023-08-15  7:21     ` Andreas Dannenberg via buildroot
2023-06-22 16:02 ` [Buildroot] [PATCH v9 06/11] board/ti/am62x_sk|am64x_sk: switch to TI SDK v8.6 sources Andreas Dannenberg via buildroot
2023-06-25 13:54   ` Yann E. MORIN
2023-06-25 14:33     ` Arnout Vandecappelle via buildroot
2023-06-25 15:22       ` Peter Korsgaard
2023-06-25 18:59         ` Arnout Vandecappelle via buildroot
2023-06-25 19:14           ` Peter Korsgaard [this message]
2023-06-25 19:36       ` Yann E. MORIN
2023-06-22 16:02 ` [Buildroot] [PATCH v9 07/11] package/ti-core-secdev-k3: new package Andreas Dannenberg via buildroot
2023-06-23  3:48   ` Patrick Oppenlander
2023-06-23 14:53     ` Andreas Dannenberg via buildroot
2023-06-24  0:32       ` Patrick Oppenlander
2023-06-24  1:11         ` Andreas Dannenberg via buildroot
2023-06-24  4:09           ` Patrick Oppenlander
2023-06-25  7:55       ` Yann E. MORIN
2023-06-25 13:26         ` Arnout Vandecappelle via buildroot
2023-06-22 16:02 ` [Buildroot] [PATCH v9 08/11] board/ti/am62x_sk|am64x_sk: switch to HS-FS device variants Andreas Dannenberg via buildroot
2023-06-22 16:02 ` [Buildroot] [PATCH v9 09/11] package/ti-rogue-km: new package Andreas Dannenberg via buildroot
2023-06-25  8:59   ` Yann E. MORIN
2023-08-18 17:30     ` Bryce Johnson
2023-06-22 16:02 ` [Buildroot] [PATCH v9 10/11] package/ti-rogue-um: " Andreas Dannenberg via buildroot
2023-06-23  7:30   ` François Perrad
2023-06-23 14:59     ` Andreas Dannenberg via buildroot
2023-06-25  5:37       ` François Perrad
2023-06-25 10:15         ` Yann E. MORIN
2023-06-27  2:02           ` Andreas Dannenberg via buildroot
2023-08-22 15:15           ` Thomas Petazzoni via buildroot
2023-06-27 22:48       ` Andreas Dannenberg via buildroot
2023-08-22 10:40     ` Thomas Petazzoni via buildroot
2023-06-22 16:02 ` [Buildroot] [PATCH v9 11/11] configs/am62x_sk_defconfig: enable IMG Rogue graphics driver Andreas Dannenberg via buildroot
2023-06-23  4:02 ` [Buildroot] [PATCH v9 00/11] add support for TI's AM64x and AM62x boards Patrick Oppenlander
2023-06-23 15:04   ` Andreas Dannenberg via buildroot
2023-08-22 10:14 ` Thomas Petazzoni via buildroot
2023-08-22 18:05   ` Thomas Petazzoni via buildroot

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=87v8fbnylo.fsf@48ers.dk \
    --to=peter@korsgaard.com \
    --cc=arnout@mind.be \
    --cc=buildroot@buildroot.org \
    --cc=dannenberg@ti.com \
    --cc=romain.naour@smile.fr \
    --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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).