Re: [PATCH v10 2/7] i3c: master: use i3c_master_register only for main master

From: Alexandre Belloni <alexandre.belloni@bootlin.com>
To: Parshuram Raju Thombare <pthombar@cadence.com>
Cc: Nicolas Pitre <nico@fluxnic.net>,
	"slongerbeam@gmail.com" <slongerbeam@gmail.com>,
	"vitor.soares@synopsys.com" <vitor.soares@synopsys.com>,
	"praneeth@ti.com" <praneeth@ti.com>,
	Milind Parab <mparab@cadence.com>,
	"linux-i3c@lists.infradead.org" <linux-i3c@lists.infradead.org>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH v10 2/7] i3c: master: use i3c_master_register only for main master
Date: Wed, 9 Jun 2021 01:21:00 +0200	[thread overview]
Message-ID: <YL/7XP1fs/2bRiTM@piout.net> (raw)
In-Reply-To: <DM5PR07MB319618AFCB415501CD33741FC1CD0@DM5PR07MB3196.namprd07.prod.outlook.com>

On 08/12/2020 05:47:51+0000, Parshuram Raju Thombare wrote:
> >This looks a bit confusing. Here you're rolling back detailss in
> >i3c_primary_master_register() that were factored out in
> >i3c_master_init(). If i3c_master_init() is successful, then you
> >shouldn't be undoing its things openly in i3c_primary_master_register().
> >Instead, there should be another function that does the reverse of
> >i3c_master_init() here.
> 
> Every function do its cleanup in case of failures.
> And if any failure occur after successful i3c_master_init(), we have
> function i3c_master_bus_cleanup() which does the major cleanup.
> 

The point from Nicolas here was that the workqueue is allocated in
i3c_master_init so you should have a function to destroy it instead of
having to do that separately in i3c_primary_master_register. The same is
true for the put_device. Or you have to ensure i3c_masterdev_release
is called when i3c_primary_master_register fails.

-- 
Alexandre Belloni, co-owner and COO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com