Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Mon, 10 Oct 2022 08:25:22 +0800
From:   Baoquan He <bhe@redhat.com>
To:     Kefeng Wang <wangkefeng.wang@huawei.com>
Cc:     linux-kernel@vger.kernel.org, linux-mm@kvack.org,
        akpm@linux-foundation.org, hch@infradead.org,
        agordeev@linux.ibm.com, christophe.leroy@csgroup.eu,
        schnelle@linux.ibm.com, David.Laight@aculab.com, shorne@gmail.com,
        Arnd Bergmann <arnd@arndb.de>, linux-arch@vger.kernel.org,
        linux-arm-kernel@lists.infradead.org
Subject: Re: [PATCH v3 03/11] mm/ioremap: change the return value of
 io[re|un]map_allowed and rename
Message-ID: <Y0NmcspddDQICfTi@MiWiFi-R3L-srv>
References: <20221009103114.149036-1-bhe@redhat.com>
 <20221009103114.149036-4-bhe@redhat.com>
 <10ada8f0-0931-b6a6-e240-fc8b500e578d@huawei.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <10ada8f0-0931-b6a6-e240-fc8b500e578d@huawei.com>
Precedence: bulk

On 10/09/22 at 07:13pm, Kefeng Wang wrote:
> 
> On 2022/10/9 18:31, Baoquan He wrote:
> > Currently, hooks ioremap_allowed() and iounmap_allowed() are used to
> > check if it's qualified to do ioremap, and now this is done on ARM64.
> > However, in oder to convert more architectures to take GENERIC_IOREMAP
> > method, several more things need be done in those two hooks:
> >   1) The io address mapping need be handled specifically on architectures,
> >      e.g arc, ia64, s390;
> >   2) The original physical address passed into ioremap_prot() need be
> >      fixed up, e.g arc;
> >   3) The 'prot' passed into ioremap_prot() need be adjusted, e.g on arc
> >      and xtensa.
> > 
> > To handle these three issues,
> > 
> >   1) Rename ioremap_allowed() and iounmap_allowed() to arch_ioremap()
> >      and arch_iounmap() since the old name can't reflect their
> >      functionality after change;
> >   2) Change the return value of arch_ioremap() so that arch can add
> >      specifical io address mapping handling inside and return the maped
> >      address. Now their returned value means:
> >      ===
> >      arch_ioremap() return a bool,
> pointer?

Right, I forgot fixing it again. Thanks.

> >        - IS_ERR means return an error
> >        - 0 means continue to remap
> >        - a non-zero, non-IS_ERR pointer is returned directly
> >      arch_iounmap() return a bool,
> >        - true means continue to vunmap
> >        - false means skip vunmap and return directly
> ...
> >   /*
> > diff --git a/include/asm-generic/io.h b/include/asm-generic/io.h
> > index a68f8fbf423b..2ae16906f3be 100644
> > --- a/include/asm-generic/io.h
> > +++ b/include/asm-generic/io.h
> > @@ -1049,25 +1049,26 @@ static inline void iounmap(volatile void __iomem *addr)
> >   /*
> >    * Arch code can implement the following two hooks when using GENERIC_IOREMAP
> > - * ioremap_allowed() return a bool,
> > - *   - true means continue to remap
> > - *   - false means skip remap and return directly
> > - * iounmap_allowed() return a bool,
> > + * arch_ioremap() return a bool,
> ditto...

Will change.

> >   	area = get_vm_area_caller(size, VM_IOREMAP,
> >   			__builtin_return_address(0));
> >   	if (!area)
> > @@ -52,7 +57,7 @@ void iounmap(volatile void __iomem *addr)
> >   {
> >   	void *vaddr = (void *)((unsigned long)addr & PAGE_MASK);
> > -	if (!iounmap_allowed(vaddr))
> > +	if (!arch_iounmap((void __iomem *)addr))
> vaddr?

No, it's intentional. Alexander suggested this, both of you discussed
this in v1, see below thread.

https://lore.kernel.org/all/Yu4mYxpV0GWRTjQp@li-4a3a4a4c-28e5-11b2-a85c-a8d192c6f089.ibm.com/T/#u

> >   		return;
> >   	if (is_vmalloc_addr(vaddr))
>