[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <abc0dce9-b64f-451e-8315-ef9e39d0503b@ti.com>
Date: Wed, 23 Oct 2024 12:01:30 -0500
From: Andrew Davis <afd@...com>
To: Jiri Slaby <jirislaby@...nel.org>, Christoph Hellwig <hch@....de>,
<jassisinghbrar@...il.com>
CC: <stefani@...bold.net>, <gregkh@...uxfoundation.org>,
<linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v2] kfifo: don't include dma-mapping.h in kfifo.h
On 10/23/24 1:30 AM, Jiri Slaby wrote:
> On 23. 10. 24, 7:53, Christoph Hellwig wrote:
>> Nothing in kfifo.h directly needs dma-mapping.h, only two macros
>> use DMA_MAPPING_ERROR when actually instantiated. Drop the
>> dma-mapping.h include to reduce include bloat.
>>
>> Add an explicity <linux/io.h> include to drivers/mailbox/omap-mailbox.c
>> as that file uses __raw_readl and __raw_writel through a complicated
>> include chain involving <linux/dma-mapping.h>
>>
>> Fixes: d52b761e4b1a ("kfifo: add kfifo_dma_out_prepare_mapped()")
>> Signed-off-by: Christoph Hellwig <hch@....de>
>> ---
>>
>> Changes since v1:
>> - improve the commit log
>>
>> drivers/mailbox/omap-mailbox.c | 1 +
>> include/linux/kfifo.h | 1 -
>> samples/kfifo/dma-example.c | 1 +
>> 3 files changed, 2 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/mailbox/omap-mailbox.c b/drivers/mailbox/omap-mailbox.c
>> index 6797770474a5..680243751d62 100644
>> --- a/drivers/mailbox/omap-mailbox.c
>> +++ b/drivers/mailbox/omap-mailbox.c
>> @@ -15,6 +15,7 @@
>> #include <linux/slab.h>
>> #include <linux/kfifo.h>
>
> Funnily enough, this driver does not use kfifo since:
> commit 3f58c1f4206f37d0af4595a9046c76016334b301
> Author: Andrew Davis <afd@...com>
> Date: Wed Apr 10 08:59:42 2024 -0500
>
> mailbox: omap: Remove kernel FIFO message queuing
>
>> #include <linux/err.h>
>> +#include <linux/io.h>
>
> So this should have been revealed at that point (while removing kfifo.h in that commit too).
>
> Perhaps Andrew can remove the include now?
>
Yes, I'll send a patch dropping kfifo.h (unless you want to just do it
as part of this patch).
The addition of io.h is still correct here as we seem to have been
including it indirectly before, which is always a bit hacky.
So for this patch,
Acked-by: Andrew Davis <afd@...com>
> thanks,
Powered by blists - more mailing lists