linux-ext4 - Re: [PATCH 2/4] fs/buffer: parse IOCB_DONTCACHE flag in block_write

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <d698490f3ee35889d8922f392079846b647cd47e.camel@dubeyko.com>
Date: Fri, 04 Jul 2025 10:29:27 -0700
From: Viacheslav Dubeyko <slava@...eyko.com>
To: Yangtao Li <frank.li@...o.com>, axboe@...nel.dk,
 aivazian.tigran@...il.com, 	viro@...iv.linux.org.uk, brauner@...nel.org,
 jack@...e.cz, linkinjeon@...nel.org, 	sj1557.seo@...sung.com,
 yuezhang.mo@...y.com, glaubitz@...sik.fu-berlin.de, 	shaggy@...nel.org,
 konishi.ryusuke@...il.com, 	almaz.alexandrovich@...agon-software.com,
 me@...copeland.com, 	willy@...radead.org, josef@...icpanda.com,
 kovalev@...linux.org, dave@...olabs.net, 	mhocko@...e.com,
 chentaotao@...iglobal.com
Cc: linux-block@...r.kernel.org, linux-kernel@...r.kernel.org, 
	linux-fsdevel@...r.kernel.org, linux-ext4@...r.kernel.org, 
	jfs-discussion@...ts.sourceforge.net, linux-nilfs@...r.kernel.org, 
	ntfs3@...ts.linux.dev, linux-karma-devel@...ts.sourceforge.net, 
	bpf@...r.kernel.org
Subject: Re: [PATCH 2/4] fs/buffer: parse IOCB_DONTCACHE flag in
 block_write_begin()

On Thu, 2025-06-26 at 11:30 -0600, Yangtao Li wrote:
> When iocb flags passes IOCB_DONTCACHE, use FGP_DONTCACHE mode to get
> folio.
> 
> Signed-off-by: Yangtao Li <frank.li@...o.com>
> ---
>  fs/buffer.c | 7 +++++--
>  1 file changed, 5 insertions(+), 2 deletions(-)
> 
> diff --git a/fs/buffer.c b/fs/buffer.c
> index f2b7b30a76ca..0ed80b62feea 100644
> --- a/fs/buffer.c
> +++ b/fs/buffer.c
> @@ -2251,11 +2251,14 @@ int block_write_begin(struct kiocb *iocb,
> struct address_space *mapping, loff_t
>  		unsigned len, struct folio **foliop, get_block_t
> *get_block)
>  {
>  	pgoff_t index = pos >> PAGE_SHIFT;
> +	fgf_t fgp = FGP_WRITEBEGIN;
>  	struct folio *folio;
>  	int status;
>  
> -	folio = __filemap_get_folio(mapping, index, FGP_WRITEBEGIN,
> -			mapping_gfp_mask(mapping));
> +	if (iocb->ki_flags & IOCB_DONTCACHE)
> +		fgp |= FGP_DONTCACHE;
> +
> +	folio = __filemap_get_folio(mapping, index, fgp,
> mapping_gfp_mask(mapping));
>  	if (IS_ERR(folio))
>  		return PTR_ERR(folio);
>  

Correct me if I am wrong. As far as I can see, the first patch depends
from  second one. It means that if somebody applies the first patch
but, somehow, don't apply the second one, then nobody will be able to
compile the kernel code. Am I correct?

Why do we need to make this modification and, then, touch other file
systems? What the justification of this? Why do we need to make this
modification at the first place?

Thanks,
Slava.