linux-kernel - Re: [RFC PATCH 1/1] mm: provide one common K(x) macro

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: <5529272.KFOknHQvy8@natalenko.name>
Date:   Wed, 01 Sep 2021 12:50:40 +0200
From:   Oleksandr Natalenko <oleksandr@...alenko.name>
To:     Michal Hocko <mhocko@...e.com>
Cc:     linux-kernel@...r.kernel.org, linux-mm@...ck.org,
        Andrew Morton <akpm@...ux-foundation.org>,
        Vlastimil Babka <vbabka@...e.cz>,
        Matthew Wilcox <willy@...radead.org>,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        Miaohe Lin <linmiaohe@...wei.com>,
        Mel Gorman <mgorman@...hsingularity.net>,
        Stephen Rothwell <sfr@...b.auug.org.au>
Subject: Re: [RFC PATCH 1/1] mm: provide one common K(x) macro

Hello.

On středa 1. září 2021 12:31:36 CEST Michal Hocko wrote:
> On Wed 01-09-21 11:21:49, Oleksandr Natalenko wrote:
> > There are various places where the K(x) macro is defined. This commit
> > gets rid of multiple definitions and provides a common one.
> > 
> > This doesn't solve open-coding this macro in various other places. This
> > should be addressed by another subsequent commit.
> 
> Why is this an improvement? You are adding a header file for a single
> macro which sounds like an overkill.

I agree a separate header file is an overkill for just one #define, hence
still looking for a suggestion on a better place for it.

> The overall net outcome is added
> lines of code.

Not always. There are some long statements like:

```
seq_printf(seq, ",size=%luk",
        sbinfo->max_blocks << (PAGE_SHIFT - 10));
```

that are split into two lines. With the macro those take one line only:

```
seq_printf(seq, ",size=%luk", K(sbinfo->max_blocks));
```

As of now (counting unposted open-coding replacements) the grand total is:

```
31 files changed, 104 insertions(+), 90 deletions(-)
```

which is not that horrible.

> It is not like K() or any of its variant is adding a
> maintenance burden due to code duplication. So why do we want to change
> the existing state?

For me it's about readability. Compare, for instance:

```
seq_put_decimal_ull_width(m, str, (val) << (PAGE_SHIFT-10), 8)
```

and

```
seq_put_decimal_ull_width(m, str, K(val), 8)
```

It's a small yet visible difference.

Thanks.

-- 
Oleksandr Natalenko (post-factum)