cocci.inria.fr archive mirror
 help / color / mirror / Atom feed
From: julia.lawall@lip6.fr (Julia Lawall)
To: cocci@systeme.lip6.fr
Subject: [Cocci] __asm statements confuse spatch
Date: Wed, 17 Oct 2018 07:25:58 +0200 (CEST)	[thread overview]
Message-ID: <alpine.DEB.2.21.1810170720110.2985@hadrien> (raw)
In-Reply-To: <CAOZdJXWvGOK3M0w_2N3ZXoLWSeJaJfC-UeHhfo6cG_qyLOFxJw@mail.gmail.com>



On Tue, 16 Oct 2018, Timur Tabi wrote:

> I'm trying to modify a Windows .c file, and it contains several __asm
> (or _asm or asm) statements that confuse spatch.  They look like this:
>
>     _asm {mov ax, ss}
>     __asm mov uRetval,eax                // Just keep 32 bits.
>     __asm {
>         PAUSE
>         PAUSE
>     }
>
> And so on.  Is there a way to get spatch to ignore these statements?

Linux uses __asm__ ( ... ), which is what Coccinelle recognizes.  I can
probably add _asm and __asm with the braces.  On the other hand, the
second case, with no delimiter seems awkward.  Does that occur a lot?
Basically it's not clear how to parse it.  I could have __asm eat up
everything until the end of the line, but then the third case won't work.

>
> Another problem I've having with the source file is that it has
> inconsistent usage of braces, and sometimes spatch wants to add
> unnecessary braces that look off.  For example, this:
>
>         if (...)
>             DBG_PRINTF((...));
>         else
>             DBG_PRINTF((...));
>         }
>
> (the } belongs to some if-statement much earlier in code somewhere) becomes:
>
>         if (...) {
>             NV_PRINTF(...);
>         }
>         else {
>             NV_PRINTF(...);
>         }
>         }
>
> I really don't want spatch to add the braces.

I don't think this has anything to do with the trailing }.  Coccinelle
knows which brace goes with what, independent of the indentation.
Something about your rule is making it unsure whether the changed code is
in a branch by itself, or whether you have added multiple statements.

For example, if your rule is

- A;
+ B;
+ C;

and the code is if (x) A;, then the braces are needed.  Spatch is a bit
conservative about this, ie it adds brace unless it is clear that there is
a replacement of a single statement by another one.

You could try to track down the problem by making a minimal semantic
patch and C code that show the problem, or just add some rules to clean
up afterwards.

julia

> _______________________________________________
> Cocci mailing list
> Cocci at systeme.lip6.fr
> https://systeme.lip6.fr/mailman/listinfo/cocci
>

  reply	other threads:[~2018-10-17  5:25 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-10-16 22:49 [Cocci] __asm statements confuse spatch Timur Tabi
2018-10-17  5:25 ` Julia Lawall [this message]
2018-10-17 12:15   ` Timur Tabi
2018-10-17 12:26     ` Julia Lawall

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=alpine.DEB.2.21.1810170720110.2985@hadrien \
    --to=julia.lawall@lip6.fr \
    --cc=cocci@systeme.lip6.fr \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).