[lkml]   [2010]   [Nov]   [3]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
    Patch in this message
    SubjectRe: [PATCH] fix CAST detection to not screw with parens handling
    On Wed,  3 Nov 2010 13:04:33 +0100
    Florian Mickler <> wrote:

    > Casts have to be handled after the last type that is followed by an
    > opening parenthesis is handled.

    That is the wrong fix. I realized now that with that patch we would
    not claim anything as a CAST anymore. Better is probably to only claim
    a CAST if av_pending is not set. Andy, would that work? It seems to be
    alright... Do you have some tests for checkpatch?

    Testing it with the reported line and some other random samples it
    seems to be alright.


    commit 11ed611c647420ea27124faf269d724b4fd3ade4
    Author: Florian Mickler <>
    Date: Wed Nov 3 15:54:26 2010 +0100 fix CAST detection

    We should only claim that something is a cast if we did not encouter a
    token before, that did set av_pending.

    This fixes the operator * in the line below to be detected as
    binary (vs unary).

    kmalloc(sizeof(struct alphatrack_ocmd) * true_size, GFP_KERNEL);

    Reported-by: Audun Hoem <>
    Signed-off-by: Florian Mickler <>

    diff --git a/scripts/ b/scripts/
    index 90b54d4..06f5c44 100755
    --- a/scripts/
    +++ b/scripts/
    @@ -859,7 +859,7 @@ sub annotate_values {
    $av_preprocessor = 0;

    - } elsif ($cur =~ /^(\(\s*$Type\s*)\)/) {
    + } elsif ($cur =~ /^(\(\s*$Type\s*)\)/ && $av_pending eq '_') {
    print "CAST($1)\n" if ($dbg_values > 1);
    push(@av_paren_type, $type);
    $type = 'C';

     \ /
      Last update: 2010-11-03 16:23    [W:0.019 / U:2.896 seconds]
    ©2003-2016 Jasper Spaans. hosted at Digital OceanAdvertise on this site