On 2016/04/26 11:11, Steven Rostedt wrote:
> On Sun, 10 Apr 2016 17:06:33 -0700
> Benjamin Poirier <bpoir...@suse.com> wrote:
>
> > Note that this may change the resulting .config, causing it to have fewer
> > symbols turned on. Before this patch we incorrectly identified some symbols
> > as not having a prompt and needing to be selected by something else.
> >
> > Also fix the whitespace repeat after "tristate".
> >
> > Signed-off-by: Benjamin Poirier <bpoir...@suse.com>
> > ---
> > scripts/kconfig/streamline_config.pl | 2 +-
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/scripts/kconfig/streamline_config.pl
> > b/scripts/kconfig/streamline_config.pl
> > index bbc160c..d465672 100755
> > --- a/scripts/kconfig/streamline_config.pl
> > +++ b/scripts/kconfig/streamline_config.pl
> > @@ -237,7 +237,7 @@ sub read_kconfig {
> > }
> >
> > # configs without prompts must be selected
> > - } elsif ($state ne "NONE" && /^\s*tristate\s\S/) {
> > + } elsif ($state ne "NONE" && /^\s*(tristate\s+\S|prompt\b)/) {
>
> I'm curious. What modules have tristate and a specified prompt?
Actually, I didn't find config symbols that are tristate and have a separate
prompt. However, the following tristate symbols were not recognized by the
original expression because they are defined with two spaces between
"tristate" and the prompt string (sigh):
IP6_NF_RAW
IP_NF_RAW
NETFILTER_XT_MATCH_COMMENT
NETFILTER_XT_MATCH_CONNBYTES
NETFILTER_XT_MATCH_CONNMARK
NETFILTER_XT_MATCH_REALM
NETFILTER_XT_MATCH_SCTP
NETFILTER_XT_MATCH_STRING
NETFILTER_XT_TARGET_CONNMARK
NETFILTER_XT_TARGET_IDLETIMER
NETFILTER_XT_TARGET_NOTRACK
NETFILTER_XT_TARGET_TRACE
NF_CT_NETLINK_TIMEOUT
USB_EHCI_HCD_AT91
USB_EHCI_HCD_ORION
So they were not added to %prompts and now they are.
>
> That is, what modules are removed with this patch?
>
None, it turns out. Because it is the case that none of the above symbols are
selected by other symbols; the second condition in
if (!defined($prompts{$config}) && defined($selects{$config})) {
$process_selects{$config} = 1;
}
is always false for these symbols.
I've updated the patch subject and log accordingly.
