From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Date: Thu, 30 Sep 2010 13:49:53 +0300 From: Johan Hedberg To: Rafal Michalski Cc: linux-bluetooth@vger.kernel.org Subject: Re: [PATCH] Fix handlig for gboolean mce_bt_set field Message-ID: <20100930104953.GA13544@jh-x301> References: <1285842160-4127-1-git-send-email-michalski.raf@gmail.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <1285842160-4127-1-git-send-email-michalski.raf@gmail.com> Sender: linux-bluetooth-owner@vger.kernel.org List-ID: Hi Rafal, On Thu, Sep 30, 2010, Rafal Michalski wrote: > Expression (sigvalue & MCE_RADIO_STATE_BLUETOOTH) is invalid because it makes > wrong assumption about that it evaluates to TRUE, when bit > MCE_RADIO_STATE_BLUETOOTH is set. In this case, assignment > mce_bt_set = (sigvalue & MCE_RADIO_STATE_BLUETOOTH) makes that gboolean > "mce_bt_set" receives value equal 8 instead of TRUE (equal 1). So condition in > "if (mce_bt_set == powered)" statement will never evaluate to true because > gboolean "powered" receives only TRUE (equal 1) or FALSE (equal 0) value. > We should remember that gboolean type is not bool type in C++ sense. > It's simply typedef of gint (int), so assignmnet: > mce_bt_set = (sigvalue & MCE_RADIO_STATE_BLUETOOTH) is valid for compiler but > in our case it makes a bug. So normalization (double negation) in assignment: > mce_bt_set = !!(sigvalue & MCE_RADIO_STATE_BLUETOOTH) preserves that > "mce_bt_set" will receive only TRUE or FALSE value. > --- > plugins/maemo6.c | 2 +- > 1 files changed, 1 insertions(+), 1 deletions(-) Thanks. The patch has been pushed upstream with a much simplified commit message (the above is really quite complicated to read compared to the simplicity of the actual fix). Johan