From mboxrd@z Thu Jan 1 00:00:00 1970 From: Boaz Harrosh Subject: Re: direct-io: squelch maybe-uninitialized warning in do_direct_IO() Date: Thu, 17 Jul 2014 12:40:44 +0300 Message-ID: <53C79A1C.5060402@gmail.com> References: <1404452632-10912-1-git-send-email-pramod.gurav.etc@gmail.com> <20140713115022.GA6054@infradead.org> <53C6B199.2070606@gmail.com> <20140716175803.GA3631@infradead.org> <1405536124.4357.17.camel@x41> <53C79949.7010201@gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: Borislav Petkov , Sam Ravnborg , pramod.gurav.etc@gmail.com, Jason Cooper , Markus Mayer , linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org To: Paul Bolle , Christoph Hellwig , viro@zeniv.linux.org.uk Return-path: In-Reply-To: <53C79949.7010201@gmail.com> Sender: linux-kernel-owner@vger.kernel.org List-Id: linux-fsdevel.vger.kernel.org On 07/17/2014 12:37 PM, Boaz Harrosh wrote: > From: Paul Bolle >=20 > The following warnings: >=20 > fs/direct-io.c: In function =E2=80=98__blockdev_direct_IO=E2=80=99: > fs/direct-io.c:1011:12: warning: =E2=80=98to=E2=80=99 may be used u= ninitialized in this function [-Wmaybe-uninitialized] > fs/direct-io.c:913:16: note: =E2=80=98to=E2=80=99 was declared here > fs/direct-io.c:1011:12: warning: =E2=80=98from=E2=80=99 may be used= uninitialized in this function [-Wmaybe-uninitialized] > fs/direct-io.c:913:10: note: =E2=80=98from=E2=80=99 was declared he= re >=20 > are false positive because dio_get_page() either fails, or sets both > 'from' and 'to'. >=20 > Maybe it's better to move initializing "to" and "from" out of > dio_get_page(). That _might_ make it easier for both the the reader a= nd > the compiler to understand what's going on. Something like this: >=20 > Christoph Hellwig said ... > The fix of moving the code defintively looks nicer, while I think > uninitialized_var is horrible wart that won't get anywhere near my co= de. >=20 > Boaz Harrosh I agree with Christoph and Paul >=20 So Here, everyone likes this one (I think) please ACK/Review and let us= merge it. > Signed-off-by: Paul Bolle Paul this is your patch I put you as Signed-off-by please acknowledge > Signed-off-by: Boaz Harrosh Thanks Boaz > --- > fs/direct-io.c | 14 +++++++------- > 1 file changed, 7 insertions(+), 7 deletions(-) >=20 > diff --git a/fs/direct-io.c b/fs/direct-io.c > index 98040ba..2f024fc 100644 > --- a/fs/direct-io.c > +++ b/fs/direct-io.c > @@ -198,9 +198,8 @@ static inline int dio_refill_pages(struct dio *di= o, struct dio_submit *sdio) > * L1 cache. > */ > static inline struct page *dio_get_page(struct dio *dio, > - struct dio_submit *sdio, size_t *from, size_t *to) > + struct dio_submit *sdio) > { > - int n; > if (dio_pages_present(sdio) =3D=3D 0) { > int ret; > =20 > @@ -209,10 +208,7 @@ static inline struct page *dio_get_page(struct d= io *dio, > return ERR_PTR(ret); > BUG_ON(dio_pages_present(sdio) =3D=3D 0); > } > - n =3D sdio->head++; > - *from =3D n ? 0 : sdio->from; > - *to =3D (n =3D=3D sdio->tail - 1) ? sdio->to : PAGE_SIZE; > - return dio->pages[n]; > + return dio->pages[sdio->head]; > } > =20 > /** > @@ -911,11 +907,15 @@ static int do_direct_IO(struct dio *dio, struct= dio_submit *sdio, > while (sdio->block_in_file < sdio->final_block_in_request) { > struct page *page; > size_t from, to; > - page =3D dio_get_page(dio, sdio, &from, &to); > + > + page =3D dio_get_page(dio, sdio); > if (IS_ERR(page)) { > ret =3D PTR_ERR(page); > goto out; > } > + from =3D sdio->head ? 0 : sdio->from; > + to =3D (sdio->head =3D=3D sdio->tail - 1) ? sdio->to : PAGE_SIZE; > + sdio->head++; > =20 > while (from < to) { > unsigned this_chunk_bytes; /* # of bytes mapped */ >=20