| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Date: Fri, 20 Mar 2015 17:27:22 +0000
From: Javi Merino <javi.merino@....com>
To: Alex Elder <elder@...e.org>
Cc: "akpm@...ux-foundation.org" <akpm@...ux-foundation.org>,
"intel-gfx@...ts.freedesktop.org" <intel-gfx@...ts.freedesktop.org>,
"dri-devel@...ts.freedesktop.org" <dri-devel@...ts.freedesktop.org>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
Antti Palosaari <crope@....fi>,
Mauro Carvalho Chehab <mchehab@....samsung.com>
Subject: Re: [PATCH 4/4] media: cxd2820r: use DIV_ROUND_CLOSEST_ULL()
On Fri, Mar 20, 2015 at 01:51:36PM +0000, Alex Elder wrote:
> On 03/20/2015 06:14 AM, Javi Merino wrote:
> > Now that the kernel provides DIV_ROUND_CLOSEST_ULL(), drop the internal
> > implementation and use the kernel one.
> >
> > Cc: Antti Palosaari <crope@....fi>
> > Cc: Mauro Carvalho Chehab <mchehab@....samsung.com>
> > Signed-off-by: Javi Merino <javi.merino@....com>
> > ---
> > I've only compile-tested it, I don't have the hardware to run it.
> >
> > drivers/media/dvb-frontends/cxd2820r_c.c | 2 +-
> > drivers/media/dvb-frontends/cxd2820r_core.c | 6 ------
> > drivers/media/dvb-frontends/cxd2820r_priv.h | 2 --
> > drivers/media/dvb-frontends/cxd2820r_t.c | 2 +-
> > drivers/media/dvb-frontends/cxd2820r_t2.c | 2 +-
> > 5 files changed, 3 insertions(+), 11 deletions(-)
> >
> > diff --git a/drivers/media/dvb-frontends/cxd2820r_c.c b/drivers/media/dvb-frontends/cxd2820r_c.c
> > index 149fdca3fb44..72b0e2db3aab 100644
> > --- a/drivers/media/dvb-frontends/cxd2820r_c.c
> > +++ b/drivers/media/dvb-frontends/cxd2820r_c.c
> > @@ -79,7 +79,7 @@ int cxd2820r_set_frontend_c(struct dvb_frontend *fe)
> >
> > num = if_freq / 1000; /* Hz => kHz */
> > num *= 0x4000;
> > - if_ctl = 0x4000 - cxd2820r_div_u64_round_closest(num, 41000);
> > + if_ctl = 0x4000 - DIV_ROUND_CLOSEST_ULL(num, 41000);
> > buf[0] = (if_ctl >> 8) & 0x3f;
> > buf[1] = (if_ctl >> 0) & 0xff;
> >
> > diff --git a/drivers/media/dvb-frontends/cxd2820r_core.c b/drivers/media/dvb-frontends/cxd2820r_core.c
> > index 422e84bbb008..490e090048ef 100644
> > --- a/drivers/media/dvb-frontends/cxd2820r_core.c
> > +++ b/drivers/media/dvb-frontends/cxd2820r_core.c
> > @@ -244,12 +244,6 @@ error:
> > return ret;
> > }
> >
> > -/* 64 bit div with round closest, like DIV_ROUND_CLOSEST but 64 bit */
> > -u32 cxd2820r_div_u64_round_closest(u64 dividend, u32 divisor)
> > -{
> > - return div_u64(dividend + (divisor / 2), divisor);
> > -}
>
> Technically, I'd say this has a bug, because the result
> needs to be 64 bits wide or your results might be much
> different from what might be desired.
>
> Practically though, I'm pretty sure all callers provide
> values that ensure the result is valid.
All the callers are substituted in this patch so we can make sure that
they are all correct.
> I only call attention because this patch changes the return
> type of the function that gets called to do the calculation.
I'm not sure I follow. Do you mean that this:
if_ctl = 0x4000 - DIV_ROUND_CLOSEST_ULL(num, 41000);
Should actually be:
if_ctl = 0x4000 - (u32)DIV_ROUND_CLOSEST_ULL(num, 41000);
?
if_ctl is a u16 so I don't think you gain anything by doing that.
> > -
> > static int cxd2820r_set_frontend(struct dvb_frontend *fe)
> > {
> > struct cxd2820r_priv *priv = fe->demodulator_priv;
> > diff --git a/drivers/media/dvb-frontends/cxd2820r_priv.h b/drivers/media/dvb-frontends/cxd2820r_priv.h
> > index 7ff5f60c83e1..4b428959b16e 100644
> > --- a/drivers/media/dvb-frontends/cxd2820r_priv.h
> > +++ b/drivers/media/dvb-frontends/cxd2820r_priv.h
> > @@ -64,8 +64,6 @@ int cxd2820r_wr_reg_mask(struct cxd2820r_priv *priv, u32 reg, u8 val,
> > int cxd2820r_wr_regs(struct cxd2820r_priv *priv, u32 reginfo, u8 *val,
> > int len);
> >
> > -u32 cxd2820r_div_u64_round_closest(u64 dividend, u32 divisor);
> > -
> > int cxd2820r_wr_regs(struct cxd2820r_priv *priv, u32 reginfo, u8 *val,
> > int len);
> >
> > diff --git a/drivers/media/dvb-frontends/cxd2820r_t.c b/drivers/media/dvb-frontends/cxd2820r_t.c
> > index 51401d036530..008cb2ac8480 100644
> > --- a/drivers/media/dvb-frontends/cxd2820r_t.c
> > +++ b/drivers/media/dvb-frontends/cxd2820r_t.c
> > @@ -103,7 +103,7 @@ int cxd2820r_set_frontend_t(struct dvb_frontend *fe)
> >
> > num = if_freq / 1000; /* Hz => kHz */
> > num *= 0x1000000;
> > - if_ctl = cxd2820r_div_u64_round_closest(num, 41000);
> > + if_ctl = DIV_ROUND_CLOSEST_ULL(num, 41000);
if_ctl is a u32, so you get the same behavior that you were getting
before: the downcasting of u64 to u32 happened in
cxd2820r_div_u64_round_closest(), now it happens here.
> > buf[0] = ((if_ctl >> 16) & 0xff);
> > buf[1] = ((if_ctl >> 8) & 0xff);
> > buf[2] = ((if_ctl >> 0) & 0xff);
> > diff --git a/drivers/media/dvb-frontends/cxd2820r_t2.c b/drivers/media/dvb-frontends/cxd2820r_t2.c
> > index 9c0c4f42175c..35fe364c7182 100644
> > --- a/drivers/media/dvb-frontends/cxd2820r_t2.c
> > +++ b/drivers/media/dvb-frontends/cxd2820r_t2.c
> > @@ -120,7 +120,7 @@ int cxd2820r_set_frontend_t2(struct dvb_frontend *fe)
> >
> > num = if_freq / 1000; /* Hz => kHz */
> > num *= 0x1000000;
> > - if_ctl = cxd2820r_div_u64_round_closest(num, 41000);
> > + if_ctl = DIV_ROUND_CLOSEST_ULL(num, 41000);
Same here if_ctl is a u32.
Cheers,
Javi
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists