All of lore.kernel.org
 help / color / mirror / Atom feed
From: Patrick Boettcher <patrick.boettcher@posteo.de>
To: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Cc: "Linux Media Mailing List" <linux-media@vger.kernel.org>,
	"Mauro Carvalho Chehab" <mchehab@infradead.org>,
	"Andy Lutomirski" <luto@amacapital.net>,
	"Johannes Stezenbach" <js@linuxtv.org>,
	"Jiri Kosina" <jikos@kernel.org>,
	"Linux Kernel Mailing List" <linux-kernel@vger.kernel.org>,
	"Andy Lutomirski" <luto@kernel.org>,
	"Michael Krufky" <mkrufky@linuxtv.org>,
	"Mauro Carvalho Chehab" <mchehab@kernel.org>,
	"Jörg Otte" <jrg.otte@gmail.com>
Subject: Re: [PATCH 05/26] cinergyT2-fe: don't do DMA on stack
Date: Mon, 10 Oct 2016 08:39:15 +0200	[thread overview]
Message-ID: <20161010083915.392a6d95@posteo.de> (raw)
In-Reply-To: <9d415b35135ba2d457df12859b64eacbc05bc2e4.1475860773.git.mchehab@s-opensource.com>

On Fri,  7 Oct 2016 14:24:15 -0300
Mauro Carvalho Chehab <mchehab@s-opensource.com> wrote:

> The USB control messages require DMA to work. We cannot pass
> a stack-allocated buffer, as it is not warranted that the
> stack would be into a DMA enabled area.
> 
> Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
> ---
>  drivers/media/usb/dvb-usb/cinergyT2-fe.c | 45
> ++++++++++++++++---------------- 1 file changed, 23 insertions(+), 22
> deletions(-)
> 
> diff --git a/drivers/media/usb/dvb-usb/cinergyT2-fe.c
> b/drivers/media/usb/dvb-usb/cinergyT2-fe.c index
> fd8edcb56e61..03ba66ef1f28 100644 ---
> a/drivers/media/usb/dvb-usb/cinergyT2-fe.c +++
> b/drivers/media/usb/dvb-usb/cinergyT2-fe.c @@ -139,6 +139,9 @@ static
> uint16_t compute_tps(struct dtv_frontend_properties *op) struct
> cinergyt2_fe_state { struct dvb_frontend fe;
>  	struct dvb_usb_device *d;
> +
> +	unsigned char data[64];
> +
>  	struct dvbt_get_status_msg status;
>  };
>  
> @@ -146,28 +149,28 @@ static int cinergyt2_fe_read_status(struct
> dvb_frontend *fe, enum fe_status *status)
>  {
>  	struct cinergyt2_fe_state *state = fe->demodulator_priv;
> -	struct dvbt_get_status_msg result;
> -	u8 cmd[] = { CINERGYT2_EP1_GET_TUNER_STATUS };
>  	int ret;
>  
> -	ret = dvb_usb_generic_rw(state->d, cmd, sizeof(cmd), (u8
> *)&result,
> -			sizeof(result), 0);
> +	state->data[0] = CINERGYT2_EP1_GET_TUNER_STATUS;
> +
> +	ret = dvb_usb_generic_rw(state->d, state->data, 1,
> +				 state->data, sizeof(state->status),
> 0); if (ret < 0)
>  		return ret;
>  
> -	state->status = result;
> +	memcpy(&state->status, state->data, sizeof(state->status));
>  
>  	*status = 0;
>  
> -	if (0xffff - le16_to_cpu(result.gain) > 30)
> +	if (0xffff - le16_to_cpu(state->status.gain) > 30)
>  		*status |= FE_HAS_SIGNAL;
> -	if (result.lock_bits & (1 << 6))
> +	if (state->status.lock_bits & (1 << 6))
>  		*status |= FE_HAS_LOCK;
> -	if (result.lock_bits & (1 << 5))
> +	if (state->status.lock_bits & (1 << 5))
>  		*status |= FE_HAS_SYNC;
> -	if (result.lock_bits & (1 << 4))
> +	if (state->status.lock_bits & (1 << 4))
>  		*status |= FE_HAS_CARRIER;
> -	if (result.lock_bits & (1 << 1))
> +	if (state->status.lock_bits & (1 << 1))
>  		*status |= FE_HAS_VITERBI;
>  
>  	if ((*status & (FE_HAS_CARRIER | FE_HAS_VITERBI |
> FE_HAS_SYNC)) != @@ -232,31 +235,29 @@ static int
> cinergyt2_fe_set_frontend(struct dvb_frontend *fe) {
>  	struct dtv_frontend_properties *fep =
> &fe->dtv_property_cache; struct cinergyt2_fe_state *state =
> fe->demodulator_priv;
> -	struct dvbt_set_parameters_msg param;
> -	char result[2];
> +	struct dvbt_set_parameters_msg *param = (void *)state->data;
>  	int err;
>  
> -	param.cmd = CINERGYT2_EP1_SET_TUNER_PARAMETERS;
> -	param.tps = cpu_to_le16(compute_tps(fep));
> -	param.freq = cpu_to_le32(fep->frequency / 1000);
> -	param.flags = 0;
> +	param->cmd = CINERGYT2_EP1_SET_TUNER_PARAMETERS;
> +	param->tps = cpu_to_le16(compute_tps(fep));
> +	param->freq = cpu_to_le32(fep->frequency / 1000);
> +	param->flags = 0;
>  
>  	switch (fep->bandwidth_hz) {
>  	default:
>  	case 8000000:
> -		param.bandwidth = 8;
> +		param->bandwidth = 8;
>  		break;
>  	case 7000000:
> -		param.bandwidth = 7;
> +		param->bandwidth = 7;
>  		break;
>  	case 6000000:
> -		param.bandwidth = 6;
> +		param->bandwidth = 6;
>  		break;
>  	}
>  
> -	err = dvb_usb_generic_rw(state->d,
> -			(char *)&param, sizeof(param),
> -			result, sizeof(result), 0);
> +	err = dvb_usb_generic_rw(state->d, state->data,
> sizeof(*param),
> +				 state->data, 2, 0);
>  	if (err < 0)
>  		err("cinergyt2_fe_set_frontend() Failed! err=%d\n",
> err); 

Reviewed-By: Patrick Boettcher <patrick.boettcher@posteo.de>

  reply	other threads:[~2016-10-10  6:39 UTC|newest]

Thread overview: 42+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-10-07 17:24 [PATCH 00/26] Don't use stack for DMA transers on dvb-usb drivers Mauro Carvalho Chehab
2016-10-07 17:24 ` [PATCH 01/26] af9005: don't do DMA on stack Mauro Carvalho Chehab
2016-10-07 17:24 ` [PATCH 02/26] cinergyT2-core: " Mauro Carvalho Chehab
2016-10-10  6:39   ` Patrick Boettcher
2016-10-07 17:24 ` [PATCH 03/26] cinergyT2-core:: handle error code on RC query Mauro Carvalho Chehab
2016-10-07 17:24 ` [PATCH 04/26] cinergyT2-fe: cache stats at cinergyt2_fe_read_status() Mauro Carvalho Chehab
2016-10-07 17:24 ` [PATCH 05/26] cinergyT2-fe: don't do DMA on stack Mauro Carvalho Chehab
2016-10-10  6:39   ` Patrick Boettcher [this message]
2016-10-07 17:24 ` [PATCH 06/26] cxusb: " Mauro Carvalho Chehab
2016-10-10  6:38   ` Patrick Boettcher
2016-10-07 17:24 ` [PATCH 07/26] dib0700: be sure that dib0700_ctrl_rd() users can do DMA Mauro Carvalho Chehab
2016-10-10  6:38   ` Patrick Boettcher
2016-10-07 17:24 ` [PATCH 08/26] dib0700_core: don't use stack on I2C reads Mauro Carvalho Chehab
2016-10-10  6:37   ` Patrick Boettcher
2016-10-07 17:24 ` [PATCH 09/26] dibusb: don't do DMA on stack Mauro Carvalho Chehab
2016-10-10  6:37   ` Patrick Boettcher
2016-10-07 17:24 ` [PATCH 10/26] dibusb: handle error code on RC query Mauro Carvalho Chehab
2016-10-07 17:24 ` [PATCH 11/26] digitv: don't do DMA on stack Mauro Carvalho Chehab
2016-10-10  6:36   ` Patrick Boettcher
2016-10-07 17:24 ` [PATCH 12/26] dtt200u-fe: " Mauro Carvalho Chehab
2016-10-07 17:24 ` [PATCH 13/26] dtt200u-fe: handle errors on USB control messages Mauro Carvalho Chehab
2016-10-07 17:24 ` [PATCH 14/26] dtt200u: don't do DMA on stack Mauro Carvalho Chehab
2016-10-10  6:36   ` Patrick Boettcher
2016-10-07 17:24 ` [PATCH 15/26] dtt200u: handle USB control message errors Mauro Carvalho Chehab
2016-10-07 17:24 ` [PATCH 16/26] dtv5100: : don't do DMA on stack Mauro Carvalho Chehab
2016-10-07 17:24 ` [PATCH 17/26] gp8psk: " Mauro Carvalho Chehab
2016-10-07 17:24 ` [PATCH 18/26] gp8psk: don't go past the buffer size Mauro Carvalho Chehab
2016-10-07 17:24 ` [PATCH 19/26] nova-t-usb2: don't do DMA on stack Mauro Carvalho Chehab
2016-10-10  6:35   ` Patrick Boettcher
2016-10-07 17:24 ` [PATCH 20/26] pctv452e: " Mauro Carvalho Chehab
2016-10-10  6:35   ` Patrick Boettcher
2016-10-07 17:24 ` [PATCH 21/26] pctv452e: don't call BUG_ON() on non-fatal error Mauro Carvalho Chehab
2016-10-08 10:11   ` [PATCH v2 " Mauro Carvalho Chehab
2016-10-07 17:24 ` [PATCH 22/26] technisat-usb2: use DMA buffers for I2C transfers Mauro Carvalho Chehab
2016-10-10  6:34   ` Patrick Boettcher
2016-10-07 17:24 ` [PATCH 23/26] dvb-usb: warn if return value for USB read/write routines is not checked Mauro Carvalho Chehab
2016-10-07 17:24 ` [PATCH 24/26] nova-t-usb2: handle error code on RC query Mauro Carvalho Chehab
2016-10-07 17:24 ` [PATCH 25/26] dw2102: return error if su3000_power_ctrl() fails Mauro Carvalho Chehab
2016-10-07 17:24 ` [PATCH 26/26] digitv: handle error code on RC query Mauro Carvalho Chehab
2016-10-10 11:24 ` [PATCH 00/26] Don't use stack for DMA transers on dvb-usb drivers Antti Palosaari
2016-10-10 11:44   ` Michael Ira Krufky
2016-10-11 10:12     ` Mauro Carvalho Chehab

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20161010083915.392a6d95@posteo.de \
    --to=patrick.boettcher@posteo.de \
    --cc=jikos@kernel.org \
    --cc=jrg.otte@gmail.com \
    --cc=js@linuxtv.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-media@vger.kernel.org \
    --cc=luto@amacapital.net \
    --cc=luto@kernel.org \
    --cc=mchehab@infradead.org \
    --cc=mchehab@kernel.org \
    --cc=mchehab@s-opensource.com \
    --cc=mkrufky@linuxtv.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.