2022-04-07 20:34:40

by Michael Walle

[permalink] [raw]
Subject: [PATCH v2] i2c: at91: use dma safe buffers

The supplied buffer might be on the stack and we get the following error
message:
[ 3.312058] at91_i2c e0070600.i2c: rejecting DMA map of vmalloc memory

Use i2c_{get,put}_dma_safe_msg_buf() to get a DMA-able memory region if
necessary.

Fixes: 60937b2cdbf9 ("i2c: at91: add dma support")
Signed-off-by: Michael Walle <[email protected]>
Reviewed-by: Codrin Ciubotariu <[email protected]>
---
changes since v1:
- remove extra empty line
- add fixes tag as suggested by Codrin

drivers/i2c/busses/i2c-at91-master.c | 11 +++++++++++
1 file changed, 11 insertions(+)

diff --git a/drivers/i2c/busses/i2c-at91-master.c b/drivers/i2c/busses/i2c-at91-master.c
index b0eae94909f4..5eca3b3bb609 100644
--- a/drivers/i2c/busses/i2c-at91-master.c
+++ b/drivers/i2c/busses/i2c-at91-master.c
@@ -656,6 +656,7 @@ static int at91_twi_xfer(struct i2c_adapter *adap, struct i2c_msg *msg, int num)
unsigned int_addr_flag = 0;
struct i2c_msg *m_start = msg;
bool is_read;
+ u8 *dma_buf;

dev_dbg(&adap->dev, "at91_xfer: processing %d messages:\n", num);

@@ -703,7 +704,17 @@ static int at91_twi_xfer(struct i2c_adapter *adap, struct i2c_msg *msg, int num)
dev->msg = m_start;
dev->recv_len_abort = false;

+ if (dev->use_dma) {
+ dma_buf = i2c_get_dma_safe_msg_buf(m_start, 1);
+ if (!dma_buf) {
+ ret = -ENOMEM;
+ goto out;
+ }
+ dev->buf = dma_buf;
+ }
+
ret = at91_do_twi_transfer(dev);
+ i2c_put_dma_safe_msg_buf(dma_buf, m_start, !ret);

ret = (ret < 0) ? ret : num;
out:
--
2.30.2


2022-05-03 00:53:01

by Michael Walle

[permalink] [raw]
Subject: Re: [PATCH v2] i2c: at91: use dma safe buffers

Am 2022-04-07 17:08, schrieb Michael Walle:
> The supplied buffer might be on the stack and we get the following
> error
> message:
> [ 3.312058] at91_i2c e0070600.i2c: rejecting DMA map of vmalloc
> memory
>
> Use i2c_{get,put}_dma_safe_msg_buf() to get a DMA-able memory region if
> necessary.
>
> Fixes: 60937b2cdbf9 ("i2c: at91: add dma support")
> Signed-off-by: Michael Walle <[email protected]>
> Reviewed-by: Codrin Ciubotariu <[email protected]>

Wolfgang, will you pick this one?

-michael

2022-05-05 03:49:19

by Wolfram Sang

[permalink] [raw]
Subject: Re: [PATCH v2] i2c: at91: use dma safe buffers

On Thu, Apr 07, 2022 at 05:08:28PM +0200, Michael Walle wrote:
> The supplied buffer might be on the stack and we get the following error
> message:
> [ 3.312058] at91_i2c e0070600.i2c: rejecting DMA map of vmalloc memory
>
> Use i2c_{get,put}_dma_safe_msg_buf() to get a DMA-able memory region if
> necessary.
>
> Fixes: 60937b2cdbf9 ("i2c: at91: add dma support")
> Signed-off-by: Michael Walle <[email protected]>
> Reviewed-by: Codrin Ciubotariu <[email protected]>

Applied to for-next, thanks!

BTW, my name is Wolfram.


Attachments:
(No filename) (561.00 B)
signature.asc (849.00 B)
Download all attachments

2022-05-05 15:45:20

by Michael Walle

[permalink] [raw]
Subject: Re: [PATCH v2] i2c: at91: use dma safe buffers

Am 2022-05-04 22:23, schrieb Wolfram Sang:
> BTW, my name is Wolfram.

Sorry :/

-michael

2022-05-05 16:09:30

by Nathan Chancellor

[permalink] [raw]
Subject: Re: [PATCH v2] i2c: at91: use dma safe buffers

On Thu, May 05, 2022 at 05:11:30PM +0200, Michael Walle wrote:
> [+ Wolfram]
>
> Am 2022-05-05 16:51, schrieb Nathan Chancellor:
> > On Thu, Apr 07, 2022 at 05:08:28PM +0200, Michael Walle wrote:
> > > The supplied buffer might be on the stack and we get the following
> > > error
> > > message:
> > > [ 3.312058] at91_i2c e0070600.i2c: rejecting DMA map of vmalloc
> > > memory
> > >
> > > Use i2c_{get,put}_dma_safe_msg_buf() to get a DMA-able memory region
> > > if
> > > necessary.
> > >
> > > Fixes: 60937b2cdbf9 ("i2c: at91: add dma support")
> > > Signed-off-by: Michael Walle <[email protected]>
> > > Reviewed-by: Codrin Ciubotariu <[email protected]>
> > > ---
> > > changes since v1:
> > > - remove extra empty line
> > > - add fixes tag as suggested by Codrin
> > >
> > > drivers/i2c/busses/i2c-at91-master.c | 11 +++++++++++
> > > 1 file changed, 11 insertions(+)
> > >
> > > diff --git a/drivers/i2c/busses/i2c-at91-master.c
> > > b/drivers/i2c/busses/i2c-at91-master.c
> > > index b0eae94909f4..5eca3b3bb609 100644
> > > --- a/drivers/i2c/busses/i2c-at91-master.c
> > > +++ b/drivers/i2c/busses/i2c-at91-master.c
> > > @@ -656,6 +656,7 @@ static int at91_twi_xfer(struct i2c_adapter
> > > *adap, struct i2c_msg *msg, int num)
> > > unsigned int_addr_flag = 0;
> > > struct i2c_msg *m_start = msg;
> > > bool is_read;
> > > + u8 *dma_buf;
> > >
> > > dev_dbg(&adap->dev, "at91_xfer: processing %d messages:\n", num);
> > >
> > > @@ -703,7 +704,17 @@ static int at91_twi_xfer(struct i2c_adapter
> > > *adap, struct i2c_msg *msg, int num)
> > > dev->msg = m_start;
> > > dev->recv_len_abort = false;
> > >
> > > + if (dev->use_dma) {
> > > + dma_buf = i2c_get_dma_safe_msg_buf(m_start, 1);
> > > + if (!dma_buf) {
> > > + ret = -ENOMEM;
> > > + goto out;
> > > + }
> > > + dev->buf = dma_buf;
> > > + }
> > > +
> > > ret = at91_do_twi_transfer(dev);
> > > + i2c_put_dma_safe_msg_buf(dma_buf, m_start, !ret);
> > >
> > > ret = (ret < 0) ? ret : num;
> > > out:
> > > --
> > > 2.30.2
> > >
> > >
> >
> > This change as commit 03fbb903c8bf ("i2c: at91: use dma safe buffers")
> > causes the following clang warning:
>
> Ugh..
>
> > drivers/i2c/busses/i2c-at91-master.c:707:6: error: variable 'dma_buf'
> > is used uninitialized whenever 'if' condition is false
> > [-Werror,-Wsometimes-uninitialized]
> > if (dev->use_dma) {
> > ^~~~~~~~~~~~
> > drivers/i2c/busses/i2c-at91-master.c:717:27: note: uninitialized use
> > occurs here
> > i2c_put_dma_safe_msg_buf(dma_buf, m_start, !ret);
> > ^~~~~~~
> > drivers/i2c/busses/i2c-at91-master.c:707:2: note: remove the 'if' if
> > its condition is always true
> > if (dev->use_dma) {
> > ^~~~~~~~~~~~~~~~~~
> > drivers/i2c/busses/i2c-at91-master.c:659:13: note: initialize the
> > variable 'dma_buf' to silence this warning
> > u8 *dma_buf;
> > ^
> > = NULL
> > 1 error generated.
> >
> > Should this variable be initialized or should the call to
> > i2c_put_dma_safe_msg_buf() be moved into the if statement?
>
> Initializing dma_buf with NULL should to the trick because
> i2c_put_dma_safe_msg_buf() is handling NULL gracefully. No
> need for an extra "if (dma_buf)".
>
> Will you take care or should I send a patch?

I can send a patch, thank you for the input!

Cheers,
Nathan

2022-05-07 05:03:39

by Michael Walle

[permalink] [raw]
Subject: Re: [PATCH v2] i2c: at91: use dma safe buffers

[+ Wolfram]

Am 2022-05-05 16:51, schrieb Nathan Chancellor:
> On Thu, Apr 07, 2022 at 05:08:28PM +0200, Michael Walle wrote:
>> The supplied buffer might be on the stack and we get the following
>> error
>> message:
>> [ 3.312058] at91_i2c e0070600.i2c: rejecting DMA map of vmalloc
>> memory
>>
>> Use i2c_{get,put}_dma_safe_msg_buf() to get a DMA-able memory region
>> if
>> necessary.
>>
>> Fixes: 60937b2cdbf9 ("i2c: at91: add dma support")
>> Signed-off-by: Michael Walle <[email protected]>
>> Reviewed-by: Codrin Ciubotariu <[email protected]>
>> ---
>> changes since v1:
>> - remove extra empty line
>> - add fixes tag as suggested by Codrin
>>
>> drivers/i2c/busses/i2c-at91-master.c | 11 +++++++++++
>> 1 file changed, 11 insertions(+)
>>
>> diff --git a/drivers/i2c/busses/i2c-at91-master.c
>> b/drivers/i2c/busses/i2c-at91-master.c
>> index b0eae94909f4..5eca3b3bb609 100644
>> --- a/drivers/i2c/busses/i2c-at91-master.c
>> +++ b/drivers/i2c/busses/i2c-at91-master.c
>> @@ -656,6 +656,7 @@ static int at91_twi_xfer(struct i2c_adapter *adap,
>> struct i2c_msg *msg, int num)
>> unsigned int_addr_flag = 0;
>> struct i2c_msg *m_start = msg;
>> bool is_read;
>> + u8 *dma_buf;
>>
>> dev_dbg(&adap->dev, "at91_xfer: processing %d messages:\n", num);
>>
>> @@ -703,7 +704,17 @@ static int at91_twi_xfer(struct i2c_adapter
>> *adap, struct i2c_msg *msg, int num)
>> dev->msg = m_start;
>> dev->recv_len_abort = false;
>>
>> + if (dev->use_dma) {
>> + dma_buf = i2c_get_dma_safe_msg_buf(m_start, 1);
>> + if (!dma_buf) {
>> + ret = -ENOMEM;
>> + goto out;
>> + }
>> + dev->buf = dma_buf;
>> + }
>> +
>> ret = at91_do_twi_transfer(dev);
>> + i2c_put_dma_safe_msg_buf(dma_buf, m_start, !ret);
>>
>> ret = (ret < 0) ? ret : num;
>> out:
>> --
>> 2.30.2
>>
>>
>
> This change as commit 03fbb903c8bf ("i2c: at91: use dma safe buffers")
> causes the following clang warning:

Ugh..

> drivers/i2c/busses/i2c-at91-master.c:707:6: error: variable 'dma_buf'
> is used uninitialized whenever 'if' condition is false
> [-Werror,-Wsometimes-uninitialized]
> if (dev->use_dma) {
> ^~~~~~~~~~~~
> drivers/i2c/busses/i2c-at91-master.c:717:27: note: uninitialized use
> occurs here
> i2c_put_dma_safe_msg_buf(dma_buf, m_start, !ret);
> ^~~~~~~
> drivers/i2c/busses/i2c-at91-master.c:707:2: note: remove the 'if' if
> its condition is always true
> if (dev->use_dma) {
> ^~~~~~~~~~~~~~~~~~
> drivers/i2c/busses/i2c-at91-master.c:659:13: note: initialize the
> variable 'dma_buf' to silence this warning
> u8 *dma_buf;
> ^
> = NULL
> 1 error generated.
>
> Should this variable be initialized or should the call to
> i2c_put_dma_safe_msg_buf() be moved into the if statement?

Initializing dma_buf with NULL should to the trick because
i2c_put_dma_safe_msg_buf() is handling NULL gracefully. No
need for an extra "if (dma_buf)".

Will you take care or should I send a patch?

-michael

2022-05-09 03:28:06

by Nathan Chancellor

[permalink] [raw]
Subject: Re: [PATCH v2] i2c: at91: use dma safe buffers

On Thu, Apr 07, 2022 at 05:08:28PM +0200, Michael Walle wrote:
> The supplied buffer might be on the stack and we get the following error
> message:
> [ 3.312058] at91_i2c e0070600.i2c: rejecting DMA map of vmalloc memory
>
> Use i2c_{get,put}_dma_safe_msg_buf() to get a DMA-able memory region if
> necessary.
>
> Fixes: 60937b2cdbf9 ("i2c: at91: add dma support")
> Signed-off-by: Michael Walle <[email protected]>
> Reviewed-by: Codrin Ciubotariu <[email protected]>
> ---
> changes since v1:
> - remove extra empty line
> - add fixes tag as suggested by Codrin
>
> drivers/i2c/busses/i2c-at91-master.c | 11 +++++++++++
> 1 file changed, 11 insertions(+)
>
> diff --git a/drivers/i2c/busses/i2c-at91-master.c b/drivers/i2c/busses/i2c-at91-master.c
> index b0eae94909f4..5eca3b3bb609 100644
> --- a/drivers/i2c/busses/i2c-at91-master.c
> +++ b/drivers/i2c/busses/i2c-at91-master.c
> @@ -656,6 +656,7 @@ static int at91_twi_xfer(struct i2c_adapter *adap, struct i2c_msg *msg, int num)
> unsigned int_addr_flag = 0;
> struct i2c_msg *m_start = msg;
> bool is_read;
> + u8 *dma_buf;
>
> dev_dbg(&adap->dev, "at91_xfer: processing %d messages:\n", num);
>
> @@ -703,7 +704,17 @@ static int at91_twi_xfer(struct i2c_adapter *adap, struct i2c_msg *msg, int num)
> dev->msg = m_start;
> dev->recv_len_abort = false;
>
> + if (dev->use_dma) {
> + dma_buf = i2c_get_dma_safe_msg_buf(m_start, 1);
> + if (!dma_buf) {
> + ret = -ENOMEM;
> + goto out;
> + }
> + dev->buf = dma_buf;
> + }
> +
> ret = at91_do_twi_transfer(dev);
> + i2c_put_dma_safe_msg_buf(dma_buf, m_start, !ret);
>
> ret = (ret < 0) ? ret : num;
> out:
> --
> 2.30.2
>
>

This change as commit 03fbb903c8bf ("i2c: at91: use dma safe buffers")
causes the following clang warning:

drivers/i2c/busses/i2c-at91-master.c:707:6: error: variable 'dma_buf' is used uninitialized whenever 'if' condition is false [-Werror,-Wsometimes-uninitialized]
if (dev->use_dma) {
^~~~~~~~~~~~
drivers/i2c/busses/i2c-at91-master.c:717:27: note: uninitialized use occurs here
i2c_put_dma_safe_msg_buf(dma_buf, m_start, !ret);
^~~~~~~
drivers/i2c/busses/i2c-at91-master.c:707:2: note: remove the 'if' if its condition is always true
if (dev->use_dma) {
^~~~~~~~~~~~~~~~~~
drivers/i2c/busses/i2c-at91-master.c:659:13: note: initialize the variable 'dma_buf' to silence this warning
u8 *dma_buf;
^
= NULL
1 error generated.

Should this variable be initialized or should the call to
i2c_put_dma_safe_msg_buf() be moved into the if statement?

Cheers,
Nathan