[PATCH] net: off by one, try #2

[PATCH] net: off by one, try #2

[Roel Kluin]

Better?

-------------------->8----------------8<-----------------------
With while (x++ < n) { ... } x can reach n+1.

Signed-off-by: Roel Kluin <roel.kluin@gmail.com>
---
 drivers/net/3c505.c        |    4 ++--
 drivers/net/irda/mcs7780.c |    2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/net/3c505.c b/drivers/net/3c505.c
index 6124605..fb74846 100644
--- a/drivers/net/3c505.c
+++ b/drivers/net/3c505.c
@@ -500,9 +500,9 @@ static bool receive_pcb(struct net_device *dev, pcb_struct * pcb)
 		pcb->data.raw[i++] = inb_command(dev->base_addr);
 		if (i > MAX_PCB_DATA)
 			INVALID_PCB_MSG(i);
-	} while ((stat & ASF_PCB_MASK) != ASF_PCB_END && j < 20000);
+	} while ((stat & ASF_PCB_MASK) != ASF_PCB_END && j <= 20000);
 	spin_unlock_irqrestore(&adapter->lock, flags);
-	if (j >= 20000) {
+	if (j > 20000) {
 		TIMEOUT_MSG(__LINE__);
 		return false;
 	}
diff --git a/drivers/net/irda/mcs7780.c b/drivers/net/irda/mcs7780.c
index 7eafdca..b4f4f19 100644
--- a/drivers/net/irda/mcs7780.c
+++ b/drivers/net/irda/mcs7780.c
@@ -585,7 +585,7 @@ static int mcs_speed_change(struct mcs_cb *mcs)
 		mcs_get_reg(mcs, MCS_RESV_REG, &rval);
 	} while(cnt++ < 100 && (rval & MCS_IRINTX));
 
-	if(cnt >= 100) {
+	if(cnt > 100) {
 		IRDA_ERROR("unable to change speed\n");
 		ret = -EIO;
 		goto error;

Re: [PATCH] net: off by one, try #2

[Jarek Poplawski]

On 11-02-2009 11:01, Roel Kluin wrote:
> Better?

Yes, but it looks like here is even more...

> 
> -------------------->8----------------8<-----------------------
> With while (x++ < n) { ... } x can reach n+1.
> 
> Signed-off-by: Roel Kluin <roel.kluin@gmail.com>
> ---
>  drivers/net/3c505.c        |    4 ++--
>  drivers/net/irda/mcs7780.c |    2 +-
>  2 files changed, 3 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/net/3c505.c b/drivers/net/3c505.c
> index 6124605..fb74846 100644
> --- a/drivers/net/3c505.c
> +++ b/drivers/net/3c505.c
> @@ -500,9 +500,9 @@ static bool receive_pcb(struct net_device *dev, pcb_struct * pcb)
>  		pcb->data.raw[i++] = inb_command(dev->base_addr);
>  		if (i > MAX_PCB_DATA)
>  			INVALID_PCB_MSG(i);

i is also misused here and array can be overriden, so additional
break/return is needed.

Thanks,
Jarek P.

> -	} while ((stat & ASF_PCB_MASK) != ASF_PCB_END && j < 20000);
> +	} while ((stat & ASF_PCB_MASK) != ASF_PCB_END && j <= 20000);
>  	spin_unlock_irqrestore(&adapter->lock, flags);
> -	if (j >= 20000) {
> +	if (j > 20000) {
>  		TIMEOUT_MSG(__LINE__);
>  		return false;
>  	}
> diff --git a/drivers/net/irda/mcs7780.c b/drivers/net/irda/mcs7780.c
> index 7eafdca..b4f4f19 100644
> --- a/drivers/net/irda/mcs7780.c
> +++ b/drivers/net/irda/mcs7780.c
> @@ -585,7 +585,7 @@ static int mcs_speed_change(struct mcs_cb *mcs)
>  		mcs_get_reg(mcs, MCS_RESV_REG, &rval);
>  	} while(cnt++ < 100 && (rval & MCS_IRINTX));
>  
> -	if(cnt >= 100) {
> +	if(cnt > 100) {
>  		IRDA_ERROR("unable to change speed\n");
>  		ret = -EIO;
>  		goto error;

Re: [PATCH] net: off by one, try #2

[Roel Kluin]

>> With while (x++ < n) { ... } x can reach n+1.
> 
> Yes, but it looks like here is even more...
> i is also misused here and array can be overriden, so additional
> break/return is needed.

Thanks, is this how it should be?
-------------------->8----------------8<-----------------------

With while (x++ < n) { ... } x can reach n+1. As Jarek Poplawski pointed out, array
pcb->data.raw was not correctly used.

Signed-off-by: Roel Kluin <roel.kluin@gmail.com>
---
diff --git a/drivers/net/3c505.c b/drivers/net/3c505.c
index 6124605..4cf3050 100644
--- a/drivers/net/3c505.c
+++ b/drivers/net/3c505.c
@@ -497,12 +497,15 @@ static bool receive_pcb(struct net_device *dev, pcb_struct * pcb)
 	do {
 		j = 0;
 		while (((stat = get_status(dev->base_addr)) & ACRF) == 0 && j++ < 20000);
-		pcb->data.raw[i++] = inb_command(dev->base_addr);
-		if (i > MAX_PCB_DATA)
-			INVALID_PCB_MSG(i);
-	} while ((stat & ASF_PCB_MASK) != ASF_PCB_END && j < 20000);
+		pcb->data.raw[i] = inb_command(dev->base_addr);
+	} while (++i < MAX_PCB_DATA && (stat & ASF_PCB_MASK) != ASF_PCB_END && j <= 20000);
+
 	spin_unlock_irqrestore(&adapter->lock, flags);
-	if (j >= 20000) {
+	if (i >= MAX_PCB_DATA) {
+		INVALID_PCB_MSG(i);
+		return false;
+	}
+	if (j > 20000) {
 		TIMEOUT_MSG(__LINE__);
 		return false;
 	}

Re: [PATCH] net: off by one, try #2

[Jarek Poplawski]

On Wed, Feb 11, 2009 at 03:22:52PM +0100, Roel Kluin wrote:
> >> With while (x++ < n) { ... } x can reach n+1.
> > 
> > Yes, but it looks like here is even more...
> > i is also misused here and array can be overriden, so additional
> > break/return is needed.
> 
> Thanks, is this how it should be?

It looks (almost) OK to me. :-) Except this > 80 line.

On the other hand, I wonder if it's not a good time to make it more
readable; I mean the first while (): length, ";", and maybe ++j
similarly to i now?

BTW, I hope you remember about irda.

Jarek P.

> -------------------->8----------------8<-----------------------
> 
> With while (x++ < n) { ... } x can reach n+1. As Jarek Poplawski pointed out, array
> pcb->data.raw was not correctly used.
> 
> Signed-off-by: Roel Kluin <roel.kluin@gmail.com>
> ---
> diff --git a/drivers/net/3c505.c b/drivers/net/3c505.c
> index 6124605..4cf3050 100644
> --- a/drivers/net/3c505.c
> +++ b/drivers/net/3c505.c
> @@ -497,12 +497,15 @@ static bool receive_pcb(struct net_device *dev, pcb_struct * pcb)
>  	do {
>  		j = 0;
>  		while (((stat = get_status(dev->base_addr)) & ACRF) == 0 && j++ < 20000);
> -		pcb->data.raw[i++] = inb_command(dev->base_addr);
> -		if (i > MAX_PCB_DATA)
> -			INVALID_PCB_MSG(i);
> -	} while ((stat & ASF_PCB_MASK) != ASF_PCB_END && j < 20000);
> +		pcb->data.raw[i] = inb_command(dev->base_addr);
> +	} while (++i < MAX_PCB_DATA && (stat & ASF_PCB_MASK) != ASF_PCB_END && j <= 20000);
> +
>  	spin_unlock_irqrestore(&adapter->lock, flags);
> -	if (j >= 20000) {
> +	if (i >= MAX_PCB_DATA) {
> +		INVALID_PCB_MSG(i);
> +		return false;
> +	}
> +	if (j > 20000) {
>  		TIMEOUT_MSG(__LINE__);
>  		return false;
>  	}

Re: [PATCH] net: off by one, try #2

[Roel Kluin]

Jarek Poplawski wrote:
> On Wed, Feb 11, 2009 at 03:22:52PM +0100, Roel Kluin wrote:
>>>> With while (x++ < n) { ... } x can reach n+1.
>>> Yes, but it looks like here is even more...
>>> i is also misused here and array can be overriden, so additional
>>> break/return is needed.
>> Thanks, is this how it should be?
> 
> It looks (almost) OK to me. :-) Except this > 80 line.
> 
> On the other hand, I wonder if it's not a good time to make it more
> readable; I mean the first while (): length, ";", and maybe ++j
> similarly to i now?
> 
> BTW, I hope you remember about irda.
> 
> Jarek P.

Thanks again for reviewing, how's this? with irda this time.
-------------------->8----------------8<-----------------------

With while (x++ < n) { ... } x can reach n+1. As Jarek Poplawski pointed out, array
pcb->data.raw was not correctly used.

Signed-off-by: Roel Kluin <roel.kluin@gmail.com>
---
diff --git a/drivers/net/3c505.c b/drivers/net/3c505.c
index 6124605..4ade64a 100644
--- a/drivers/net/3c505.c
+++ b/drivers/net/3c505.c
@@ -493,15 +493,21 @@ static bool receive_pcb(struct net_device *dev, pcb_struct * pcb)
 	}
 	/* read the data */
 	spin_lock_irqsave(&adapter->lock, flags);
-	i = 0;
-	do {
-		j = 0;
-		while (((stat = get_status(dev->base_addr)) & ACRF) == 0 && j++ < 20000);
-		pcb->data.raw[i++] = inb_command(dev->base_addr);
-		if (i > MAX_PCB_DATA)
-			INVALID_PCB_MSG(i);
-	} while ((stat & ASF_PCB_MASK) != ASF_PCB_END && j < 20000);
+	for (i = 0; i < MAX_PCB_DATA; i++) {
+		for (j = 0; j < 20000; j++) {
+			stat = get_status(dev->base_addr);
+			if (stat & ACRF)
+				break;
+		}
+		pcb->data.raw[i] = inb_command(dev->base_addr);
+		if (stat & ASF_PCB_MASK == ASF_PCB_END || j >= 20000)
+			break;
+	}
 	spin_unlock_irqrestore(&adapter->lock, flags);
+	if (i >= MAX_PCB_DATA) {
+		INVALID_PCB_MSG(i);
+		return false;
+	}
 	if (j >= 20000) {
 		TIMEOUT_MSG(__LINE__);
 		return false;
diff --git a/drivers/net/irda/mcs7780.c b/drivers/net/irda/mcs7780.c
index 7eafdca..85e88da 100644
--- a/drivers/net/irda/mcs7780.c
+++ b/drivers/net/irda/mcs7780.c
@@ -585,7 +585,7 @@ static int mcs_speed_change(struct mcs_cb *mcs)
 		mcs_get_reg(mcs, MCS_RESV_REG, &rval);
 	} while(cnt++ < 100 && (rval & MCS_IRINTX));
 
-	if(cnt >= 100) {
+	if (cnt > 100) {
 		IRDA_ERROR("unable to change speed\n");
 		ret = -EIO;
 		goto error;

Re: [PATCH] net: off by one, try #2

[Jarek Poplawski]

On Wed, Feb 11, 2009 at 08:33:25PM +0100, Roel Kluin wrote:
> Jarek Poplawski wrote:
> > On Wed, Feb 11, 2009 at 03:22:52PM +0100, Roel Kluin wrote:
> >>>> With while (x++ < n) { ... } x can reach n+1.
> >>> Yes, but it looks like here is even more...
> >>> i is also misused here and array can be overriden, so additional
> >>> break/return is needed.
> >> Thanks, is this how it should be?
> > 
> > It looks (almost) OK to me. :-) Except this > 80 line.
> > 
> > On the other hand, I wonder if it's not a good time to make it more
> > readable; I mean the first while (): length, ";", and maybe ++j
> > similarly to i now?
> > 
> > BTW, I hope you remember about irda.
> > 
> > Jarek P.
> 
> Thanks again for reviewing, how's this? with irda this time.

Hmm... except one bug: "stat & ..." needs "()"...

Roel, please resend once more, but separated: something could go to
stable, but not necessarily together. And add more description,
especially to this 3c505 (here is more changes than 'while (x++ < n)').

Anyway, I hope David will like these "off by one" patches again,
especially after this, and sun3lance one. Good catches!

Cheers,
Jarek P.

> -------------------->8----------------8<-----------------------
> 
> With while (x++ < n) { ... } x can reach n+1. As Jarek Poplawski pointed out, array
> pcb->data.raw was not correctly used.
> 
> Signed-off-by: Roel Kluin <roel.kluin@gmail.com>
> ---
> diff --git a/drivers/net/3c505.c b/drivers/net/3c505.c
> index 6124605..4ade64a 100644
> --- a/drivers/net/3c505.c
> +++ b/drivers/net/3c505.c
> @@ -493,15 +493,21 @@ static bool receive_pcb(struct net_device *dev, pcb_struct * pcb)
>  	}
>  	/* read the data */
>  	spin_lock_irqsave(&adapter->lock, flags);
> -	i = 0;
> -	do {
> -		j = 0;
> -		while (((stat = get_status(dev->base_addr)) & ACRF) == 0 && j++ < 20000);
> -		pcb->data.raw[i++] = inb_command(dev->base_addr);
> -		if (i > MAX_PCB_DATA)
> -			INVALID_PCB_MSG(i);
> -	} while ((stat & ASF_PCB_MASK) != ASF_PCB_END && j < 20000);
> +	for (i = 0; i < MAX_PCB_DATA; i++) {
> +		for (j = 0; j < 20000; j++) {
> +			stat = get_status(dev->base_addr);
> +			if (stat & ACRF)
> +				break;
> +		}
> +		pcb->data.raw[i] = inb_command(dev->base_addr);
> +		if (stat & ASF_PCB_MASK == ASF_PCB_END || j >= 20000)
> +			break;
> +	}
>  	spin_unlock_irqrestore(&adapter->lock, flags);
> +	if (i >= MAX_PCB_DATA) {
> +		INVALID_PCB_MSG(i);
> +		return false;
> +	}
>  	if (j >= 20000) {
>  		TIMEOUT_MSG(__LINE__);
>  		return false;
> diff --git a/drivers/net/irda/mcs7780.c b/drivers/net/irda/mcs7780.c
> index 7eafdca..85e88da 100644
> --- a/drivers/net/irda/mcs7780.c
> +++ b/drivers/net/irda/mcs7780.c
> @@ -585,7 +585,7 @@ static int mcs_speed_change(struct mcs_cb *mcs)
>  		mcs_get_reg(mcs, MCS_RESV_REG, &rval);
>  	} while(cnt++ < 100 && (rval & MCS_IRINTX));
>  
> -	if(cnt >= 100) {
> +	if (cnt > 100) {
>  		IRDA_ERROR("unable to change speed\n");
>  		ret = -EIO;
>  		goto error;
> 

Re: [PATCH] net: off by one, try #2

[Jarek Poplawski]

On Wed, Feb 11, 2009 at 09:27:55PM +0100, Jarek Poplawski wrote:
> On Wed, Feb 11, 2009 at 08:33:25PM +0100, Roel Kluin wrote:
...

One more thing below...

> > Signed-off-by: Roel Kluin <roel.kluin@gmail.com>
> > ---
> > diff --git a/drivers/net/3c505.c b/drivers/net/3c505.c
> > index 6124605..4ade64a 100644
> > --- a/drivers/net/3c505.c
> > +++ b/drivers/net/3c505.c
> > @@ -493,15 +493,21 @@ static bool receive_pcb(struct net_device *dev, pcb_struct * pcb)
> >  	}
> >  	/* read the data */
> >  	spin_lock_irqsave(&adapter->lock, flags);
> > -	i = 0;
> > -	do {
> > -		j = 0;
> > -		while (((stat = get_status(dev->base_addr)) & ACRF) == 0 && j++ < 20000);
> > -		pcb->data.raw[i++] = inb_command(dev->base_addr);
> > -		if (i > MAX_PCB_DATA)
> > -			INVALID_PCB_MSG(i);
> > -	} while ((stat & ASF_PCB_MASK) != ASF_PCB_END && j < 20000);
> > +	for (i = 0; i < MAX_PCB_DATA; i++) {
> > +		for (j = 0; j < 20000; j++) {
> > +			stat = get_status(dev->base_addr);
> > +			if (stat & ACRF)
> > +				break;
> > +		}
> > +		pcb->data.raw[i] = inb_command(dev->base_addr);
> > +		if (stat & ASF_PCB_MASK == ASF_PCB_END || j >= 20000)
> > +			break;

Here is no ++i after breaking the loop, so no need for [--i] later;
please, check this more.

Jarek P.

> > +	}
> >  	spin_unlock_irqrestore(&adapter->lock, flags);
> > +	if (i >= MAX_PCB_DATA) {
> > +		INVALID_PCB_MSG(i);
> > +		return false;
> > +	}
> >  	if (j >= 20000) {
> >  		TIMEOUT_MSG(__LINE__);
> >  		return false;
> > diff --git a/drivers/net/irda/mcs7780.c b/drivers/net/irda/mcs7780.c
> > index 7eafdca..85e88da 100644
> > --- a/drivers/net/irda/mcs7780.c
> > +++ b/drivers/net/irda/mcs7780.c
> > @@ -585,7 +585,7 @@ static int mcs_speed_change(struct mcs_cb *mcs)
> >  		mcs_get_reg(mcs, MCS_RESV_REG, &rval);
> >  	} while(cnt++ < 100 && (rval & MCS_IRINTX));
> >  
> > -	if(cnt >= 100) {
> > +	if (cnt > 100) {
> >  		IRDA_ERROR("unable to change speed\n");
> >  		ret = -EIO;
> >  		goto error;
> > 

[PATCH] 3c505: do not set pcb->data.raw beyond its size

[Roel Kluin]

Many thanks, Jarek,

Is this changelog ok?
------------------------------>8----------------8<------------------------------

Ensure that we do not set pcb->data.raw beyond its size, print an error message
and return false if we attempt to. A timout message was printed one too early.

Signed-off-by: Roel Kluin <roel.kluin@gmail.com>
---
diff --git a/drivers/net/3c505.c b/drivers/net/3c505.c
index 6124605..a8107f9 100644
--- a/drivers/net/3c505.c
+++ b/drivers/net/3c505.c
@@ -493,21 +493,27 @@ static bool receive_pcb(struct net_device *dev, pcb_struct * pcb)
 	}
 	/* read the data */
 	spin_lock_irqsave(&adapter->lock, flags);
-	i = 0;
-	do {
-		j = 0;
-		while (((stat = get_status(dev->base_addr)) & ACRF) == 0 && j++ < 20000);
-		pcb->data.raw[i++] = inb_command(dev->base_addr);
-		if (i > MAX_PCB_DATA)
-			INVALID_PCB_MSG(i);
-	} while ((stat & ASF_PCB_MASK) != ASF_PCB_END && j < 20000);
+	for (i = 0; i < MAX_PCB_DATA; i++) {
+		for (j = 0; j < 20000; j++) {
+			stat = get_status(dev->base_addr);
+			if (stat & ACRF)
+				break;
+		}
+		pcb->data.raw[i] = inb_command(dev->base_addr);
+		if ((stat & ASF_PCB_MASK) == ASF_PCB_END || j >= 20000)
+			break;
+	}
 	spin_unlock_irqrestore(&adapter->lock, flags);
+	if (i >= MAX_PCB_DATA) {
+		INVALID_PCB_MSG(i);
+		return false;
+	}
 	if (j >= 20000) {
 		TIMEOUT_MSG(__LINE__);
 		return false;
 	}
-	/* woops, the last "data" byte was really the length! */
-	total_length = pcb->data.raw[--i];
+	/* the last "data" byte was really the length! */
+	total_length = pcb->data.raw[i];
 
 	/* safety check total length vs data length */
 	if (total_length != (pcb->length + 2)) {

Re: [PATCH] 3c505: do not set pcb->data.raw beyond its size

[Jarek Poplawski]

On Wed, Feb 11, 2009 at 11:55:40PM +0100, Roel Kluin wrote:
> Many thanks, Jarek,
> 
> Is this changelog ok?

Yes, the changelog and patch look OK to me.

Thanks,
Jarek P.

> ------------------------------>8----------------8<------------------------------
> 
> Ensure that we do not set pcb->data.raw beyond its size, print an error message
> and return false if we attempt to. A timout message was printed one too early.
> 
> Signed-off-by: Roel Kluin <roel.kluin@gmail.com>
> ---
> diff --git a/drivers/net/3c505.c b/drivers/net/3c505.c
> index 6124605..a8107f9 100644
> --- a/drivers/net/3c505.c
> +++ b/drivers/net/3c505.c
> @@ -493,21 +493,27 @@ static bool receive_pcb(struct net_device *dev, pcb_struct * pcb)
>  	}
>  	/* read the data */
>  	spin_lock_irqsave(&adapter->lock, flags);
> -	i = 0;
> -	do {
> -		j = 0;
> -		while (((stat = get_status(dev->base_addr)) & ACRF) == 0 && j++ < 20000);
> -		pcb->data.raw[i++] = inb_command(dev->base_addr);
> -		if (i > MAX_PCB_DATA)
> -			INVALID_PCB_MSG(i);
> -	} while ((stat & ASF_PCB_MASK) != ASF_PCB_END && j < 20000);
> +	for (i = 0; i < MAX_PCB_DATA; i++) {
> +		for (j = 0; j < 20000; j++) {
> +			stat = get_status(dev->base_addr);
> +			if (stat & ACRF)
> +				break;
> +		}
> +		pcb->data.raw[i] = inb_command(dev->base_addr);
> +		if ((stat & ASF_PCB_MASK) == ASF_PCB_END || j >= 20000)
> +			break;
> +	}
>  	spin_unlock_irqrestore(&adapter->lock, flags);
> +	if (i >= MAX_PCB_DATA) {
> +		INVALID_PCB_MSG(i);
> +		return false;
> +	}
>  	if (j >= 20000) {
>  		TIMEOUT_MSG(__LINE__);
>  		return false;
>  	}
> -	/* woops, the last "data" byte was really the length! */
> -	total_length = pcb->data.raw[--i];
> +	/* the last "data" byte was really the length! */
> +	total_length = pcb->data.raw[i];
>  
>  	/* safety check total length vs data length */
>  	if (total_length != (pcb->length + 2)) {

Re: [PATCH] 3c505: do not set pcb->data.raw beyond its size

[David Miller]

From: Jarek Poplawski <jarkao2@gmail.com>
Date: Thu, 12 Feb 2009 06:37:15 +0000

> On Wed, Feb 11, 2009 at 11:55:40PM +0100, Roel Kluin wrote:
> > Many thanks, Jarek,
> > 
> > Is this changelog ok?
> 
> Yes, the changelog and patch look OK to me.

Applied, thanks everyone.

[PATCH] IRDA: cnt is off by 1

[Roel Kluin]

Jarek Poplawski wrote:

> Roel, please resend once more, but separated: something could go to
> stable, but not necessarily together. 

Ok, Here's for irda:

------------------------------>8----------------8<------------------------------
If no prior break occurs, cnt reaches 101 after the loop, so we are still able
to change speed when cnt has become 100.

Signed-off-by: Roel Kluin <roel.kluin@gmail.com>

diff --git a/drivers/net/irda/mcs7780.c b/drivers/net/irda/mcs7780.c
index 7eafdca..85e88da 100644
--- a/drivers/net/irda/mcs7780.c
+++ b/drivers/net/irda/mcs7780.c
@@ -585,7 +585,7 @@ static int mcs_speed_change(struct mcs_cb *mcs)
 		mcs_get_reg(mcs, MCS_RESV_REG, &rval);
 	} while(cnt++ < 100 && (rval & MCS_IRINTX));
 
-	if(cnt >= 100) {
+	if (cnt > 100) {
 		IRDA_ERROR("unable to change speed\n");
 		ret = -EIO;
 		goto error;

Re: [PATCH] IRDA: cnt is off by 1

[David Miller]

From: Roel Kluin <roel.kluin@gmail.com>
Date: Wed, 11 Feb 2009 22:22:21 +0100

> If no prior break occurs, cnt reaches 101 after the loop, so we are still able
> to change speed when cnt has become 100.
> 
> Signed-off-by: Roel Kluin <roel.kluin@gmail.com>

Applied, thanks Roel.