Skip to contentSkip to conversation

igc: Avoid dereference of ptr_err in igc_clean_rx_irq()

RFC PATCH In review Simon Horman 3 participants 6 messages Jun 15, 2023

Subsystems (4)
lore
Share & tools
lore

Simon Horman

author reviewer
#0 lore raw

RFC PATCH 1/1 igc: Avoid dereference of ptr_err in igc_clean_rx_irq()

In igc_clean_rx_irq() the result of a call to igc_xdp_run_prog() is assigned to the skb local variable. This may be an ERR_PTR.

A little later the following is executed, which seems to be a possible dereference of an ERR_PTR.

	total_bytes += skb->len;

Avoid this problem by continuing the loop in which all of the above occurs once the handling of the NULL case completes.

This proposed fix is speculative - I do not have deep knowledge of this driver. And I am concerned about the effect of skipping the following logic:

  igc_put_rx_buffer(rx_ring, rx_buffer, rx_buffer_pgcnt);
  cleaned_count++;

Flagged by Smatch as:

.../igc_main.c:2467 igc_xdp_run_prog() warn: passing zero to 'ERR_PTR'

Compile tested only.

Fixes 26575105d6ed ("igc: Add initial XDP support")
---
 drivers/net/ethernet/intel/igc/igc_main.c | 1 +
 1 file changed, 1 insertion(+)
drivers/net/ethernet/intel/igc/igc_main.c +1 -0
index 88145c30c919..b58c8a674bd1 100644
@@ -2586,6 +2586,7 @@ static int igc_clean_rx_irq(struct igc_q_vector *q_vector, const int budget)
 
 			total_packets++;
 			total_bytes += size;
+			continue;
 		} else if (skb)
 			igc_add_rx_frag(rx_ring, rx_buffer, skb, size);
 		else if (ring_uses_build_skb(rx_ring))
#1 lore raw

On Thu, Jun 15, 2023 at 11:45:36AM +0200, Simon Horman wrote:

Hi Simon,

Quoted 11 lines

In igc_clean_rx_irq() the result of a call to igc_xdp_run_prog() is assigned to the skb local variable. This may be an ERR_PTR.

A little later the following is executed, which seems to be a possible dereference of an ERR_PTR.

	total_bytes += skb->len;

Avoid this problem by continuing the loop in which all of the above occurs once the handling of the NULL case completes.

Quoted 6 lines

This proposed fix is speculative - I do not have deep knowledge of this driver. And I am concerned about the effect of skipping the following logic:

  igc_put_rx_buffer(rx_ring, rx_buffer, rx_buffer_pgcnt);
  cleaned_count++;

this will break - you have to recycle the buffer to have it going.

Quoted 4 lines

Flagged by Smatch as:

.../igc_main.c:2467 igc_xdp_run_prog() warn: passing zero to 'ERR_PTR'

how about PTR_ERR_OR_ZERO() ? this would silence smatch and is not an intrusive change. another way is to get rid of ERR_PTR() around skb/xdp run result but i think the former would be just fine.

Quoted 23 lines

Compile tested only.

Fixes 26575105d6ed ("igc: Add initial XDP support")
---
 drivers/net/ethernet/intel/igc/igc_main.c | 1 +
 1 file changed, 1 insertion(+)
drivers/net/ethernet/intel/igc/igc_main.c +1 -0
index 88145c30c919..b58c8a674bd1 100644
@@ -2586,6 +2586,7 @@ static int igc_clean_rx_irq(struct igc_q_vector *q_vector, const int budget)
 
 			total_packets++;
 			total_bytes += size;
+			continue;
 		} else if (skb)
 			igc_add_rx_frag(rx_ring, rx_buffer, skb, size);
 		else if (ring_uses_build_skb(rx_ring))
author reviewer
#2 lore raw
Quoted 1 line · Maciej Fijalkowski

On Thu, Jun 15, 2023 at 11:45:36AM +0200, Simon Horman wrote:

Hi Marciej,

Quoted 13 lines

Hi Simon,

Quoted 11 lines

In igc_clean_rx_irq() the result of a call to igc_xdp_run_prog() is assigned to the skb local variable. This may be an ERR_PTR.

A little later the following is executed, which seems to be a possible dereference of an ERR_PTR.

	total_bytes += skb->len;

Avoid this problem by continuing the loop in which all of the above occurs once the handling of the NULL case completes.

Quoted 8 lines
Quoted 6 lines

This proposed fix is speculative - I do not have deep knowledge of this driver. And I am concerned about the effect of skipping the following logic:

  igc_put_rx_buffer(rx_ring, rx_buffer, rx_buffer_pgcnt);
  cleaned_count++;

this will break - you have to recycle the buffer to have it going.

Thanks. As I said I wasn't sure about the fix: it was a strawman.

Quoted 7 lines
Quoted 3 lines

Flagged by Smatch as:

.../igc_main.c:2467 igc_xdp_run_prog() warn: passing zero to 'ERR_PTR'

how about PTR_ERR_OR_ZERO() ? this would silence smatch and is not an intrusive change. another way is to get rid of ERR_PTR() around skb/xdp run result but i think the former would be just fine.

Sorry, there were two warnings. And I accidently trimmed the one that is more relevant instead of the one that is less relevant. I do agree the one above does not appear to be a bug.

But I am concerned abut this one:

.../igc_main.c:2618 igc_clean_rx_irq() error: 'skb' dereferencing possible ERR_PTR()

If skb is an error pointer, e.g. ERR_PTR(-IGC_XDP_PASS), and it is dereferenced, that would be a problem, right?

Perhaps I'm missing something obvious and this can't occur. But it does seem possible to me.

Quoted 25 lines
Quoted 24 lines

Compile tested only.

Fixes 26575105d6ed ("igc: Add initial XDP support")
---
 drivers/net/ethernet/intel/igc/igc_main.c | 1 +
 1 file changed, 1 insertion(+)
drivers/net/ethernet/intel/igc/igc_main.c +1 -0
index 88145c30c919..b58c8a674bd1 100644
@@ -2586,6 +2586,7 @@ static int igc_clean_rx_irq(struct igc_q_vector *q_vector, const int budget)
 
 			total_packets++;
 			total_bytes += size;
+			continue;
 		} else if (skb)
 			igc_add_rx_frag(rx_ring, rx_buffer, skb, size);
 		else if (ring_uses_build_skb(rx_ring))
#3 lore raw
Quoted 41 lines · Simon Horman
Quoted 1 line · Maciej Fijalkowski

On Thu, Jun 15, 2023 at 11:45:36AM +0200, Simon Horman wrote:

Hi Marciej,

Quoted 13 lines

Hi Simon,

Quoted 11 lines

In igc_clean_rx_irq() the result of a call to igc_xdp_run_prog() is assigned to the skb local variable. This may be an ERR_PTR.

A little later the following is executed, which seems to be a possible dereference of an ERR_PTR.

	total_bytes += skb->len;

Avoid this problem by continuing the loop in which all of the above occurs once the handling of the NULL case completes.

Quoted 8 lines
Quoted 6 lines

This proposed fix is speculative - I do not have deep knowledge of this driver. And I am concerned about the effect of skipping the following logic:

  igc_put_rx_buffer(rx_ring, rx_buffer, rx_buffer_pgcnt);
  cleaned_count++;

this will break - you have to recycle the buffer to have it going.

Thanks. As I said I wasn't sure about the fix: it was a strawman.

Quoted 7 lines
Quoted 3 lines

Flagged by Smatch as:

.../igc_main.c:2467 igc_xdp_run_prog() warn: passing zero to 'ERR_PTR'

how about PTR_ERR_OR_ZERO() ? this would silence smatch and is not an intrusive change. another way is to get rid of ERR_PTR() around skb/xdp run result but i think the former would be just fine.

Sorry, there were two warnings. And I accidently trimmed the one that is more relevant instead of the one that is less relevant. I do agree the one above does not appear to be a bug.

Quoted 6 lines

But I am concerned abut this one:

.../igc_main.c:2618 igc_clean_rx_irq() error: 'skb' dereferencing possible ERR_PTR()

If skb is an error pointer, e.g. ERR_PTR(-IGC_XDP_PASS), and it is dereferenced, that would be a problem, right?

IGC_XDP_PASS is 0. -0 is still 0 right?

this means skb is NULL and igc_{build,construct}_skb() will init it. For ERR_PTR, igc_cleanup_headers() does IS_ERR() against it and continues. So you will get to line 2618 only for valid skb, it just happens that logic is written in a way that skb is supposed to carry XDP return code. We removed this in ice for example but i40e works like that for many years without issues, AFAICT...

Quoted 29 lines

Perhaps I'm missing something obvious and this can't occur. But it does seem possible to me.

Quoted 25 lines
Quoted 24 lines

Compile tested only.

Fixes 26575105d6ed ("igc: Add initial XDP support")
---
 drivers/net/ethernet/intel/igc/igc_main.c | 1 +
 1 file changed, 1 insertion(+)
drivers/net/ethernet/intel/igc/igc_main.c +1 -0
index 88145c30c919..b58c8a674bd1 100644
@@ -2586,6 +2586,7 @@ static int igc_clean_rx_irq(struct igc_q_vector *q_vector, const int budget)
 
 			total_packets++;
 			total_bytes += size;
+			continue;
 		} else if (skb)
 			igc_add_rx_frag(rx_ring, rx_buffer, skb, size);
 		else if (ring_uses_build_skb(rx_ring))
author reviewer
#4 lore raw
Quoted 42 lines · Maciej Fijalkowski
Quoted 41 lines · Simon Horman
Quoted 1 line · Maciej Fijalkowski

On Thu, Jun 15, 2023 at 11:45:36AM +0200, Simon Horman wrote:

Hi Marciej,

Quoted 13 lines

Hi Simon,

Quoted 11 lines

In igc_clean_rx_irq() the result of a call to igc_xdp_run_prog() is assigned to the skb local variable. This may be an ERR_PTR.

A little later the following is executed, which seems to be a possible dereference of an ERR_PTR.

	total_bytes += skb->len;

Avoid this problem by continuing the loop in which all of the above occurs once the handling of the NULL case completes.

Quoted 8 lines
Quoted 6 lines

This proposed fix is speculative - I do not have deep knowledge of this driver. And I am concerned about the effect of skipping the following logic:

  igc_put_rx_buffer(rx_ring, rx_buffer, rx_buffer_pgcnt);
  cleaned_count++;

this will break - you have to recycle the buffer to have it going.

Thanks. As I said I wasn't sure about the fix: it was a strawman.

Quoted 7 lines
Quoted 3 lines

Flagged by Smatch as:

.../igc_main.c:2467 igc_xdp_run_prog() warn: passing zero to 'ERR_PTR'

how about PTR_ERR_OR_ZERO() ? this would silence smatch and is not an intrusive change. another way is to get rid of ERR_PTR() around skb/xdp run result but i think the former would be just fine.

Sorry, there were two warnings. And I accidently trimmed the one that is more relevant instead of the one that is less relevant. I do agree the one above does not appear to be a bug.

Quoted 8 lines
Quoted 6 lines

But I am concerned abut this one:

.../igc_main.c:2618 igc_clean_rx_irq() error: 'skb' dereferencing possible ERR_PTR()

If skb is an error pointer, e.g. ERR_PTR(-IGC_XDP_PASS), and it is dereferenced, that would be a problem, right?

IGC_XDP_PASS is 0. -0 is still 0 right?

Yes, I missed that point.
Though I could have chosen a different value which is not zero.

Quoted 6 lines

this means skb is NULL and igc_{build,construct}_skb() will init it. For ERR_PTR, igc_cleanup_headers() does IS_ERR() against it and continues. So you will get to line 2618 only for valid skb, it just happens that logic is written in a way that skb is supposed to carry XDP return code. We removed this in ice for example but i40e works like that for many years without issues, AFAICT...

Thanks. I now see that the key point I was missing is the IS_ERR() check in igc_cleanup_headers().

I agree this is not a bug.

#5 lore raw
In reply to Simon Horman

The original code is okay. Passing zero to ERR_PTR() is intentional.

Quoted 7 lines · Simon Horman

In igc_clean_rx_irq() the result of a call to igc_xdp_run_prog() is assigned to the skb local variable. This may be an ERR_PTR.

A little later the following is executed, which seems to be a possible dereference of an ERR_PTR.

	total_bytes += skb->len;

There is an IS_ERR() check in igc_cleanup_headers() which prevents this. Sort of tricky to see. Do you have the cross function database set up? If so then Smatch shouldn't warn about this dereference.

Quoted 8 lines

Avoid this problem by continuing the loop in which all of the above occurs once the handling of the NULL case completes.

This proposed fix is speculative - I do not have deep knowledge of this driver. And I am concerned about the effect of skipping the following logic:

Quoted 6 lines
  igc_put_rx_buffer(rx_ring, rx_buffer, rx_buffer_pgcnt);
  cleaned_count++;

Flagged by Smatch as:

.../igc_main.c:2467 igc_xdp_run_prog() warn: passing zero to 'ERR_PTR'

Linus once complained to me that this check is bogus and passing zero to ERR_PTR() is fine and an intended use case. But actually this test does really find a lot of bugs. I think for new warnings it is less than 10% false positives. But we fix the bugs so warnings which are over three month old are probably 97% false positives.

regards,
dan carpenter