improvements to SSN and TSN handling

PATCH In review Marcelo Ricardo Leitner 3 participants 6 messages Sep 20, 2016 – Sep 23, 2016

Subsystems (3)

Networking Network drivers SCTP protocol

lore

Share & tools

lore

Marcelo Ricardo Leitner

author maintainer

2016-09-20 21:19 UTC #0 lore raw

cover improvements to SSN and TSN handling

First patch fixes a potential issue made visible by the second one noticed by David Laight and is a preparation for the next one.

The second patch changes how SSN, TSN and ASCONF serials are compared so they can use typecheck() and are more like time_before() macro.

Marcelo Ricardo Leitner (2):
sctp: fix the handling of SACK Gap Ack blocks
sctp: improve how SSN, TSN and ASCONF serial are compared

 include/net/sctp/sm.h | 94 ++++++++++-----------------------------------------
 net/sctp/outqueue.c   | 11 +++---
 2 files changed, 24 insertions(+), 81 deletions(-)

Signature · 1 line

2.7.4

Marcelo Ricardo Leitner

author maintainer

2016-09-20 21:19 UTC #1 lore raw

PATCH 1/2 sctp: fix the handling of SACK Gap Ack blocks

sctp_acked() is using 32bit arithmetics on 16bits vars, via TSN_lte() macros, which is weird and confusing.

Once the offset to ctsn is calculated, all wrapping is already handled and thus to verify the Gap Ack blocks we can just use pure less/big-or-equal than checks.

Also, rename gap variable to tsn_offset, so it's more meaningful, as it doesn't point to any gap at all.

Even so, I don't think this discrepancy resulted in any practical bug.

This patch is a preparation for the next one, which will introduce typecheck() for TSN_lte() macros and would cause a compile error here.

Reported-by David Laight <[email protected]>

Signed-off-by Marcelo Ricardo Leitner <[email protected]>

---
 net/sctp/outqueue.c | 11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)

net/sctp/outqueue.c +6 -5

index 8c3f446d965c030376d0cfafd2c64b9a946d2cc1..3ec6da8bbb5360187007935838717885c97f9e91 100644
@@ -1719,7 +1719,7 @@ static int sctp_acked(struct sctp_sackhdr *sack, __u32 tsn)
 {
 	int i;
 	sctp_sack_variable_t *frags;
-	__u16 gap;
+	__u16 tsn_offset, blocks;
 	__u32 ctsn = ntohl(sack->cum_tsn_ack);
 
 	if (TSN_lte(tsn, ctsn))

@@ -1738,10 +1738,11 @@ static int sctp_acked(struct sctp_sackhdr *sack, __u32 tsn)
 	 */
 
 	frags = sack->variable;
-	gap = tsn - ctsn;
-	for (i = 0; i < ntohs(sack->num_gap_ack_blocks); ++i) {
-		if (TSN_lte(ntohs(frags[i].gab.start), gap) &&
-		    TSN_lte(gap, ntohs(frags[i].gab.end)))
+	blocks = ntohs(sack->num_gap_ack_blocks);
+	tsn_offset = tsn - ctsn;
+	for (i = 0; i < blocks; ++i) {
+		if (tsn_offset >= ntohs(frags[i].gab.start) &&
+		    tsn_offset <= ntohs(frags[i].gab.end))
 			goto pass;
 	}

Signature · 1 line

2.7.4

Marcelo Ricardo Leitner

author maintainer

2016-09-20 21:19 UTC #4 lore raw

PATCH 2/2 sctp: improve how SSN, TSN and ASCONF serial are compared

In reply to Marcelo Ricardo Leitner

Make it similar to time_before() macros:
- easier to understand
- make use of typecheck() to avoid working on unexpected variable types (made the issue on previous patch visible)
- for _[lg]te versions, slighly faster, as the compiler used to generate a sequence of cmp/je/cmp/js instructions and now it's sub/test/jle (for _lte):

Before, for sctp_outq_sack:

	if (primary->cacc.changeover_active) {
    1f01:	80 b9 84 02 00 00 00 	cmpb   $0x0,0x284(%rcx)
    1f08:	74 6e                	je     1f78 <sctp_outq_sack+0xe8>
		u8 clear_cycling = 0;

		if (TSN_lte(primary->cacc.next_tsn_at_change, sack_ctsn)) {
    1f0a:	8b 81 80 02 00 00    	mov    0x280(%rcx),%eax
	return ((s) - (t)) & TSN_SIGN_BIT;
}

static inline int TSN_lte(__u32 s, __u32 t)

{
	return ((s) = (t)) || (((s) - (t)) & TSN_SIGN_BIT);
    1f10:	8b 7d bc             	mov    -0x44(%rbp),%edi
    1f13:	39 c7                	cmp    %eax,%edi
    1f15:	74 25                	je     1f3c <sctp_outq_sack+0xac>
    1f17:	39 f8                	cmp    %edi,%eax
    1f19:	78 21                	js     1f3c <sctp_outq_sack+0xac>
			primary->cacc.changeover_active = 0;

After:

	if (primary->cacc.changeover_active) {
    1ee7:	80 b9 84 02 00 00 00 	cmpb   $0x0,0x284(%rcx)
    1eee:	74 73                	je     1f63 <sctp_outq_sack+0xf3>
		u8 clear_cycling = 0;

		if (TSN_lte(primary->cacc.next_tsn_at_change, sack_ctsn)) {
    1ef0:	8b 81 80 02 00 00    	mov    0x280(%rcx),%eax
    1ef6:	2b 45 b4             	sub    -0x4c(%rbp),%eax
    1ef9:	85 c0                	test   %eax,%eax
    1efb:	7e 26                	jle    1f23 <sctp_outq_sack+0xb3>
			primary->cacc.changeover_active = 0;

*_lt() generated pretty much the same code.
Tested with gcc (GCC) 6.1.1 20160621.

This patch also removes SSN_lte as it is not used and cleanups some comments.

Signed-off-by Marcelo Ricardo Leitner <[email protected]>

---
 include/net/sctp/sm.h | 94 ++++++++++-----------------------------------------
 1 file changed, 18 insertions(+), 76 deletions(-)

include/net/sctp/sm.h +18 ~~-76~~

index bafe2a0ab9085f24e17038516c55c00cfddd02f4..ca6c971dd74aede829d4512ddf71006520c78f47 100644
@@ -307,85 +307,27 @@ static inline __u16 sctp_data_size(struct sctp_chunk *chunk)
 }
 
 /* Compare two TSNs */
+#define TSN_lt(a,b)	\
+	(typecheck(__u32, a) && \
+	 typecheck(__u32, b) && \
+	 ((__s32)((a) - (b)) < 0))
 
-/* RFC 1982 - Serial Number Arithmetic
- *
- * 2. Comparison
- *  Then, s1 is said to be equal to s2 if and only if i1 is equal to i2,
- *  in all other cases, s1 is not equal to s2.
- *
- * s1 is said to be less than s2 if, and only if, s1 is not equal to s2,
- * and
- *
- *      (i1 < i2 and i2 - i1 < 2^(SERIAL_BITS - 1)) or
- *      (i1 > i2 and i1 - i2 > 2^(SERIAL_BITS - 1))
- *
- * s1 is said to be greater than s2 if, and only if, s1 is not equal to
- * s2, and
- *
- *      (i1 < i2 and i2 - i1 > 2^(SERIAL_BITS - 1)) or
- *      (i1 > i2 and i1 - i2 < 2^(SERIAL_BITS - 1))
- */
-
-/*
- * RFC 2960
- *  1.6 Serial Number Arithmetic
- *
- * Comparisons and arithmetic on TSNs in this document SHOULD use Serial
- * Number Arithmetic as defined in [RFC1982] where SERIAL_BITS = 32.
- */
-
-enum {
-	TSN_SIGN_BIT = (1<<31)
-};
-
-static inline int TSN_lt(__u32 s, __u32 t)
-{
-	return ((s) - (t)) & TSN_SIGN_BIT;
-}
-
-static inline int TSN_lte(__u32 s, __u32 t)
-{
-	return ((s) = (t)) || (((s) - (t)) & TSN_SIGN_BIT);
-}
+#define TSN_lte(a,b)	\
+	(typecheck(__u32, a) && \
+	 typecheck(__u32, b) && \
+	 ((__s32)((a) - (b)) <= 0))
 
 /* Compare two SSNs */
-
-/*
- * RFC 2960
- *  1.6 Serial Number Arithmetic
- *
- * Comparisons and arithmetic on Stream Sequence Numbers in this document
- * SHOULD use Serial Number Arithmetic as defined in [RFC1982] where
- * SERIAL_BITS = 16.
- */
-enum {
-	SSN_SIGN_BIT = (1<<15)
-};
-
-static inline int SSN_lt(__u16 s, __u16 t)
-{
-	return ((s) - (t)) & SSN_SIGN_BIT;
-}
-
-static inline int SSN_lte(__u16 s, __u16 t)
-{
-	return ((s) = (t)) || (((s) - (t)) & SSN_SIGN_BIT);
-}
-
-/*
- * ADDIP 3.1.1
- * The valid range of Serial Number is from 0 to 4294967295 (2**32 - 1). Serial
- * Numbers wrap back to 0 after reaching 4294967295.
- */
-enum {
-	ADDIP_SERIAL_SIGN_BIT = (1<<31)
-};
-
-static inline int ADDIP_SERIAL_gte(__u32 s, __u32 t)
-{
-	return ((s) = (t)) || (((t) - (s)) & ADDIP_SERIAL_SIGN_BIT);
-}
+#define SSN_lt(a,b)		\
+	(typecheck(__u16, a) && \
+	 typecheck(__u16, b) && \
+	 ((__s16)((a) - (b)) < 0))
+
+/* ADDIP 3.1.1 */
+#define ADDIP_SERIAL_gte(a,b)	\
+	(typecheck(__u32, a) && \
+	 typecheck(__u32, b) && \
+	 ((__s32)((b) - (a)) <= 0))
 
 /* Check VTAG of the packet matches the sender's own tag. */
 static inline int

Signature · 1 line

2.7.4

David Laight

2016-09-21 10:21 UTC #2 lore raw

RE: [PATCH next 1/2] sctp: fix the handling of SACK Gap Ack blocks

In reply to Marcelo Ricardo Leitner

From: Marcelo Ricardo Leitner

Quoted 4 lines

Sent: 20 September 2016 22:19
sctp_acked() is using 32bit arithmetics on 16bits vars, via TSN_lte() macros, which is weird and confusing.

Quoted 8 lines

Once the offset to ctsn is calculated, all wrapping is already handled and thus to verify the Gap Ack blocks we can just use pure less/big-or-equal than checks.
Also, rename gap variable to tsn_offset, so it's more meaningful, as it doesn't point to any gap at all.
Even so, I don't think this discrepancy resulted in any practical bug.

I think it might if gab.start/end are greater than 32767

...

net/sctp/outqueue.c static int sctp_acked(struct sctp_sackhdr *sack, __u32 tsn) · line 1745

Quoted 9 lines

net/sctp/outqueue.c +5 -4

-	gap = tsn - ctsn;
-	for (i = 0; i < ntohs(sack->num_gap_ack_blocks); ++i) {
-		if (TSN_lte(ntohs(frags[i].gab.start), gap) &&
-		    TSN_lte(gap, ntohs(frags[i].gab.end)))
+	blocks = ntohs(sack->num_gap_ack_blocks);
+	tsn_offset = tsn - ctsn;
+	for (i = 0; i < blocks; ++i) {
+		if (tsn_offset >= ntohs(frags[i].gab.start) &&
+		    tsn_offset <= ntohs(frags[i].gab.end))

...

David

'Marcelo Ricardo Leitner'

author maintainer

2016-09-21 11:01 UTC #3 lore raw

Re: [PATCH next 1/2] sctp: fix the handling of SACK Gap Ack blocks

Quoted 9 lines · David Laight

From: Marcelo Ricardo Leitner
Quoted 8 lines
Sent: 20 September 2016 22:19
sctp_acked() is using 32bit arithmetics on 16bits vars, via TSN_lte() macros, which is weird and confusing.
Once the offset to ctsn is calculated, all wrapping is already handled and thus to verify the Gap Ack blocks we can just use pure less/big-or-equal than checks.

Quoted 6 lines

Quoted 4 lines
Also, rename gap variable to tsn_offset, so it's more meaningful, as it doesn't point to any gap at all.
Even so, I don't think this discrepancy resulted in any practical bug.
I think it might if gab.start/end are greater than 32767

Why? Mind sharing a case that it goes wrong?

Marcelo

Quoted 20 lines

...
net/sctp/outqueue.c static int sctp_acked(struct sctp_sackhdr *sack, __u32 tsn) · line 1745
Quoted 9 lines
net/sctp/outqueue.c +5 -4
-	gap = tsn - ctsn;
-	for (i = 0; i < ntohs(sack->num_gap_ack_blocks); ++i) {
-		if (TSN_lte(ntohs(frags[i].gab.start), gap) &&
-		    TSN_lte(gap, ntohs(frags[i].gab.end)))
+	blocks = ntohs(sack->num_gap_ack_blocks);
+	tsn_offset = tsn - ctsn;
+	for (i = 0; i < blocks; ++i) {
+		if (tsn_offset >= ntohs(frags[i].gab.start) &&
+		    tsn_offset <= ntohs(frags[i].gab.end))
...
David
--
To unsubscribe from this list: send the line "unsubscribe linux-sctp" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

David Miller

maintainer

2016-09-23 10:55 UTC #5 lore raw

In reply to Marcelo Ricardo Leitner

From: Marcelo Ricardo Leitner <[email protected]>
Date: Tue, 20 Sep 2016 18:19:12 -0300

Quoted 5 lines

First patch fixes a potential issue made visible by the second one noticed by David Laight and is a preparation for the next one.
The second patch changes how SSN, TSN and ASCONF serials are compared so they can use typecheck() and are more like time_before() macro.

Series applied, thanks.