Skip to contentSkip to conversation

doc: updated incorrect value for IP frag max fragments

PATCH Superseded Euan Bourke 5 participants 11 messages Dec 21, 2023 – Oct 9, 2024 1 message not in the archive

lore
Share & tools
lore

A message that is not in the archive (sent off-list or not yet received): https://inbox.dpdk.org/dev/[email protected]/

#1 lore raw

PATCH v2 1/2 doc: updated incorrect value for IP frag max fragments

Docs for IP Fragment said RTE_LIBRTE_IP_FRAG_MAX_FRAGS was 4 by default, however this was changed to 8.

Documentation has been updated to account for this, including a snippet of the code where RTE_LIBRTE_IP_FRAG_MAX_FRAGS is defined to ensure the documentation stays up to date.

---
 .mailmap                                             | 1 +
 config/rte_config.h                                  | 4 +++-
 doc/guides/prog_guide/ip_fragment_reassembly_lib.rst | 7 ++++++-
 3 files changed, 10 insertions(+), 2 deletions(-)
.mailmap +1 -0
index ab0742a382..528bc68a30 100644
@@ -379,6 +379,7 @@ Eric Zhang <[email protected]>
 Erik Gabriel Carrillo <[email protected]>
 Erik Ziegenbalg <[email protected]>
 Erlu Chen <[email protected]>
+Euan Bourke <[email protected]>
 Eugenio Pérez <[email protected]>
 Eugeny Parshutin <[email protected]>
 Evan Swanson <[email protected]>
config/rte_config.h +3 -1
index da265d7dd2..a3f92354f8 100644
@@ -84,8 +84,10 @@
 /* rawdev defines */
 #define RTE_RAWDEV_MAX_DEVS 64
 
-/* ip_fragmentation defines */
+/* ip_fragmentation defines 8< */
 #define RTE_LIBRTE_IP_FRAG_MAX_FRAG 8
+/* >8 End of ip_fragmentation defines */
+
 // RTE_LIBRTE_IP_FRAG_TBL_STAT is not set
 
 /* rte_power defines */
doc/guides/prog_guide/ip_fragment_reassembly_lib.rst +6 -1
index 314d4adbb8..5010472246 100644
@@ -43,7 +43,12 @@ Note that all update/lookup operations on Fragment Table are not thread safe.
 So if different execution contexts (threads/processes) will access the same table simultaneously,
 then some external syncing mechanism have to be provided.
 
-Each table entry can hold information about packets consisting of up to RTE_LIBRTE_IP_FRAG_MAX (by default: 4) fragments.
+Each table entry can hold information about packets of up to ``RTE_LIBRTE_IP_FRAG_MAX_FRAGS`` fragments,
+where ``RTE_LIBRTE_IP_FRAG_MAX_FRAGS`` defaults to:
+
+.. literalinclude:: ../../../config/rte_config.h
+    :start-after: ip_fragmentation defines 8<
+    :end-before: >8 End of ip_fragmentation defines
 
 Code example, that demonstrates creation of a new Fragment table:
 
Signature · 1 line
2.34.1
#2 lore raw

PATCH v2 2/2 ip_frag: updated name for IP frag define

Removed LIBRTE from name as its an old prefix.

---
 app/test/test_reassembly_perf.c                      | 2 +-
 config/rte_config.h                                  | 2 +-
 doc/guides/prog_guide/ip_fragment_reassembly_lib.rst | 4 ++--
 doc/guides/sample_app_ug/ip_reassembly.rst           | 4 ++--
 examples/ip_fragmentation/main.c                     | 2 +-
 examples/ip_reassembly/main.c                        | 2 +-
 examples/ipsec-secgw/ipsec_worker.h                  | 2 +-
 lib/ip_frag/ip_reassembly.h                          | 2 +-
 lib/ip_frag/rte_ip_frag.h                            | 2 +-
 9 files changed, 11 insertions(+), 11 deletions(-)
app/test/test_reassembly_perf.c +1 -1
index 3912179022..805ae2fe9d 100644
@@ -20,7 +20,7 @@
 #define MAX_FLOWS	    (1024 * 32)
 #define MAX_BKTS	    MAX_FLOWS
 #define MAX_ENTRIES_PER_BKT 16
-#define MAX_FRAGMENTS	    RTE_LIBRTE_IP_FRAG_MAX_FRAG
+#define MAX_FRAGMENTS	    RTE_IP_FRAG_MAX_FRAG
 #define MIN_FRAGMENTS	    2
 #define MAX_PKTS	    (MAX_FLOWS * MAX_FRAGMENTS)
 
config/rte_config.h +1 -1
index a3f92354f8..1f4567b0d2 100644
@@ -85,7 +85,7 @@
 #define RTE_RAWDEV_MAX_DEVS 64
 
 /* ip_fragmentation defines 8< */
-#define RTE_LIBRTE_IP_FRAG_MAX_FRAG 8
+#define RTE_IP_FRAG_MAX_FRAG 8
 /* >8 End of ip_fragmentation defines */
 
 // RTE_LIBRTE_IP_FRAG_TBL_STAT is not set
doc/guides/prog_guide/ip_fragment_reassembly_lib.rst +2 -2
index 5010472246..176186e3ea 100644
@@ -43,8 +43,8 @@ Note that all update/lookup operations on Fragment Table are not thread safe.
 So if different execution contexts (threads/processes) will access the same table simultaneously,
 then some external syncing mechanism have to be provided.
 
-Each table entry can hold information about packets of up to ``RTE_LIBRTE_IP_FRAG_MAX_FRAGS`` fragments,
-where ``RTE_LIBRTE_IP_FRAG_MAX_FRAGS`` defaults to:
+Each table entry can hold information about packets of up to ``RTE_IP_FRAG_MAX_FRAGS`` fragments,
+where ``RTE_IP_FRAG_MAX_FRAGS`` defaults to:
 
 .. literalinclude:: ../../../config/rte_config.h
     :start-after: ip_fragmentation defines 8<
doc/guides/sample_app_ug/ip_reassembly.rst +2 -2
index 5280bf4ea0..9cf4fb3f7d 100644
@@ -135,7 +135,7 @@ Fragment table maintains information about already received fragments of the pac
 Each IP packet is uniquely identified by triple <Source IP address>, <Destination IP address>, <ID>.
 To avoid lock contention, each RX queue has its own Fragment Table,
 e.g. the application can't handle the situation when different fragments of the same packet arrive through different RX queues.
-Each table entry can hold information about packet consisting of up to RTE_LIBRTE_IP_FRAG_MAX_FRAGS fragments.
+Each table entry can hold information about packet consisting of up to RTE_IP_FRAG_MAX_FRAGS fragments.
 
 .. literalinclude:: ../../../examples/ip_reassembly/main.c
     :language: c
@@ -147,7 +147,7 @@ Mempools Initialization
 ~~~~~~~~~~~~~~~~~~~~~~~
 
 The reassembly application demands a lot of mbuf's to be allocated.
-At any given time up to (2 \* max_flow_num \* RTE_LIBRTE_IP_FRAG_MAX_FRAGS \* <maximum number of mbufs per packet>)
+At any given time up to (2 \* max_flow_num \* RTE_IP_FRAG_MAX_FRAGS \* <maximum number of mbufs per packet>)
 can be stored inside Fragment Table waiting for remaining fragments.
 To keep mempool size under reasonable limits and to avoid situation when one RX queue can starve other queues,
 each RX queue uses its own mempool.
examples/ip_fragmentation/main.c +1 -1
index 744a1aa9b4..1e4471891b 100644
@@ -71,7 +71,7 @@
 /*
  * Max number of fragments per packet expected - defined by config file.
  */
-#define	MAX_PACKET_FRAG RTE_LIBRTE_IP_FRAG_MAX_FRAG
+#define	MAX_PACKET_FRAG RTE_IP_FRAG_MAX_FRAG
 
 #define NB_MBUF   8192
 
examples/ip_reassembly/main.c +1 -1
index bd0b1d31de..16607d99f3 100644
@@ -69,7 +69,7 @@
 #define	MIN_FLOW_TTL	1
 #define	DEF_FLOW_TTL	MS_PER_S
 
-#define MAX_FRAG_NUM RTE_LIBRTE_IP_FRAG_MAX_FRAG
+#define MAX_FRAG_NUM RTE_IP_FRAG_MAX_FRAG
 
 /* Should be power of two. */
 #define	IP_FRAG_TBL_BUCKET_ENTRIES	16
examples/ipsec-secgw/ipsec_worker.h +1 -1
index ac980b8bcf..918e6b5200 100644
@@ -291,7 +291,7 @@ send_fragment_packet(struct lcore_conf *qconf, struct rte_mbuf *m,
 	len = tbl->len;
 
 	/* free space for new fragments */
-	if (len + RTE_LIBRTE_IP_FRAG_MAX_FRAG >=  RTE_DIM(tbl->m_table)) {
+	if (len + RTE_IP_FRAG_MAX_FRAG >=  RTE_DIM(tbl->m_table)) {
 		send_burst(qconf, len, port);
 		len = 0;
 	}
lib/ip_frag/ip_reassembly.h +1 -1
index a9f97aefca..77c22ca9d1 100644
@@ -16,7 +16,7 @@ enum {
 	IP_LAST_FRAG_IDX,    /* index of last fragment */
 	IP_FIRST_FRAG_IDX,   /* index of first fragment */
 	IP_MIN_FRAG_NUM,     /* minimum number of fragments */
-	IP_MAX_FRAG_NUM = RTE_LIBRTE_IP_FRAG_MAX_FRAG,
+	IP_MAX_FRAG_NUM = RTE_IP_FRAG_MAX_FRAG,
 	/* maximum number of fragments per packet */
 };
 
lib/ip_frag/rte_ip_frag.h +1 -1
index 2ad318096b..6465ec316b 100644
@@ -32,7 +32,7 @@ struct rte_mbuf;
 
 /** death row size in mbufs */
 #define RTE_IP_FRAG_DEATH_ROW_MBUF_LEN \
-	(RTE_IP_FRAG_DEATH_ROW_LEN * (RTE_LIBRTE_IP_FRAG_MAX_FRAG + 1))
+	(RTE_IP_FRAG_DEATH_ROW_LEN * (RTE_IP_FRAG_MAX_FRAG + 1))
 
 /** mbuf death row (packets to be freed) */
 struct rte_ip_frag_death_row {
Signature · 1 line
2.34.1
#3 lore raw
In reply to Euan Bourke

Hello,

Quoted 9 lines · Euan Bourke

Docs for IP Fragment said RTE_LIBRTE_IP_FRAG_MAX_FRAGS was 4 by default, however this was changed to 8.

Documentation has been updated to account for this, including a snippet of the code where RTE_LIBRTE_IP_FRAG_MAX_FRAGS is defined to ensure the documentation stays up to date.

Fixes tag?
Signature · 1 line
David Marchand
#4 lore raw
In reply to Euan Bourke

21/12/2023 12:32, Euan Bourke:

Quoted 4 lines
config/rte_config.h +2 -1
-/* ip_fragmentation defines */
+/* ip_fragmentation defines 8< */
 #define RTE_LIBRTE_IP_FRAG_MAX_FRAG 8
+/* >8 End of ip_fragmentation defines */

[...]

doc/guides/prog_guide/ip_fragment_reassembly_lib.rst Note that all update/lookup operations on Fragment Table are not thread safe. · line 51

Quoted 3 lines

+.. literalinclude:: ../../../config/rte_config.h

doc/guides/prog_guide/ip_fragment_reassembly_lib.rst +2 -0
+    :start-after: ip_fragmentation defines 8<
+    :end-before: >8 End of ip_fragmentation defines

In v1 I asked to use "start-at". You didn't reply. It is more convenient for a one-line include. Don't you agree?

#5 lore raw
Quoted 6 lines · Thomas Monjalon

21/12/2023 12:32, Euan Bourke:

Quoted 4 lines
config/rte_config.h +2 -1
-/* ip_fragmentation defines */
+/* ip_fragmentation defines 8< */
  #define RTE_LIBRTE_IP_FRAG_MAX_FRAG 8
+/* >8 End of ip_fragmentation defines */

[...]

Quoted 9 lines

doc/guides/prog_guide/ip_fragment_reassembly_lib.rst Note that all update/lookup operations on Fragment Table are not thread safe. · line 51

Quoted 3 lines

+.. literalinclude:: ../../../config/rte_config.h

doc/guides/prog_guide/ip_fragment_reassembly_lib.rst +2 -0
+    :start-after: ip_fragmentation defines 8<
+    :end-before: >8 End of ip_fragmentation defines

In v1 I asked to use "start-at". You didn't reply. It is more convenient for a one-line include. Don't you agree?

Apologies, as I was updating the name of the declare in the next patch, I thought this way would be better, will use "start-at" in next version.

#6 lore raw

PATCH v3 1/2 doc: updated incorrect value for IP frag max fragments

In reply to Euan Bourke

Docs for IP Fragment said RTE_LIBRTE_IP_FRAG_MAX_FRAG was 4 by default, however this was changed to 8.

Documentation has been updated to account for this, including a snippet of code where RTE_LIBRTE_IP_FRAG_MAX_FRAG is defined to ensure documentation stays up to date.

Fixes f8e0f8ce9030 ("ip_frag: increase default maximum of fragments")
---
 .mailmap                                             | 1 +
 doc/guides/prog_guide/ip_fragment_reassembly_lib.rst | 8 +++++++-
 2 files changed, 8 insertions(+), 1 deletion(-)
.mailmap +1 -0
index ab0742a382..528bc68a30 100644
@@ -379,6 +379,7 @@ Eric Zhang <[email protected]>
 Erik Gabriel Carrillo <[email protected]>
 Erik Ziegenbalg <[email protected]>
 Erlu Chen <[email protected]>
+Euan Bourke <[email protected]>
 Eugenio Pérez <[email protected]>
 Eugeny Parshutin <[email protected]>
 Evan Swanson <[email protected]>
doc/guides/prog_guide/ip_fragment_reassembly_lib.rst +7 -1
index 314d4adbb8..f3ed90d700 100644
@@ -43,7 +43,13 @@ Note that all update/lookup operations on Fragment Table are not thread safe.
 So if different execution contexts (threads/processes) will access the same table simultaneously,
 then some external syncing mechanism have to be provided.
 
-Each table entry can hold information about packets consisting of up to RTE_LIBRTE_IP_FRAG_MAX (by default: 4) fragments.
+Each table entry can hold information about packets of up to ``RTE_LIBRTE_IP_FRAG_MAX_FRAG`` fragments,
+where ``RTE_LIBRTE_IP_FRAG_MAX_FRAG`` defaults to:
+
+.. literalinclude:: ../../../config/rte_config.h
+    :language: c
+    :start-at: #define RTE_LIBRTE_IP_FRAG_MAX_FRAG
+    :lines: 1
 
 Code example, that demonstrates creation of a new Fragment table:
 
Signature · 1 line
2.34.1
#7 lore raw

PATCH v3 2/2 ip_frag: updated name for IP frag max frag define

Removed LIBRTE from name as its an old prefix.

---
 app/test/test_reassembly_perf.c                      | 2 +-
 config/rte_config.h                                  | 2 +-
 doc/guides/prog_guide/ip_fragment_reassembly_lib.rst | 8 ++++----
 doc/guides/sample_app_ug/ip_reassembly.rst           | 4 ++--
 examples/ip_fragmentation/main.c                     | 2 +-
 examples/ip_reassembly/main.c                        | 2 +-
 examples/ipsec-secgw/ipsec_worker.h                  | 2 +-
 lib/ip_frag/ip_reassembly.h                          | 2 +-
 lib/ip_frag/rte_ip_frag.h                            | 2 +-
 9 files changed, 13 insertions(+), 13 deletions(-)
app/test/test_reassembly_perf.c +1 -1
index 3912179022..805ae2fe9d 100644
@@ -20,7 +20,7 @@
 #define MAX_FLOWS	    (1024 * 32)
 #define MAX_BKTS	    MAX_FLOWS
 #define MAX_ENTRIES_PER_BKT 16
-#define MAX_FRAGMENTS	    RTE_LIBRTE_IP_FRAG_MAX_FRAG
+#define MAX_FRAGMENTS	    RTE_IP_FRAG_MAX_FRAG
 #define MIN_FRAGMENTS	    2
 #define MAX_PKTS	    (MAX_FLOWS * MAX_FRAGMENTS)
 
config/rte_config.h +1 -1
index da265d7dd2..e2fa2a58fa 100644
@@ -85,7 +85,7 @@
 #define RTE_RAWDEV_MAX_DEVS 64
 
 /* ip_fragmentation defines */
-#define RTE_LIBRTE_IP_FRAG_MAX_FRAG 8
+#define RTE_IP_FRAG_MAX_FRAG 8
 // RTE_LIBRTE_IP_FRAG_TBL_STAT is not set
 
 /* rte_power defines */
doc/guides/prog_guide/ip_fragment_reassembly_lib.rst +4 -4
index f3ed90d700..2d422309bb 100644
@@ -43,12 +43,12 @@ Note that all update/lookup operations on Fragment Table are not thread safe.
 So if different execution contexts (threads/processes) will access the same table simultaneously,
 then some external syncing mechanism have to be provided.
 
-Each table entry can hold information about packets of up to ``RTE_LIBRTE_IP_FRAG_MAX_FRAG`` fragments,
-where ``RTE_LIBRTE_IP_FRAG_MAX_FRAG`` defaults to:
+Each table entry can hold information about packets of up to ``RTE_IP_FRAG_MAX_FRAG`` fragments,
+where ``RTE_IP_FRAG_MAX_FRAG`` defaults to:
 
 .. literalinclude:: ../../../config/rte_config.h
     :language: c
-    :start-at: #define RTE_LIBRTE_IP_FRAG_MAX_FRAG
+    :start-at: #define RTE_IP_FRAG_MAX_FRAG
     :lines: 1
 
 Code example, that demonstrates creation of a new Fragment table:
@@ -69,7 +69,7 @@ Also, entries that resides in the table longer then <max_cycles> are considered
 and could be removed/replaced by the new ones.
 
 Note that reassembly demands a lot of mbuf's to be allocated.
-At any given time up to (2 \* bucket_entries \* RTE_LIBRTE_IP_FRAG_MAX \* <maximum number of mbufs per packet>)
+At any given time up to (2 \* bucket_entries \* RTE_IP_FRAG_MAX \* <maximum number of mbufs per packet>)
 can be stored inside Fragment Table waiting for remaining fragments.
 
 Packet Reassembly
doc/guides/sample_app_ug/ip_reassembly.rst +2 -2
index 5280bf4ea0..d9aaa4b5c4 100644
@@ -135,7 +135,7 @@ Fragment table maintains information about already received fragments of the pac
 Each IP packet is uniquely identified by triple <Source IP address>, <Destination IP address>, <ID>.
 To avoid lock contention, each RX queue has its own Fragment Table,
 e.g. the application can't handle the situation when different fragments of the same packet arrive through different RX queues.
-Each table entry can hold information about packet consisting of up to RTE_LIBRTE_IP_FRAG_MAX_FRAGS fragments.
+Each table entry can hold information about packet consisting of up to RTE_IP_FRAG_MAX_FRAGS fragments.
 
 .. literalinclude:: ../../../examples/ip_reassembly/main.c
     :language: c
@@ -147,7 +147,7 @@ Mempools Initialization
 ~~~~~~~~~~~~~~~~~~~~~~~
 
 The reassembly application demands a lot of mbuf's to be allocated.
-At any given time up to (2 \* max_flow_num \* RTE_LIBRTE_IP_FRAG_MAX_FRAGS \* <maximum number of mbufs per packet>)
+At any given time up to (2 \* max_flow_num \* RTE_IP_FRAG_MAX_FRAGS_FRAG \* <maximum number of mbufs per packet>)
 can be stored inside Fragment Table waiting for remaining fragments.
 To keep mempool size under reasonable limits and to avoid situation when one RX queue can starve other queues,
 each RX queue uses its own mempool.
examples/ip_fragmentation/main.c +1 -1
index 744a1aa9b4..1e4471891b 100644
@@ -71,7 +71,7 @@
 /*
  * Max number of fragments per packet expected - defined by config file.
  */
-#define	MAX_PACKET_FRAG RTE_LIBRTE_IP_FRAG_MAX_FRAG
+#define	MAX_PACKET_FRAG RTE_IP_FRAG_MAX_FRAG
 
 #define NB_MBUF   8192
 
examples/ip_reassembly/main.c +1 -1
index bd0b1d31de..16607d99f3 100644
@@ -69,7 +69,7 @@
 #define	MIN_FLOW_TTL	1
 #define	DEF_FLOW_TTL	MS_PER_S
 
-#define MAX_FRAG_NUM RTE_LIBRTE_IP_FRAG_MAX_FRAG
+#define MAX_FRAG_NUM RTE_IP_FRAG_MAX_FRAG
 
 /* Should be power of two. */
 #define	IP_FRAG_TBL_BUCKET_ENTRIES	16
examples/ipsec-secgw/ipsec_worker.h +1 -1
index ac980b8bcf..918e6b5200 100644
@@ -291,7 +291,7 @@ send_fragment_packet(struct lcore_conf *qconf, struct rte_mbuf *m,
 	len = tbl->len;
 
 	/* free space for new fragments */
-	if (len + RTE_LIBRTE_IP_FRAG_MAX_FRAG >=  RTE_DIM(tbl->m_table)) {
+	if (len + RTE_IP_FRAG_MAX_FRAG >=  RTE_DIM(tbl->m_table)) {
 		send_burst(qconf, len, port);
 		len = 0;
 	}
lib/ip_frag/ip_reassembly.h +1 -1
index a9f97aefca..77c22ca9d1 100644
@@ -16,7 +16,7 @@ enum {
 	IP_LAST_FRAG_IDX,    /* index of last fragment */
 	IP_FIRST_FRAG_IDX,   /* index of first fragment */
 	IP_MIN_FRAG_NUM,     /* minimum number of fragments */
-	IP_MAX_FRAG_NUM = RTE_LIBRTE_IP_FRAG_MAX_FRAG,
+	IP_MAX_FRAG_NUM = RTE_IP_FRAG_MAX_FRAG,
 	/* maximum number of fragments per packet */
 };
 
lib/ip_frag/rte_ip_frag.h +1 -1
index 2ad318096b..6465ec316b 100644
@@ -32,7 +32,7 @@ struct rte_mbuf;
 
 /** death row size in mbufs */
 #define RTE_IP_FRAG_DEATH_ROW_MBUF_LEN \
-	(RTE_IP_FRAG_DEATH_ROW_LEN * (RTE_LIBRTE_IP_FRAG_MAX_FRAG + 1))
+	(RTE_IP_FRAG_DEATH_ROW_LEN * (RTE_IP_FRAG_MAX_FRAG + 1))
 
 /** mbuf death row (packets to be freed) */
 struct rte_ip_frag_death_row {
Signature · 1 line
2.34.1
#8 lore raw

Re: [PATCH v3 2/2] ip_frag: updated name for IP frag max frag define

Quoted 62 lines · Euan Bourke

Removed LIBRTE from name as its an old prefix.

---
 app/test/test_reassembly_perf.c                      | 2 +-
 config/rte_config.h                                  | 2 +-
 doc/guides/prog_guide/ip_fragment_reassembly_lib.rst | 8 ++++----
 doc/guides/sample_app_ug/ip_reassembly.rst           | 4 ++--
 examples/ip_fragmentation/main.c                     | 2 +-
 examples/ip_reassembly/main.c                        | 2 +-
 examples/ipsec-secgw/ipsec_worker.h                  | 2 +-
 lib/ip_frag/ip_reassembly.h                          | 2 +-
 lib/ip_frag/rte_ip_frag.h                            | 2 +-
 9 files changed, 13 insertions(+), 13 deletions(-)
app/test/test_reassembly_perf.c +1 -1
index 3912179022..805ae2fe9d 100644
@@ -20,7 +20,7 @@
 #define MAX_FLOWS	    (1024 * 32)
 #define MAX_BKTS	    MAX_FLOWS
 #define MAX_ENTRIES_PER_BKT 16
-#define MAX_FRAGMENTS	    RTE_LIBRTE_IP_FRAG_MAX_FRAG
+#define MAX_FRAGMENTS	    RTE_IP_FRAG_MAX_FRAG
 #define MIN_FRAGMENTS	    2
 #define MAX_PKTS	    (MAX_FLOWS * MAX_FRAGMENTS)
 
config/rte_config.h +1 -1
index da265d7dd2..e2fa2a58fa 100644
@@ -85,7 +85,7 @@
 #define RTE_RAWDEV_MAX_DEVS 64
 
 /* ip_fragmentation defines */
-#define RTE_LIBRTE_IP_FRAG_MAX_FRAG 8
+#define RTE_IP_FRAG_MAX_FRAG 8
 // RTE_LIBRTE_IP_FRAG_TBL_STAT is not set
 
 /* rte_power defines */
doc/guides/prog_guide/ip_fragment_reassembly_lib.rst +3 -3
index f3ed90d700..2d422309bb 100644
@@ -43,12 +43,12 @@ Note that all update/lookup operations on Fragment Table are not thread safe.
 So if different execution contexts (threads/processes) will access the same table simultaneously,
 then some external syncing mechanism have to be provided.
 
-Each table entry can hold information about packets of up to ``RTE_LIBRTE_IP_FRAG_MAX_FRAG`` fragments,
-where ``RTE_LIBRTE_IP_FRAG_MAX_FRAG`` defaults to:
+Each table entry can hold information about packets of up to ``RTE_IP_FRAG_MAX_FRAG`` fragments,
+where ``RTE_IP_FRAG_MAX_FRAG`` defaults to:
 
 .. literalinclude:: ../../../config/rte_config.h
     :language: c
-    :start-at: #define RTE_LIBRTE_IP_FRAG_MAX_FRAG
+    :start-at: #define RTE_IP_FRAG_MAX_FRAG
     :lines: 1
 
 Code example, that demonstrates creation of a new Fragment table:

doc/guides/prog_guide/ip_fragment_reassembly_lib.rst Also, entries that resides in the table longer then <max_cycles> are considered · line 72

Quoted 5 lines
doc/guides/prog_guide/ip_fragment_reassembly_lib.rst +1 -1
@@ -69,7 +69,7 @@ Also, entries that resides in the table longer then <max_cycles> are considered
 and could be removed/replaced by the new ones.
 
 Note that reassembly demands a lot of mbuf's to be allocated.
-At any given time up to (2 \* bucket_entries \* RTE_LIBRTE_IP_FRAG_MAX \* <maximum number of mbufs per packet>)
+At any given time up to (2 \* bucket_entries \* RTE_IP_FRAG_MAX \* <maximum number of mbufs per packet>)

Should be RTE_IP_FRAG_MAX_FRAG. You should check all other replacements to ensure the correct define name is being used. I see a couple of other occurances in this patch where the name is wrong.

doc/guides/prog_guide/ip_fragment_reassembly_lib.rst Also, entries that resides in the table longer then <max_cycles> are considered · line 74

Quoted 2 lines
doc/guides/prog_guide/ip_fragment_reassembly_lib.rst +0 -0
 can be stored inside Fragment Table waiting for remaining fragments.
 

<snip for brevity>

#9 lore raw

PATCH v4 1/2 doc: updated incorrect value for IP frag max fragments

In reply to Euan Bourke

Docs for IP Fragment said RTE_LIBRTE_IP_FRAG_MAX_FRAG was 4 by default, however this was changed to 8.

Documentation has been updated to account for this, including a snippet of code where RTE_LIBRTE_IP_FRAG_MAX_FRAG is defined to ensure documentation stays up to date.

Fixes f8e0f8ce9030 ("ip_frag: increase default maximum of fragments")
---
 .mailmap                                             | 1 +
 doc/guides/prog_guide/ip_fragment_reassembly_lib.rst | 8 +++++++-
 2 files changed, 8 insertions(+), 1 deletion(-)
.mailmap +1 -0
index ab0742a382..528bc68a30 100644
@@ -379,6 +379,7 @@ Eric Zhang <[email protected]>
 Erik Gabriel Carrillo <[email protected]>
 Erik Ziegenbalg <[email protected]>
 Erlu Chen <[email protected]>
+Euan Bourke <[email protected]>
 Eugenio Pérez <[email protected]>
 Eugeny Parshutin <[email protected]>
 Evan Swanson <[email protected]>
doc/guides/prog_guide/ip_fragment_reassembly_lib.rst +7 -1
index 314d4adbb8..f3ed90d700 100644
@@ -43,7 +43,13 @@ Note that all update/lookup operations on Fragment Table are not thread safe.
 So if different execution contexts (threads/processes) will access the same table simultaneously,
 then some external syncing mechanism have to be provided.
 
-Each table entry can hold information about packets consisting of up to RTE_LIBRTE_IP_FRAG_MAX (by default: 4) fragments.
+Each table entry can hold information about packets of up to ``RTE_LIBRTE_IP_FRAG_MAX_FRAG`` fragments,
+where ``RTE_LIBRTE_IP_FRAG_MAX_FRAG`` defaults to:
+
+.. literalinclude:: ../../../config/rte_config.h
+    :language: c
+    :start-at: #define RTE_LIBRTE_IP_FRAG_MAX_FRAG
+    :lines: 1
 
 Code example, that demonstrates creation of a new Fragment table:
 
Signature · 1 line
2.34.1
#10 lore raw

PATCH v4 2/2 ip_frag: updated name for IP frag define

Removed LIBRTE from name as its an old prefix.

---
 app/test/test_reassembly_perf.c                      | 2 +-
 config/rte_config.h                                  | 2 +-
 doc/guides/prog_guide/ip_fragment_reassembly_lib.rst | 8 ++++----
 doc/guides/sample_app_ug/ip_reassembly.rst           | 4 ++--
 examples/ip_fragmentation/main.c                     | 2 +-
 examples/ip_reassembly/main.c                        | 2 +-
 examples/ipsec-secgw/ipsec_worker.h                  | 2 +-
 lib/ip_frag/ip_reassembly.h                          | 2 +-
 lib/ip_frag/rte_ip_frag.h                            | 2 +-
 9 files changed, 13 insertions(+), 13 deletions(-)
app/test/test_reassembly_perf.c +1 -1
index 3912179022..805ae2fe9d 100644
@@ -20,7 +20,7 @@
 #define MAX_FLOWS	    (1024 * 32)
 #define MAX_BKTS	    MAX_FLOWS
 #define MAX_ENTRIES_PER_BKT 16
-#define MAX_FRAGMENTS	    RTE_LIBRTE_IP_FRAG_MAX_FRAG
+#define MAX_FRAGMENTS	    RTE_IP_FRAG_MAX_FRAG
 #define MIN_FRAGMENTS	    2
 #define MAX_PKTS	    (MAX_FLOWS * MAX_FRAGMENTS)
 
config/rte_config.h +1 -1
index da265d7dd2..e2fa2a58fa 100644
@@ -85,7 +85,7 @@
 #define RTE_RAWDEV_MAX_DEVS 64
 
 /* ip_fragmentation defines */
-#define RTE_LIBRTE_IP_FRAG_MAX_FRAG 8
+#define RTE_IP_FRAG_MAX_FRAG 8
 // RTE_LIBRTE_IP_FRAG_TBL_STAT is not set
 
 /* rte_power defines */
doc/guides/prog_guide/ip_fragment_reassembly_lib.rst +4 -4
index f3ed90d700..d826854890 100644
@@ -43,12 +43,12 @@ Note that all update/lookup operations on Fragment Table are not thread safe.
 So if different execution contexts (threads/processes) will access the same table simultaneously,
 then some external syncing mechanism have to be provided.
 
-Each table entry can hold information about packets of up to ``RTE_LIBRTE_IP_FRAG_MAX_FRAG`` fragments,
-where ``RTE_LIBRTE_IP_FRAG_MAX_FRAG`` defaults to:
+Each table entry can hold information about packets of up to ``RTE_IP_FRAG_MAX_FRAG`` fragments,
+where ``RTE_IP_FRAG_MAX_FRAG`` defaults to:
 
 .. literalinclude:: ../../../config/rte_config.h
     :language: c
-    :start-at: #define RTE_LIBRTE_IP_FRAG_MAX_FRAG
+    :start-at: #define RTE_IP_FRAG_MAX_FRAG
     :lines: 1
 
 Code example, that demonstrates creation of a new Fragment table:
@@ -69,7 +69,7 @@ Also, entries that resides in the table longer then <max_cycles> are considered
 and could be removed/replaced by the new ones.
 
 Note that reassembly demands a lot of mbuf's to be allocated.
-At any given time up to (2 \* bucket_entries \* RTE_LIBRTE_IP_FRAG_MAX \* <maximum number of mbufs per packet>)
+At any given time up to (2 \* bucket_entries \* RTE_IP_FRAG_MAX_FRAG \* <maximum number of mbufs per packet>)
 can be stored inside Fragment Table waiting for remaining fragments.
 
 Packet Reassembly
doc/guides/sample_app_ug/ip_reassembly.rst +2 -2
index 5280bf4ea0..44684a39e2 100644
@@ -135,7 +135,7 @@ Fragment table maintains information about already received fragments of the pac
 Each IP packet is uniquely identified by triple <Source IP address>, <Destination IP address>, <ID>.
 To avoid lock contention, each RX queue has its own Fragment Table,
 e.g. the application can't handle the situation when different fragments of the same packet arrive through different RX queues.
-Each table entry can hold information about packet consisting of up to RTE_LIBRTE_IP_FRAG_MAX_FRAGS fragments.
+Each table entry can hold information about packet consisting of up to RTE_IP_FRAG_MAX_FRAG fragments.
 
 .. literalinclude:: ../../../examples/ip_reassembly/main.c
     :language: c
@@ -147,7 +147,7 @@ Mempools Initialization
 ~~~~~~~~~~~~~~~~~~~~~~~
 
 The reassembly application demands a lot of mbuf's to be allocated.
-At any given time up to (2 \* max_flow_num \* RTE_LIBRTE_IP_FRAG_MAX_FRAGS \* <maximum number of mbufs per packet>)
+At any given time up to (2 \* max_flow_num \* RTE_IP_FRAG_MAX_FRAG \* <maximum number of mbufs per packet>)
 can be stored inside Fragment Table waiting for remaining fragments.
 To keep mempool size under reasonable limits and to avoid situation when one RX queue can starve other queues,
 each RX queue uses its own mempool.
examples/ip_fragmentation/main.c +1 -1
index 744a1aa9b4..1e4471891b 100644
@@ -71,7 +71,7 @@
 /*
  * Max number of fragments per packet expected - defined by config file.
  */
-#define	MAX_PACKET_FRAG RTE_LIBRTE_IP_FRAG_MAX_FRAG
+#define	MAX_PACKET_FRAG RTE_IP_FRAG_MAX_FRAG
 
 #define NB_MBUF   8192
 
examples/ip_reassembly/main.c +1 -1
index bd0b1d31de..16607d99f3 100644
@@ -69,7 +69,7 @@
 #define	MIN_FLOW_TTL	1
 #define	DEF_FLOW_TTL	MS_PER_S
 
-#define MAX_FRAG_NUM RTE_LIBRTE_IP_FRAG_MAX_FRAG
+#define MAX_FRAG_NUM RTE_IP_FRAG_MAX_FRAG
 
 /* Should be power of two. */
 #define	IP_FRAG_TBL_BUCKET_ENTRIES	16
examples/ipsec-secgw/ipsec_worker.h +1 -1
index ac980b8bcf..918e6b5200 100644
@@ -291,7 +291,7 @@ send_fragment_packet(struct lcore_conf *qconf, struct rte_mbuf *m,
 	len = tbl->len;
 
 	/* free space for new fragments */
-	if (len + RTE_LIBRTE_IP_FRAG_MAX_FRAG >=  RTE_DIM(tbl->m_table)) {
+	if (len + RTE_IP_FRAG_MAX_FRAG >=  RTE_DIM(tbl->m_table)) {
 		send_burst(qconf, len, port);
 		len = 0;
 	}
lib/ip_frag/ip_reassembly.h +1 -1
index a9f97aefca..77c22ca9d1 100644
@@ -16,7 +16,7 @@ enum {
 	IP_LAST_FRAG_IDX,    /* index of last fragment */
 	IP_FIRST_FRAG_IDX,   /* index of first fragment */
 	IP_MIN_FRAG_NUM,     /* minimum number of fragments */
-	IP_MAX_FRAG_NUM = RTE_LIBRTE_IP_FRAG_MAX_FRAG,
+	IP_MAX_FRAG_NUM = RTE_IP_FRAG_MAX_FRAG,
 	/* maximum number of fragments per packet */
 };
 
lib/ip_frag/rte_ip_frag.h +1 -1
index 2ad318096b..6465ec316b 100644
@@ -32,7 +32,7 @@ struct rte_mbuf;
 
 /** death row size in mbufs */
 #define RTE_IP_FRAG_DEATH_ROW_MBUF_LEN \
-	(RTE_IP_FRAG_DEATH_ROW_LEN * (RTE_LIBRTE_IP_FRAG_MAX_FRAG + 1))
+	(RTE_IP_FRAG_DEATH_ROW_LEN * (RTE_IP_FRAG_MAX_FRAG + 1))
 
 /** mbuf death row (packets to be freed) */
 struct rte_ip_frag_death_row {
Signature · 1 line
2.34.1
#11 lore raw

Re: [PATCH v4 2/2] ip_frag: updated name for IP frag define

config/rte_config.h line 90

Quoted 12 lines · On Mon, 8 Jan 2024 11:36:14 +0000 Euan Bourke <[email protected]> wrote:
config/rte_config.h +1 -1
index da265d7dd2..e2fa2a58fa 100644
@@ -85,7 +85,7 @@
 #define RTE_RAWDEV_MAX_DEVS 64
 
 /* ip_fragmentation defines */
-#define RTE_LIBRTE_IP_FRAG_MAX_FRAG 8
+#define RTE_IP_FRAG_MAX_FRAG 8
 // RTE_LIBRTE_IP_FRAG_TBL_STAT is not set
 

Any change to a visible define should be documented as an API change in the release notes.

Also if changing the name for MAX_FRAG shouldn't the name for the statistics be changed as well as in: RTE_IP_FRAG_TBL_STAT?