* [PATCH bpf-next 1/5] bpf: net: Remove TC_AT_INGRESS_OFFSET and SKB_MONO_DELIVERY_TIME_OFFSET macro
2022-03-09 9:04 [PATCH bpf-next 0/5] bpf: Follow up on bpf __sk_buff->tstamp Martin KaFai Lau
@ 2022-03-09 9:04 ` Martin KaFai Lau
2022-03-09 9:04 ` [PATCH bpf-next 2/5] bpf: Simplify insn rewrite on BPF_READ __sk_buff->tstamp Martin KaFai Lau
` (4 subsequent siblings)
5 siblings, 0 replies; 7+ messages in thread
From: Martin KaFai Lau @ 2022-03-09 9:04 UTC (permalink / raw)
To: bpf
Cc: Alexei Starovoitov, Andrii Nakryiko, Daniel Borkmann, kernel-team,
netdev
This patch removes the TC_AT_INGRESS_OFFSET and
SKB_MONO_DELIVERY_TIME_OFFSET macros. Instead, PKT_VLAN_PRESENT_OFFSET
is used because all of them are at the same offset. Comment is added to
make it clear that changing the position of tc_at_ingress or
mono_delivery_time will require to adjust the defined macros.
The earlier discussion can be found here:
https://lore.kernel.org/bpf/419d994e-ff61-7c11-0ec7-11fefcb0186e@iogearbox.net/
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
---
include/linux/skbuff.h | 10 +++++-----
net/core/filter.c | 14 +++++++-------
2 files changed, 12 insertions(+), 12 deletions(-)
diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index 2be263184d1e..7b0cb10e70cb 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -960,10 +960,10 @@ struct sk_buff {
__u8 csum_complete_sw:1;
__u8 csum_level:2;
__u8 dst_pending_confirm:1;
- __u8 mono_delivery_time:1;
+ __u8 mono_delivery_time:1; /* See SKB_MONO_DELIVERY_TIME_MASK */
#ifdef CONFIG_NET_CLS_ACT
__u8 tc_skip_classify:1;
- __u8 tc_at_ingress:1;
+ __u8 tc_at_ingress:1; /* See TC_AT_INGRESS_MASK */
#endif
#ifdef CONFIG_IPV6_NDISC_NODETYPE
__u8 ndisc_nodetype:2;
@@ -1062,7 +1062,9 @@ struct sk_buff {
#endif
#define PKT_TYPE_OFFSET offsetof(struct sk_buff, __pkt_type_offset)
-/* if you move pkt_vlan_present around you also must adapt these constants */
+/* if you move pkt_vlan_present, tc_at_ingress, or mono_delivery_time
+ * around, you also must adapt these constants.
+ */
#ifdef __BIG_ENDIAN_BITFIELD
#define PKT_VLAN_PRESENT_BIT 7
#define TC_AT_INGRESS_MASK (1 << 0)
@@ -1073,8 +1075,6 @@ struct sk_buff {
#define SKB_MONO_DELIVERY_TIME_MASK (1 << 5)
#endif
#define PKT_VLAN_PRESENT_OFFSET offsetof(struct sk_buff, __pkt_vlan_present_offset)
-#define TC_AT_INGRESS_OFFSET offsetof(struct sk_buff, __pkt_vlan_present_offset)
-#define SKB_MONO_DELIVERY_TIME_OFFSET offsetof(struct sk_buff, __pkt_vlan_present_offset)
#ifdef __KERNEL__
/*
diff --git a/net/core/filter.c b/net/core/filter.c
index 88767f7da150..738a294a3c82 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -8896,7 +8896,7 @@ static struct bpf_insn *bpf_convert_dtime_type_read(const struct bpf_insn *si,
__u8 tmp_reg = BPF_REG_AX;
*insn++ = BPF_LDX_MEM(BPF_B, tmp_reg, skb_reg,
- SKB_MONO_DELIVERY_TIME_OFFSET);
+ PKT_VLAN_PRESENT_OFFSET);
*insn++ = BPF_ALU32_IMM(BPF_AND, tmp_reg,
SKB_MONO_DELIVERY_TIME_MASK);
*insn++ = BPF_JMP32_IMM(BPF_JEQ, tmp_reg, 0, 2);
@@ -8912,7 +8912,7 @@ static struct bpf_insn *bpf_convert_dtime_type_read(const struct bpf_insn *si,
*insn++ = BPF_JMP_A(IS_ENABLED(CONFIG_NET_CLS_ACT) ? 6 : 1);
#ifdef CONFIG_NET_CLS_ACT
- *insn++ = BPF_LDX_MEM(BPF_B, tmp_reg, skb_reg, TC_AT_INGRESS_OFFSET);
+ *insn++ = BPF_LDX_MEM(BPF_B, tmp_reg, skb_reg, PKT_VLAN_PRESENT_OFFSET);
*insn++ = BPF_ALU32_IMM(BPF_AND, tmp_reg, TC_AT_INGRESS_MASK);
*insn++ = BPF_JMP32_IMM(BPF_JEQ, tmp_reg, 0, 2);
/* At ingress, value_reg = 0 */
@@ -8959,14 +8959,14 @@ static struct bpf_insn *bpf_convert_tstamp_read(const struct bpf_prog *prog,
if (!prog->delivery_time_access) {
__u8 tmp_reg = BPF_REG_AX;
- *insn++ = BPF_LDX_MEM(BPF_B, tmp_reg, skb_reg, TC_AT_INGRESS_OFFSET);
+ *insn++ = BPF_LDX_MEM(BPF_B, tmp_reg, skb_reg, PKT_VLAN_PRESENT_OFFSET);
*insn++ = BPF_ALU32_IMM(BPF_AND, tmp_reg, TC_AT_INGRESS_MASK);
*insn++ = BPF_JMP32_IMM(BPF_JEQ, tmp_reg, 0, 5);
/* @ingress, read __sk_buff->tstamp as the (rcv) timestamp,
* so check the skb->mono_delivery_time.
*/
*insn++ = BPF_LDX_MEM(BPF_B, tmp_reg, skb_reg,
- SKB_MONO_DELIVERY_TIME_OFFSET);
+ PKT_VLAN_PRESENT_OFFSET);
*insn++ = BPF_ALU32_IMM(BPF_AND, tmp_reg,
SKB_MONO_DELIVERY_TIME_MASK);
*insn++ = BPF_JMP32_IMM(BPF_JEQ, tmp_reg, 0, 2);
@@ -8992,18 +8992,18 @@ static struct bpf_insn *bpf_convert_tstamp_write(const struct bpf_prog *prog,
if (!prog->delivery_time_access) {
__u8 tmp_reg = BPF_REG_AX;
- *insn++ = BPF_LDX_MEM(BPF_B, tmp_reg, skb_reg, TC_AT_INGRESS_OFFSET);
+ *insn++ = BPF_LDX_MEM(BPF_B, tmp_reg, skb_reg, PKT_VLAN_PRESENT_OFFSET);
*insn++ = BPF_ALU32_IMM(BPF_AND, tmp_reg, TC_AT_INGRESS_MASK);
*insn++ = BPF_JMP32_IMM(BPF_JEQ, tmp_reg, 0, 3);
/* Writing __sk_buff->tstamp at ingress as the (rcv) timestamp.
* Clear the skb->mono_delivery_time.
*/
*insn++ = BPF_LDX_MEM(BPF_B, tmp_reg, skb_reg,
- SKB_MONO_DELIVERY_TIME_OFFSET);
+ PKT_VLAN_PRESENT_OFFSET);
*insn++ = BPF_ALU32_IMM(BPF_AND, tmp_reg,
~SKB_MONO_DELIVERY_TIME_MASK);
*insn++ = BPF_STX_MEM(BPF_B, skb_reg, tmp_reg,
- SKB_MONO_DELIVERY_TIME_OFFSET);
+ PKT_VLAN_PRESENT_OFFSET);
}
#endif
--
2.30.2
^ permalink raw reply related [flat|nested] 7+ messages in thread* [PATCH bpf-next 2/5] bpf: Simplify insn rewrite on BPF_READ __sk_buff->tstamp
2022-03-09 9:04 [PATCH bpf-next 0/5] bpf: Follow up on bpf __sk_buff->tstamp Martin KaFai Lau
2022-03-09 9:04 ` [PATCH bpf-next 1/5] bpf: net: Remove TC_AT_INGRESS_OFFSET and SKB_MONO_DELIVERY_TIME_OFFSET macro Martin KaFai Lau
@ 2022-03-09 9:04 ` Martin KaFai Lau
2022-03-09 9:05 ` [PATCH bpf-next 3/5] bpf: Simplify insn rewrite on BPF_WRITE __sk_buff->tstamp Martin KaFai Lau
` (3 subsequent siblings)
5 siblings, 0 replies; 7+ messages in thread
From: Martin KaFai Lau @ 2022-03-09 9:04 UTC (permalink / raw)
To: bpf
Cc: Alexei Starovoitov, Andrii Nakryiko, Daniel Borkmann, kernel-team,
netdev
The skb->tc_at_ingress and skb->mono_delivery_time are at the same
byte offset. Thus, only one BPF_LDX_MEM(BPF_B) is needed
and both bits can be tested together.
/* BPF_READ: a = __sk_buff->tstamp */
if (skb->tc_at_ingress && skb->mono_delivery_time)
a = 0;
else
a = skb->tstamp;
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
---
net/core/filter.c | 21 +++++++++++----------
1 file changed, 11 insertions(+), 10 deletions(-)
diff --git a/net/core/filter.c b/net/core/filter.c
index 738a294a3c82..2c83d1f38704 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -8956,21 +8956,22 @@ static struct bpf_insn *bpf_convert_tstamp_read(const struct bpf_prog *prog,
__u8 skb_reg = si->src_reg;
#ifdef CONFIG_NET_CLS_ACT
+ /* If the delivery_time_type is read,
+ * the bpf prog is aware the tstamp could have delivery time.
+ * Thus, read skb->tstamp as is if delivery_time_access is true.
+ */
if (!prog->delivery_time_access) {
+ /* AX is needed because src_reg and dst_reg could be the same */
__u8 tmp_reg = BPF_REG_AX;
*insn++ = BPF_LDX_MEM(BPF_B, tmp_reg, skb_reg, PKT_VLAN_PRESENT_OFFSET);
- *insn++ = BPF_ALU32_IMM(BPF_AND, tmp_reg, TC_AT_INGRESS_MASK);
- *insn++ = BPF_JMP32_IMM(BPF_JEQ, tmp_reg, 0, 5);
- /* @ingress, read __sk_buff->tstamp as the (rcv) timestamp,
- * so check the skb->mono_delivery_time.
- */
- *insn++ = BPF_LDX_MEM(BPF_B, tmp_reg, skb_reg,
- PKT_VLAN_PRESENT_OFFSET);
*insn++ = BPF_ALU32_IMM(BPF_AND, tmp_reg,
- SKB_MONO_DELIVERY_TIME_MASK);
- *insn++ = BPF_JMP32_IMM(BPF_JEQ, tmp_reg, 0, 2);
- /* skb->mono_delivery_time is set, read 0 as the (rcv) timestamp. */
+ TC_AT_INGRESS_MASK | SKB_MONO_DELIVERY_TIME_MASK);
+ *insn++ = BPF_JMP32_IMM(BPF_JNE, tmp_reg,
+ TC_AT_INGRESS_MASK | SKB_MONO_DELIVERY_TIME_MASK, 2);
+ /* skb->tc_at_ingress && skb->mono_delivery_time,
+ * read 0 as the (rcv) timestamp.
+ */
*insn++ = BPF_MOV64_IMM(value_reg, 0);
*insn++ = BPF_JMP_A(1);
}
--
2.30.2
^ permalink raw reply related [flat|nested] 7+ messages in thread* [PATCH bpf-next 3/5] bpf: Simplify insn rewrite on BPF_WRITE __sk_buff->tstamp
2022-03-09 9:04 [PATCH bpf-next 0/5] bpf: Follow up on bpf __sk_buff->tstamp Martin KaFai Lau
2022-03-09 9:04 ` [PATCH bpf-next 1/5] bpf: net: Remove TC_AT_INGRESS_OFFSET and SKB_MONO_DELIVERY_TIME_OFFSET macro Martin KaFai Lau
2022-03-09 9:04 ` [PATCH bpf-next 2/5] bpf: Simplify insn rewrite on BPF_READ __sk_buff->tstamp Martin KaFai Lau
@ 2022-03-09 9:05 ` Martin KaFai Lau
2022-03-09 9:05 ` [PATCH bpf-next 4/5] bpf: Remove BPF_SKB_DELIVERY_TIME_NONE and rename s/delivery_time_/tstamp_/ Martin KaFai Lau
` (2 subsequent siblings)
5 siblings, 0 replies; 7+ messages in thread
From: Martin KaFai Lau @ 2022-03-09 9:05 UTC (permalink / raw)
To: bpf
Cc: Alexei Starovoitov, Andrii Nakryiko, Daniel Borkmann, kernel-team,
netdev
BPF_JMP32_IMM(BPF_JSET) is used to save a BPF_ALU32_IMM(BPF_AND).
The skb->tc_at_ingress and skb->mono_delivery_time are at the same
offset, so only one BPF_LDX_MEM(BPF_B) is needed.
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
---
net/core/filter.c | 26 ++++++++++++++------------
1 file changed, 14 insertions(+), 12 deletions(-)
diff --git a/net/core/filter.c b/net/core/filter.c
index 2c83d1f38704..f914e4b13b18 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -8990,25 +8990,27 @@ static struct bpf_insn *bpf_convert_tstamp_write(const struct bpf_prog *prog,
__u8 skb_reg = si->dst_reg;
#ifdef CONFIG_NET_CLS_ACT
+ /* If the delivery_time_type is read,
+ * the bpf prog is aware the tstamp could have delivery time.
+ * Thus, write skb->tstamp as is if delivery_time_access is true.
+ * Otherwise, writing at ingress will have to clear the
+ * mono_delivery_time bit also.
+ */
if (!prog->delivery_time_access) {
__u8 tmp_reg = BPF_REG_AX;
*insn++ = BPF_LDX_MEM(BPF_B, tmp_reg, skb_reg, PKT_VLAN_PRESENT_OFFSET);
- *insn++ = BPF_ALU32_IMM(BPF_AND, tmp_reg, TC_AT_INGRESS_MASK);
- *insn++ = BPF_JMP32_IMM(BPF_JEQ, tmp_reg, 0, 3);
- /* Writing __sk_buff->tstamp at ingress as the (rcv) timestamp.
- * Clear the skb->mono_delivery_time.
- */
- *insn++ = BPF_LDX_MEM(BPF_B, tmp_reg, skb_reg,
- PKT_VLAN_PRESENT_OFFSET);
- *insn++ = BPF_ALU32_IMM(BPF_AND, tmp_reg,
- ~SKB_MONO_DELIVERY_TIME_MASK);
- *insn++ = BPF_STX_MEM(BPF_B, skb_reg, tmp_reg,
- PKT_VLAN_PRESENT_OFFSET);
+ /* Writing __sk_buff->tstamp as ingress, goto <clear> */
+ *insn++ = BPF_JMP32_IMM(BPF_JSET, tmp_reg, TC_AT_INGRESS_MASK, 1);
+ /* goto <store> */
+ *insn++ = BPF_JMP_A(2);
+ /* <clear>: mono_delivery_time */
+ *insn++ = BPF_ALU32_IMM(BPF_AND, tmp_reg, ~SKB_MONO_DELIVERY_TIME_MASK);
+ *insn++ = BPF_STX_MEM(BPF_B, skb_reg, tmp_reg, PKT_VLAN_PRESENT_OFFSET);
}
#endif
- /* skb->tstamp = tstamp */
+ /* <store>: skb->tstamp = tstamp */
*insn++ = BPF_STX_MEM(BPF_DW, skb_reg, value_reg,
offsetof(struct sk_buff, tstamp));
return insn;
--
2.30.2
^ permalink raw reply related [flat|nested] 7+ messages in thread* [PATCH bpf-next 4/5] bpf: Remove BPF_SKB_DELIVERY_TIME_NONE and rename s/delivery_time_/tstamp_/
2022-03-09 9:04 [PATCH bpf-next 0/5] bpf: Follow up on bpf __sk_buff->tstamp Martin KaFai Lau
` (2 preceding siblings ...)
2022-03-09 9:05 ` [PATCH bpf-next 3/5] bpf: Simplify insn rewrite on BPF_WRITE __sk_buff->tstamp Martin KaFai Lau
@ 2022-03-09 9:05 ` Martin KaFai Lau
2022-03-09 9:05 ` [PATCH bpf-next 5/5] bpf: selftests: Update tests after s/delivery_time/tstamp/ change in bpf.h Martin KaFai Lau
2022-03-10 22:00 ` [PATCH bpf-next 0/5] bpf: Follow up on bpf __sk_buff->tstamp patchwork-bot+netdevbpf
5 siblings, 0 replies; 7+ messages in thread
From: Martin KaFai Lau @ 2022-03-09 9:05 UTC (permalink / raw)
To: bpf
Cc: Alexei Starovoitov, Andrii Nakryiko, Daniel Borkmann, kernel-team,
netdev
This patch is to simplify the uapi bpf.h regarding to the tstamp type
and use a similar way as the kernel to describe the value stored
in __sk_buff->tstamp.
My earlier thought was to avoid describing the semantic and
clock base for the rcv timestamp until there is more clarity
on the use case, so the __sk_buff->delivery_time_type naming instead
of __sk_buff->tstamp_type.
With some thoughts, it can reuse the UNSPEC naming. This patch first
removes BPF_SKB_DELIVERY_TIME_NONE and also
rename BPF_SKB_DELIVERY_TIME_UNSPEC to BPF_SKB_TSTAMP_UNSPEC
and BPF_SKB_DELIVERY_TIME_MONO to BPF_SKB_TSTAMP_DELIVERY_MONO.
The semantic of BPF_SKB_TSTAMP_DELIVERY_MONO is the same:
__sk_buff->tstamp has delivery time in mono clock base.
BPF_SKB_TSTAMP_UNSPEC means __sk_buff->tstamp has the (rcv)
tstamp at ingress and the delivery time at egress. At egress,
the clock base could be found from skb->sk->sk_clockid.
__sk_buff->tstamp == 0 naturally means NONE, so NONE is not needed.
With BPF_SKB_TSTAMP_UNSPEC for the rcv tstamp at ingress,
the __sk_buff->delivery_time_type is also renamed to __sk_buff->tstamp_type
which was also suggested in the earlier discussion:
https://lore.kernel.org/bpf/b181acbe-caf8-502d-4b7b-7d96b9fc5d55@iogearbox.net/
The above will then make __sk_buff->tstamp and __sk_buff->tstamp_type
the same as its kernel skb->tstamp and skb->mono_delivery_time
counter part.
The internal kernel function bpf_skb_convert_dtime_type_read() is then
renamed to bpf_skb_convert_tstamp_type_read() and it can be simplified
with the BPF_SKB_DELIVERY_TIME_NONE gone. A BPF_ALU32_IMM(BPF_AND)
insn is also saved by using BPF_JMP32_IMM(BPF_JSET).
The bpf helper bpf_skb_set_delivery_time() is also renamed to
bpf_skb_set_tstamp(). The arg name is changed from dtime
to tstamp also. It only allows setting tstamp 0 for
BPF_SKB_TSTAMP_UNSPEC and it could be relaxed later
if there is use case to change mono delivery time to
non mono.
prog->delivery_time_access is also renamed to prog->tstamp_type_access.
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
---
include/linux/filter.h | 2 +-
include/uapi/linux/bpf.h | 40 ++++++++--------
net/core/filter.c | 88 +++++++++++++---------------------
tools/include/uapi/linux/bpf.h | 40 ++++++++--------
4 files changed, 77 insertions(+), 93 deletions(-)
diff --git a/include/linux/filter.h b/include/linux/filter.h
index 9bf26307247f..05ed9bd31b45 100644
--- a/include/linux/filter.h
+++ b/include/linux/filter.h
@@ -573,7 +573,7 @@ struct bpf_prog {
enforce_expected_attach_type:1, /* Enforce expected_attach_type checking at attach time */
call_get_stack:1, /* Do we call bpf_get_stack() or bpf_get_stackid() */
call_get_func_ip:1, /* Do we call get_func_ip() */
- delivery_time_access:1; /* Accessed __sk_buff->delivery_time_type */
+ tstamp_type_access:1; /* Accessed __sk_buff->tstamp_type */
enum bpf_prog_type type; /* Type of BPF program */
enum bpf_attach_type expected_attach_type; /* For some prog types */
u32 len; /* Number of filter blocks */
diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
index 4eebea830613..65871bdbb9c8 100644
--- a/include/uapi/linux/bpf.h
+++ b/include/uapi/linux/bpf.h
@@ -5087,23 +5087,22 @@ union bpf_attr {
* 0 on success, or a negative error in case of failure. On error
* *dst* buffer is zeroed out.
*
- * long bpf_skb_set_delivery_time(struct sk_buff *skb, u64 dtime, u32 dtime_type)
+ * long bpf_skb_set_tstamp(struct sk_buff *skb, u64 tstamp, u32 tstamp_type)
* Description
- * Set a *dtime* (delivery time) to the __sk_buff->tstamp and also
- * change the __sk_buff->delivery_time_type to *dtime_type*.
+ * Change the __sk_buff->tstamp_type to *tstamp_type*
+ * and set *tstamp* to the __sk_buff->tstamp together.
*
- * When setting a delivery time (non zero *dtime*) to
- * __sk_buff->tstamp, only BPF_SKB_DELIVERY_TIME_MONO *dtime_type*
- * is supported. It is the only delivery_time_type that will be
- * kept after bpf_redirect_*().
- *
- * If there is no need to change the __sk_buff->delivery_time_type,
- * the delivery time can be directly written to __sk_buff->tstamp
+ * If there is no need to change the __sk_buff->tstamp_type,
+ * the tstamp value can be directly written to __sk_buff->tstamp
* instead.
*
- * *dtime* 0 and *dtime_type* BPF_SKB_DELIVERY_TIME_NONE
- * can be used to clear any delivery time stored in
- * __sk_buff->tstamp.
+ * BPF_SKB_TSTAMP_DELIVERY_MONO is the only tstamp that
+ * will be kept during bpf_redirect_*(). A non zero
+ * *tstamp* must be used with the BPF_SKB_TSTAMP_DELIVERY_MONO
+ * *tstamp_type*.
+ *
+ * A BPF_SKB_TSTAMP_UNSPEC *tstamp_type* can only be used
+ * with a zero *tstamp*.
*
* Only IPv4 and IPv6 skb->protocol are supported.
*
@@ -5116,7 +5115,7 @@ union bpf_attr {
* Return
* 0 on success.
* **-EINVAL** for invalid input
- * **-EOPNOTSUPP** for unsupported delivery_time_type and protocol
+ * **-EOPNOTSUPP** for unsupported protocol
*/
#define __BPF_FUNC_MAPPER(FN) \
FN(unspec), \
@@ -5311,7 +5310,7 @@ union bpf_attr {
FN(xdp_load_bytes), \
FN(xdp_store_bytes), \
FN(copy_from_user_task), \
- FN(skb_set_delivery_time), \
+ FN(skb_set_tstamp), \
/* */
/* integer value in 'imm' field of BPF_CALL instruction selects which helper
@@ -5502,9 +5501,12 @@ union { \
} __attribute__((aligned(8)))
enum {
- BPF_SKB_DELIVERY_TIME_NONE,
- BPF_SKB_DELIVERY_TIME_UNSPEC,
- BPF_SKB_DELIVERY_TIME_MONO,
+ BPF_SKB_TSTAMP_UNSPEC,
+ BPF_SKB_TSTAMP_DELIVERY_MONO, /* tstamp has mono delivery time */
+ /* For any BPF_SKB_TSTAMP_* that the bpf prog cannot handle,
+ * the bpf prog should handle it like BPF_SKB_TSTAMP_UNSPEC
+ * and try to deduce it by ingress, egress or skb->sk->sk_clockid.
+ */
};
/* user accessible mirror of in-kernel sk_buff.
@@ -5547,7 +5549,7 @@ struct __sk_buff {
__u32 gso_segs;
__bpf_md_ptr(struct bpf_sock *, sk);
__u32 gso_size;
- __u8 delivery_time_type;
+ __u8 tstamp_type;
__u32 :24; /* Padding, future use. */
__u64 hwtstamp;
};
diff --git a/net/core/filter.c b/net/core/filter.c
index f914e4b13b18..03655f2074ae 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -7388,36 +7388,36 @@ static const struct bpf_func_proto bpf_sock_ops_reserve_hdr_opt_proto = {
.arg3_type = ARG_ANYTHING,
};
-BPF_CALL_3(bpf_skb_set_delivery_time, struct sk_buff *, skb,
- u64, dtime, u32, dtime_type)
+BPF_CALL_3(bpf_skb_set_tstamp, struct sk_buff *, skb,
+ u64, tstamp, u32, tstamp_type)
{
/* skb_clear_delivery_time() is done for inet protocol */
if (skb->protocol != htons(ETH_P_IP) &&
skb->protocol != htons(ETH_P_IPV6))
return -EOPNOTSUPP;
- switch (dtime_type) {
- case BPF_SKB_DELIVERY_TIME_MONO:
- if (!dtime)
+ switch (tstamp_type) {
+ case BPF_SKB_TSTAMP_DELIVERY_MONO:
+ if (!tstamp)
return -EINVAL;
- skb->tstamp = dtime;
+ skb->tstamp = tstamp;
skb->mono_delivery_time = 1;
break;
- case BPF_SKB_DELIVERY_TIME_NONE:
- if (dtime)
+ case BPF_SKB_TSTAMP_UNSPEC:
+ if (tstamp)
return -EINVAL;
skb->tstamp = 0;
skb->mono_delivery_time = 0;
break;
default:
- return -EOPNOTSUPP;
+ return -EINVAL;
}
return 0;
}
-static const struct bpf_func_proto bpf_skb_set_delivery_time_proto = {
- .func = bpf_skb_set_delivery_time,
+static const struct bpf_func_proto bpf_skb_set_tstamp_proto = {
+ .func = bpf_skb_set_tstamp,
.gpl_only = false,
.ret_type = RET_INTEGER,
.arg1_type = ARG_PTR_TO_CTX,
@@ -7786,8 +7786,8 @@ tc_cls_act_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog)
return &bpf_tcp_gen_syncookie_proto;
case BPF_FUNC_sk_assign:
return &bpf_sk_assign_proto;
- case BPF_FUNC_skb_set_delivery_time:
- return &bpf_skb_set_delivery_time_proto;
+ case BPF_FUNC_skb_set_tstamp:
+ return &bpf_skb_set_tstamp_proto;
#endif
default:
return bpf_sk_base_func_proto(func_id);
@@ -8127,9 +8127,9 @@ static bool bpf_skb_is_valid_access(int off, int size, enum bpf_access_type type
return false;
info->reg_type = PTR_TO_SOCK_COMMON_OR_NULL;
break;
- case offsetof(struct __sk_buff, delivery_time_type):
+ case offsetof(struct __sk_buff, tstamp_type):
return false;
- case offsetofend(struct __sk_buff, delivery_time_type) ... offsetof(struct __sk_buff, hwtstamp) - 1:
+ case offsetofend(struct __sk_buff, tstamp_type) ... offsetof(struct __sk_buff, hwtstamp) - 1:
/* Explicitly prohibit access to padding in __sk_buff. */
return false;
default:
@@ -8484,14 +8484,14 @@ static bool tc_cls_act_is_valid_access(int off, int size,
break;
case bpf_ctx_range_till(struct __sk_buff, family, local_port):
return false;
- case offsetof(struct __sk_buff, delivery_time_type):
+ case offsetof(struct __sk_buff, tstamp_type):
/* The convert_ctx_access() on reading and writing
* __sk_buff->tstamp depends on whether the bpf prog
- * has used __sk_buff->delivery_time_type or not.
- * Thus, we need to set prog->delivery_time_access
+ * has used __sk_buff->tstamp_type or not.
+ * Thus, we need to set prog->tstamp_type_access
* earlier during is_valid_access() here.
*/
- ((struct bpf_prog *)prog)->delivery_time_access = 1;
+ ((struct bpf_prog *)prog)->tstamp_type_access = 1;
return size == sizeof(__u8);
}
@@ -8888,42 +8888,22 @@ static u32 flow_dissector_convert_ctx_access(enum bpf_access_type type,
return insn - insn_buf;
}
-static struct bpf_insn *bpf_convert_dtime_type_read(const struct bpf_insn *si,
- struct bpf_insn *insn)
+static struct bpf_insn *bpf_convert_tstamp_type_read(const struct bpf_insn *si,
+ struct bpf_insn *insn)
{
__u8 value_reg = si->dst_reg;
__u8 skb_reg = si->src_reg;
+ /* AX is needed because src_reg and dst_reg could be the same */
__u8 tmp_reg = BPF_REG_AX;
*insn++ = BPF_LDX_MEM(BPF_B, tmp_reg, skb_reg,
PKT_VLAN_PRESENT_OFFSET);
- *insn++ = BPF_ALU32_IMM(BPF_AND, tmp_reg,
- SKB_MONO_DELIVERY_TIME_MASK);
- *insn++ = BPF_JMP32_IMM(BPF_JEQ, tmp_reg, 0, 2);
- /* value_reg = BPF_SKB_DELIVERY_TIME_MONO */
- *insn++ = BPF_MOV32_IMM(value_reg, BPF_SKB_DELIVERY_TIME_MONO);
- *insn++ = BPF_JMP_A(IS_ENABLED(CONFIG_NET_CLS_ACT) ? 10 : 5);
-
- *insn++ = BPF_LDX_MEM(BPF_DW, tmp_reg, skb_reg,
- offsetof(struct sk_buff, tstamp));
- *insn++ = BPF_JMP_IMM(BPF_JNE, tmp_reg, 0, 2);
- /* value_reg = BPF_SKB_DELIVERY_TIME_NONE */
- *insn++ = BPF_MOV32_IMM(value_reg, BPF_SKB_DELIVERY_TIME_NONE);
- *insn++ = BPF_JMP_A(IS_ENABLED(CONFIG_NET_CLS_ACT) ? 6 : 1);
-
-#ifdef CONFIG_NET_CLS_ACT
- *insn++ = BPF_LDX_MEM(BPF_B, tmp_reg, skb_reg, PKT_VLAN_PRESENT_OFFSET);
- *insn++ = BPF_ALU32_IMM(BPF_AND, tmp_reg, TC_AT_INGRESS_MASK);
- *insn++ = BPF_JMP32_IMM(BPF_JEQ, tmp_reg, 0, 2);
- /* At ingress, value_reg = 0 */
- *insn++ = BPF_MOV32_IMM(value_reg, 0);
+ *insn++ = BPF_JMP32_IMM(BPF_JSET, tmp_reg,
+ SKB_MONO_DELIVERY_TIME_MASK, 2);
+ *insn++ = BPF_MOV32_IMM(value_reg, BPF_SKB_TSTAMP_UNSPEC);
*insn++ = BPF_JMP_A(1);
-#endif
-
- /* value_reg = BPF_SKB_DELIVERYT_TIME_UNSPEC */
- *insn++ = BPF_MOV32_IMM(value_reg, BPF_SKB_DELIVERY_TIME_UNSPEC);
+ *insn++ = BPF_MOV32_IMM(value_reg, BPF_SKB_TSTAMP_DELIVERY_MONO);
- /* 15 insns with CONFIG_NET_CLS_ACT */
return insn;
}
@@ -8956,11 +8936,11 @@ static struct bpf_insn *bpf_convert_tstamp_read(const struct bpf_prog *prog,
__u8 skb_reg = si->src_reg;
#ifdef CONFIG_NET_CLS_ACT
- /* If the delivery_time_type is read,
+ /* If the tstamp_type is read,
* the bpf prog is aware the tstamp could have delivery time.
- * Thus, read skb->tstamp as is if delivery_time_access is true.
+ * Thus, read skb->tstamp as is if tstamp_type_access is true.
*/
- if (!prog->delivery_time_access) {
+ if (!prog->tstamp_type_access) {
/* AX is needed because src_reg and dst_reg could be the same */
__u8 tmp_reg = BPF_REG_AX;
@@ -8990,13 +8970,13 @@ static struct bpf_insn *bpf_convert_tstamp_write(const struct bpf_prog *prog,
__u8 skb_reg = si->dst_reg;
#ifdef CONFIG_NET_CLS_ACT
- /* If the delivery_time_type is read,
+ /* If the tstamp_type is read,
* the bpf prog is aware the tstamp could have delivery time.
- * Thus, write skb->tstamp as is if delivery_time_access is true.
+ * Thus, write skb->tstamp as is if tstamp_type_access is true.
* Otherwise, writing at ingress will have to clear the
* mono_delivery_time bit also.
*/
- if (!prog->delivery_time_access) {
+ if (!prog->tstamp_type_access) {
__u8 tmp_reg = BPF_REG_AX;
*insn++ = BPF_LDX_MEM(BPF_B, tmp_reg, skb_reg, PKT_VLAN_PRESENT_OFFSET);
@@ -9329,8 +9309,8 @@ static u32 bpf_convert_ctx_access(enum bpf_access_type type,
insn = bpf_convert_tstamp_read(prog, si, insn);
break;
- case offsetof(struct __sk_buff, delivery_time_type):
- insn = bpf_convert_dtime_type_read(si, insn);
+ case offsetof(struct __sk_buff, tstamp_type):
+ insn = bpf_convert_tstamp_type_read(si, insn);
break;
case offsetof(struct __sk_buff, gso_segs):
diff --git a/tools/include/uapi/linux/bpf.h b/tools/include/uapi/linux/bpf.h
index 4eebea830613..65871bdbb9c8 100644
--- a/tools/include/uapi/linux/bpf.h
+++ b/tools/include/uapi/linux/bpf.h
@@ -5087,23 +5087,22 @@ union bpf_attr {
* 0 on success, or a negative error in case of failure. On error
* *dst* buffer is zeroed out.
*
- * long bpf_skb_set_delivery_time(struct sk_buff *skb, u64 dtime, u32 dtime_type)
+ * long bpf_skb_set_tstamp(struct sk_buff *skb, u64 tstamp, u32 tstamp_type)
* Description
- * Set a *dtime* (delivery time) to the __sk_buff->tstamp and also
- * change the __sk_buff->delivery_time_type to *dtime_type*.
+ * Change the __sk_buff->tstamp_type to *tstamp_type*
+ * and set *tstamp* to the __sk_buff->tstamp together.
*
- * When setting a delivery time (non zero *dtime*) to
- * __sk_buff->tstamp, only BPF_SKB_DELIVERY_TIME_MONO *dtime_type*
- * is supported. It is the only delivery_time_type that will be
- * kept after bpf_redirect_*().
- *
- * If there is no need to change the __sk_buff->delivery_time_type,
- * the delivery time can be directly written to __sk_buff->tstamp
+ * If there is no need to change the __sk_buff->tstamp_type,
+ * the tstamp value can be directly written to __sk_buff->tstamp
* instead.
*
- * *dtime* 0 and *dtime_type* BPF_SKB_DELIVERY_TIME_NONE
- * can be used to clear any delivery time stored in
- * __sk_buff->tstamp.
+ * BPF_SKB_TSTAMP_DELIVERY_MONO is the only tstamp that
+ * will be kept during bpf_redirect_*(). A non zero
+ * *tstamp* must be used with the BPF_SKB_TSTAMP_DELIVERY_MONO
+ * *tstamp_type*.
+ *
+ * A BPF_SKB_TSTAMP_UNSPEC *tstamp_type* can only be used
+ * with a zero *tstamp*.
*
* Only IPv4 and IPv6 skb->protocol are supported.
*
@@ -5116,7 +5115,7 @@ union bpf_attr {
* Return
* 0 on success.
* **-EINVAL** for invalid input
- * **-EOPNOTSUPP** for unsupported delivery_time_type and protocol
+ * **-EOPNOTSUPP** for unsupported protocol
*/
#define __BPF_FUNC_MAPPER(FN) \
FN(unspec), \
@@ -5311,7 +5310,7 @@ union bpf_attr {
FN(xdp_load_bytes), \
FN(xdp_store_bytes), \
FN(copy_from_user_task), \
- FN(skb_set_delivery_time), \
+ FN(skb_set_tstamp), \
/* */
/* integer value in 'imm' field of BPF_CALL instruction selects which helper
@@ -5502,9 +5501,12 @@ union { \
} __attribute__((aligned(8)))
enum {
- BPF_SKB_DELIVERY_TIME_NONE,
- BPF_SKB_DELIVERY_TIME_UNSPEC,
- BPF_SKB_DELIVERY_TIME_MONO,
+ BPF_SKB_TSTAMP_UNSPEC,
+ BPF_SKB_TSTAMP_DELIVERY_MONO, /* tstamp has mono delivery time */
+ /* For any BPF_SKB_TSTAMP_* that the bpf prog cannot handle,
+ * the bpf prog should handle it like BPF_SKB_TSTAMP_UNSPEC
+ * and try to deduce it by ingress, egress or skb->sk->sk_clockid.
+ */
};
/* user accessible mirror of in-kernel sk_buff.
@@ -5547,7 +5549,7 @@ struct __sk_buff {
__u32 gso_segs;
__bpf_md_ptr(struct bpf_sock *, sk);
__u32 gso_size;
- __u8 delivery_time_type;
+ __u8 tstamp_type;
__u32 :24; /* Padding, future use. */
__u64 hwtstamp;
};
--
2.30.2
^ permalink raw reply related [flat|nested] 7+ messages in thread* [PATCH bpf-next 5/5] bpf: selftests: Update tests after s/delivery_time/tstamp/ change in bpf.h
2022-03-09 9:04 [PATCH bpf-next 0/5] bpf: Follow up on bpf __sk_buff->tstamp Martin KaFai Lau
` (3 preceding siblings ...)
2022-03-09 9:05 ` [PATCH bpf-next 4/5] bpf: Remove BPF_SKB_DELIVERY_TIME_NONE and rename s/delivery_time_/tstamp_/ Martin KaFai Lau
@ 2022-03-09 9:05 ` Martin KaFai Lau
2022-03-10 22:00 ` [PATCH bpf-next 0/5] bpf: Follow up on bpf __sk_buff->tstamp patchwork-bot+netdevbpf
5 siblings, 0 replies; 7+ messages in thread
From: Martin KaFai Lau @ 2022-03-09 9:05 UTC (permalink / raw)
To: bpf
Cc: Alexei Starovoitov, Andrii Nakryiko, Daniel Borkmann, kernel-team,
netdev
The previous patch made the follow changes:
- s/delivery_time_type/tstamp_type/
- s/bpf_skb_set_delivery_time/bpf_skb_set_tstamp/
- BPF_SKB_DELIVERY_TIME_* to BPF_SKB_TSTAMP_*
This patch is to change the test_tc_dtime.c to reflect the above.
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
---
.../selftests/bpf/progs/test_tc_dtime.c | 38 +++++++++----------
1 file changed, 19 insertions(+), 19 deletions(-)
diff --git a/tools/testing/selftests/bpf/progs/test_tc_dtime.c b/tools/testing/selftests/bpf/progs/test_tc_dtime.c
index 9d9e8e17b8a0..06f300d06dbd 100644
--- a/tools/testing/selftests/bpf/progs/test_tc_dtime.c
+++ b/tools/testing/selftests/bpf/progs/test_tc_dtime.c
@@ -174,13 +174,13 @@ int egress_host(struct __sk_buff *skb)
return TC_ACT_OK;
if (skb_proto(skb_type) == IPPROTO_TCP) {
- if (skb->delivery_time_type == BPF_SKB_DELIVERY_TIME_MONO &&
+ if (skb->tstamp_type == BPF_SKB_TSTAMP_DELIVERY_MONO &&
skb->tstamp)
inc_dtimes(EGRESS_ENDHOST);
else
inc_errs(EGRESS_ENDHOST);
} else {
- if (skb->delivery_time_type == BPF_SKB_DELIVERY_TIME_UNSPEC &&
+ if (skb->tstamp_type == BPF_SKB_TSTAMP_UNSPEC &&
skb->tstamp)
inc_dtimes(EGRESS_ENDHOST);
else
@@ -204,7 +204,7 @@ int ingress_host(struct __sk_buff *skb)
if (!skb_type)
return TC_ACT_OK;
- if (skb->delivery_time_type == BPF_SKB_DELIVERY_TIME_MONO &&
+ if (skb->tstamp_type == BPF_SKB_TSTAMP_DELIVERY_MONO &&
skb->tstamp == EGRESS_FWDNS_MAGIC)
inc_dtimes(INGRESS_ENDHOST);
else
@@ -226,7 +226,7 @@ int ingress_fwdns_prio100(struct __sk_buff *skb)
return TC_ACT_OK;
/* delivery_time is only available to the ingress
- * if the tc-bpf checks the skb->delivery_time_type.
+ * if the tc-bpf checks the skb->tstamp_type.
*/
if (skb->tstamp == EGRESS_ENDHOST_MAGIC)
inc_errs(INGRESS_FWDNS_P100);
@@ -250,7 +250,7 @@ int egress_fwdns_prio100(struct __sk_buff *skb)
return TC_ACT_OK;
/* delivery_time is always available to egress even
- * the tc-bpf did not use the delivery_time_type.
+ * the tc-bpf did not use the tstamp_type.
*/
if (skb->tstamp == INGRESS_FWDNS_MAGIC)
inc_dtimes(EGRESS_FWDNS_P100);
@@ -278,9 +278,9 @@ int ingress_fwdns_prio101(struct __sk_buff *skb)
if (skb_proto(skb_type) == IPPROTO_UDP)
expected_dtime = 0;
- if (skb->delivery_time_type) {
+ if (skb->tstamp_type) {
if (fwdns_clear_dtime() ||
- skb->delivery_time_type != BPF_SKB_DELIVERY_TIME_MONO ||
+ skb->tstamp_type != BPF_SKB_TSTAMP_DELIVERY_MONO ||
skb->tstamp != expected_dtime)
inc_errs(INGRESS_FWDNS_P101);
else
@@ -290,14 +290,14 @@ int ingress_fwdns_prio101(struct __sk_buff *skb)
inc_errs(INGRESS_FWDNS_P101);
}
- if (skb->delivery_time_type == BPF_SKB_DELIVERY_TIME_MONO) {
+ if (skb->tstamp_type == BPF_SKB_TSTAMP_DELIVERY_MONO) {
skb->tstamp = INGRESS_FWDNS_MAGIC;
} else {
- if (bpf_skb_set_delivery_time(skb, INGRESS_FWDNS_MAGIC,
- BPF_SKB_DELIVERY_TIME_MONO))
+ if (bpf_skb_set_tstamp(skb, INGRESS_FWDNS_MAGIC,
+ BPF_SKB_TSTAMP_DELIVERY_MONO))
inc_errs(SET_DTIME);
- if (!bpf_skb_set_delivery_time(skb, INGRESS_FWDNS_MAGIC,
- BPF_SKB_DELIVERY_TIME_UNSPEC))
+ if (!bpf_skb_set_tstamp(skb, INGRESS_FWDNS_MAGIC,
+ BPF_SKB_TSTAMP_UNSPEC))
inc_errs(SET_DTIME);
}
@@ -320,9 +320,9 @@ int egress_fwdns_prio101(struct __sk_buff *skb)
/* Should have handled in prio100 */
return TC_ACT_SHOT;
- if (skb->delivery_time_type) {
+ if (skb->tstamp_type) {
if (fwdns_clear_dtime() ||
- skb->delivery_time_type != BPF_SKB_DELIVERY_TIME_MONO ||
+ skb->tstamp_type != BPF_SKB_TSTAMP_DELIVERY_MONO ||
skb->tstamp != INGRESS_FWDNS_MAGIC)
inc_errs(EGRESS_FWDNS_P101);
else
@@ -332,14 +332,14 @@ int egress_fwdns_prio101(struct __sk_buff *skb)
inc_errs(EGRESS_FWDNS_P101);
}
- if (skb->delivery_time_type == BPF_SKB_DELIVERY_TIME_MONO) {
+ if (skb->tstamp_type == BPF_SKB_TSTAMP_DELIVERY_MONO) {
skb->tstamp = EGRESS_FWDNS_MAGIC;
} else {
- if (bpf_skb_set_delivery_time(skb, EGRESS_FWDNS_MAGIC,
- BPF_SKB_DELIVERY_TIME_MONO))
+ if (bpf_skb_set_tstamp(skb, EGRESS_FWDNS_MAGIC,
+ BPF_SKB_TSTAMP_DELIVERY_MONO))
inc_errs(SET_DTIME);
- if (!bpf_skb_set_delivery_time(skb, EGRESS_FWDNS_MAGIC,
- BPF_SKB_DELIVERY_TIME_UNSPEC))
+ if (!bpf_skb_set_tstamp(skb, INGRESS_FWDNS_MAGIC,
+ BPF_SKB_TSTAMP_UNSPEC))
inc_errs(SET_DTIME);
}
--
2.30.2
^ permalink raw reply related [flat|nested] 7+ messages in thread* Re: [PATCH bpf-next 0/5] bpf: Follow up on bpf __sk_buff->tstamp
2022-03-09 9:04 [PATCH bpf-next 0/5] bpf: Follow up on bpf __sk_buff->tstamp Martin KaFai Lau
` (4 preceding siblings ...)
2022-03-09 9:05 ` [PATCH bpf-next 5/5] bpf: selftests: Update tests after s/delivery_time/tstamp/ change in bpf.h Martin KaFai Lau
@ 2022-03-10 22:00 ` patchwork-bot+netdevbpf
5 siblings, 0 replies; 7+ messages in thread
From: patchwork-bot+netdevbpf @ 2022-03-10 22:00 UTC (permalink / raw)
To: Martin KaFai Lau; +Cc: bpf, ast, andrii, daniel, kernel-team, netdev
Hello:
This series was applied to bpf/bpf-next.git (master)
by Daniel Borkmann <daniel@iogearbox.net>:
On Wed, 9 Mar 2022 01:04:44 -0800 you wrote:
> This set is a follow up on the bpf side based on discussion [0].
>
> Patch 1 is to remove some skbuff macros that are used in bpf filter.c
>
> Patch 2 and 3 are to simplify the bpf insn rewrite on __sk_buff->tstamp.
>
> Patch 4 is to simplify the bpf uapi by modeling the __sk_buff->tstamp
> and __sk_buff->tstamp_type (was delivery_time_type) the same as its kernel
> counter part skb->tstamp and skb->mono_delivery_time.
>
> [...]
Here is the summary with links:
- [bpf-next,1/5] bpf: net: Remove TC_AT_INGRESS_OFFSET and SKB_MONO_DELIVERY_TIME_OFFSET macro
https://git.kernel.org/bpf/bpf-next/c/3b5d4ddf8fe1
- [bpf-next,2/5] bpf: Simplify insn rewrite on BPF_READ __sk_buff->tstamp
https://git.kernel.org/bpf/bpf-next/c/539de9328e3a
- [bpf-next,3/5] bpf: Simplify insn rewrite on BPF_WRITE __sk_buff->tstamp
https://git.kernel.org/bpf/bpf-next/c/9d90db97e4d4
- [bpf-next,4/5] bpf: Remove BPF_SKB_DELIVERY_TIME_NONE and rename s/delivery_time_/tstamp_/
https://git.kernel.org/bpf/bpf-next/c/9bb984f28d5b
- [bpf-next,5/5] bpf: selftests: Update tests after s/delivery_time/tstamp/ change in bpf.h
https://git.kernel.org/bpf/bpf-next/c/3daf0896f3f9
You are awesome, thank you!
--
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html
^ permalink raw reply [flat|nested] 7+ messages in thread