Several conflicts, seemingly all over the place. I used Stephen Rothwell's sample resolutions for many of these, if not just to double check my own work, so definitely the credit largely goes to him. The NFP conflict consisted of a bug fix (moving operations past the rhashtable operation) while chaning the initial argument in the function call in the moved code. The net/dsa/master.c conflict had to do with a bug fix intermixing of making dsa_master_set_mtu() static with the fixing of the tagging attribute location. cls_flower had a conflict because the dup reject fix from Or overlapped with the addition of port range classifiction. __set_phy_supported()'s conflict was relatively easy to resolve because Andrew fixed it in both trees, so it was just a matter of taking the net-next copy. Or at least I think it was :-) Joe Stringer's fix to the handling of netns id 0 in bpf_sk_lookup() intermixed with changes on how the sdif and caller_net are calculated in these code paths in net-next. The remaining BPF conflicts were largely about the addition of the __bpf_md_ptr stuff in 'net' overlapping with adjustments and additions to the relevant data structure where the MD pointer macros are used.
Showing the changes to include/uapi/linux (2 of 576 files). Show all
Unified · Split
| @@ -2177,7 +2177,7 @@ union bpf_attr { · 2176 unmodified lines | ||
| 2177 | 2177 | * Return |
| 2178 | 2178 | * 0 on success, or a negative error in case of failure. |
| 2179 | 2179 | * |
| 2180 | − * struct bpf_sock *bpf_sk_lookup_tcp(void *ctx, struct bpf_sock_tuple *tuple, u32 tuple_size, u32 netns, u64 flags) | |
| 2180 | + * struct bpf_sock *bpf_sk_lookup_tcp(void *ctx, struct bpf_sock_tuple *tuple, u32 tuple_size, u64 netns, u64 flags) | |
| 2181 | 2181 | * Description |
| 2182 | 2182 | * Look for TCP socket matching *tuple*, optionally in a child |
| 2183 | 2183 | * network namespace *netns*. The return value must be checked, |
| @@ -2194,12 +2194,14 @@ union bpf_attr { · 10 unmodified lines | ||
| 2194 | 2194 | * **sizeof**\ (*tuple*\ **->ipv6**) |
| 2195 | 2195 | * Look for an IPv6 socket. |
| 2196 | 2196 | * |
| 2197 | − * If the *netns* is zero, then the socket lookup table in the | |
| 2198 | − * netns associated with the *ctx* will be used. For the TC hooks, | |
| 2199 | − * this in the netns of the device in the skb. For socket hooks, | |
| 2200 | − * this in the netns of the socket. If *netns* is non-zero, then | |
| 2201 | − * it specifies the ID of the netns relative to the netns | |
| 2202 | − * associated with the *ctx*. | |
| 2197 | + * If the *netns* is a negative signed 32-bit integer, then the | |
| 2198 | + * socket lookup table in the netns associated with the *ctx* will | |
| 2199 | + * will be used. For the TC hooks, this is the netns of the device | |
| 2200 | + * in the skb. For socket hooks, this is the netns of the socket. | |
| 2201 | + * If *netns* is any other signed 32-bit value greater than or | |
| 2202 | + * equal to zero then it specifies the ID of the netns relative to | |
| 2203 | + * the netns associated with the *ctx*. *netns* values beyond the | |
| 2204 | + * range of 32-bit integers are reserved for future use. | |
| 2203 | 2205 | * |
| 2204 | 2206 | * All values for *flags* are reserved for future usage, and must |
| 2205 | 2207 | * be left at zero. |
| @@ -2208,10 +2210,10 @@ union bpf_attr { · 2 unmodified lines | ||
| 2208 | 2210 | * **CONFIG_NET** configuration option. |
| 2209 | 2211 | * Return |
| 2210 | 2212 | * Pointer to *struct bpf_sock*, or NULL in case of failure. |
| 2211 | − * For sockets with reuseport option, *struct bpf_sock* | |
| 2212 | − * return is from reuse->socks[] using hash of the packet. | |
| 2213 | + * For sockets with reuseport option, the *struct bpf_sock* | |
| 2214 | + * result is from reuse->socks[] using the hash of the tuple. | |
| 2213 | 2215 | * |
| 2214 | − * struct bpf_sock *bpf_sk_lookup_udp(void *ctx, struct bpf_sock_tuple *tuple, u32 tuple_size, u32 netns, u64 flags) | |
| 2216 | + * struct bpf_sock *bpf_sk_lookup_udp(void *ctx, struct bpf_sock_tuple *tuple, u32 tuple_size, u64 netns, u64 flags) | |
| 2215 | 2217 | * Description |
| 2216 | 2218 | * Look for UDP socket matching *tuple*, optionally in a child |
| 2217 | 2219 | * network namespace *netns*. The return value must be checked, |
| @@ -2228,12 +2230,14 @@ union bpf_attr { · 10 unmodified lines | ||
| 2228 | 2230 | * **sizeof**\ (*tuple*\ **->ipv6**) |
| 2229 | 2231 | * Look for an IPv6 socket. |
| 2230 | 2232 | * |
| 2231 | − * If the *netns* is zero, then the socket lookup table in the | |
| 2232 | − * netns associated with the *ctx* will be used. For the TC hooks, | |
| 2233 | − * this in the netns of the device in the skb. For socket hooks, | |
| 2234 | − * this in the netns of the socket. If *netns* is non-zero, then | |
| 2235 | − * it specifies the ID of the netns relative to the netns | |
| 2236 | − * associated with the *ctx*. | |
| 2233 | + * If the *netns* is a negative signed 32-bit integer, then the | |
| 2234 | + * socket lookup table in the netns associated with the *ctx* will | |
| 2235 | + * will be used. For the TC hooks, this is the netns of the device | |
| 2236 | + * in the skb. For socket hooks, this is the netns of the socket. | |
| 2237 | + * If *netns* is any other signed 32-bit value greater than or | |
| 2238 | + * equal to zero then it specifies the ID of the netns relative to | |
| 2239 | + * the netns associated with the *ctx*. *netns* values beyond the | |
| 2240 | + * range of 32-bit integers are reserved for future use. | |
| 2237 | 2241 | * |
| 2238 | 2242 | * All values for *flags* are reserved for future usage, and must |
| 2239 | 2243 | * be left at zero. |
| @@ -2242,8 +2246,8 @@ union bpf_attr { · 2 unmodified lines | ||
| 2242 | 2246 | * **CONFIG_NET** configuration option. |
| 2243 | 2247 | * Return |
| 2244 | 2248 | * Pointer to *struct bpf_sock*, or NULL in case of failure. |
| 2245 | − * For sockets with reuseport option, *struct bpf_sock* | |
| 2246 | − * return is from reuse->socks[] using hash of the packet. | |
| 2249 | + * For sockets with reuseport option, the *struct bpf_sock* | |
| 2250 | + * result is from reuse->socks[] using the hash of the tuple. | |
| 2247 | 2251 | * |
| 2248 | 2252 | * int bpf_sk_release(struct bpf_sock *sk) |
| 2249 | 2253 | * Description |
| @@ -2430,6 +2434,9 @@ enum bpf_func_id { · 180 unmodified lines | ||
| 2430 | 2434 | /* BPF_FUNC_perf_event_output for sk_buff input context. */ |
| 2431 | 2435 | #define BPF_F_CTXLEN_MASK (0xfffffULL << 32) |
| 2432 | 2436 | |
| 2437 | +/* Current network namespace */ | |
| 2438 | +#define BPF_F_CURRENT_NETNS (-1L) | |
| 2439 | + | |
| 2433 | 2440 | /* Mode for BPF_FUNC_skb_adjust_room helper. */ |
| 2434 | 2441 | enum bpf_adj_room_mode { |
| 2435 | 2442 | BPF_ADJ_ROOM_NET, |
| @@ -2447,6 +2454,12 @@ enum bpf_lwt_encap_mode { · 11 unmodified lines | ||
| 2447 | 2454 | BPF_LWT_ENCAP_SEG6_INLINE |
| 2448 | 2455 | }; |
| 2449 | 2456 | |
| 2457 | +#define __bpf_md_ptr(type, name) \ | |
| 2458 | +union { \ | |
| 2459 | + type name; \ | |
| 2460 | + __u64 :64; \ | |
| 2461 | +} __attribute__((aligned(8))) | |
| 2462 | + | |
| 2450 | 2463 | /* user accessible mirror of in-kernel sk_buff. |
| 2451 | 2464 | * new fields can only be added to the end of this structure |
| 2452 | 2465 | */ |
| @@ -2481,7 +2494,7 @@ struct __sk_buff { · 28 unmodified lines | ||
| 2481 | 2494 | /* ... here. */ |
| 2482 | 2495 | |
| 2483 | 2496 | __u32 data_meta; |
| 2484 | − struct bpf_flow_keys *flow_keys; | |
| 2497 | + __bpf_md_ptr(struct bpf_flow_keys *, flow_keys); | |
| 2485 | 2498 | __u64 tstamp; |
| 2486 | 2499 | }; |
| 2487 | 2500 | |
| @@ -2598,8 +2611,8 @@ enum sk_action { · 110 unmodified lines | ||
| 2598 | 2611 | * be added to the end of this structure |
| 2599 | 2612 | */ |
| 2600 | 2613 | struct sk_msg_md { |
| 2601 | − void *data; | |
| 2602 | − void *data_end; | |
| 2614 | + __bpf_md_ptr(void *, data); | |
| 2615 | + __bpf_md_ptr(void *, data_end); | |
| 2603 | 2616 | |
| 2604 | 2617 | __u32 family; |
| 2605 | 2618 | __u32 remote_ip4; /* Stored in network byte order */ |
| @@ -2615,8 +2628,9 @@ struct sk_reuseport_md { · 9 unmodified lines | ||
| 2615 | 2628 | * Start of directly accessible data. It begins from |
| 2616 | 2629 | * the tcp/udp header. |
| 2617 | 2630 | */ |
| 2618 | − void *data; | |
| 2619 | − void *data_end; /* End of directly accessible data */ | |
| 2631 | + __bpf_md_ptr(void *, data); | |
| 2632 | + /* End of directly accessible data */ | |
| 2633 | + __bpf_md_ptr(void *, data_end); | |
| 2620 | 2634 | /* |
| 2621 | 2635 | * Total length of packet (starting from the tcp/udp header). |
| 2622 | 2636 | * Note that the directly accessible bytes (data_end - data) |
| @@ -212,6 +212,7 @@ struct prctl_mm_map { · 211 unmodified lines | ||
| 212 | 212 | #define PR_SET_SPECULATION_CTRL 53 |
| 213 | 213 | /* Speculation control variants */ |
| 214 | 214 | # define PR_SPEC_STORE_BYPASS 0 |
| 215 | +# define PR_SPEC_INDIRECT_BRANCH 1 | |
| 215 | 216 | /* Return and control values for PR_SET/GET_SPECULATION_CTRL */ |
| 216 | 217 | # define PR_SPEC_NOT_AFFECTED 0 |
| 217 | 218 | # define PR_SPEC_PRCTL (1UL << 0) |