[PATCH bpf-next v3 0/8] bpf: document eBPF helpers and add a script to generate man page

[PATCH bpf-next v3 0/8] bpf: document eBPF helpers and add a script to generate man page

[Quentin Monnet]

eBPF helper functions can be called from within eBPF programs to perform
a variety of tasks that would be otherwise hard or impossible to do with
eBPF itself. There is a growing number of such helper functions in the
kernel, but documentation is scarce. The main user space header file
does contain a short commented description of most helpers, but it is
somewhat outdated and not complete. It is more a "cheat sheet" than a
real documentation accessible to new eBPF developers.

This commit attempts to improve the situation by replacing the existing
overview for the helpers with a more developed description. Furthermore,
a Python script is added to generate a manual page for eBPF helpers. The
workflow is the following, and requires the rst2man utility:

    $ ./scripts/bpf_helpers_doc.py \
            --filename include/uapi/linux/bpf.h > /tmp/bpf-helpers.rst
    $ rst2man /tmp/bpf-helpers.rst > /tmp/bpf-helpers.7
    $ man /tmp/bpf-helpers.7

The objective is to keep all documentation related to the helpers in a
single place, and to be able to generate from here a manual page that
could be packaged in the man-pages repository and shipped with most
distributions.

Additionally, parsing the prototypes of the helper functions could
hopefully be reused, with a different Printer object, to generate
header files needed in some eBPF-related projects.

Regarding the description of each helper, it comprises several items:

- The function prototype.
- A description of the function and of its arguments (except for a
  couple of cases, when there are no arguments and the return value
  makes the function usage really obvious).
- A description of return values (if not void).

Additional items such as the list of compatible eBPF program and map
types for each helper, Linux kernel version that introduced the helper,
GPL-only restriction, and commit hash could be added in the future, but
it was decided on the mailing list to leave them aside for now.

For several helpers, descriptions are inspired (at times, nearly copied)
from the commit logs introducing them in the kernel--Many thanks to
their respective authors! They were completed as much as possible, the
objective being to have something easily accessible even for people just
starting with eBPF. There is probably a bit more work to do in this
direction for some helpers.

Some RST formatting is used in the descriptions (not in function
prototypes, to keep them readable, but the Python script provided in
order to generate the RST for the manual page does add formatting to
prototypes, to produce something pretty) to get "bold" and "italics" in
manual pages. Hopefully, the descriptions in bpf.h file remains
perfectly readable. Note that the few trailing white spaces are
intentional, removing them would break paragraphs for rst2man.

The descriptions should ideally be updated each time someone adds a new
helper, or updates the behaviour (new socket option supported, ...) or
the interface (new flags available, ...) of existing ones.

To ease the review process, the documentation has been split into several
patches.

RFC v2 -> PATCH v3:
Several fixes on man page header and footer, and helpers documentation.
Please refer to individual patches for details.

RFC v1 -> RFC v2:
- Remove "For" (compatible program and map types), "Since" (minimal
  Linux kernel version required), "GPL only" sections and commit hashes
  for the helpers.
- Add comment on top of the description list to explain how this
  documentation is supposed to be processed.
- Update Python script accordingly (remove the same sections, and remove
  paragraphs on program types and GPL restrictions from man page
  header).
- Split series into several patches.

Cc: linux-doc@vger.kernel.org
Cc: linux-man@vger.kernel.org

Quentin Monnet (8):
  bpf: add script and prepare bpf.h for new helpers documentation
  bpf: add documentation for eBPF helpers (01-11)
  bpf: add documentation for eBPF helpers (12-22)
  bpf: add documentation for eBPF helpers (23-32)
  bpf: add documentation for eBPF helpers (33-41)
  bpf: add documentation for eBPF helpers (42-50)
  bpf: add documentation for eBPF helpers (51-57)
  bpf: add documentation for eBPF helpers (58-64)

 include/uapi/linux/bpf.h   | 1604 +++++++++++++++++++++++++++++++++-----------
 scripts/bpf_helpers_doc.py |  434 ++++++++++++
 2 files changed, 1658 insertions(+), 380 deletions(-)
 create mode 100755 scripts/bpf_helpers_doc.py

-- 
2.14.1

[PATCH bpf-next v3 1/8] bpf: add script and prepare bpf.h for new helpers documentation

[Quentin Monnet]

Remove previous "overview" of eBPF helpers from user bpf.h header.
Replace it by a comment explaining how to process the new documentation
(to come in following patches) with a Python script to produce RST, then
man page documentation.

Also add the aforementioned Python script under scripts/. It is used to
process include/uapi/linux/bpf.h and to extract helper descriptions, to
turn it into a RST document that can further be processed with rst2man
to produce a man page. The script takes one "--filename <path/to/file>"
option. If the script is launched from scripts/ in the kernel root
directory, it should be able to find the location of the header to
parse, and "--filename <path/to/file>" is then optional. If it cannot
find the file, then the option becomes mandatory. RST-formatted
documentation is printed to standard output.

Typical workflow for producing the final man page would be:

    $ ./scripts/bpf_helpers_doc.py \
            --filename include/uapi/linux/bpf.h > /tmp/bpf-helpers.rst
    $ rst2man /tmp/bpf-helpers.rst > /tmp/bpf-helpers.7
    $ man /tmp/bpf-helpers.7

Note that the tool kernel-doc cannot be used to document eBPF helpers,
whose signatures are not available directly in the header files
(pre-processor directives are used to produce them at the beginning of
the compilation process).

v3:
- Change license for man page.
- Remove "for safety reasons" from man page header text.
- Change "packets metadata" to "packets" in man page header text.
- Move and fix comment on helpers introducing no overhead.
- Remove "NOTES" section from man page footer.
- Add "LICENSE" section to man page footer.
- Edit description of file include/uapi/linux/bpf.h in man page footer.

Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>
---
 include/uapi/linux/bpf.h   | 406 ++----------------------------------------
 scripts/bpf_helpers_doc.py | 434 +++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 450 insertions(+), 390 deletions(-)
 create mode 100755 scripts/bpf_helpers_doc.py

diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
index c5ec89732a8d..45f77f01e672 100644
--- a/include/uapi/linux/bpf.h
+++ b/include/uapi/linux/bpf.h
@@ -365,396 +365,22 @@ union bpf_attr {
 	} raw_tracepoint;
 } __attribute__((aligned(8)));
 
-/* BPF helper function descriptions:
- *
- * void *bpf_map_lookup_elem(&map, &key)
- *     Return: Map value or NULL
- *
- * int bpf_map_update_elem(&map, &key, &value, flags)
- *     Return: 0 on success or negative error
- *
- * int bpf_map_delete_elem(&map, &key)
- *     Return: 0 on success or negative error
- *
- * int bpf_probe_read(void *dst, int size, void *src)
- *     Return: 0 on success or negative error
- *
- * u64 bpf_ktime_get_ns(void)
- *     Return: current ktime
- *
- * int bpf_trace_printk(const char *fmt, int fmt_size, ...)
- *     Return: length of buffer written or negative error
- *
- * u32 bpf_prandom_u32(void)
- *     Return: random value
- *
- * u32 bpf_raw_smp_processor_id(void)
- *     Return: SMP processor ID
- *
- * int bpf_skb_store_bytes(skb, offset, from, len, flags)
- *     store bytes into packet
- *     @skb: pointer to skb
- *     @offset: offset within packet from skb->mac_header
- *     @from: pointer where to copy bytes from
- *     @len: number of bytes to store into packet
- *     @flags: bit 0 - if true, recompute skb->csum
- *             other bits - reserved
- *     Return: 0 on success or negative error
- *
- * int bpf_l3_csum_replace(skb, offset, from, to, flags)
- *     recompute IP checksum
- *     @skb: pointer to skb
- *     @offset: offset within packet where IP checksum is located
- *     @from: old value of header field
- *     @to: new value of header field
- *     @flags: bits 0-3 - size of header field
- *             other bits - reserved
- *     Return: 0 on success or negative error
- *
- * int bpf_l4_csum_replace(skb, offset, from, to, flags)
- *     recompute TCP/UDP checksum
- *     @skb: pointer to skb
- *     @offset: offset within packet where TCP/UDP checksum is located
- *     @from: old value of header field
- *     @to: new value of header field
- *     @flags: bits 0-3 - size of header field
- *             bit 4 - is pseudo header
- *             other bits - reserved
- *     Return: 0 on success or negative error
- *
- * int bpf_tail_call(ctx, prog_array_map, index)
- *     jump into another BPF program
- *     @ctx: context pointer passed to next program
- *     @prog_array_map: pointer to map which type is BPF_MAP_TYPE_PROG_ARRAY
- *     @index: 32-bit index inside array that selects specific program to run
- *     Return: 0 on success or negative error
- *
- * int bpf_clone_redirect(skb, ifindex, flags)
- *     redirect to another netdev
- *     @skb: pointer to skb
- *     @ifindex: ifindex of the net device
- *     @flags: bit 0 - if set, redirect to ingress instead of egress
- *             other bits - reserved
- *     Return: 0 on success or negative error
- *
- * u64 bpf_get_current_pid_tgid(void)
- *     Return: current->tgid << 32 | current->pid
- *
- * u64 bpf_get_current_uid_gid(void)
- *     Return: current_gid << 32 | current_uid
- *
- * int bpf_get_current_comm(char *buf, int size_of_buf)
- *     stores current->comm into buf
- *     Return: 0 on success or negative error
- *
- * u32 bpf_get_cgroup_classid(skb)
- *     retrieve a proc's classid
- *     @skb: pointer to skb
- *     Return: classid if != 0
- *
- * int bpf_skb_vlan_push(skb, vlan_proto, vlan_tci)
- *     Return: 0 on success or negative error
- *
- * int bpf_skb_vlan_pop(skb)
- *     Return: 0 on success or negative error
- *
- * int bpf_skb_get_tunnel_key(skb, key, size, flags)
- * int bpf_skb_set_tunnel_key(skb, key, size, flags)
- *     retrieve or populate tunnel metadata
- *     @skb: pointer to skb
- *     @key: pointer to 'struct bpf_tunnel_key'
- *     @size: size of 'struct bpf_tunnel_key'
- *     @flags: room for future extensions
- *     Return: 0 on success or negative error
- *
- * u64 bpf_perf_event_read(map, flags)
- *     read perf event counter value
- *     @map: pointer to perf_event_array map
- *     @flags: index of event in the map or bitmask flags
- *     Return: value of perf event counter read or error code
- *
- * int bpf_redirect(ifindex, flags)
- *     redirect to another netdev
- *     @ifindex: ifindex of the net device
- *     @flags:
- *	  cls_bpf:
- *          bit 0 - if set, redirect to ingress instead of egress
- *          other bits - reserved
- *	  xdp_bpf:
- *	    all bits - reserved
- *     Return: cls_bpf: TC_ACT_REDIRECT on success or TC_ACT_SHOT on error
- *	       xdp_bfp: XDP_REDIRECT on success or XDP_ABORT on error
- * int bpf_redirect_map(map, key, flags)
- *     redirect to endpoint in map
- *     @map: pointer to dev map
- *     @key: index in map to lookup
- *     @flags: --
- *     Return: XDP_REDIRECT on success or XDP_ABORT on error
- *
- * u32 bpf_get_route_realm(skb)
- *     retrieve a dst's tclassid
- *     @skb: pointer to skb
- *     Return: realm if != 0
- *
- * int bpf_perf_event_output(ctx, map, flags, data, size)
- *     output perf raw sample
- *     @ctx: struct pt_regs*
- *     @map: pointer to perf_event_array map
- *     @flags: index of event in the map or bitmask flags
- *     @data: data on stack to be output as raw data
- *     @size: size of data
- *     Return: 0 on success or negative error
- *
- * int bpf_get_stackid(ctx, map, flags)
- *     walk user or kernel stack and return id
- *     @ctx: struct pt_regs*
- *     @map: pointer to stack_trace map
- *     @flags: bits 0-7 - numer of stack frames to skip
- *             bit 8 - collect user stack instead of kernel
- *             bit 9 - compare stacks by hash only
- *             bit 10 - if two different stacks hash into the same stackid
- *                      discard old
- *             other bits - reserved
- *     Return: >= 0 stackid on success or negative error
- *
- * s64 bpf_csum_diff(from, from_size, to, to_size, seed)
- *     calculate csum diff
- *     @from: raw from buffer
- *     @from_size: length of from buffer
- *     @to: raw to buffer
- *     @to_size: length of to buffer
- *     @seed: optional seed
- *     Return: csum result or negative error code
- *
- * int bpf_skb_get_tunnel_opt(skb, opt, size)
- *     retrieve tunnel options metadata
- *     @skb: pointer to skb
- *     @opt: pointer to raw tunnel option data
- *     @size: size of @opt
- *     Return: option size
- *
- * int bpf_skb_set_tunnel_opt(skb, opt, size)
- *     populate tunnel options metadata
- *     @skb: pointer to skb
- *     @opt: pointer to raw tunnel option data
- *     @size: size of @opt
- *     Return: 0 on success or negative error
- *
- * int bpf_skb_change_proto(skb, proto, flags)
- *     Change protocol of the skb. Currently supported is v4 -> v6,
- *     v6 -> v4 transitions. The helper will also resize the skb. eBPF
- *     program is expected to fill the new headers via skb_store_bytes
- *     and lX_csum_replace.
- *     @skb: pointer to skb
- *     @proto: new skb->protocol type
- *     @flags: reserved
- *     Return: 0 on success or negative error
- *
- * int bpf_skb_change_type(skb, type)
- *     Change packet type of skb.
- *     @skb: pointer to skb
- *     @type: new skb->pkt_type type
- *     Return: 0 on success or negative error
- *
- * int bpf_skb_under_cgroup(skb, map, index)
- *     Check cgroup2 membership of skb
- *     @skb: pointer to skb
- *     @map: pointer to bpf_map in BPF_MAP_TYPE_CGROUP_ARRAY type
- *     @index: index of the cgroup in the bpf_map
- *     Return:
- *       == 0 skb failed the cgroup2 descendant test
- *       == 1 skb succeeded the cgroup2 descendant test
- *        < 0 error
- *
- * u32 bpf_get_hash_recalc(skb)
- *     Retrieve and possibly recalculate skb->hash.
- *     @skb: pointer to skb
- *     Return: hash
- *
- * u64 bpf_get_current_task(void)
- *     Returns current task_struct
- *     Return: current
- *
- * int bpf_probe_write_user(void *dst, void *src, int len)
- *     safely attempt to write to a location
- *     @dst: destination address in userspace
- *     @src: source address on stack
- *     @len: number of bytes to copy
- *     Return: 0 on success or negative error
- *
- * int bpf_current_task_under_cgroup(map, index)
- *     Check cgroup2 membership of current task
- *     @map: pointer to bpf_map in BPF_MAP_TYPE_CGROUP_ARRAY type
- *     @index: index of the cgroup in the bpf_map
- *     Return:
- *       == 0 current failed the cgroup2 descendant test
- *       == 1 current succeeded the cgroup2 descendant test
- *        < 0 error
- *
- * int bpf_skb_change_tail(skb, len, flags)
- *     The helper will resize the skb to the given new size, to be used f.e.
- *     with control messages.
- *     @skb: pointer to skb
- *     @len: new skb length
- *     @flags: reserved
- *     Return: 0 on success or negative error
- *
- * int bpf_skb_pull_data(skb, len)
- *     The helper will pull in non-linear data in case the skb is non-linear
- *     and not all of len are part of the linear section. Only needed for
- *     read/write with direct packet access.
- *     @skb: pointer to skb
- *     @len: len to make read/writeable
- *     Return: 0 on success or negative error
- *
- * s64 bpf_csum_update(skb, csum)
- *     Adds csum into skb->csum in case of CHECKSUM_COMPLETE.
- *     @skb: pointer to skb
- *     @csum: csum to add
- *     Return: csum on success or negative error
- *
- * void bpf_set_hash_invalid(skb)
- *     Invalidate current skb->hash.
- *     @skb: pointer to skb
- *
- * int bpf_get_numa_node_id()
- *     Return: Id of current NUMA node.
- *
- * int bpf_skb_change_head()
- *     Grows headroom of skb and adjusts MAC header offset accordingly.
- *     Will extends/reallocae as required automatically.
- *     May change skb data pointer and will thus invalidate any check
- *     performed for direct packet access.
- *     @skb: pointer to skb
- *     @len: length of header to be pushed in front
- *     @flags: Flags (unused for now)
- *     Return: 0 on success or negative error
- *
- * int bpf_xdp_adjust_head(xdp_md, delta)
- *     Adjust the xdp_md.data by delta
- *     @xdp_md: pointer to xdp_md
- *     @delta: An positive/negative integer to be added to xdp_md.data
- *     Return: 0 on success or negative on error
- *
- * int bpf_probe_read_str(void *dst, int size, const void *unsafe_ptr)
- *     Copy a NUL terminated string from unsafe address. In case the string
- *     length is smaller than size, the target is not padded with further NUL
- *     bytes. In case the string length is larger than size, just count-1
- *     bytes are copied and the last byte is set to NUL.
- *     @dst: destination address
- *     @size: maximum number of bytes to copy, including the trailing NUL
- *     @unsafe_ptr: unsafe address
- *     Return:
- *       > 0 length of the string including the trailing NUL on success
- *       < 0 error
- *
- * u64 bpf_get_socket_cookie(skb)
- *     Get the cookie for the socket stored inside sk_buff.
- *     @skb: pointer to skb
- *     Return: 8 Bytes non-decreasing number on success or 0 if the socket
- *     field is missing inside sk_buff
- *
- * u32 bpf_get_socket_uid(skb)
- *     Get the owner uid of the socket stored inside sk_buff.
- *     @skb: pointer to skb
- *     Return: uid of the socket owner on success or overflowuid if failed.
- *
- * u32 bpf_set_hash(skb, hash)
- *     Set full skb->hash.
- *     @skb: pointer to skb
- *     @hash: hash to set
- *
- * int bpf_setsockopt(bpf_socket, level, optname, optval, optlen)
- *     Calls setsockopt. Not all opts are available, only those with
- *     integer optvals plus TCP_CONGESTION.
- *     Supported levels: SOL_SOCKET and IPPROTO_TCP
- *     @bpf_socket: pointer to bpf_socket
- *     @level: SOL_SOCKET or IPPROTO_TCP
- *     @optname: option name
- *     @optval: pointer to option value
- *     @optlen: length of optval in bytes
- *     Return: 0 or negative error
- *
- * int bpf_getsockopt(bpf_socket, level, optname, optval, optlen)
- *     Calls getsockopt. Not all opts are available.
- *     Supported levels: IPPROTO_TCP
- *     @bpf_socket: pointer to bpf_socket
- *     @level: IPPROTO_TCP
- *     @optname: option name
- *     @optval: pointer to option value
- *     @optlen: length of optval in bytes
- *     Return: 0 or negative error
- *
- * int bpf_sock_ops_cb_flags_set(bpf_sock_ops, flags)
- *     Set callback flags for sock_ops
- *     @bpf_sock_ops: pointer to bpf_sock_ops_kern struct
- *     @flags: flags value
- *     Return: 0 for no error
- *             -EINVAL if there is no full tcp socket
- *             bits in flags that are not supported by current kernel
- *
- * int bpf_skb_adjust_room(skb, len_diff, mode, flags)
- *     Grow or shrink room in sk_buff.
- *     @skb: pointer to skb
- *     @len_diff: (signed) amount of room to grow/shrink
- *     @mode: operation mode (enum bpf_adj_room_mode)
- *     @flags: reserved for future use
- *     Return: 0 on success or negative error code
- *
- * int bpf_sk_redirect_map(map, key, flags)
- *     Redirect skb to a sock in map using key as a lookup key for the
- *     sock in map.
- *     @map: pointer to sockmap
- *     @key: key to lookup sock in map
- *     @flags: reserved for future use
- *     Return: SK_PASS
- *
- * int bpf_sock_map_update(skops, map, key, flags)
- *	@skops: pointer to bpf_sock_ops
- *	@map: pointer to sockmap to update
- *	@key: key to insert/update sock in map
- *	@flags: same flags as map update elem
- *
- * int bpf_xdp_adjust_meta(xdp_md, delta)
- *     Adjust the xdp_md.data_meta by delta
- *     @xdp_md: pointer to xdp_md
- *     @delta: An positive/negative integer to be added to xdp_md.data_meta
- *     Return: 0 on success or negative on error
- *
- * int bpf_perf_event_read_value(map, flags, buf, buf_size)
- *     read perf event counter value and perf event enabled/running time
- *     @map: pointer to perf_event_array map
- *     @flags: index of event in the map or bitmask flags
- *     @buf: buf to fill
- *     @buf_size: size of the buf
- *     Return: 0 on success or negative error code
- *
- * int bpf_perf_prog_read_value(ctx, buf, buf_size)
- *     read perf prog attached perf event counter and enabled/running time
- *     @ctx: pointer to ctx
- *     @buf: buf to fill
- *     @buf_size: size of the buf
- *     Return : 0 on success or negative error code
- *
- * int bpf_override_return(pt_regs, rc)
- *	@pt_regs: pointer to struct pt_regs
- *	@rc: the return value to set
- *
- * int bpf_msg_redirect_map(map, key, flags)
- *     Redirect msg to a sock in map using key as a lookup key for the
- *     sock in map.
- *     @map: pointer to sockmap
- *     @key: key to lookup sock in map
- *     @flags: reserved for future use
- *     Return: SK_PASS
- *
- * int bpf_bind(ctx, addr, addr_len)
- *     Bind socket to address. Only binding to IP is supported, no port can be
- *     set in addr.
- *     @ctx: pointer to context of type bpf_sock_addr
- *     @addr: pointer to struct sockaddr to bind socket to
- *     @addr_len: length of sockaddr structure
- *     Return: 0 on success or negative error code
+/* The description below is an attempt at providing documentation to eBPF
+ * developers about the multiple available eBPF helper functions. It can be
+ * parsed and used to produce a manual page. The workflow is the following,
+ * and requires the rst2man utility:
+ *
+ *     $ ./scripts/bpf_helpers_doc.py \
+ *             --filename include/uapi/linux/bpf.h > /tmp/bpf-helpers.rst
+ *     $ rst2man /tmp/bpf-helpers.rst > /tmp/bpf-helpers.7
+ *     $ man /tmp/bpf-helpers.7
+ *
+ * Note that in order to produce this external documentation, some RST
+ * formatting is used in the descriptions to get "bold" and "italics" in
+ * manual pages. Also note that the few trailing white spaces are
+ * intentional, removing them would break paragraphs for rst2man.
+ *
+ * Start of BPF helper function descriptions:
  */
 #define __BPF_FUNC_MAPPER(FN)		\
 	FN(unspec),			\
diff --git a/scripts/bpf_helpers_doc.py b/scripts/bpf_helpers_doc.py
new file mode 100755
index 000000000000..57857d597237
--- /dev/null
+++ b/scripts/bpf_helpers_doc.py
@@ -0,0 +1,434 @@
+#!/usr/bin/python3
+#
+# Copyright (C) 2018 Netronome Systems, Inc.
+#
+# This software is licensed under the GNU General License Version 2,
+# June 1991 as shown in the file COPYING in the top-level directory of this
+# source tree.
+#
+# THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS"
+# WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING,
+# BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
+# FOR A PARTICULAR PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE
+# OF THE PROGRAM IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME
+# THE COST OF ALL NECESSARY SERVICING, REPAIR OR CORRECTION.
+
+# In case user attempts to run with Python 2.
+from __future__ import print_function
+
+import argparse
+import re
+import sys, os
+
+class NoHelperFound(BaseException):
+    pass
+
+class ParsingError(BaseException):
+    def __init__(self, line='<line not provided>', reader=None):
+        if reader:
+            BaseException.__init__(self,
+                                   'Error at file offset %d, parsing line: %s' %
+                                   (reader.tell(), line))
+        else:
+            BaseException.__init__(self, 'Error parsing line: %s' % line)
+
+class Helper(object):
+    """
+    An object representing the description of an eBPF helper function.
+    @proto: function prototype of the helper function
+    @desc: textual description of the helper function
+    @ret: description of the return value of the helper function
+    """
+    def __init__(self, proto='', desc='', ret=''):
+        self.proto = proto
+        self.desc = desc
+        self.ret = ret
+
+    def proto_break_down(self):
+        """
+        Break down helper function protocol into smaller chunks: return type,
+        name, distincts arguments.
+        """
+        arg_re = re.compile('^((const )?(struct )?(\w+|...))( (\**)(\w+))?$')
+        res = {}
+        proto_re = re.compile('^(.+) (\**)(\w+)\(((([^,]+)(, )?){1,5})\)$')
+
+        capture = proto_re.match(self.proto)
+        res['ret_type'] = capture.group(1)
+        res['ret_star'] = capture.group(2)
+        res['name']     = capture.group(3)
+        res['args'] = []
+
+        args    = capture.group(4).split(', ')
+        for a in args:
+            capture = arg_re.match(a)
+            res['args'].append({
+                'type' : capture.group(1),
+                'star' : capture.group(6),
+                'name' : capture.group(7)
+            })
+
+        return res
+
+class HeaderParser(object):
+    """
+    An object used to parse a file in order to extract the documentation of a
+    list of eBPF helper functions. All the helpers that can be retrieved are
+    stored as Helper object, in the self.helpers() array.
+    @filename: name of file to parse, usually include/uapi/linux/bpf.h in the
+               kernel tree
+    """
+    def __init__(self, filename):
+        self.reader = open(filename, 'r')
+        self.line = ''
+        self.helpers = []
+
+    def parse_helper(self):
+        proto    = self.parse_proto()
+        desc     = self.parse_desc()
+        ret      = self.parse_ret()
+        return Helper(proto=proto, desc=desc, ret=ret)
+
+    def parse_proto(self):
+        # Argument can be of shape:
+        #   - "void"
+        #   - "type  name"
+        #   - "type *name"
+        #   - Same as above, with "const" and/or "struct" in front of type
+        #   - "..." (undefined number of arguments, for bpf_trace_printk())
+        # There is at least one term ("void"), and at most five arguments.
+        p = re.compile('^ \* ((.+) \**\w+\((((const )?(struct )?(\w+|\.\.\.)( \**\w+)?)(, )?){1,5}\))$')
+        capture = p.match(self.line)
+        if not capture:
+            raise NoHelperFound
+        self.line = self.reader.readline()
+        return capture.group(1)
+
+    def parse_desc(self):
+        p = re.compile('^ \* \tDescription$')
+        capture = p.match(self.line)
+        if not capture:
+            # Helper can have empty description and we might be parsing another
+            # attribute: return but do not consume.
+            return ''
+        # Description can be several lines, some of them possibly empty, and it
+        # stops when another subsection title is met.
+        desc = ''
+        while True:
+            self.line = self.reader.readline()
+            if self.line == ' *\n':
+                desc += '\n'
+            else:
+                p = re.compile('^ \* \t\t(.*)')
+                capture = p.match(self.line)
+                if capture:
+                    desc += capture.group(1) + '\n'
+                else:
+                    break
+        return desc
+
+    def parse_ret(self):
+        p = re.compile('^ \* \tReturn$')
+        capture = p.match(self.line)
+        if not capture:
+            # Helper can have empty retval and we might be parsing another
+            # attribute: return but do not consume.
+            return ''
+        # Return value description can be several lines, some of them possibly
+        # empty, and it stops when another subsection title is met.
+        ret = ''
+        while True:
+            self.line = self.reader.readline()
+            if self.line == ' *\n':
+                ret += '\n'
+            else:
+                p = re.compile('^ \* \t\t(.*)')
+                capture = p.match(self.line)
+                if capture:
+                    ret += capture.group(1) + '\n'
+                else:
+                    break
+        return ret
+
+    def run(self):
+        # Advance to start of helper function descriptions.
+        offset = self.reader.read().find('* Start of BPF helper function descriptions:')
+        if offset == -1:
+            raise Exception('Could not find start of eBPF helper descriptions list')
+        self.reader.seek(offset)
+        self.reader.readline()
+        self.reader.readline()
+        self.line = self.reader.readline()
+
+        while True:
+            try:
+                helper = self.parse_helper()
+                self.helpers.append(helper)
+            except NoHelperFound:
+                break
+
+        self.reader.close()
+        print('Parsed description of %d helper function(s)' % len(self.helpers),
+              file=sys.stderr)
+
+###############################################################################
+
+class Printer(object):
+    """
+    A generic class for printers. Printers should be created with an array of
+    Helper objects, and implement a way to print them in the desired fashion.
+    @helpers: array of Helper objects to print to standard output
+    """
+    def __init__(self, helpers):
+        self.helpers = helpers
+
+    def print_header(self):
+        pass
+
+    def print_footer(self):
+        pass
+
+    def print_one(self, helper):
+        pass
+
+    def print_all(self):
+        self.print_header()
+        for helper in self.helpers:
+            self.print_one(helper)
+        self.print_footer()
+
+class PrinterRST(Printer):
+    """
+    A printer for dumping collected information about helpers as a ReStructured
+    Text page compatible with the rst2man program, which can be used to
+    generate a manual page for the helpers.
+    @helpers: array of Helper objects to print to standard output
+    """
+    def print_header(self):
+        header = '''\
+.. Copyright (C) All BPF authors and contributors from 2014 to present.
+.. See git log include/uapi/linux/bpf.h in kernel tree for details.
+.. 
+.. %%%LICENSE_START(VERBATIM)
+.. Permission is granted to make and distribute verbatim copies of this
+.. manual provided the copyright notice and this permission notice are
+.. preserved on all copies.
+.. 
+.. Permission is granted to copy and distribute modified versions of this
+.. manual under the conditions for verbatim copying, provided that the
+.. entire resulting derived work is distributed under the terms of a
+.. permission notice identical to this one.
+.. 
+.. Since the Linux kernel and libraries are constantly changing, this
+.. manual page may be incorrect or out-of-date.  The author(s) assume no
+.. responsibility for errors or omissions, or for damages resulting from
+.. the use of the information contained herein.  The author(s) may not
+.. have taken the same level of care in the production of this manual,
+.. which is licensed free of charge, as they might when working
+.. professionally.
+.. 
+.. Formatted or processed versions of this manual, if unaccompanied by
+.. the source, must acknowledge the copyright and authors of this work.
+.. %%%LICENSE_END
+.. 
+.. Please do not edit this file. It was generated from the documentation
+.. located in file include/uapi/linux/bpf.h of the Linux kernel sources
+.. (helpers description), and from scripts/bpf_helpers_doc.py in the same
+.. repository (header and footer).
+
+===========
+BPF-HELPERS
+===========
+-------------------------------------------------------------------------------
+list of eBPF helper functions
+-------------------------------------------------------------------------------
+
+:Manual section: 7
+
+DESCRIPTION
+===========
+
+The extended Berkeley Packet Filter (eBPF) subsystem consists in programs
+written in a pseudo-assembly language, then attached to one of the several
+kernel hooks and run in reaction of specific events. This framework differs
+from the older, "classic" BPF (or "cBPF") in several aspects, one of them being
+the ability to call special functions (or "helpers") from within a program.
+These functions are restricted to a white-list of helpers defined in the
+kernel.
+
+These helpers are used by eBPF programs to interact with the system, or with
+the context in which they work. For instance, they can be used to print
+debugging messages, to get the time since the system was booted, to interact
+with eBPF maps, or to manipulate network packets. Since there are several eBPF
+program types, and that they do not run in the same context, each program type
+can only call a subset of those helpers.
+
+Due to eBPF conventions, a helper can not have more than five arguments.
+
+Internally, eBPF programs call directly into the compiled helper functions
+without requiring any foreign-function interface. As a result, calling helpers
+introduces no overhead, thus offering excellent performance.
+
+This document is an attempt to list and document the helpers available to eBPF
+developers. They are sorted by chronological order (the oldest helpers in the
+kernel at the top).
+
+HELPERS
+=======
+'''
+        print(header)
+
+    def print_footer(self):
+        footer = '''
+EXAMPLES
+========
+
+Example usage for most of the eBPF helpers listed in this manual page are
+available within the Linux kernel sources, at the following locations:
+
+* *samples/bpf/*
+* *tools/testing/selftests/bpf/*
+
+LICENSE
+=======
+
+Helpers related to tracing and monitoring functions are for "GPL-only" eBPF
+programs. These functions can be used to inspect the internals of the Linux
+kernel, which is under the GNU Privacy License (GPL): so eBPF programs must
+themselves be compatible with this license to be allowed to use such helpers
+(The licensing rules are the same as for kernel modules, so that also dual
+licenses, such as "Dual BSD/GPL", may be used).
+
+This requires the programs are loaded with the correct license string passed
+(via **attr**) to the **bpf**\ () system call, and generally translates into
+the C source code of such programs containing a line similar to the following:
+
+::
+
+	char ____license[] __attribute__((section("license"), used)) = "GPL";
+
+Helpers restricted to networking usage are free to use in non-GPL programs,
+because the packets on which they operate are just abstract bytes on the wire.
+
+IMPLEMENTATION
+==============
+
+This manual page is an effort to document the existing eBPF helper functions.
+But as of this writing, the BPF sub-system is under heavy development. New eBPF
+program or map types are added, along with new helper functions. Some helpers
+are occasionally made available for additional program types. So in spite of
+the efforts of the community, this page might not be up-to-date. If you want to
+check by yourself what helper functions exist in your kernel, or what types of
+programs they can support, here are some files among the kernel tree that you
+may be interested in:
+
+* *include/uapi/linux/bpf.h* is the main BPF header. It contains the full list
+  of all helper functions, as well as many other BPF definitions including most
+  of the flags, structs or constants used by the helpers.
+* *net/core/filter.c* contains the definition of most network-related helper
+  functions, and the list of program types from which they can be used.
+* *kernel/trace/bpf_trace.c* is the equivalent for most tracing program-related
+  helpers.
+* *kernel/bpf/verifier.c* contains the functions used to check that valid types
+  of eBPF maps are used with a given helper function.
+* *kernel/bpf/* directory contains other files in which additional helpers are
+  defined (for cgroups, sockmaps, etc.).
+
+Compatibility between helper functions and program types can generally be found
+in the files where helper functions are defined. Look for the **struct
+bpf_func_proto** objects and for functions returning them: these functions
+contain a list of helpers that a given program type can call. Note that the
+**default:** label of the **switch ... case** used to filter helpers can call
+other functions, themselves allowing access to additional helpers. The
+requirement for GPL license is also in those **struct bpf_func_proto**.
+
+Compatibility between helper functions and map types can be found in the
+**check_map_func_compatibility**\ () function in file *kernel/bpf/verifier.c*.
+
+Helper functions that invalidate the checks on **data** and **data_end**
+pointers for network processing are listed in function
+**bpf_helper_changes_pkt_data**\ () in file *net/core/filter.c*.
+
+SEE ALSO
+========
+
+**bpf**\ (2),
+**cgroups**\ (7),
+**ip**\ (8),
+**perf_event_open**\ (2),
+**sendmsg**\ (2),
+**socket**\ (7),
+**tc-bpf**\ (8)'''
+        print(footer)
+
+    def print_proto(self, helper):
+        """
+        Format function protocol with bold and italics markers. This makes RST
+        file less readable, but gives nice results in the manual page.
+        """
+        proto = helper.proto_break_down()
+
+        print('**%s %s%s(' % (proto['ret_type'],
+                              proto['ret_star'].replace('*', '\\*'),
+                              proto['name']),
+              end='')
+
+        comma = ''
+        for a in proto['args']:
+            one_arg = '{}{}'.format(comma, a['type'])
+            if a['name']:
+                if a['star']:
+                    one_arg += ' {}**\ '.format(a['star'].replace('*', '\\*'))
+                else:
+                    one_arg += '** '
+                one_arg += '*{}*\\ **'.format(a['name'])
+            comma = ', '
+            print(one_arg, end='')
+
+        print(')**')
+
+    def print_one(self, helper):
+        self.print_proto(helper)
+
+        if (helper.desc):
+            print('\tDescription')
+            # Do not strip all newline characters: formatted code at the end of
+            # a section must be followed by a blank line.
+            for line in re.sub('\n$', '', helper.desc, count=1).split('\n'):
+                print('{}{}'.format('\t\t' if line else '', line))
+
+        if (helper.ret):
+            print('\tReturn')
+            for line in helper.ret.rstrip().split('\n'):
+                print('{}{}'.format('\t\t' if line else '', line))
+
+        print('')
+
+###############################################################################
+
+# If script is launched from scripts/ from kernel tree and can access
+# ../include/uapi/linux/bpf.h, use it as a default name for the file to parse,
+# otherwise the --filename argument will be required from the command line.
+script = os.path.abspath(sys.argv[0])
+linuxRoot = os.path.dirname(os.path.dirname(script))
+bpfh = os.path.join(linuxRoot, 'include/uapi/linux/bpf.h')
+
+argParser = argparse.ArgumentParser(description="""
+Parse eBPF header file and generate documentation for eBPF helper functions.
+The RST-formatted output produced can be turned into a manual page with the
+rst2man utility.
+""")
+if (os.path.isfile(bpfh)):
+    argParser.add_argument('--filename', help='path to include/uapi/linux/bpf.h',
+                           default=bpfh)
+else:
+    argParser.add_argument('--filename', help='path to include/uapi/linux/bpf.h')
+args = argParser.parse_args()
+
+# Parse file.
+headerParser = HeaderParser(args.filename)
+headerParser.run()
+
+# Print formatted output to standard output.
+printer = PrinterRST(headerParser.helpers)
+printer.print_all()
-- 
2.14.1

[PATCH bpf-next v3 2/8] bpf: add documentation for eBPF helpers (01-11)

[Quentin Monnet]

Add documentation for eBPF helper functions to bpf.h user header file.
This documentation can be parsed with the Python script provided in
another commit of the patch series, in order to provide a RST document
that can later be converted into a man page.

The objective is to make the documentation easily understandable and
accessible to all eBPF developers, including beginners.

This patch contains descriptions for the following helper functions, all
written by Alexei:

- bpf_map_lookup_elem()
- bpf_map_update_elem()
- bpf_map_delete_elem()
- bpf_probe_read()
- bpf_ktime_get_ns()
- bpf_trace_printk()
- bpf_skb_store_bytes()
- bpf_l3_csum_replace()
- bpf_l4_csum_replace()
- bpf_tail_call()
- bpf_clone_redirect()

v3:
- bpf_map_lookup_elem(): Fix description of restrictions for flags
  related to the existence of the entry.
- bpf_trace_printk(): State that trace_pipe can be configured. Fix
  return value in case an unknown format specifier is met. Add a note on
  kernel log notice when the helper is used. Edit example.
- bpf_tail_call(): Improve comment on stack inheritance.
- bpf_clone_redirect(): Improve description of BPF_F_INGRESS flag.

Cc: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>
---
 include/uapi/linux/bpf.h | 210 +++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 210 insertions(+)

diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
index 45f77f01e672..02b7d522b3c0 100644
--- a/include/uapi/linux/bpf.h
+++ b/include/uapi/linux/bpf.h
@@ -381,6 +381,216 @@ union bpf_attr {
  * intentional, removing them would break paragraphs for rst2man.
  *
  * Start of BPF helper function descriptions:
+ *
+ * void *bpf_map_lookup_elem(struct bpf_map *map, void *key)
+ * 	Description
+ * 		Perform a lookup in *map* for an entry associated to *key*.
+ * 	Return
+ * 		Map value associated to *key*, or **NULL** if no entry was
+ * 		found.
+ *
+ * int bpf_map_update_elem(struct bpf_map *map, void *key, void *value, u64 flags)
+ * 	Description
+ * 		Add or update the value of the entry associated to *key* in
+ * 		*map* with *value*. *flags* is one of:
+ *
+ * 		**BPF_NOEXIST**
+ * 			The entry for *key* must not exist in the map.
+ * 		**BPF_EXIST**
+ * 			The entry for *key* must already exist in the map.
+ * 		**BPF_ANY**
+ * 			No condition on the existence of the entry for *key*.
+ *
+ * 		Flag value **BPF_NOEXIST** cannot be used for maps of types
+ * 		**BPF_MAP_TYPE_ARRAY** or **BPF_MAP_TYPE_PERCPU_ARRAY**  (all
+ * 		elements always exist), the helper would return an error.
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
+ *
+ * int bpf_map_delete_elem(struct bpf_map *map, void *key)
+ * 	Description
+ * 		Delete entry with *key* from *map*.
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
+ *
+ * int bpf_probe_read(void *dst, u32 size, const void *src)
+ * 	Description
+ * 		For tracing programs, safely attempt to read *size* bytes from
+ * 		address *src* and store the data in *dst*.
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
+ *
+ * u64 bpf_ktime_get_ns(void)
+ * 	Description
+ * 		Return the time elapsed since system boot, in nanoseconds.
+ * 	Return
+ * 		Current *ktime*.
+ *
+ * int bpf_trace_printk(const char *fmt, u32 fmt_size, ...)
+ * 	Description
+ * 		This helper is a "printk()-like" facility for debugging. It
+ * 		prints a message defined by format *fmt* (of size *fmt_size*)
+ * 		to file *\/sys/kernel/debug/tracing/trace* from DebugFS, if
+ * 		available. It can take up to three additional **u64**
+ * 		arguments (as an eBPF helpers, the total number of arguments is
+ * 		limited to five).
+ *
+ * 		Each time the helper is called, it appends a line to the trace.
+ * 		The format of the trace is customizable, and the exact output
+ * 		one will get depends on the options set in
+ * 		*\/sys/kernel/debug/tracing/trace_options* (see also the
+ * 		*README* file under the same directory). However, it usually
+ * 		defaults to something like:
+ *
+ * 		::
+ *
+ * 			telnet-470   [001] .N.. 419421.045894: 0x00000001: <formatted msg>
+ *
+ * 		In the above:
+ *
+ * 			* ``telnet`` is the name of the current task.
+ * 			* ``470`` is the PID of the current task.
+ * 			* ``001`` is the CPU number on which the task is
+ * 			  running.
+ * 			* In ``.N..``, each character refers to a set of
+ * 			  options (whether irqs are enabled, scheduling
+ * 			  options, whether hard/softirqs are running, level of
+ * 			  preempt_disabled respectively). **N** means that
+ * 			  **TIF_NEED_RESCHED** and **PREEMPT_NEED_RESCHED**
+ * 			  are set.
+ * 			* ``419421.045894`` is a timestamp.
+ * 			* ``0x00000001`` is a fake value used by BPF for the
+ * 			  instruction pointer register.
+ * 			* ``<formatted msg>`` is the message formatted with
+ * 			  *fmt*.
+ *
+ * 		The conversion specifiers supported by *fmt* are similar, but
+ * 		more limited than for printk(). They are **%d**, **%i**,
+ * 		**%u**, **%x**, **%ld**, **%li**, **%lu**, **%lx**, **%lld**,
+ * 		**%lli**, **%llu**, **%llx**, **%p**, **%s**. No modifier (size
+ * 		of field, padding with zeroes, etc.) is available, and the
+ * 		helper will return **-EINVAL** (but print nothing) if it
+ * 		encounters an unknown specifier.
+ *
+ * 		Also, note that **bpf_trace_printk**\ () is slow, and should
+ * 		only be used for debugging purposes. For this reason, a notice
+ * 		bloc (spanning several lines) is printed to kernel logs and
+ * 		states that the helper should not be used "for production use"
+ * 		the first time this helper is used (or more precisely, when
+ * 		**trace_printk**\ () buffers are allocated). For passing values
+ * 		to user space, perf events should be preferred.
+ * 	Return
+ * 		The number of bytes written to the buffer, or a negative error
+ * 		in case of failure.
+ *
+ * int bpf_skb_store_bytes(struct sk_buff *skb, u32 offset, const void *from, u32 len, u64 flags)
+ * 	Description
+ * 		Store *len* bytes from address *from* into the packet
+ * 		associated to *skb*, at *offset*. *flags* are a combination of
+ * 		**BPF_F_RECOMPUTE_CSUM** (automatically recompute the
+ * 		checksum for the packet after storing the bytes) and
+ * 		**BPF_F_INVALIDATE_HASH** (set *skb*\ **->hash**, *skb*\
+ * 		**->swhash** and *skb*\ **->l4hash** to 0).
+ *
+ * 		A call to this helper is susceptible to change data from the
+ * 		packet. Therefore, at load time, all checks on pointers
+ * 		previously done by the verifier are invalidated and must be
+ * 		performed again.
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
+ *
+ * int bpf_l3_csum_replace(struct sk_buff *skb, u32 offset, u64 from, u64 to, u64 size)
+ * 	Description
+ * 		Recompute the IP checksum for the packet associated to *skb*.
+ * 		Computation is incremental, so the helper must know the former
+ * 		value of the header field that was modified (*from*), the new
+ * 		value of this field (*to*), and the number of bytes (2 or 4)
+ * 		for this field, stored in *size*. Alternatively, it is possible
+ * 		to store the difference between the previous and the new values
+ * 		of the header field in *to*, by setting *from* and *size* to 0.
+ * 		For both methods, *offset* indicates the location of the IP
+ * 		checksum within the packet.
+ *
+ * 		A call to this helper is susceptible to change data from the
+ * 		packet. Therefore, at load time, all checks on pointers
+ * 		previously done by the verifier are invalidated and must be
+ * 		performed again.
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
+ *
+ * int bpf_l4_csum_replace(struct sk_buff *skb, u32 offset, u64 from, u64 to, u64 flags)
+ * 	Description
+ * 		Recompute the TCP or UDP checksum for the packet associated to
+ * 		*skb*. Computation is incremental, so the helper must know the
+ * 		former value of the header field that was modified (*from*),
+ * 		the new value of this field (*to*), and the number of bytes (2
+ * 		or 4) for this field, stored on the lowest four bits of
+ * 		*flags*. Alternatively, it is possible to store the difference
+ * 		between the previous and the new values of the header field in
+ * 		*to*, by setting *from* and the four lowest bits of *flags* to
+ * 		0. For both methods, *offset* indicates the location of the IP
+ * 		checksum within the packet. In addition to the size of the
+ * 		field, *flags* can be added (bitwise OR) actual flags. With
+ * 		**BPF_F_MARK_MANGLED_0**, a null checksum is left untouched
+ * 		(unless **BPF_F_MARK_ENFORCE** is added as well), and for
+ * 		updates resulting in a null checksum the value is set to
+ * 		**CSUM_MANGLED_0** instead. Flag **BPF_F_PSEUDO_HDR**
+ * 		indicates the checksum is to be computed against a
+ * 		pseudo-header.
+ *
+ * 		A call to this helper is susceptible to change data from the
+ * 		packet. Therefore, at load time, all checks on pointers
+ * 		previously done by the verifier are invalidated and must be
+ * 		performed again.
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
+ *
+ * int bpf_tail_call(void *ctx, struct bpf_map *prog_array_map, u32 index)
+ * 	Description
+ * 		This special helper is used to trigger a "tail call", or in
+ * 		other words, to jump into another eBPF program. The same stack
+ * 		frame is used (but values on stack and in registers for the
+ * 		caller are not accessible to the callee). This mechanism allows
+ * 		for program chaining, either for raising the maximum number of
+ * 		available eBPF instructions, or to execute given programs in
+ * 		conditional blocks. For security reasons, there is an upper
+ * 		limit to the number of successive tail calls that can be
+ * 		performed.
+ *
+ * 		Upon call of this helper, the program attempts to jump into a
+ * 		program referenced at index *index* in *prog_array_map*, a
+ * 		special map of type **BPF_MAP_TYPE_PROG_ARRAY**, and passes
+ * 		*ctx*, a pointer to the context.
+ *
+ * 		If the call succeeds, the kernel immediately runs the first
+ * 		instruction of the new program. This is not a function call,
+ * 		and it never goes back to the previous program. If the call
+ * 		fails, then the helper has no effect, and the caller continues
+ * 		to run its own instructions. A call can fail if the destination
+ * 		program for the jump does not exist (i.e. *index* is superior
+ * 		to the number of entries in *prog_array_map*), or if the
+ * 		maximum number of tail calls has been reached for this chain of
+ * 		programs. This limit is defined in the kernel by the macro
+ * 		**MAX_TAIL_CALL_CNT** (not accessible to user space), which
+ * 		is currently set to 32.
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
+ *
+ * int bpf_clone_redirect(struct sk_buff *skb, u32 ifindex, u64 flags)
+ * 	Description
+ * 		Clone and redirect the packet associated to *skb* to another
+ * 		net device of index *ifindex*. Both ingress and egress
+ * 		interfaces can be used for redirection. The **BPF_F_INGRESS**
+ * 		value in *flags* is used to make the distinction (ingress path
+ * 		is selected if the flag is present, egress path otherwise).
+ * 		This is the only flag supported for now.
+ *
+ * 		A call to this helper is susceptible to change data from the
+ * 		packet. Therefore, at load time, all checks on pointers
+ * 		previously done by the verifier are invalidated and must be
+ * 		performed again.
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
  */
 #define __BPF_FUNC_MAPPER(FN)		\
 	FN(unspec),			\
-- 
2.14.1

Re: [PATCH bpf-next v3 2/8] bpf: add documentation for eBPF helpers (01-11)

[Alexei Starovoitov]

On Tue, Apr 17, 2018 at 03:34:32PM +0100, Quentin Monnet wrote:
> Add documentation for eBPF helper functions to bpf.h user header file.
> This documentation can be parsed with the Python script provided in
> another commit of the patch series, in order to provide a RST document
> that can later be converted into a man page.
> 
> The objective is to make the documentation easily understandable and
> accessible to all eBPF developers, including beginners.
> 
> This patch contains descriptions for the following helper functions, all
> written by Alexei:
> 
> - bpf_map_lookup_elem()
> - bpf_map_update_elem()
> - bpf_map_delete_elem()
> - bpf_probe_read()
> - bpf_ktime_get_ns()
> - bpf_trace_printk()
> - bpf_skb_store_bytes()
> - bpf_l3_csum_replace()
> - bpf_l4_csum_replace()
> - bpf_tail_call()
> - bpf_clone_redirect()
> 
> v3:
> - bpf_map_lookup_elem(): Fix description of restrictions for flags
>   related to the existence of the entry.
> - bpf_trace_printk(): State that trace_pipe can be configured. Fix
>   return value in case an unknown format specifier is met. Add a note on
>   kernel log notice when the helper is used. Edit example.
> - bpf_tail_call(): Improve comment on stack inheritance.
> - bpf_clone_redirect(): Improve description of BPF_F_INGRESS flag.
> 
> Cc: Alexei Starovoitov <ast@kernel.org>
> Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>

Acked-by: Alexei Starovoitov <ast@kernel.org>

Re: [PATCH bpf-next v3 2/8] bpf: add documentation for eBPF helpers (01-11)

[Daniel Borkmann]

On 04/17/2018 04:34 PM, Quentin Monnet wrote:
> Add documentation for eBPF helper functions to bpf.h user header file.
> This documentation can be parsed with the Python script provided in
> another commit of the patch series, in order to provide a RST document
> that can later be converted into a man page.
> 
> The objective is to make the documentation easily understandable and
> accessible to all eBPF developers, including beginners.
> 
> This patch contains descriptions for the following helper functions, all
> written by Alexei:
> 
> - bpf_map_lookup_elem()
> - bpf_map_update_elem()
> - bpf_map_delete_elem()
> - bpf_probe_read()
> - bpf_ktime_get_ns()
> - bpf_trace_printk()
> - bpf_skb_store_bytes()
> - bpf_l3_csum_replace()
> - bpf_l4_csum_replace()
> - bpf_tail_call()
> - bpf_clone_redirect()
> 
> v3:
> - bpf_map_lookup_elem(): Fix description of restrictions for flags
>   related to the existence of the entry.
> - bpf_trace_printk(): State that trace_pipe can be configured. Fix
>   return value in case an unknown format specifier is met. Add a note on
>   kernel log notice when the helper is used. Edit example.
> - bpf_tail_call(): Improve comment on stack inheritance.
> - bpf_clone_redirect(): Improve description of BPF_F_INGRESS flag.
> 
> Cc: Alexei Starovoitov <ast@kernel.org>
> Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>

Thanks for doing all this work, Quentin!

Just some small improvements while reading over it:

> ---
>  include/uapi/linux/bpf.h | 210 +++++++++++++++++++++++++++++++++++++++++++++++
>  1 file changed, 210 insertions(+)
> 
> diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
> index 45f77f01e672..02b7d522b3c0 100644
> --- a/include/uapi/linux/bpf.h
> +++ b/include/uapi/linux/bpf.h
> @@ -381,6 +381,216 @@ union bpf_attr {
>   * intentional, removing them would break paragraphs for rst2man.
>   *
>   * Start of BPF helper function descriptions:
> + *
> + * void *bpf_map_lookup_elem(struct bpf_map *map, void *key)

const void *key

> + * 	Description
> + * 		Perform a lookup in *map* for an entry associated to *key*.
> + * 	Return
> + * 		Map value associated to *key*, or **NULL** if no entry was
> + * 		found.
> + *
> + * int bpf_map_update_elem(struct bpf_map *map, void *key, void *value, u64 flags)

const void *key, const void *value

> + * 	Description
> + * 		Add or update the value of the entry associated to *key* in
> + * 		*map* with *value*. *flags* is one of:
> + *
> + * 		**BPF_NOEXIST**
> + * 			The entry for *key* must not exist in the map.
> + * 		**BPF_EXIST**
> + * 			The entry for *key* must already exist in the map.
> + * 		**BPF_ANY**
> + * 			No condition on the existence of the entry for *key*.
> + *
> + * 		Flag value **BPF_NOEXIST** cannot be used for maps of types
> + * 		**BPF_MAP_TYPE_ARRAY** or **BPF_MAP_TYPE_PERCPU_ARRAY**  (all
> + * 		elements always exist), the helper would return an error.
> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
> + *
> + * int bpf_map_delete_elem(struct bpf_map *map, void *key)

const void *key

> + * 	Description
> + * 		Delete entry with *key* from *map*.
> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
> + *
> + * int bpf_probe_read(void *dst, u32 size, const void *src)
> + * 	Description
> + * 		For tracing programs, safely attempt to read *size* bytes from
> + * 		address *src* and store the data in *dst*.
> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
> + *
> + * u64 bpf_ktime_get_ns(void)
> + * 	Description
> + * 		Return the time elapsed since system boot, in nanoseconds.
> + * 	Return
> + * 		Current *ktime*.
> + *
> + * int bpf_trace_printk(const char *fmt, u32 fmt_size, ...)
> + * 	Description
> + * 		This helper is a "printk()-like" facility for debugging. It
> + * 		prints a message defined by format *fmt* (of size *fmt_size*)
> + * 		to file *\/sys/kernel/debug/tracing/trace* from DebugFS, if
> + * 		available. It can take up to three additional **u64**
> + * 		arguments (as an eBPF helpers, the total number of arguments is
> + * 		limited to five).
> + *
> + * 		Each time the helper is called, it appends a line to the trace.
> + * 		The format of the trace is customizable, and the exact output
> + * 		one will get depends on the options set in
> + * 		*\/sys/kernel/debug/tracing/trace_options* (see also the
> + * 		*README* file under the same directory). However, it usually
> + * 		defaults to something like:
> + *
> + * 		::
> + *
> + * 			telnet-470   [001] .N.. 419421.045894: 0x00000001: <formatted msg>
> + *
> + * 		In the above:
> + *
> + * 			* ``telnet`` is the name of the current task.
> + * 			* ``470`` is the PID of the current task.
> + * 			* ``001`` is the CPU number on which the task is
> + * 			  running.
> + * 			* In ``.N..``, each character refers to a set of
> + * 			  options (whether irqs are enabled, scheduling
> + * 			  options, whether hard/softirqs are running, level of
> + * 			  preempt_disabled respectively). **N** means that
> + * 			  **TIF_NEED_RESCHED** and **PREEMPT_NEED_RESCHED**
> + * 			  are set.
> + * 			* ``419421.045894`` is a timestamp.
> + * 			* ``0x00000001`` is a fake value used by BPF for the
> + * 			  instruction pointer register.
> + * 			* ``<formatted msg>`` is the message formatted with
> + * 			  *fmt*.
> + *
> + * 		The conversion specifiers supported by *fmt* are similar, but
> + * 		more limited than for printk(). They are **%d**, **%i**,
> + * 		**%u**, **%x**, **%ld**, **%li**, **%lu**, **%lx**, **%lld**,
> + * 		**%lli**, **%llu**, **%llx**, **%p**, **%s**. No modifier (size
> + * 		of field, padding with zeroes, etc.) is available, and the
> + * 		helper will return **-EINVAL** (but print nothing) if it
> + * 		encounters an unknown specifier.
> + *
> + * 		Also, note that **bpf_trace_printk**\ () is slow, and should
> + * 		only be used for debugging purposes. For this reason, a notice
> + * 		bloc (spanning several lines) is printed to kernel logs and
> + * 		states that the helper should not be used "for production use"
> + * 		the first time this helper is used (or more precisely, when
> + * 		**trace_printk**\ () buffers are allocated). For passing values
> + * 		to user space, perf events should be preferred.
> + * 	Return
> + * 		The number of bytes written to the buffer, or a negative error
> + * 		in case of failure.
> + *
> + * int bpf_skb_store_bytes(struct sk_buff *skb, u32 offset, const void *from, u32 len, u64 flags)
> + * 	Description
> + * 		Store *len* bytes from address *from* into the packet
> + * 		associated to *skb*, at *offset*. *flags* are a combination of
> + * 		**BPF_F_RECOMPUTE_CSUM** (automatically recompute the
> + * 		checksum for the packet after storing the bytes) and
> + * 		**BPF_F_INVALIDATE_HASH** (set *skb*\ **->hash**, *skb*\
> + * 		**->swhash** and *skb*\ **->l4hash** to 0).
> + *
> + * 		A call to this helper is susceptible to change data from the
> + * 		packet. Therefore, at load time, all checks on pointers

Perhaps: to change the underlying packet buffer (it's not the data itself but
a potential reallocation to e.g. unclone the skb which otherwise would lead to
a use-after-free if not forced to be invalidated from verifier).

> + * 		previously done by the verifier are invalidated and must be
> + * 		performed again.

I would add something like "if used in combination with direct packet access"
where it's the only place this comment is relevant.

> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
> + *
> + * int bpf_l3_csum_replace(struct sk_buff *skb, u32 offset, u64 from, u64 to, u64 size)
> + * 	Description
> + * 		Recompute the IP checksum for the packet associated to *skb*.

Perhaps we could say something like 'L3 (e.g. IP)'. The helper itself has zero
knowledge of the underlying protocol used. It's solely replacing the csum and
updating skb->csum when necessary.

> + * 		Computation is incremental, so the helper must know the former
> + * 		value of the header field that was modified (*from*), the new
> + * 		value of this field (*to*), and the number of bytes (2 or 4)
> + * 		for this field, stored in *size*. Alternatively, it is possible
> + * 		to store the difference between the previous and the new values
> + * 		of the header field in *to*, by setting *from* and *size* to 0.

I would add that this works in combination with csum_diff() helper, allows for
more flexibility and to handle sizes larger than 2 or 4.

> + * 		For both methods, *offset* indicates the location of the IP
> + * 		checksum within the packet.
> + *
> + * 		A call to this helper is susceptible to change data from the
> + * 		packet. Therefore, at load time, all checks on pointers
> + * 		previously done by the verifier are invalidated and must be
> + * 		performed again.
> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
> + *
> + * int bpf_l4_csum_replace(struct sk_buff *skb, u32 offset, u64 from, u64 to, u64 flags)
> + * 	Description
> + * 		Recompute the TCP or UDP checksum for the packet associated to

See prior comment, L4 (e.g. TCP, UDP or ICMP).

> + * 		*skb*. Computation is incremental, so the helper must know the
> + * 		former value of the header field that was modified (*from*),
> + * 		the new value of this field (*to*), and the number of bytes (2
> + * 		or 4) for this field, stored on the lowest four bits of
> + * 		*flags*. Alternatively, it is possible to store the difference
> + * 		between the previous and the new values of the header field in
> + * 		*to*, by setting *from* and the four lowest bits of *flags* to

Same reference for csum_diff().

> + * 		0. For both methods, *offset* indicates the location of the IP
> + * 		checksum within the packet. In addition to the size of the
> + * 		field, *flags* can be added (bitwise OR) actual flags. With
> + * 		**BPF_F_MARK_MANGLED_0**, a null checksum is left untouched
> + * 		(unless **BPF_F_MARK_ENFORCE** is added as well), and for
> + * 		updates resulting in a null checksum the value is set to
> + * 		**CSUM_MANGLED_0** instead. Flag **BPF_F_PSEUDO_HDR**
> + * 		indicates the checksum is to be computed against a
> + * 		pseudo-header.
> + *
> + * 		A call to this helper is susceptible to change data from the
> + * 		packet. Therefore, at load time, all checks on pointers
> + * 		previously done by the verifier are invalidated and must be
> + * 		performed again.
> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
> + *
> + * int bpf_tail_call(void *ctx, struct bpf_map *prog_array_map, u32 index)
> + * 	Description
> + * 		This special helper is used to trigger a "tail call", or in
> + * 		other words, to jump into another eBPF program. The same stack
> + * 		frame is used (but values on stack and in registers for the
> + * 		caller are not accessible to the callee). This mechanism allows
> + * 		for program chaining, either for raising the maximum number of
> + * 		available eBPF instructions, or to execute given programs in
> + * 		conditional blocks. For security reasons, there is an upper
> + * 		limit to the number of successive tail calls that can be
> + * 		performed.
> + *
> + * 		Upon call of this helper, the program attempts to jump into a
> + * 		program referenced at index *index* in *prog_array_map*, a
> + * 		special map of type **BPF_MAP_TYPE_PROG_ARRAY**, and passes
> + * 		*ctx*, a pointer to the context.
> + *
> + * 		If the call succeeds, the kernel immediately runs the first
> + * 		instruction of the new program. This is not a function call,
> + * 		and it never goes back to the previous program. If the call

Nit: s/goes back/returns/

> + * 		fails, then the helper has no effect, and the caller continues
> + * 		to run its own instructions. A call can fail if the destination

Maybe: s/to run its own/to run its subsequent/

> + * 		program for the jump does not exist (i.e. *index* is superior
> + * 		to the number of entries in *prog_array_map*), or if the
> + * 		maximum number of tail calls has been reached for this chain of
> + * 		programs. This limit is defined in the kernel by the macro
> + * 		**MAX_TAIL_CALL_CNT** (not accessible to user space), which
> + * 		is currently set to 32.
> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
> + *
> + * int bpf_clone_redirect(struct sk_buff *skb, u32 ifindex, u64 flags)
> + * 	Description
> + * 		Clone and redirect the packet associated to *skb* to another
> + * 		net device of index *ifindex*. Both ingress and egress
> + * 		interfaces can be used for redirection. The **BPF_F_INGRESS**
> + * 		value in *flags* is used to make the distinction (ingress path
> + * 		is selected if the flag is present, egress path otherwise).
> + * 		This is the only flag supported for now.

Would probably make sense to describe the relation to bpf_redirect() helper
in one sentence, meaning, that bpf_clone_redirect() has the associated cost
duplicating the skb but the helper can be used out of the BPF prog whereas
this is not the case with bpf_redirect() which is therefore more efficient
but handled through an action code where the redirect happens after the BPF
prog returned.

> + * 		A call to this helper is susceptible to change data from the
> + * 		packet. Therefore, at load time, all checks on pointers
> + * 		previously done by the verifier are invalidated and must be
> + * 		performed again.
> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
>   */
>  #define __BPF_FUNC_MAPPER(FN)		\
>  	FN(unspec),			\
> 

[PATCH bpf-next v3 3/8] bpf: add documentation for eBPF helpers (12-22)

[Quentin Monnet]

Add documentation for eBPF helper functions to bpf.h user header file.
This documentation can be parsed with the Python script provided in
another commit of the patch series, in order to provide a RST document
that can later be converted into a man page.

The objective is to make the documentation easily understandable and
accessible to all eBPF developers, including beginners.

This patch contains descriptions for the following helper functions, all
written by Alexei:

- bpf_get_current_pid_tgid()
- bpf_get_current_uid_gid()
- bpf_get_current_comm()
- bpf_skb_vlan_push()
- bpf_skb_vlan_pop()
- bpf_skb_get_tunnel_key()
- bpf_skb_set_tunnel_key()
- bpf_redirect()
- bpf_perf_event_output()
- bpf_get_stackid()
- bpf_get_current_task()

v3:
- bpf_skb_get_tunnel_key(): Change and improve description and example.
- bpf_redirect(): Improve description of BPF_F_INGRESS flag.
- bpf_perf_event_output(): Fix first sentence of description. Delete
  wrong statement on context being evaluated as a struct pt_reg. Remove
  the long yet incomplete example.
- bpf_get_stackid(): Add a note about PERF_MAX_STACK_DEPTH being
  configurable.

Cc: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>
---
 include/uapi/linux/bpf.h | 225 +++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 225 insertions(+)

diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
index 02b7d522b3c0..c59bf5b28164 100644
--- a/include/uapi/linux/bpf.h
+++ b/include/uapi/linux/bpf.h
@@ -591,6 +591,231 @@ union bpf_attr {
  * 		performed again.
  * 	Return
  * 		0 on success, or a negative error in case of failure.
+ *
+ * u64 bpf_get_current_pid_tgid(void)
+ * 	Return
+ * 		A 64-bit integer containing the current tgid and pid, and
+ * 		created as such:
+ * 		*current_task*\ **->tgid << 32 \|**
+ * 		*current_task*\ **->pid**.
+ *
+ * u64 bpf_get_current_uid_gid(void)
+ * 	Return
+ * 		A 64-bit integer containing the current GID and UID, and
+ * 		created as such: *current_gid* **<< 32 \|** *current_uid*.
+ *
+ * int bpf_get_current_comm(char *buf, u32 size_of_buf)
+ * 	Description
+ * 		Copy the **comm** attribute of the current task into *buf* of
+ * 		*size_of_buf*. The **comm** attribute contains the name of
+ * 		the executable (excluding the path) for the current task. The
+ * 		*size_of_buf* must be strictly positive. On success, the
+ * 		helper makes sure that the *buf* is NUL-terminated. On failure,
+ * 		it is filled with zeroes.
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
+ *
+ * int bpf_skb_vlan_push(struct sk_buff *skb, __be16 vlan_proto, u16 vlan_tci)
+ * 	Description
+ * 		Push a *vlan_tci* (VLAN tag control information) of protocol
+ * 		*vlan_proto* to the packet associated to *skb*, then update
+ * 		the checksum. Note that if *vlan_proto* is different from
+ * 		**ETH_P_8021Q** and **ETH_P_8021AD**, it is considered to
+ * 		be **ETH_P_8021Q**.
+ *
+ * 		A call to this helper is susceptible to change data from the
+ * 		packet. Therefore, at load time, all checks on pointers
+ * 		previously done by the verifier are invalidated and must be
+ * 		performed again.
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
+ *
+ * int bpf_skb_vlan_pop(struct sk_buff *skb)
+ * 	Description
+ * 		Pop a VLAN header from the packet associated to *skb*.
+ *
+ * 		A call to this helper is susceptible to change data from the
+ * 		packet. Therefore, at load time, all checks on pointers
+ * 		previously done by the verifier are invalidated and must be
+ * 		performed again.
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
+ *
+ * int bpf_skb_get_tunnel_key(struct sk_buff *skb, struct bpf_tunnel_key *key, u32 size, u64 flags)
+ * 	Description
+ * 		Get tunnel metadata. This helper takes a pointer *key* to an
+ * 		empty **struct bpf_tunnel_key** of **size**, that will be
+ * 		filled with tunnel metadata for the packet associated to *skb*.
+ * 		The *flags* can be set to **BPF_F_TUNINFO_IPV6**, which
+ * 		indicates that the tunnel is based on IPv6 protocol instead of
+ * 		IPv4.
+ *
+ * 		The **struct bpf_tunnel_key** is an object that generalizes the
+ * 		principal parameters used by various tunneling protocols into a
+ * 		single struct. This way, it can be used to easily make a
+ * 		decision based on the contents of the encapsulation header,
+ * 		"summarized" in this struct. In particular, it holds the IP
+ * 		address of the remote end (IPv4 or IPv6, depending on the case)
+ * 		in *key*\ **->remote_ipv4** or *key*\ **->remote_ipv6**.
+ *
+ * 		Let's imagine that the following code is part of a program
+ * 		attached to the TC ingress interface, on one end of a GRE
+ * 		tunnel, and is supposed to filter out all messages coming from
+ * 		remote ends with IPv4 address other than 10.0.0.1:
+ *
+ * 		::
+ *
+ * 			int ret;
+ * 			struct bpf_tunnel_key key = {};
+ * 			
+ * 			ret = bpf_skb_get_tunnel_key(skb, &key, sizeof(key), 0);
+ * 			if (ret < 0)
+ * 				return TC_ACT_SHOT;	// drop packet
+ * 			
+ * 			if (key.remote_ipv4 != 0x0a000001)
+ * 				return TC_ACT_SHOT;	// drop packet
+ * 			
+ * 			return TC_ACT_OK;		// accept packet
+ *
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
+ *
+ * int bpf_skb_set_tunnel_key(struct sk_buff *skb, struct bpf_tunnel_key *key, u32 size, u64 flags)
+ * 	Description
+ * 		Populate tunnel metadata for packet associated to *skb.* The
+ * 		tunnel metadata is set to the contents of *key*, of *size*. The
+ * 		*flags* can be set to a combination of the following values:
+ *
+ * 		**BPF_F_TUNINFO_IPV6**
+ * 			Indicate that the tunnel is based on IPv6 protocol
+ * 			instead of IPv4.
+ * 		**BPF_F_ZERO_CSUM_TX**
+ * 			For IPv4 packets, add a flag to tunnel metadata
+ * 			indicating that checksum computation should be skipped
+ * 			and checksum set to zeroes.
+ * 		**BPF_F_DONT_FRAGMENT**
+ * 			Add a flag to tunnel metadata indicating that the
+ * 			packet should not be fragmented.
+ * 		**BPF_F_SEQ_NUMBER**
+ * 			Add a flag to tunnel metadata indicating that a
+ * 			sequence number should be added to tunnel header before
+ * 			sending the packet. This flag was added for GRE
+ * 			encapsulation, but might be used with other protocols
+ * 			as well in the future.
+ *
+ * 		Here is a typical usage on the transmit path:
+ *
+ * 		::
+ *
+ * 			struct bpf_tunnel_key key;
+ * 			     populate key ...
+ * 			bpf_skb_set_tunnel_key(skb, &key, sizeof(key), 0);
+ * 			bpf_clone_redirect(skb, vxlan_dev_ifindex, 0);
+ *
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
+ *
+ * int bpf_redirect(u32 ifindex, u64 flags)
+ * 	Description
+ * 		Redirect the packet to another net device of index *ifindex*.
+ * 		This helper is somewhat similar to **bpf_clone_redirect**\
+ * 		(), except that the packet is not cloned, which provides
+ * 		increased performance.
+ *
+ * 		Save for XDP, both ingress and egress interfaces can be used
+ * 		for redirection. The **BPF_F_INGRESS** value in *flags* is used
+ * 		to make the distinction (ingress path is selected if the flag
+ * 		is present, egress path otherwise). Currently, XDP only
+ * 		supports redirection to the egress interface, and accepts no
+ * 		flag at all.
+ * 	Return
+ * 		For XDP, the helper returns **XDP_REDIRECT** on success or
+ * 		**XDP_ABORT** on error. For other program types, the values
+ * 		are **TC_ACT_REDIRECT** on success or **TC_ACT_SHOT** on
+ * 		error.
+ *
+ * int bpf_perf_event_output(struct pt_reg *ctx, struct bpf_map *map, u64 flags, void *data, u64 size)
+ * 	Description
+ * 		Write raw *data* blob into a special BPF perf event held by
+ * 		*map* of type **BPF_MAP_TYPE_PERF_EVENT_ARRAY**. This perf
+ * 		event must have the following attributes: **PERF_SAMPLE_RAW**
+ * 		as **sample_type**, **PERF_TYPE_SOFTWARE** as **type**, and
+ * 		**PERF_COUNT_SW_BPF_OUTPUT** as **config**.
+ *
+ * 		The *flags* are used to indicate the index in *map* for which
+ * 		the value must be put, masked with **BPF_F_INDEX_MASK**.
+ * 		Alternatively, *flags* can be set to **BPF_F_CURRENT_CPU**
+ * 		to indicate that the index of the current CPU core should be
+ * 		used.
+ *
+ * 		The value to write, of *size*, is passed through eBPF stack and
+ * 		pointed by *data*.
+ *
+ * 		The context of the program *ctx* needs also be passed to the
+ * 		helper.
+ *
+ * 		On user space, a program willing to read the values needs to
+ * 		call **perf_event_open**\ () on the perf event (either for
+ * 		one or for all CPUs) and to store the file descriptor into the
+ * 		*map*. This must be done before the eBPF program can send data
+ * 		into it. An example is available in file
+ * 		*samples/bpf/trace_output_user.c* in the Linux kernel source
+ * 		tree (the eBPF program counterpart is in
+ * 		*samples/bpf/trace_output_kern.c*).
+ *
+ * 		**bpf_perf_event_output**\ () achieves better performance
+ * 		than **bpf_trace_printk**\ () for sharing data with user
+ * 		space, and is much better suitable for streaming data from eBPF
+ * 		programs.
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
+ *
+ * int bpf_get_stackid(struct pt_reg *ctx, struct bpf_map *map, u64 flags)
+ * 	Description
+ * 		Walk a user or a kernel stack and return its id. To achieve
+ * 		this, the helper needs *ctx*, which is a pointer to the context
+ * 		on which the tracing program is executed, and a pointer to a
+ * 		*map* of type **BPF_MAP_TYPE_STACK_TRACE**.
+ *
+ * 		The last argument, *flags*, holds the number of stack frames to
+ * 		skip (from 0 to 255), masked with
+ * 		**BPF_F_SKIP_FIELD_MASK**. The next bits can be used to set
+ * 		a combination of the following flags:
+ *
+ * 		**BPF_F_USER_STACK**
+ * 			Collect a user space stack instead of a kernel stack.
+ * 		**BPF_F_FAST_STACK_CMP**
+ * 			Compare stacks by hash only.
+ * 		**BPF_F_REUSE_STACKID**
+ * 			If two different stacks hash into the same *stackid*,
+ * 			discard the old one.
+ *
+ * 		The stack id retrieved is a 32 bit long integer handle which
+ * 		can be further combined with other data (including other stack
+ * 		ids) and used as a key into maps. This can be useful for
+ * 		generating a variety of graphs (such as flame graphs or off-cpu
+ * 		graphs).
+ *
+ * 		For walking a stack, this helper is an improvement over
+ * 		**bpf_probe_read**\ (), which can be used with unrolled loops
+ * 		but is not efficient and consumes a lot of eBPF instructions.
+ * 		Instead, **bpf_get_stackid**\ () can collect up to
+ * 		**PERF_MAX_STACK_DEPTH** both kernel and user frames. Note that
+ * 		this limit can be controlled with the **sysctl** program, and
+ * 		that it should be manually increased in order to profile long
+ * 		user stacks (such as stacks for Java programs). To do so, use:
+ *
+ * 		::
+ *
+ * 			# sysctl kernel.perf_event_max_stack=<new value>
+ *
+ * 	Return
+ * 		The positive or null stack id on success, or a negative error
+ * 		in case of failure.
+ *
+ * u64 bpf_get_current_task(void)
+ * 	Return
+ * 		A pointer to the current task struct.
  */
 #define __BPF_FUNC_MAPPER(FN)		\
 	FN(unspec),			\
-- 
2.14.1

Re: [PATCH bpf-next v3 3/8] bpf: add documentation for eBPF helpers (12-22)

[Alexei Starovoitov]

On Tue, Apr 17, 2018 at 03:34:33PM +0100, Quentin Monnet wrote:
> Add documentation for eBPF helper functions to bpf.h user header file.
> This documentation can be parsed with the Python script provided in
> another commit of the patch series, in order to provide a RST document
> that can later be converted into a man page.
> 
> The objective is to make the documentation easily understandable and
> accessible to all eBPF developers, including beginners.
> 
> This patch contains descriptions for the following helper functions, all
> written by Alexei:
> 
> - bpf_get_current_pid_tgid()
> - bpf_get_current_uid_gid()
> - bpf_get_current_comm()
> - bpf_skb_vlan_push()
> - bpf_skb_vlan_pop()
> - bpf_skb_get_tunnel_key()
> - bpf_skb_set_tunnel_key()
> - bpf_redirect()
> - bpf_perf_event_output()
> - bpf_get_stackid()
> - bpf_get_current_task()
> 
> v3:
> - bpf_skb_get_tunnel_key(): Change and improve description and example.
> - bpf_redirect(): Improve description of BPF_F_INGRESS flag.
> - bpf_perf_event_output(): Fix first sentence of description. Delete
>   wrong statement on context being evaluated as a struct pt_reg. Remove
>   the long yet incomplete example.
> - bpf_get_stackid(): Add a note about PERF_MAX_STACK_DEPTH being
>   configurable.
> 
> Cc: Alexei Starovoitov <ast@kernel.org>
> Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>

looks great.
Acked-by: Alexei Starovoitov <ast@kernel.org>

Re: [PATCH bpf-next v3 3/8] bpf: add documentation for eBPF helpers (12-22)

[Daniel Borkmann]

On 04/17/2018 04:34 PM, Quentin Monnet wrote:
> Add documentation for eBPF helper functions to bpf.h user header file.
> This documentation can be parsed with the Python script provided in
> another commit of the patch series, in order to provide a RST document
> that can later be converted into a man page.
> 
> The objective is to make the documentation easily understandable and
> accessible to all eBPF developers, including beginners.
> 
> This patch contains descriptions for the following helper functions, all
> written by Alexei:
> 
> - bpf_get_current_pid_tgid()
> - bpf_get_current_uid_gid()
> - bpf_get_current_comm()
> - bpf_skb_vlan_push()
> - bpf_skb_vlan_pop()
> - bpf_skb_get_tunnel_key()
> - bpf_skb_set_tunnel_key()
> - bpf_redirect()
> - bpf_perf_event_output()
> - bpf_get_stackid()
> - bpf_get_current_task()
> 
> v3:
> - bpf_skb_get_tunnel_key(): Change and improve description and example.
> - bpf_redirect(): Improve description of BPF_F_INGRESS flag.
> - bpf_perf_event_output(): Fix first sentence of description. Delete
>   wrong statement on context being evaluated as a struct pt_reg. Remove
>   the long yet incomplete example.
> - bpf_get_stackid(): Add a note about PERF_MAX_STACK_DEPTH being
>   configurable.
> 
> Cc: Alexei Starovoitov <ast@kernel.org>
> Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>
> ---
>  include/uapi/linux/bpf.h | 225 +++++++++++++++++++++++++++++++++++++++++++++++
>  1 file changed, 225 insertions(+)
> 
> diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
> index 02b7d522b3c0..c59bf5b28164 100644
> --- a/include/uapi/linux/bpf.h
> +++ b/include/uapi/linux/bpf.h
> @@ -591,6 +591,231 @@ union bpf_attr {
>   * 		performed again.
>   * 	Return
>   * 		0 on success, or a negative error in case of failure.
> + *
> + * u64 bpf_get_current_pid_tgid(void)
> + * 	Return
> + * 		A 64-bit integer containing the current tgid and pid, and
> + * 		created as such:
> + * 		*current_task*\ **->tgid << 32 \|**
> + * 		*current_task*\ **->pid**.
> + *
> + * u64 bpf_get_current_uid_gid(void)
> + * 	Return
> + * 		A 64-bit integer containing the current GID and UID, and
> + * 		created as such: *current_gid* **<< 32 \|** *current_uid*.
> + *
> + * int bpf_get_current_comm(char *buf, u32 size_of_buf)
> + * 	Description
> + * 		Copy the **comm** attribute of the current task into *buf* of
> + * 		*size_of_buf*. The **comm** attribute contains the name of
> + * 		the executable (excluding the path) for the current task. The
> + * 		*size_of_buf* must be strictly positive. On success, the
> + * 		helper makes sure that the *buf* is NUL-terminated. On failure,
> + * 		it is filled with zeroes.
> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
> + *
> + * int bpf_skb_vlan_push(struct sk_buff *skb, __be16 vlan_proto, u16 vlan_tci)
> + * 	Description
> + * 		Push a *vlan_tci* (VLAN tag control information) of protocol
> + * 		*vlan_proto* to the packet associated to *skb*, then update
> + * 		the checksum. Note that if *vlan_proto* is different from
> + * 		**ETH_P_8021Q** and **ETH_P_8021AD**, it is considered to
> + * 		be **ETH_P_8021Q**.
> + *
> + * 		A call to this helper is susceptible to change data from the
> + * 		packet. Therefore, at load time, all checks on pointers
> + * 		previously done by the verifier are invalidated and must be
> + * 		performed again.
> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
> + *
> + * int bpf_skb_vlan_pop(struct sk_buff *skb)
> + * 	Description
> + * 		Pop a VLAN header from the packet associated to *skb*.
> + *
> + * 		A call to this helper is susceptible to change data from the
> + * 		packet. Therefore, at load time, all checks on pointers
> + * 		previously done by the verifier are invalidated and must be
> + * 		performed again.
> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
> + *
> + * int bpf_skb_get_tunnel_key(struct sk_buff *skb, struct bpf_tunnel_key *key, u32 size, u64 flags)
> + * 	Description
> + * 		Get tunnel metadata. This helper takes a pointer *key* to an
> + * 		empty **struct bpf_tunnel_key** of **size**, that will be
> + * 		filled with tunnel metadata for the packet associated to *skb*.
> + * 		The *flags* can be set to **BPF_F_TUNINFO_IPV6**, which
> + * 		indicates that the tunnel is based on IPv6 protocol instead of
> + * 		IPv4.
> + *
> + * 		The **struct bpf_tunnel_key** is an object that generalizes the
> + * 		principal parameters used by various tunneling protocols into a
> + * 		single struct. This way, it can be used to easily make a
> + * 		decision based on the contents of the encapsulation header,
> + * 		"summarized" in this struct. In particular, it holds the IP
> + * 		address of the remote end (IPv4 or IPv6, depending on the case)
> + * 		in *key*\ **->remote_ipv4** or *key*\ **->remote_ipv6**.

I would also mention the tunnel_id which is typically mapped to a vni, allowing
to make this id programmable together with bpf_skb_set_tunnel_key() helper.

> + * 		Let's imagine that the following code is part of a program
> + * 		attached to the TC ingress interface, on one end of a GRE
> + * 		tunnel, and is supposed to filter out all messages coming from
> + * 		remote ends with IPv4 address other than 10.0.0.1:
> + *
> + * 		::
> + *
> + * 			int ret;
> + * 			struct bpf_tunnel_key key = {};
> + * 			
> + * 			ret = bpf_skb_get_tunnel_key(skb, &key, sizeof(key), 0);
> + * 			if (ret < 0)
> + * 				return TC_ACT_SHOT;	// drop packet
> + * 			
> + * 			if (key.remote_ipv4 != 0x0a000001)
> + * 				return TC_ACT_SHOT;	// drop packet
> + * 			
> + * 			return TC_ACT_OK;		// accept packet

Lets also add a small sentence that this interface can be used with all
encap devs that can operate in 'collect metadata' mode, where instead of
having one netdevice per specific configuration, the 'collect metadata'
mode only requires a single device where the configuration can be extracted
from those BPF helpers. Could also mentioned this can be used together with
vxlan, geneve, gre and ipip tunnels.

> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
> + *
> + * int bpf_skb_set_tunnel_key(struct sk_buff *skb, struct bpf_tunnel_key *key, u32 size, u64 flags)
> + * 	Description
> + * 		Populate tunnel metadata for packet associated to *skb.* The
> + * 		tunnel metadata is set to the contents of *key*, of *size*. The
> + * 		*flags* can be set to a combination of the following values:
> + *
> + * 		**BPF_F_TUNINFO_IPV6**
> + * 			Indicate that the tunnel is based on IPv6 protocol
> + * 			instead of IPv4.
> + * 		**BPF_F_ZERO_CSUM_TX**
> + * 			For IPv4 packets, add a flag to tunnel metadata
> + * 			indicating that checksum computation should be skipped
> + * 			and checksum set to zeroes.
> + * 		**BPF_F_DONT_FRAGMENT**
> + * 			Add a flag to tunnel metadata indicating that the
> + * 			packet should not be fragmented.
> + * 		**BPF_F_SEQ_NUMBER**
> + * 			Add a flag to tunnel metadata indicating that a
> + * 			sequence number should be added to tunnel header before
> + * 			sending the packet. This flag was added for GRE
> + * 			encapsulation, but might be used with other protocols
> + * 			as well in the future.
> + *
> + * 		Here is a typical usage on the transmit path:
> + *
> + * 		::
> + *
> + * 			struct bpf_tunnel_key key;
> + * 			     populate key ...
> + * 			bpf_skb_set_tunnel_key(skb, &key, sizeof(key), 0);
> + * 			bpf_clone_redirect(skb, vxlan_dev_ifindex, 0);

See above, maybe this can just reference bpf_skb_get_tunnel_key() from here.

> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
> + *
> + * int bpf_redirect(u32 ifindex, u64 flags)
> + * 	Description
> + * 		Redirect the packet to another net device of index *ifindex*.
> + * 		This helper is somewhat similar to **bpf_clone_redirect**\
> + * 		(), except that the packet is not cloned, which provides
> + * 		increased performance.
> + *
> + * 		Save for XDP, both ingress and egress interfaces can be used

s/Save/Same/ ?

> + * 		for redirection. The **BPF_F_INGRESS** value in *flags* is used

(In XDP case, BPF_F_INGRESS cannot be used.)

> + * 		to make the distinction (ingress path is selected if the flag
> + * 		is present, egress path otherwise). Currently, XDP only
> + * 		supports redirection to the egress interface, and accepts no
> + * 		flag at all.
> + * 	Return
> + * 		For XDP, the helper returns **XDP_REDIRECT** on success or
> + * 		**XDP_ABORT** on error. For other program types, the values
> + * 		are **TC_ACT_REDIRECT** on success or **TC_ACT_SHOT** on
> + * 		error.
> + *
> + * int bpf_perf_event_output(struct pt_reg *ctx, struct bpf_map *map, u64 flags, void *data, u64 size)
> + * 	Description
> + * 		Write raw *data* blob into a special BPF perf event held by
> + * 		*map* of type **BPF_MAP_TYPE_PERF_EVENT_ARRAY**. This perf
> + * 		event must have the following attributes: **PERF_SAMPLE_RAW**
> + * 		as **sample_type**, **PERF_TYPE_SOFTWARE** as **type**, and
> + * 		**PERF_COUNT_SW_BPF_OUTPUT** as **config**.
> + *
> + * 		The *flags* are used to indicate the index in *map* for which
> + * 		the value must be put, masked with **BPF_F_INDEX_MASK**.
> + * 		Alternatively, *flags* can be set to **BPF_F_CURRENT_CPU**
> + * 		to indicate that the index of the current CPU core should be
> + * 		used.
> + *
> + * 		The value to write, of *size*, is passed through eBPF stack and
> + * 		pointed by *data*.
> + *
> + * 		The context of the program *ctx* needs also be passed to the
> + * 		helper.
> + *
> + * 		On user space, a program willing to read the values needs to
> + * 		call **perf_event_open**\ () on the perf event (either for
> + * 		one or for all CPUs) and to store the file descriptor into the
> + * 		*map*. This must be done before the eBPF program can send data
> + * 		into it. An example is available in file
> + * 		*samples/bpf/trace_output_user.c* in the Linux kernel source
> + * 		tree (the eBPF program counterpart is in
> + * 		*samples/bpf/trace_output_kern.c*).
> + *
> + * 		**bpf_perf_event_output**\ () achieves better performance
> + * 		than **bpf_trace_printk**\ () for sharing data with user
> + * 		space, and is much better suitable for streaming data from eBPF
> + * 		programs.

Would also mentioned that this helper can be used out of tc and XDP BPF
programs as well and allows for passing i) only custom structs, ii) only
packet payload, or iii) a combination of both to user space listeners.

> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
> + *
> + * int bpf_get_stackid(struct pt_reg *ctx, struct bpf_map *map, u64 flags)
> + * 	Description
> + * 		Walk a user or a kernel stack and return its id. To achieve
> + * 		this, the helper needs *ctx*, which is a pointer to the context
> + * 		on which the tracing program is executed, and a pointer to a
> + * 		*map* of type **BPF_MAP_TYPE_STACK_TRACE**.
> + *
> + * 		The last argument, *flags*, holds the number of stack frames to
> + * 		skip (from 0 to 255), masked with
> + * 		**BPF_F_SKIP_FIELD_MASK**. The next bits can be used to set
> + * 		a combination of the following flags:
> + *
> + * 		**BPF_F_USER_STACK**
> + * 			Collect a user space stack instead of a kernel stack.
> + * 		**BPF_F_FAST_STACK_CMP**
> + * 			Compare stacks by hash only.
> + * 		**BPF_F_REUSE_STACKID**
> + * 			If two different stacks hash into the same *stackid*,
> + * 			discard the old one.
> + *
> + * 		The stack id retrieved is a 32 bit long integer handle which
> + * 		can be further combined with other data (including other stack
> + * 		ids) and used as a key into maps. This can be useful for
> + * 		generating a variety of graphs (such as flame graphs or off-cpu
> + * 		graphs).
> + *
> + * 		For walking a stack, this helper is an improvement over
> + * 		**bpf_probe_read**\ (), which can be used with unrolled loops
> + * 		but is not efficient and consumes a lot of eBPF instructions.
> + * 		Instead, **bpf_get_stackid**\ () can collect up to
> + * 		**PERF_MAX_STACK_DEPTH** both kernel and user frames. Note that
> + * 		this limit can be controlled with the **sysctl** program, and
> + * 		that it should be manually increased in order to profile long
> + * 		user stacks (such as stacks for Java programs). To do so, use:
> + *
> + * 		::
> + *
> + * 			# sysctl kernel.perf_event_max_stack=<new value>
> + *
> + * 	Return
> + * 		The positive or null stack id on success, or a negative error
> + * 		in case of failure.
> + *
> + * u64 bpf_get_current_task(void)
> + * 	Return
> + * 		A pointer to the current task struct.
>   */
>  #define __BPF_FUNC_MAPPER(FN)		\
>  	FN(unspec),			\
> 

[PATCH bpf-next v3 4/8] bpf: add documentation for eBPF helpers (23-32)

[Quentin Monnet]

Add documentation for eBPF helper functions to bpf.h user header file.
This documentation can be parsed with the Python script provided in
another commit of the patch series, in order to provide a RST document
that can later be converted into a man page.

The objective is to make the documentation easily understandable and
accessible to all eBPF developers, including beginners.

This patch contains descriptions for the following helper functions, all
written by Daniel:

- bpf_get_prandom_u32()
- bpf_get_smp_processor_id()
- bpf_get_cgroup_classid()
- bpf_get_route_realm()
- bpf_skb_load_bytes()
- bpf_csum_diff()
- bpf_skb_get_tunnel_opt()
- bpf_skb_set_tunnel_opt()
- bpf_skb_change_proto()
- bpf_skb_change_type()

v3:
- bpf_get_prandom_u32(): Fix helper name :(. Add description, including
  a note on the internal random state.
- bpf_get_smp_processor_id(): Add description, including a note on the
  processor id remaining stable during program run.
- bpf_get_cgroup_classid(): State that CONFIG_CGROUP_NET_CLASSID is
  required to use the helper. Add a reference to related documentation.
  State that placing a task in net_cls controller disables cgroup-bpf.
- bpf_get_route_realm(): State that CONFIG_CGROUP_NET_CLASSID is
  required to use this helper.
- bpf_skb_load_bytes(): Fix comment on current use cases for the helper.

Cc: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>
---
 include/uapi/linux/bpf.h | 152 +++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 152 insertions(+)

diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
index c59bf5b28164..d748f65a8f58 100644
--- a/include/uapi/linux/bpf.h
+++ b/include/uapi/linux/bpf.h
@@ -483,6 +483,23 @@ union bpf_attr {
  * 		The number of bytes written to the buffer, or a negative error
  * 		in case of failure.
  *
+ * u32 bpf_get_prandom_u32(void)
+ * 	Description
+ * 		Get a pseudo-random number. Note that this helper uses its own
+ * 		pseudo-random internal state, and cannot be used to infer the
+ * 		seed of other random functions in the kernel.
+ * 	Return
+ * 		A random 32-bit unsigned value.
+ *
+ * u32 bpf_get_smp_processor_id(void)
+ * 	Description
+ * 		Get the SMP (Symmetric multiprocessing) processor id. Note that
+ * 		all programs run with preemption disabled, which means that the
+ * 		SMP processor id is stable during all the execution of the
+ * 		program.
+ * 	Return
+ * 		The SMP id of the processor running the program.
+ *
  * int bpf_skb_store_bytes(struct sk_buff *skb, u32 offset, const void *from, u32 len, u64 flags)
  * 	Description
  * 		Store *len* bytes from address *from* into the packet
@@ -615,6 +632,27 @@ union bpf_attr {
  * 	Return
  * 		0 on success, or a negative error in case of failure.
  *
+ * u32 bpf_get_cgroup_classid(struct sk_buff *skb)
+ * 	Description
+ * 		Retrieve the classid for the current task, i.e. for the
+ * 		net_cls (network classifier) cgroup to which *skb* belongs.
+ *
+ * 		This helper is only available is the kernel was compiled with
+ * 		the **CONFIG_CGROUP_NET_CLASSID** configuration option set to
+ * 		"**y**" or to "**m**".
+ *
+ * 		Note that placing a task into the net_cls controller completely
+ * 		disables the execution of eBPF programs with the cgroup.
+ *
+ * 		Also note that, in the above description, the "network
+ * 		classifier" cgroup does not designate a generic classifier, but
+ * 		a particular mechanism that provides an interface to tag
+ * 		network packets with a specific class identifier. See also the
+ * 		related kernel documentation, available from the Linux sources
+ * 		in file *Documentation/cgroup-v1/net_cls.txt*.
+ * 	Return
+ * 		The classid, or 0 for the default unconfigured classid.
+ *
  * int bpf_skb_vlan_push(struct sk_buff *skb, __be16 vlan_proto, u16 vlan_tci)
  * 	Description
  * 		Push a *vlan_tci* (VLAN tag control information) of protocol
@@ -734,6 +772,16 @@ union bpf_attr {
  * 		are **TC_ACT_REDIRECT** on success or **TC_ACT_SHOT** on
  * 		error.
  *
+ * u32 bpf_get_route_realm(struct sk_buff *skb)
+ * 	Description
+ * 		Retrieve the realm or the route, that is to say the
+ * 		**tclassid** field of the destination for the *skb*. This
+ * 		helper is available only if the kernel was compiled with
+ * 		**CONFIG_IP_ROUTE_CLASSID** configuration option.
+ * 	Return
+ * 		The realm of the route for the packet associated to *sdb*, or 0
+ * 		if none was found.
+ *
  * int bpf_perf_event_output(struct pt_reg *ctx, struct bpf_map *map, u64 flags, void *data, u64 size)
  * 	Description
  * 		Write raw *data* blob into a special BPF perf event held by
@@ -770,6 +818,23 @@ union bpf_attr {
  * 	Return
  * 		0 on success, or a negative error in case of failure.
  *
+ * int bpf_skb_load_bytes(const struct sk_buff *skb, u32 offset, void *to, u32 len)
+ * 	Description
+ * 		This helper was provided as an easy way to load data from a
+ * 		packet. It can be used to load *len* bytes from *offset* from
+ * 		the packet associated to *skb*, into the buffer pointed by
+ * 		*to*.
+ *
+ * 		Since Linux 4.7, usage of this helper has mostly been replaced
+ * 		by "direct packet access", enabling packet data to be
+ * 		manipulated with *skb*\ **->data** and *skb*\ **->data_end**
+ * 		pointing respectively to the first byte of packet data and to
+ * 		the byte after the last byte of packet data. However, it
+ * 		remains useful if one wishes to read large quantities of data
+ * 		at once from a packet.
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
+ *
  * int bpf_get_stackid(struct pt_reg *ctx, struct bpf_map *map, u64 flags)
  * 	Description
  * 		Walk a user or a kernel stack and return its id. To achieve
@@ -813,6 +878,93 @@ union bpf_attr {
  * 		The positive or null stack id on success, or a negative error
  * 		in case of failure.
  *
+ * s64 bpf_csum_diff(__be32 *from, u32 from_size, __be32 *to, u32 to_size, __wsum seed)
+ * 	Description
+ * 		Compute a checksum difference, from the raw buffer pointed by
+ * 		*from*, of length *from_size* (that must be a multiple of 4),
+ * 		towards the raw buffer pointed by *to*, of size *to_size*
+ * 		(same remark). An optional *seed* can be added to the value.
+ *
+ * 		This is flexible enough to be used in several ways:
+ *
+ * 		* With *from_size* == 0, *to_size* > 0 and *seed* set to
+ * 		  checksum, it can be used when pushing new data.
+ * 		* With *from_size* > 0, *to_size* == 0 and *seed* set to
+ * 		  checksum, it can be used when removing data from a packet.
+ * 		* With *from_size* > 0, *to_size* > 0 and *seed* set to 0, it
+ * 		  can be used to compute a diff. Note that *from_size* and
+ * 		  *to_size* do not need to be equal.
+ * 	Return
+ * 		The checksum result, or a negative error code in case of
+ * 		failure.
+ *
+ * int bpf_skb_get_tunnel_opt(struct sk_buff *skb, u8 *opt, u32 size)
+ * 	Description
+ * 		Retrieve tunnel options metadata for the packet associated to
+ * 		*skb*, and store the raw tunnel option data to the buffer *opt*
+ * 		of *size*.
+ * 	Return
+ * 		The size of the option data retrieved.
+ *
+ * int bpf_skb_set_tunnel_opt(struct sk_buff *skb, u8 *opt, u32 size)
+ * 	Description
+ * 		Set tunnel options metadata for the packet associated to *skb*
+ * 		to the option data contained in the raw buffer *opt* of *size*.
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
+ *
+ * int bpf_skb_change_proto(struct sk_buff *skb, __be16 proto, u64 flags)
+ * 	Description
+ * 		Change the protocol of the *skb* to *proto*. Currently
+ * 		supported are transition from IPv4 to IPv6, and from IPv6 to
+ * 		IPv4. The helper takes care of the groundwork for the
+ * 		transition, including resizing the socket buffer. The eBPF
+ * 		program is expected to fill the new headers, if any, via
+ * 		**skb_store_bytes**\ () and to recompute the checksums with
+ * 		**bpf_l3_csum_replace**\ () and **bpf_l4_csum_replace**\
+ * 		().
+ *
+ * 		Internally, the GSO type is marked as dodgy so that headers are
+ * 		checked and segments are recalculated by the GSO/GRO engine.
+ * 		The size for GSO target is adapted as well.
+ *
+ * 		All values for *flags* are reserved for future usage, and must
+ * 		be left at zero.
+ *
+ * 		A call to this helper is susceptible to change data from the
+ * 		packet. Therefore, at load time, all checks on pointers
+ * 		previously done by the verifier are invalidated and must be
+ * 		performed again.
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
+ *
+ * int bpf_skb_change_type(struct sk_buff *skb, u32 type)
+ * 	Description
+ * 		Change the packet type for the packet associated to *skb*. This
+ * 		comes down to setting *skb*\ **->pkt_type** to *type*, except
+ * 		the eBPF program does not have a write access to *skb*\
+ * 		**->pkt_type** beside this helper. Using a helper here allows
+ * 		for graceful handling of errors.
+ *
+ * 		The major use case is to change incoming *skb*s to
+ * 		**PACKET_HOST** in a programmatic way instead of having to
+ * 		recirculate via **redirect**\ (..., **BPF_F_INGRESS**), for
+ * 		example.
+ *
+ * 		Note that *type* only allows certain values. At this time, they
+ * 		are:
+ *
+ * 		**PACKET_HOST**
+ * 			Packet is for us.
+ * 		**PACKET_BROADCAST**
+ * 			Send packet to all.
+ * 		**PACKET_MULTICAST**
+ * 			Send packet to group.
+ * 		**PACKET_OTHERHOST**
+ * 			Send packet to someone else.
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
+ *
  * u64 bpf_get_current_task(void)
  * 	Return
  * 		A pointer to the current task struct.
-- 
2.14.1

Re: [PATCH bpf-next v3 4/8] bpf: add documentation for eBPF helpers (23-32)

[Alexei Starovoitov]

On Tue, Apr 17, 2018 at 03:34:34PM +0100, Quentin Monnet wrote:
> Add documentation for eBPF helper functions to bpf.h user header file.
> This documentation can be parsed with the Python script provided in
> another commit of the patch series, in order to provide a RST document
> that can later be converted into a man page.
> 
> The objective is to make the documentation easily understandable and
> accessible to all eBPF developers, including beginners.
> 
> This patch contains descriptions for the following helper functions, all
> written by Daniel:
> 
> - bpf_get_prandom_u32()
> - bpf_get_smp_processor_id()
> - bpf_get_cgroup_classid()
> - bpf_get_route_realm()
> - bpf_skb_load_bytes()
> - bpf_csum_diff()
> - bpf_skb_get_tunnel_opt()
> - bpf_skb_set_tunnel_opt()
> - bpf_skb_change_proto()
> - bpf_skb_change_type()
> 
> v3:
> - bpf_get_prandom_u32(): Fix helper name :(. Add description, including
>   a note on the internal random state.
> - bpf_get_smp_processor_id(): Add description, including a note on the
>   processor id remaining stable during program run.
> - bpf_get_cgroup_classid(): State that CONFIG_CGROUP_NET_CLASSID is
>   required to use the helper. Add a reference to related documentation.
>   State that placing a task in net_cls controller disables cgroup-bpf.
> - bpf_get_route_realm(): State that CONFIG_CGROUP_NET_CLASSID is
>   required to use this helper.
> - bpf_skb_load_bytes(): Fix comment on current use cases for the helper.
> 
> Cc: Daniel Borkmann <daniel@iogearbox.net>
> Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>

Acked-by: Alexei Starovoitov <ast@kernel.org>

Re: [PATCH bpf-next v3 4/8] bpf: add documentation for eBPF helpers (23-32)

[Daniel Borkmann]

On 04/17/2018 04:34 PM, Quentin Monnet wrote:
> Add documentation for eBPF helper functions to bpf.h user header file.
> This documentation can be parsed with the Python script provided in
> another commit of the patch series, in order to provide a RST document
> that can later be converted into a man page.
> 
> The objective is to make the documentation easily understandable and
> accessible to all eBPF developers, including beginners.
> 
> This patch contains descriptions for the following helper functions, all
> written by Daniel:
> 
> - bpf_get_prandom_u32()
> - bpf_get_smp_processor_id()
> - bpf_get_cgroup_classid()
> - bpf_get_route_realm()
> - bpf_skb_load_bytes()
> - bpf_csum_diff()
> - bpf_skb_get_tunnel_opt()
> - bpf_skb_set_tunnel_opt()
> - bpf_skb_change_proto()
> - bpf_skb_change_type()
> 
> v3:
> - bpf_get_prandom_u32(): Fix helper name :(. Add description, including
>   a note on the internal random state.
> - bpf_get_smp_processor_id(): Add description, including a note on the
>   processor id remaining stable during program run.
> - bpf_get_cgroup_classid(): State that CONFIG_CGROUP_NET_CLASSID is
>   required to use the helper. Add a reference to related documentation.
>   State that placing a task in net_cls controller disables cgroup-bpf.
> - bpf_get_route_realm(): State that CONFIG_CGROUP_NET_CLASSID is
>   required to use this helper.
> - bpf_skb_load_bytes(): Fix comment on current use cases for the helper.
> 
> Cc: Daniel Borkmann <daniel@iogearbox.net>
> Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>
> ---
>  include/uapi/linux/bpf.h | 152 +++++++++++++++++++++++++++++++++++++++++++++++
>  1 file changed, 152 insertions(+)
> 
> diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
> index c59bf5b28164..d748f65a8f58 100644
> --- a/include/uapi/linux/bpf.h
> +++ b/include/uapi/linux/bpf.h
> @@ -483,6 +483,23 @@ union bpf_attr {
>   * 		The number of bytes written to the buffer, or a negative error
>   * 		in case of failure.
>   *
> + * u32 bpf_get_prandom_u32(void)
> + * 	Description
> + * 		Get a pseudo-random number. Note that this helper uses its own
> + * 		pseudo-random internal state, and cannot be used to infer the
> + * 		seed of other random functions in the kernel.

We should still add that this prng is not cryptographically secure.

> + * 	Return
> + * 		A random 32-bit unsigned value.
> + *
> + * u32 bpf_get_smp_processor_id(void)
> + * 	Description
> + * 		Get the SMP (Symmetric multiprocessing) processor id. Note that

Nit: s/Symmetric/symmetric/ ?

> + * 		all programs run with preemption disabled, which means that the
> + * 		SMP processor id is stable during all the execution of the
> + * 		program.
> + * 	Return
> + * 		The SMP id of the processor running the program.
> + *
>   * int bpf_skb_store_bytes(struct sk_buff *skb, u32 offset, const void *from, u32 len, u64 flags)
>   * 	Description
>   * 		Store *len* bytes from address *from* into the packet
> @@ -615,6 +632,27 @@ union bpf_attr {
>   * 	Return
>   * 		0 on success, or a negative error in case of failure.
>   *
> + * u32 bpf_get_cgroup_classid(struct sk_buff *skb)
> + * 	Description
> + * 		Retrieve the classid for the current task, i.e. for the
> + * 		net_cls (network classifier) cgroup to which *skb* belongs.
> + *
> + * 		This helper is only available is the kernel was compiled with
> + * 		the **CONFIG_CGROUP_NET_CLASSID** configuration option set to
> + * 		"**y**" or to "**m**".
> + *
> + * 		Note that placing a task into the net_cls controller completely
> + * 		disables the execution of eBPF programs with the cgroup.

I'm not sure I follow the above sentence, what do you mean by that?

I would definitely also add here that this helper is limited to cgroups v1
only, and that it works on clsact TC egress hook but not the ingress one.

> + * 		Also note that, in the above description, the "network
> + * 		classifier" cgroup does not designate a generic classifier, but
> + * 		a particular mechanism that provides an interface to tag
> + * 		network packets with a specific class identifier. See also the

The "generic classifier" part is a bit strange to parse. I would probably
leave the first part out and explain that this provides a means to tag
packets based on a user-provided ID for all traffic coming from the tasks
belonging to the related cgroup.

> + * 		related kernel documentation, available from the Linux sources
> + * 		in file *Documentation/cgroup-v1/net_cls.txt*.
> + * 	Return
> + * 		The classid, or 0 for the default unconfigured classid.
> + *
>   * int bpf_skb_vlan_push(struct sk_buff *skb, __be16 vlan_proto, u16 vlan_tci)
>   * 	Description
>   * 		Push a *vlan_tci* (VLAN tag control information) of protocol
> @@ -734,6 +772,16 @@ union bpf_attr {
>   * 		are **TC_ACT_REDIRECT** on success or **TC_ACT_SHOT** on
>   * 		error.
>   *
> + * u32 bpf_get_route_realm(struct sk_buff *skb)
> + * 	Description
> + * 		Retrieve the realm or the route, that is to say the
> + * 		**tclassid** field of the destination for the *skb*. This
> + * 		helper is available only if the kernel was compiled with
> + * 		**CONFIG_IP_ROUTE_CLASSID** configuration option.

Could mention that this is a similar user provided tag like in the net_cls
case with cgroups only that this applies to routes here (dst entries) which
hold this tag.

Also, should say that this works with clsact TC egress hook or alternatively
conventional classful egress qdiscs, but not on TC ingress. In case of clsact
TC egress hook this has the advantage that the dst entry has not been dropped
yet in the xmit path. Therefore, the dst entry does not need to be artificially
held via netif_keep_dst() for a classful qdisc until the skb is freed.

> + * 	Return
> + * 		The realm of the route for the packet associated to *sdb*, or 0

Typo: sdb

> + * 		if none was found.
> + *
>   * int bpf_perf_event_output(struct pt_reg *ctx, struct bpf_map *map, u64 flags, void *data, u64 size)
>   * 	Description
>   * 		Write raw *data* blob into a special BPF perf event held by
> @@ -770,6 +818,23 @@ union bpf_attr {
>   * 	Return
>   * 		0 on success, or a negative error in case of failure.
>   *
> + * int bpf_skb_load_bytes(const struct sk_buff *skb, u32 offset, void *to, u32 len)
> + * 	Description
> + * 		This helper was provided as an easy way to load data from a
> + * 		packet. It can be used to load *len* bytes from *offset* from
> + * 		the packet associated to *skb*, into the buffer pointed by
> + * 		*to*.
> + *
> + * 		Since Linux 4.7, usage of this helper has mostly been replaced
> + * 		by "direct packet access", enabling packet data to be
> + * 		manipulated with *skb*\ **->data** and *skb*\ **->data_end**
> + * 		pointing respectively to the first byte of packet data and to
> + * 		the byte after the last byte of packet data. However, it
> + * 		remains useful if one wishes to read large quantities of data
> + * 		at once from a packet.

I would add: s/at once from a packet/at once from a packet into the BPF stack/

> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
> + *
>   * int bpf_get_stackid(struct pt_reg *ctx, struct bpf_map *map, u64 flags)
>   * 	Description
>   * 		Walk a user or a kernel stack and return its id. To achieve
> @@ -813,6 +878,93 @@ union bpf_attr {
>   * 		The positive or null stack id on success, or a negative error
>   * 		in case of failure.
>   *
> + * s64 bpf_csum_diff(__be32 *from, u32 from_size, __be32 *to, u32 to_size, __wsum seed)
> + * 	Description
> + * 		Compute a checksum difference, from the raw buffer pointed by
> + * 		*from*, of length *from_size* (that must be a multiple of 4),
> + * 		towards the raw buffer pointed by *to*, of size *to_size*
> + * 		(same remark). An optional *seed* can be added to the value.

Wrt seed, we should explicitly mention that this can be cascaded but also that
this helper works in combination with the l3/l4 csum ones where you feed in this
diff coming from bpf_csum_diff().

> + * 		This is flexible enough to be used in several ways:
> + *
> + * 		* With *from_size* == 0, *to_size* > 0 and *seed* set to
> + * 		  checksum, it can be used when pushing new data.
> + * 		* With *from_size* > 0, *to_size* == 0 and *seed* set to
> + * 		  checksum, it can be used when removing data from a packet.
> + * 		* With *from_size* > 0, *to_size* > 0 and *seed* set to 0, it
> + * 		  can be used to compute a diff. Note that *from_size* and
> + * 		  *to_size* do not need to be equal.
> + * 	Return
> + * 		The checksum result, or a negative error code in case of
> + * 		failure.
> + *
> + * int bpf_skb_get_tunnel_opt(struct sk_buff *skb, u8 *opt, u32 size)
> + * 	Description
> + * 		Retrieve tunnel options metadata for the packet associated to
> + * 		*skb*, and store the raw tunnel option data to the buffer *opt*
> + * 		of *size*.
> + * 	Return
> + * 		The size of the option data retrieved.
> + *
> + * int bpf_skb_set_tunnel_opt(struct sk_buff *skb, u8 *opt, u32 size)
> + * 	Description
> + * 		Set tunnel options metadata for the packet associated to *skb*
> + * 		to the option data contained in the raw buffer *opt* of *size*.

Also here the same remark with collect meta data I made earlier, and as a
particular example where this can be used in combination with geneve where
this allows for pushing and retrieving (bpf_skb_get_tunnel_opt() case)
arbitrary TLVs from the BPF program that allows for full customization.

> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
> + *
> + * int bpf_skb_change_proto(struct sk_buff *skb, __be16 proto, u64 flags)
> + * 	Description
> + * 		Change the protocol of the *skb* to *proto*. Currently
> + * 		supported are transition from IPv4 to IPv6, and from IPv6 to
> + * 		IPv4. The helper takes care of the groundwork for the
> + * 		transition, including resizing the socket buffer. The eBPF
> + * 		program is expected to fill the new headers, if any, via
> + * 		**skb_store_bytes**\ () and to recompute the checksums with
> + * 		**bpf_l3_csum_replace**\ () and **bpf_l4_csum_replace**\
> + * 		().

Could mention the main use case for NAT64 out of an BPF program.

> + *
> + * 		Internally, the GSO type is marked as dodgy so that headers are
> + * 		checked and segments are recalculated by the GSO/GRO engine.
> + * 		The size for GSO target is adapted as well.
> + *
> + * 		All values for *flags* are reserved for future usage, and must
> + * 		be left at zero.
> + *
> + * 		A call to this helper is susceptible to change data from the
> + * 		packet. Therefore, at load time, all checks on pointers
> + * 		previously done by the verifier are invalidated and must be
> + * 		performed again.
> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
> + *
> + * int bpf_skb_change_type(struct sk_buff *skb, u32 type)
> + * 	Description
> + * 		Change the packet type for the packet associated to *skb*. This
> + * 		comes down to setting *skb*\ **->pkt_type** to *type*, except
> + * 		the eBPF program does not have a write access to *skb*\
> + * 		**->pkt_type** beside this helper. Using a helper here allows
> + * 		for graceful handling of errors.
> + *
> + * 		The major use case is to change incoming *skb*s to
> + * 		**PACKET_HOST** in a programmatic way instead of having to
> + * 		recirculate via **redirect**\ (..., **BPF_F_INGRESS**), for
> + * 		example.
> + *
> + * 		Note that *type* only allows certain values. At this time, they
> + * 		are:
> + *
> + * 		**PACKET_HOST**
> + * 			Packet is for us.
> + * 		**PACKET_BROADCAST**
> + * 			Send packet to all.
> + * 		**PACKET_MULTICAST**
> + * 			Send packet to group.
> + * 		**PACKET_OTHERHOST**
> + * 			Send packet to someone else.
> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
> + *
>   * u64 bpf_get_current_task(void)
>   * 	Return
>   * 		A pointer to the current task struct.
> 

Re: [PATCH bpf-next v3 4/8] bpf: add documentation for eBPF helpers (23-32)

[Quentin Monnet]

2018-04-19 13:16 UTC+0200 ~ Daniel Borkmann <daniel@iogearbox.net>
> On 04/17/2018 04:34 PM, Quentin Monnet wrote:
>> Add documentation for eBPF helper functions to bpf.h user header file.
>> This documentation can be parsed with the Python script provided in
>> another commit of the patch series, in order to provide a RST document
>> that can later be converted into a man page.
>>
>> The objective is to make the documentation easily understandable and
>> accessible to all eBPF developers, including beginners.
>>
>> This patch contains descriptions for the following helper functions, all
>> written by Daniel:
>>
>> - bpf_get_prandom_u32()
>> - bpf_get_smp_processor_id()
>> - bpf_get_cgroup_classid()
>> - bpf_get_route_realm()
>> - bpf_skb_load_bytes()
>> - bpf_csum_diff()
>> - bpf_skb_get_tunnel_opt()
>> - bpf_skb_set_tunnel_opt()
>> - bpf_skb_change_proto()
>> - bpf_skb_change_type()
>>
>> v3:
>> - bpf_get_prandom_u32(): Fix helper name :(. Add description, including
>>   a note on the internal random state.
>> - bpf_get_smp_processor_id(): Add description, including a note on the
>>   processor id remaining stable during program run.
>> - bpf_get_cgroup_classid(): State that CONFIG_CGROUP_NET_CLASSID is
>>   required to use the helper. Add a reference to related documentation.
>>   State that placing a task in net_cls controller disables cgroup-bpf.
>> - bpf_get_route_realm(): State that CONFIG_CGROUP_NET_CLASSID is
>>   required to use this helper.
>> - bpf_skb_load_bytes(): Fix comment on current use cases for the helper.
>>
>> Cc: Daniel Borkmann <daniel@iogearbox.net>
>> Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>
>> ---
>>  include/uapi/linux/bpf.h | 152 +++++++++++++++++++++++++++++++++++++++++++++++
>>  1 file changed, 152 insertions(+)
>>
>> diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
>> index c59bf5b28164..d748f65a8f58 100644
>> --- a/include/uapi/linux/bpf.h
>> +++ b/include/uapi/linux/bpf.h

[...]

>> @@ -615,6 +632,27 @@ union bpf_attr {
>>   * 	Return
>>   * 		0 on success, or a negative error in case of failure.
>>   *
>> + * u32 bpf_get_cgroup_classid(struct sk_buff *skb)
>> + * 	Description
>> + * 		Retrieve the classid for the current task, i.e. for the
>> + * 		net_cls (network classifier) cgroup to which *skb* belongs.
>> + *
>> + * 		This helper is only available is the kernel was compiled with
>> + * 		the **CONFIG_CGROUP_NET_CLASSID** configuration option set to
>> + * 		"**y**" or to "**m**".
>> + *
>> + * 		Note that placing a task into the net_cls controller completely
>> + * 		disables the execution of eBPF programs with the cgroup.
> 
> I'm not sure I follow the above sentence, what do you mean by that?

Please see comment below.

> I would definitely also add here that this helper is limited to cgroups v1
> only, and that it works on clsact TC egress hook but not the ingress one.
> 
>> + * 		Also note that, in the above description, the "network
>> + * 		classifier" cgroup does not designate a generic classifier, but
>> + * 		a particular mechanism that provides an interface to tag
>> + * 		network packets with a specific class identifier. See also the
> 
> The "generic classifier" part is a bit strange to parse. I would probably
> leave the first part out and explain that this provides a means to tag
> packets based on a user-provided ID for all traffic coming from the tasks
> belonging to the related cgroup.

In this paragraph and in the one above, I am trying to address Alexei
comments to the RFC v2:

    please add that kernel should be configured with
    CONFIG_NET_CLS_CGROUP=y|m
    and mention Documentation/cgroup-v1/net_cls.txt
    Otherwise 'network classifier' is way too generic.
    I'd also mention that placing a task into net_cls controller
    disables all of cgroup-bpf.

But I must admit I am struggling a bit on this helper. If I understand
correctly, "network classifier" is too generic and could be confused
with TC classifiers? Hence the attempt to avoid that in the second
paragraph... As for "placing a task into net_cls controller disables all
of cgroup-bpf", again if I understand correctly, I think it refers to
cgroup_sk_alloc_disable() being called in write_classid() in file
net/core/netclassid_cgroup.c. But I am not familiar with it and cannot
find a nice way to explain that in the doc :/.

>> + * 		related kernel documentation, available from the Linux sources
>> + * 		in file *Documentation/cgroup-v1/net_cls.txt*.
>> + * 	Return
>> + * 		The classid, or 0 for the default unconfigured classid.
>> + *
>>   * int bpf_skb_vlan_push(struct sk_buff *skb, __be16 vlan_proto, u16 vlan_tci)
>>   * 	Description
>>   * 		Push a *vlan_tci* (VLAN tag control information) of protocol
>> @@ -734,6 +772,16 @@ union bpf_attr {
>>   * 		are **TC_ACT_REDIRECT** on success or **TC_ACT_SHOT** on
>>   * 		error.
>>   *
>> + * u32 bpf_get_route_realm(struct sk_buff *skb)
>> + * 	Description
>> + * 		Retrieve the realm or the route, that is to say the
>> + * 		**tclassid** field of the destination for the *skb*. This
>> + * 		helper is available only if the kernel was compiled with
>> + * 		**CONFIG_IP_ROUTE_CLASSID** configuration option.
> 
> Could mention that this is a similar user provided tag like in the net_cls
> case with cgroups only that this applies to routes here (dst entries) which
> hold this tag.
> 
> Also, should say that this works with clsact TC egress hook or alternatively
> conventional classful egress qdiscs, but not on TC ingress. In case of clsact
> TC egress hook this has the advantage that the dst entry has not been dropped
> yet in the xmit path. Therefore, the dst entry does not need to be artificially
> held via netif_keep_dst() for a classful qdisc until the skb is freed.

Here I am not sure to follow, the advantage is for clsact, but this is
only for a classful qdisc that we can avoid holding the dst entry? How
does holding the entry with netif_keep_dst() translate, from a user
space point of view?

Thanks a lot for the exhaustive review!
Quentin

Re: [PATCH bpf-next v3 4/8] bpf: add documentation for eBPF helpers (23-32)

[Daniel Borkmann]

On 04/20/2018 08:54 PM, Quentin Monnet wrote:
> 2018-04-19 13:16 UTC+0200 ~ Daniel Borkmann <daniel@iogearbox.net>
>> On 04/17/2018 04:34 PM, Quentin Monnet wrote:
>>> Add documentation for eBPF helper functions to bpf.h user header file.
>>> This documentation can be parsed with the Python script provided in
>>> another commit of the patch series, in order to provide a RST document
>>> that can later be converted into a man page.
>>>
>>> The objective is to make the documentation easily understandable and
>>> accessible to all eBPF developers, including beginners.
>>>
>>> This patch contains descriptions for the following helper functions, all
>>> written by Daniel:
>>>
>>> - bpf_get_prandom_u32()
>>> - bpf_get_smp_processor_id()
>>> - bpf_get_cgroup_classid()
>>> - bpf_get_route_realm()
>>> - bpf_skb_load_bytes()
>>> - bpf_csum_diff()
>>> - bpf_skb_get_tunnel_opt()
>>> - bpf_skb_set_tunnel_opt()
>>> - bpf_skb_change_proto()
>>> - bpf_skb_change_type()
>>>
>>> v3:
>>> - bpf_get_prandom_u32(): Fix helper name :(. Add description, including
>>>   a note on the internal random state.
>>> - bpf_get_smp_processor_id(): Add description, including a note on the
>>>   processor id remaining stable during program run.
>>> - bpf_get_cgroup_classid(): State that CONFIG_CGROUP_NET_CLASSID is
>>>   required to use the helper. Add a reference to related documentation.
>>>   State that placing a task in net_cls controller disables cgroup-bpf.
>>> - bpf_get_route_realm(): State that CONFIG_CGROUP_NET_CLASSID is
>>>   required to use this helper.
>>> - bpf_skb_load_bytes(): Fix comment on current use cases for the helper.
>>>
>>> Cc: Daniel Borkmann <daniel@iogearbox.net>
>>> Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>
>>> ---
>>>  include/uapi/linux/bpf.h | 152 +++++++++++++++++++++++++++++++++++++++++++++++
>>>  1 file changed, 152 insertions(+)
>>>
>>> diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
>>> index c59bf5b28164..d748f65a8f58 100644
>>> --- a/include/uapi/linux/bpf.h
>>> +++ b/include/uapi/linux/bpf.h
> 
> [...]
> 
>>> @@ -615,6 +632,27 @@ union bpf_attr {
>>>   * 	Return
>>>   * 		0 on success, or a negative error in case of failure.
>>>   *
>>> + * u32 bpf_get_cgroup_classid(struct sk_buff *skb)
>>> + * 	Description
>>> + * 		Retrieve the classid for the current task, i.e. for the
>>> + * 		net_cls (network classifier) cgroup to which *skb* belongs.
>>> + *
>>> + * 		This helper is only available is the kernel was compiled with
>>> + * 		the **CONFIG_CGROUP_NET_CLASSID** configuration option set to
>>> + * 		"**y**" or to "**m**".
>>> + *
>>> + * 		Note that placing a task into the net_cls controller completely
>>> + * 		disables the execution of eBPF programs with the cgroup.
>>
>> I'm not sure I follow the above sentence, what do you mean by that?
> 
> Please see comment below.
> 
>> I would definitely also add here that this helper is limited to cgroups v1
>> only, and that it works on clsact TC egress hook but not the ingress one.
>>
>>> + * 		Also note that, in the above description, the "network
>>> + * 		classifier" cgroup does not designate a generic classifier, but
>>> + * 		a particular mechanism that provides an interface to tag
>>> + * 		network packets with a specific class identifier. See also the
>>
>> The "generic classifier" part is a bit strange to parse. I would probably
>> leave the first part out and explain that this provides a means to tag
>> packets based on a user-provided ID for all traffic coming from the tasks
>> belonging to the related cgroup.
> 
> In this paragraph and in the one above, I am trying to address Alexei
> comments to the RFC v2:
> 
>     please add that kernel should be configured with
>     CONFIG_NET_CLS_CGROUP=y|m
>     and mention Documentation/cgroup-v1/net_cls.txt
>     Otherwise 'network classifier' is way too generic.
>     I'd also mention that placing a task into net_cls controller
>     disables all of cgroup-bpf.
> 
> But I must admit I am struggling a bit on this helper. If I understand
> correctly, "network classifier" is too generic and could be confused
> with TC classifiers? Hence the attempt to avoid that in the second

I think if you just name it "net_cls cgroup" it should be good enough in case
the concern is that "network classifier" name would be misunderstood.

> paragraph... As for "placing a task into net_cls controller disables all
> of cgroup-bpf", again if I understand correctly, I think it refers to
> cgroup_sk_alloc_disable() being called in write_classid() in file
> net/core/netclassid_cgroup.c. But I am not familiar with it and cannot
> find a nice way to explain that in the doc :/.

Point here should be that there's cgroups v1 and v2 infra. Both are available
and you can use a mixture of them, but net_cls is v1 only whereas bpf progs on
cgroups is v2 only feature. And if you look at struct sock_cgroup_data, it's
a union for the socket, so you can only use one of them with regards to sockets.

>>> + * 		related kernel documentation, available from the Linux sources
>>> + * 		in file *Documentation/cgroup-v1/net_cls.txt*.
>>> + * 	Return
>>> + * 		The classid, or 0 for the default unconfigured classid.
>>> + *
>>>   * int bpf_skb_vlan_push(struct sk_buff *skb, __be16 vlan_proto, u16 vlan_tci)
>>>   * 	Description
>>>   * 		Push a *vlan_tci* (VLAN tag control information) of protocol
>>> @@ -734,6 +772,16 @@ union bpf_attr {
>>>   * 		are **TC_ACT_REDIRECT** on success or **TC_ACT_SHOT** on
>>>   * 		error.
>>>   *
>>> + * u32 bpf_get_route_realm(struct sk_buff *skb)
>>> + * 	Description
>>> + * 		Retrieve the realm or the route, that is to say the
>>> + * 		**tclassid** field of the destination for the *skb*. This
>>> + * 		helper is available only if the kernel was compiled with
>>> + * 		**CONFIG_IP_ROUTE_CLASSID** configuration option.
>>
>> Could mention that this is a similar user provided tag like in the net_cls
>> case with cgroups only that this applies to routes here (dst entries) which
>> hold this tag.
>>
>> Also, should say that this works with clsact TC egress hook or alternatively
>> conventional classful egress qdiscs, but not on TC ingress. In case of clsact
>> TC egress hook this has the advantage that the dst entry has not been dropped
>> yet in the xmit path. Therefore, the dst entry does not need to be artificially
>> held via netif_keep_dst() for a classful qdisc until the skb is freed.
> 
> Here I am not sure to follow, the advantage is for clsact, but this is
> only for a classful qdisc that we can avoid holding the dst entry? How

No, it's only for sch_clsact where we don't need to hold it. Take a look at
__dev_queue_xmit(), and where sch_handle_egress() is called. It's right before
the dev->priv_flags & IFF_XMIT_DST_RELEASE check where we either hold or drop
the dst from skb.

In cls_bpf_prog_from_efd() the tcf_block_netif_keep_dst() will handle this.
If you call cls_bpf e.g. out of htb via htb_classify() -> tcf_classify()
then you would have to hold it and thus call netif_keep_dst() so that the
test in __dev_queue_xmit() results in holding the dst via skb_dst_force().

Cheers,
Daniel

[PATCH bpf-next v3 5/8] bpf: add documentation for eBPF helpers (33-41)

[Quentin Monnet]

Add documentation for eBPF helper functions to bpf.h user header file.
This documentation can be parsed with the Python script provided in
another commit of the patch series, in order to provide a RST document
that can later be converted into a man page.

The objective is to make the documentation easily understandable and
accessible to all eBPF developers, including beginners.

This patch contains descriptions for the following helper functions, all
written by Daniel:

- bpf_get_hash_recalc()
- bpf_skb_change_tail()
- bpf_skb_pull_data()
- bpf_csum_update()
- bpf_set_hash_invalid()
- bpf_get_numa_node_id()
- bpf_set_hash()
- bpf_skb_adjust_room()
- bpf_xdp_adjust_meta()

Cc: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>
---
 include/uapi/linux/bpf.h | 155 +++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 155 insertions(+)

diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
index d748f65a8f58..3a40f5debac2 100644
--- a/include/uapi/linux/bpf.h
+++ b/include/uapi/linux/bpf.h
@@ -965,9 +965,164 @@ union bpf_attr {
  * 	Return
  * 		0 on success, or a negative error in case of failure.
  *
+ * u32 bpf_get_hash_recalc(struct sk_buff *skb)
+ * 	Description
+ * 		Retrieve the hash of the packet, *skb*\ **->hash**. If it is
+ * 		not set, in particular if the hash was cleared due to mangling,
+ * 		recompute this hash. Later accesses to the hash can be done
+ * 		directly with *skb*\ **->hash**.
+ *
+ * 		Calling **bpf_set_hash_invalid**\ (), changing a packet
+ * 		prototype with **bpf_skb_change_proto**\ (), or calling
+ * 		**bpf_skb_store_bytes**\ () with the
+ * 		**BPF_F_INVALIDATE_HASH** are actions susceptible to clear
+ * 		the hash and to trigger a new computation for the next call to
+ * 		**bpf_get_hash_recalc**\ ().
+ * 	Return
+ * 		The 32-bit hash.
+ *
  * u64 bpf_get_current_task(void)
  * 	Return
  * 		A pointer to the current task struct.
+ *
+ * int bpf_skb_change_tail(struct sk_buff *skb, u32 len, u64 flags)
+ * 	Description
+ * 		Resize (trim or grow) the packet associated to *skb* to the
+ * 		new *len*. The *flags* are reserved for future usage, and must
+ * 		be left at zero.
+ *
+ * 		The basic idea is that the helper performs the needed work to
+ * 		change the size of the packet, then the eBPF program rewrites
+ * 		the rest via helpers like **bpf_skb_store_bytes**\ (),
+ * 		**bpf_l3_csum_replace**\ (), **bpf_l3_csum_replace**\ ()
+ * 		and others. This helper is a slow path utility intended for
+ * 		replies with control messages. And because it is targeted for
+ * 		slow path, the helper itself can afford to be slow: it
+ * 		implicitly linearizes, unclones and drops offloads from the
+ * 		*skb*.
+ *
+ * 		A call to this helper is susceptible to change data from the
+ * 		packet. Therefore, at load time, all checks on pointers
+ * 		previously done by the verifier are invalidated and must be
+ * 		performed again.
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
+ *
+ * int bpf_skb_pull_data(struct sk_buff *skb, u32 len)
+ * 	Description
+ * 		Pull in non-linear data in case the *skb* is non-linear and not
+ * 		all of *len* are part of the linear section. Make *len* bytes
+ * 		from *skb* readable and writable. If a zero value is passed for
+ * 		*len*, then the whole length of the *skb* is pulled.
+ *
+ * 		This helper is only needed for reading and writing with direct
+ * 		packet access.
+ *
+ * 		For direct packet access, when testing that offsets to access
+ * 		are within packet boundaries (test on *skb*\ **->data_end**)
+ * 		fails, programs just bail out, or, in the direct read case, use
+ * 		**bpf_skb_load_bytes()** as an alternative to overcome this
+ * 		limitation. If such data sits in non-linear parts, it is
+ * 		possible to pull them in once with the new helper, retest and
+ * 		eventually access them.
+ *
+ * 		At the same time, this also makes sure the skb is uncloned,
+ * 		which is a necessary condition for direct write. As this needs
+ * 		to be an invariant for the write part only, the verifier
+ * 		detects writes and adds a prologue that is calling
+ * 		**bpf_skb_pull_data()** to effectively unclone the skb from the
+ * 		very beginning in case it is indeed cloned.
+ *
+ * 		A call to this helper is susceptible to change data from the
+ * 		packet. Therefore, at load time, all checks on pointers
+ * 		previously done by the verifier are invalidated and must be
+ * 		performed again.
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
+ *
+ * s64 bpf_csum_update(struct sk_buff *skb, __wsum csum)
+ * 	Description
+ * 		Add the checksum *csum* into *skb*\ **->csum** in case the
+ * 		driver fed us an IP checksum. Return an error otherwise. This
+ * 		header is intended to be used in combination with
+ * 		**bpf_csum_diff()** helper, in particular when the checksum
+ * 		needs to be updated after data has been written into the packet
+ * 		through direct packet access.
+ * 	Return
+ * 		The checksum on success, or a negative error code in case of
+ * 		failure.
+ *
+ * void bpf_set_hash_invalid(struct sk_buff *skb)
+ * 	Description
+ * 		Invalidate the current *skb*\ **->hash**. It can be used after
+ * 		mangling on headers through direct packet access, in order to
+ * 		indicate that the hash is outdated and to trigger a
+ * 		recalculation the next time the kernel tries to access this
+ * 		hash.
+ *
+ * int bpf_get_numa_node_id(void)
+ * 	Description
+ * 		Return the id of the current NUMA node. The primary use case
+ * 		for this helper is the selection of sockets for the local NUMA
+ * 		node, when the program is attached to sockets using the
+ * 		**SO_ATTACH_REUSEPORT_EBPF** option (see also **socket(7)**).
+ * 	Return
+ * 		The id of current NUMA node.
+ *
+ * u32 bpf_set_hash(struct sk_buff *skb, u32 hash)
+ * 	Description
+ * 		Set the full hash for *skb* (set the field *skb*\ **->hash**)
+ * 		to value *hash*.
+ * 	Return
+ * 		0
+ *
+ * int bpf_skb_adjust_room(struct sk_buff *skb, u32 len_diff, u32 mode, u64 flags)
+ * 	Description
+ * 		Grow or shrink the room for data in the packet associated to
+ * 		*skb* by *len_diff*, and according to the selected *mode*.
+ *
+ * 		There is a single supported mode at this time:
+ *
+ * 		* **BPF_ADJ_ROOM_NET**: Adjust room at the network layer
+ * 		  (room space is added or removed below the layer 3 header).
+ *
+ * 		All values for *flags* are reserved for future usage, and must
+ * 		be left at zero.
+ *
+ * 		A call to this helper is susceptible to change data from the
+ * 		packet. Therefore, at load time, all checks on pointers
+ * 		previously done by the verifier are invalidated and must be
+ * 		performed again.
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
+ *
+ * int bpf_xdp_adjust_meta(struct xdp_buff *xdp_md, int delta)
+ * 	Description
+ * 		Adjust the address pointed by *xdp_md*\ **->data_meta** by
+ * 		*delta* (which can be positive or negative). Note that this
+ * 		operation modifies the address stored in *xdp_md*\ **->data**,
+ * 		so the latter must be loaded only after the helper has been
+ * 		called.
+ *
+ * 		The use of *xdp_md*\ **->data_meta** is optional and programs
+ * 		are not required to use it. The rationale is that when the
+ * 		packet is processed with XDP (e.g. as DoS filter), it is
+ * 		possible to push further meta data along with it before passing
+ * 		to the stack, and to give the guarantee that an ingress eBPF
+ * 		program attached as a TC classifier on the same device can pick
+ * 		this up for further post-processing. Since TC works with socket
+ * 		buffers, it remains possible to set from XDP the **mark** or
+ * 		**priority** pointers, or other pointers for the socket buffer.
+ * 		Having this scratch space generic and programmable allows for
+ * 		more flexibility as the user is free to store whatever meta
+ * 		data they need.
+ *
+ * 		A call to this helper is susceptible to change data from the
+ * 		packet. Therefore, at load time, all checks on pointers
+ * 		previously done by the verifier are invalidated and must be
+ * 		performed again.
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
  */
 #define __BPF_FUNC_MAPPER(FN)		\
 	FN(unspec),			\
-- 
2.14.1

Re: [PATCH bpf-next v3 5/8] bpf: add documentation for eBPF helpers (33-41)

[Alexei Starovoitov]

On Tue, Apr 17, 2018 at 03:34:35PM +0100, Quentin Monnet wrote:
> Add documentation for eBPF helper functions to bpf.h user header file.
> This documentation can be parsed with the Python script provided in
> another commit of the patch series, in order to provide a RST document
> that can later be converted into a man page.
> 
> The objective is to make the documentation easily understandable and
> accessible to all eBPF developers, including beginners.
> 
> This patch contains descriptions for the following helper functions, all
> written by Daniel:
> 
> - bpf_get_hash_recalc()
> - bpf_skb_change_tail()
> - bpf_skb_pull_data()
> - bpf_csum_update()
> - bpf_set_hash_invalid()
> - bpf_get_numa_node_id()
> - bpf_set_hash()
> - bpf_skb_adjust_room()
> - bpf_xdp_adjust_meta()
> 
> Cc: Daniel Borkmann <daniel@iogearbox.net>
> Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>

Acked-by: Alexei Starovoitov <ast@kernel.org>

Re: [PATCH bpf-next v3 5/8] bpf: add documentation for eBPF helpers (33-41)

[Daniel Borkmann]

On 04/17/2018 04:34 PM, Quentin Monnet wrote:
> Add documentation for eBPF helper functions to bpf.h user header file.
> This documentation can be parsed with the Python script provided in
> another commit of the patch series, in order to provide a RST document
> that can later be converted into a man page.
> 
> The objective is to make the documentation easily understandable and
> accessible to all eBPF developers, including beginners.
> 
> This patch contains descriptions for the following helper functions, all
> written by Daniel:
> 
> - bpf_get_hash_recalc()
> - bpf_skb_change_tail()
> - bpf_skb_pull_data()
> - bpf_csum_update()
> - bpf_set_hash_invalid()
> - bpf_get_numa_node_id()
> - bpf_set_hash()
> - bpf_skb_adjust_room()
> - bpf_xdp_adjust_meta()
> 
> Cc: Daniel Borkmann <daniel@iogearbox.net>
> Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>
> ---
>  include/uapi/linux/bpf.h | 155 +++++++++++++++++++++++++++++++++++++++++++++++
>  1 file changed, 155 insertions(+)
> 
> diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
> index d748f65a8f58..3a40f5debac2 100644
> --- a/include/uapi/linux/bpf.h
> +++ b/include/uapi/linux/bpf.h
> @@ -965,9 +965,164 @@ union bpf_attr {
>   * 	Return
>   * 		0 on success, or a negative error in case of failure.
>   *
> + * u32 bpf_get_hash_recalc(struct sk_buff *skb)
> + * 	Description
> + * 		Retrieve the hash of the packet, *skb*\ **->hash**. If it is
> + * 		not set, in particular if the hash was cleared due to mangling,
> + * 		recompute this hash. Later accesses to the hash can be done
> + * 		directly with *skb*\ **->hash**.
> + *
> + * 		Calling **bpf_set_hash_invalid**\ (), changing a packet
> + * 		prototype with **bpf_skb_change_proto**\ (), or calling
> + * 		**bpf_skb_store_bytes**\ () with the
> + * 		**BPF_F_INVALIDATE_HASH** are actions susceptible to clear
> + * 		the hash and to trigger a new computation for the next call to
> + * 		**bpf_get_hash_recalc**\ ().
> + * 	Return
> + * 		The 32-bit hash.
> + *
>   * u64 bpf_get_current_task(void)
>   * 	Return
>   * 		A pointer to the current task struct.
> + *
> + * int bpf_skb_change_tail(struct sk_buff *skb, u32 len, u64 flags)
> + * 	Description
> + * 		Resize (trim or grow) the packet associated to *skb* to the
> + * 		new *len*. The *flags* are reserved for future usage, and must
> + * 		be left at zero.
> + *
> + * 		The basic idea is that the helper performs the needed work to
> + * 		change the size of the packet, then the eBPF program rewrites
> + * 		the rest via helpers like **bpf_skb_store_bytes**\ (),
> + * 		**bpf_l3_csum_replace**\ (), **bpf_l3_csum_replace**\ ()
> + * 		and others. This helper is a slow path utility intended for
> + * 		replies with control messages. And because it is targeted for
> + * 		slow path, the helper itself can afford to be slow: it
> + * 		implicitly linearizes, unclones and drops offloads from the
> + * 		*skb*.
> + *
> + * 		A call to this helper is susceptible to change data from the
> + * 		packet. Therefore, at load time, all checks on pointers
> + * 		previously done by the verifier are invalidated and must be
> + * 		performed again.
> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
> + *
> + * int bpf_skb_pull_data(struct sk_buff *skb, u32 len)
> + * 	Description
> + * 		Pull in non-linear data in case the *skb* is non-linear and not
> + * 		all of *len* are part of the linear section. Make *len* bytes
> + * 		from *skb* readable and writable. If a zero value is passed for
> + * 		*len*, then the whole length of the *skb* is pulled.
> + *
> + * 		This helper is only needed for reading and writing with direct
> + * 		packet access.
> + *
> + * 		For direct packet access, when testing that offsets to access
> + * 		are within packet boundaries (test on *skb*\ **->data_end**)
> + * 		fails, programs just bail out, or, in the direct read case, use

I would add here to why it can fail, meaning either due to invalid offsets
or due to the requested data being in non-linear parts of the skb where then
either the bpf_skb_load_bytes() can be used as you mentioned or the data
pulled in via bpf_skb_pull_data().

> + * 		**bpf_skb_load_bytes()** as an alternative to overcome this
> + * 		limitation. If such data sits in non-linear parts, it is
> + * 		possible to pull them in once with the new helper, retest and
> + * 		eventually access them.

You do this here, but maybe slightly rearranging this one paragraph a bit as
to why one would use either of the helpers would help reading flow a bit.

> + * 		At the same time, this also makes sure the skb is uncloned,
> + * 		which is a necessary condition for direct write. As this needs
> + * 		to be an invariant for the write part only, the verifier
> + * 		detects writes and adds a prologue that is calling
> + * 		**bpf_skb_pull_data()** to effectively unclone the skb from the
> + * 		very beginning in case it is indeed cloned.
> + *
> + * 		A call to this helper is susceptible to change data from the
> + * 		packet. Therefore, at load time, all checks on pointers
> + * 		previously done by the verifier are invalidated and must be
> + * 		performed again.
> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
> + *
> + * s64 bpf_csum_update(struct sk_buff *skb, __wsum csum)
> + * 	Description
> + * 		Add the checksum *csum* into *skb*\ **->csum** in case the
> + * 		driver fed us an IP checksum. Return an error otherwise. This

It's not IP checksum specifically (if that is what you meant), it's when the
driver propagates CHECKSUM_COMPLETE to the skb, where the device has supplied
the checksum of the whole packet into skb->csum. At TC ingress time, this
covers everything starting from net header offset to the end of the skb since
mac hdr skb->csum has been pulled already. Main use case indeed direct packet
access.

> + * 		header is intended to be used in combination with
> + * 		**bpf_csum_diff()** helper, in particular when the checksum
> + * 		needs to be updated after data has been written into the packet
> + * 		through direct packet access.
> + * 	Return
> + * 		The checksum on success, or a negative error code in case of
> + * 		failure.
> + *
> + * void bpf_set_hash_invalid(struct sk_buff *skb)
> + * 	Description
> + * 		Invalidate the current *skb*\ **->hash**. It can be used after
> + * 		mangling on headers through direct packet access, in order to
> + * 		indicate that the hash is outdated and to trigger a
> + * 		recalculation the next time the kernel tries to access this
> + * 		hash.

[...] hash or through the helper bpf_get_hash_recalc().

> + *
> + * int bpf_get_numa_node_id(void)
> + * 	Description
> + * 		Return the id of the current NUMA node. The primary use case
> + * 		for this helper is the selection of sockets for the local NUMA
> + * 		node, when the program is attached to sockets using the
> + * 		**SO_ATTACH_REUSEPORT_EBPF** option (see also **socket(7)**).

I would mention that this also available for other types similarly to
bpf_get_smp_processor_id() helper though. (Otherwise one might read that
this could not be the case.)

> + * 	Return
> + * 		The id of current NUMA node.
> + *
> + * u32 bpf_set_hash(struct sk_buff *skb, u32 hash)
> + * 	Description
> + * 		Set the full hash for *skb* (set the field *skb*\ **->hash**)
> + * 		to value *hash*.
> + * 	Return
> + * 		0
> + *
> + * int bpf_skb_adjust_room(struct sk_buff *skb, u32 len_diff, u32 mode, u64 flags)
> + * 	Description
> + * 		Grow or shrink the room for data in the packet associated to
> + * 		*skb* by *len_diff*, and according to the selected *mode*.
> + *
> + * 		There is a single supported mode at this time:
> + *
> + * 		* **BPF_ADJ_ROOM_NET**: Adjust room at the network layer
> + * 		  (room space is added or removed below the layer 3 header).
> + *
> + * 		All values for *flags* are reserved for future usage, and must
> + * 		be left at zero.
> + *
> + * 		A call to this helper is susceptible to change data from the
> + * 		packet. Therefore, at load time, all checks on pointers
> + * 		previously done by the verifier are invalidated and must be
> + * 		performed again.
> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
> + *
> + * int bpf_xdp_adjust_meta(struct xdp_buff *xdp_md, int delta)
> + * 	Description
> + * 		Adjust the address pointed by *xdp_md*\ **->data_meta** by
> + * 		*delta* (which can be positive or negative). Note that this
> + * 		operation modifies the address stored in *xdp_md*\ **->data**,
> + * 		so the latter must be loaded only after the helper has been
> + * 		called.
> + *
> + * 		The use of *xdp_md*\ **->data_meta** is optional and programs
> + * 		are not required to use it. The rationale is that when the
> + * 		packet is processed with XDP (e.g. as DoS filter), it is
> + * 		possible to push further meta data along with it before passing
> + * 		to the stack, and to give the guarantee that an ingress eBPF
> + * 		program attached as a TC classifier on the same device can pick
> + * 		this up for further post-processing. Since TC works with socket
> + * 		buffers, it remains possible to set from XDP the **mark** or
> + * 		**priority** pointers, or other pointers for the socket buffer.
> + * 		Having this scratch space generic and programmable allows for
> + * 		more flexibility as the user is free to store whatever meta
> + * 		data they need.
> + *
> + * 		A call to this helper is susceptible to change data from the
> + * 		packet. Therefore, at load time, all checks on pointers
> + * 		previously done by the verifier are invalidated and must be
> + * 		performed again.
> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
>   */
>  #define __BPF_FUNC_MAPPER(FN)		\
>  	FN(unspec),			\
> 

[PATCH bpf-next v3 6/8] bpf: add documentation for eBPF helpers (42-50)

[Quentin Monnet]

Add documentation for eBPF helper functions to bpf.h user header file.
This documentation can be parsed with the Python script provided in
another commit of the patch series, in order to provide a RST document
that can later be converted into a man page.

The objective is to make the documentation easily understandable and
accessible to all eBPF developers, including beginners.

This patch contains descriptions for the following helper functions:

Helper from Kaixu:
- bpf_perf_event_read()

Helpers from Martin:
- bpf_skb_under_cgroup()
- bpf_xdp_adjust_head()

Helpers from Sargun:
- bpf_probe_write_user()
- bpf_current_task_under_cgroup()

Helper from Thomas:
- bpf_skb_change_head()

Helper from Gianluca:
- bpf_probe_read_str()

Helpers from Chenbo:
- bpf_get_socket_cookie()
- bpf_get_socket_uid()

v3:
- bpf_perf_event_read(): Fix time of selection for perf event type in
  description. Remove occurences of "cores" to avoid confusion with
  "CPU".

Cc: Kaixu Xia <xiakaixu@huawei.com>
Cc: Martin KaFai Lau <kafai@fb.com>
Cc: Sargun Dhillon <sargun@sargun.me>
Cc: Thomas Graf <tgraf@suug.ch>
Cc: Gianluca Borello <g.borello@gmail.com>
Cc: Chenbo Feng <fengc@google.com>
Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>
---
 include/uapi/linux/bpf.h | 158 +++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 158 insertions(+)

diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
index 3a40f5debac2..dd79a1c82adf 100644
--- a/include/uapi/linux/bpf.h
+++ b/include/uapi/linux/bpf.h
@@ -753,6 +753,25 @@ union bpf_attr {
  * 	Return
  * 		0 on success, or a negative error in case of failure.
  *
+ * u64 bpf_perf_event_read(struct bpf_map *map, u64 flags)
+ * 	Description
+ * 		Read the value of a perf event counter. This helper relies on a
+ * 		*map* of type **BPF_MAP_TYPE_PERF_EVENT_ARRAY**. The nature of
+ * 		the perf event counter is selected when *map* is updated with
+ * 		perf event file descriptors. The *map* is an array whose size
+ * 		is the number of available CPUs, and each cell contains a value
+ * 		relative to one CPU. The value to retrieve is indicated by
+ * 		*flags*, that contains the index of the CPU to look up, masked
+ * 		with **BPF_F_INDEX_MASK**. Alternatively, *flags* can be set to
+ * 		**BPF_F_CURRENT_CPU** to indicate that the value for the
+ * 		current CPU should be retrieved.
+ *
+ * 		Note that before Linux 4.13, only hardware perf event can be
+ * 		retrieved.
+ * 	Return
+ * 		The value of the perf event counter read from the map, or a
+ * 		negative error code in case of failure.
+ *
  * int bpf_redirect(u32 ifindex, u64 flags)
  * 	Description
  * 		Redirect the packet to another net device of index *ifindex*.
@@ -965,6 +984,17 @@ union bpf_attr {
  * 	Return
  * 		0 on success, or a negative error in case of failure.
  *
+ * int bpf_skb_under_cgroup(struct sk_buff *skb, struct bpf_map *map, u32 index)
+ * 	Description
+ * 		Check whether *skb* is a descendant of the cgroup2 held by
+ * 		*map* of type **BPF_MAP_TYPE_CGROUP_ARRAY**, at *index*.
+ * 	Return
+ * 		The return value depends on the result of the test, and can be:
+ *
+ * 		* 0, if the *skb* failed the cgroup2 descendant test.
+ * 		* 1, if the *skb* succeeded the cgroup2 descendant test.
+ * 		* A negative error code, if an error occurred.
+ *
  * u32 bpf_get_hash_recalc(struct sk_buff *skb)
  * 	Description
  * 		Retrieve the hash of the packet, *skb*\ **->hash**. If it is
@@ -985,6 +1015,37 @@ union bpf_attr {
  * 	Return
  * 		A pointer to the current task struct.
  *
+ * int bpf_probe_write_user(void *dst, const void *src, u32 len)
+ * 	Description
+ * 		Attempt in a safe way to write *len* bytes from the buffer
+ * 		*src* to *dst* in memory. It only works for threads that are in
+ * 		user context.
+ *
+ * 		This helper should not be used to implement any kind of
+ * 		security mechanism because of TOC-TOU attacks, but rather to
+ * 		debug, divert, and manipulate execution of semi-cooperative
+ * 		processes.
+ *
+ * 		Keep in mind that this feature is meant for experiments, and it
+ * 		has a risk of crashing the system and running programs.
+ * 		Therefore, when an eBPF program using this helper is attached,
+ * 		a warning including PID and process name is printed to kernel
+ * 		logs.
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
+ *
+ * int bpf_current_task_under_cgroup(struct bpf_map *map, u32 index)
+ * 	Description
+ * 		Check whether the probe is being run is the context of a given
+ * 		subset of the cgroup2 hierarchy. The cgroup2 to test is held by
+ * 		*map* of type **BPF_MAP_TYPE_CGROUP_ARRAY**, at *index*.
+ * 	Return
+ * 		The return value depends on the result of the test, and can be:
+ *
+ * 		* 0, if the *skb* task belongs to the cgroup2.
+ * 		* 1, if the *skb* task does not belong to the cgroup2.
+ * 		* A negative error code, if an error occurred.
+ *
  * int bpf_skb_change_tail(struct sk_buff *skb, u32 len, u64 flags)
  * 	Description
  * 		Resize (trim or grow) the packet associated to *skb* to the
@@ -1069,6 +1130,103 @@ union bpf_attr {
  * 	Return
  * 		The id of current NUMA node.
  *
+ * int bpf_skb_change_head(struct sk_buff *skb, u32 len, u64 flags)
+ * 	Description
+ * 		Grows headroom of packet associated to *skb* and adjusts the
+ * 		offset of the MAC header accordingly, adding *len* bytes of
+ * 		space. It automatically extends and reallocates memory as
+ * 		required.
+ *
+ * 		This helper can be used on a layer 3 *skb* to push a MAC header
+ * 		for redirection into a layer 2 device.
+ *
+ * 		All values for *flags* are reserved for future usage, and must
+ * 		be left at zero.
+ *
+ * 		A call to this helper is susceptible to change data from the
+ * 		packet. Therefore, at load time, all checks on pointers
+ * 		previously done by the verifier are invalidated and must be
+ * 		performed again.
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
+ *
+ * int bpf_xdp_adjust_head(struct xdp_buff *xdp_md, int delta)
+ * 	Description
+ * 		Adjust (move) *xdp_md*\ **->data** by *delta* bytes. Note that
+ * 		it is possible to use a negative value for *delta*. This helper
+ * 		can be used to prepare the packet for pushing or popping
+ * 		headers.
+ *
+ * 		A call to this helper is susceptible to change data from the
+ * 		packet. Therefore, at load time, all checks on pointers
+ * 		previously done by the verifier are invalidated and must be
+ * 		performed again.
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
+ *
+ * int bpf_probe_read_str(void *dst, int size, const void *unsafe_ptr)
+ * 	Description
+ * 		Copy a NUL terminated string from an unsafe address
+ * 		*unsafe_ptr* to *dst*. The *size* should include the
+ * 		terminating NUL byte. In case the string length is smaller than
+ * 		*size*, the target is not padded with further NUL bytes. If the
+ * 		string length is larger than *size*, just *size*-1 bytes are
+ * 		copied and the last byte is set to NUL.
+ *
+ * 		On success, the length of the copied string is returned. This
+ * 		makes this helper useful in tracing programs for reading
+ * 		strings, and more importantly to get its length at runtime. See
+ * 		the following snippet:
+ *
+ * 		::
+ *
+ * 			SEC("kprobe/sys_open")
+ * 			void bpf_sys_open(struct pt_regs *ctx)
+ * 			{
+ * 			        char buf[PATHLEN]; // PATHLEN is defined to 256
+ * 			        int res = bpf_probe_read_str(buf, sizeof(buf),
+ * 				                             ctx->di);
+ *
+ * 				// Consume buf, for example push it to
+ * 				// userspace via bpf_perf_event_output(); we
+ * 				// can use res (the string length) as event
+ * 				// size, after checking its boundaries.
+ * 			}
+ *
+ * 		In comparison, using **bpf_probe_read()** helper here instead
+ * 		to read the string would require to estimate the length at
+ * 		compile time, and would often result in copying more memory
+ * 		than necessary.
+ *
+ * 		Another useful use case is when parsing individual process
+ * 		arguments or individual environment variables navigating
+ * 		*current*\ **->mm->arg_start** and *current*\
+ * 		**->mm->env_start**: using this helper and the return value,
+ * 		one can quickly iterate at the right offset of the memory area.
+ * 	Return
+ * 		On success, the strictly positive length of the string,
+ * 		including the trailing NUL character. On error, a negative
+ * 		value.
+ *
+ * u64 bpf_get_socket_cookie(struct sk_buff *skb)
+ * 	Description
+ * 		Retrieve the socket cookie generated by the kernel from a
+ * 		**struct sk_buff** with a known socket. If none has been set
+ * 		yet, generate a new cookie. This helper can be useful for
+ * 		monitoring per socket networking traffic statistics as it
+ * 		provides a unique socket identifier per namespace.
+ * 	Return
+ * 		A 8-byte long non-decreasing number on success, or 0 if the
+ * 		socket field is missing inside *skb*.
+ *
+ * u32 bpf_get_socket_uid(struct sk_buff *skb)
+ * 	Return
+ * 		The owner UID of the socket associated to *skb*. If the socket
+ * 		is **NULL**, or if it is not a full socket (i.e. if it is a
+ * 		time-wait or a request socket instead), **overflowuid** value
+ * 		is returned (note that **overflowuid** might also be the actual
+ * 		UID value for the socket).
+ *
  * u32 bpf_set_hash(struct sk_buff *skb, u32 hash)
  * 	Description
  * 		Set the full hash for *skb* (set the field *skb*\ **->hash**)
-- 
2.14.1

Re: [PATCH bpf-next v3 6/8] bpf: add documentation for eBPF helpers (42-50)

[Alexei Starovoitov]

On Tue, Apr 17, 2018 at 03:34:36PM +0100, Quentin Monnet wrote:
> Add documentation for eBPF helper functions to bpf.h user header file.
> This documentation can be parsed with the Python script provided in
> another commit of the patch series, in order to provide a RST document
> that can later be converted into a man page.
> 
> The objective is to make the documentation easily understandable and
> accessible to all eBPF developers, including beginners.
> 
> This patch contains descriptions for the following helper functions:
> 
> Helper from Kaixu:
> - bpf_perf_event_read()
> 
> Helpers from Martin:
> - bpf_skb_under_cgroup()
> - bpf_xdp_adjust_head()
> 
> Helpers from Sargun:
> - bpf_probe_write_user()
> - bpf_current_task_under_cgroup()
> 
> Helper from Thomas:
> - bpf_skb_change_head()
> 
> Helper from Gianluca:
> - bpf_probe_read_str()
> 
> Helpers from Chenbo:
> - bpf_get_socket_cookie()
> - bpf_get_socket_uid()
> 
> v3:
> - bpf_perf_event_read(): Fix time of selection for perf event type in
>   description. Remove occurences of "cores" to avoid confusion with
>   "CPU".
> 
> Cc: Kaixu Xia <xiakaixu@huawei.com>
> Cc: Martin KaFai Lau <kafai@fb.com>
> Cc: Sargun Dhillon <sargun@sargun.me>
> Cc: Thomas Graf <tgraf@suug.ch>
> Cc: Gianluca Borello <g.borello@gmail.com>
> Cc: Chenbo Feng <fengc@google.com>
> Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>
...
> + *
> + * u64 bpf_get_socket_cookie(struct sk_buff *skb)
> + * 	Description
> + * 		Retrieve the socket cookie generated by the kernel from a
> + * 		**struct sk_buff** with a known socket. If none has been set

this bit could use some improvement, since it reads as cookie is
generated from sk_buff, whereas it has nothing to do with this particular
sk_buff. Cookie belongs to the socket and generated for the socket.
Would be good to explain that cookie is stable for the life of the socket.

For the rest:
Acked-by: Alexei Starovoitov <ast@kernel.org>

Re: [PATCH bpf-next v3 6/8] bpf: add documentation for eBPF helpers (42-50)

[Martin KaFai Lau]

On Tue, Apr 17, 2018 at 03:34:36PM +0100, Quentin Monnet wrote:
[...]
> @@ -965,6 +984,17 @@ union bpf_attr {
>   * 	Return
>   * 		0 on success, or a negative error in case of failure.
>   *
> + * int bpf_skb_under_cgroup(struct sk_buff *skb, struct bpf_map *map, u32 index)
> + * 	Description
> + * 		Check whether *skb* is a descendant of the cgroup2 held by
> + * 		*map* of type **BPF_MAP_TYPE_CGROUP_ARRAY**, at *index*.
> + * 	Return
> + * 		The return value depends on the result of the test, and can be:
> + *
> + * 		* 0, if the *skb* failed the cgroup2 descendant test.
> + * 		* 1, if the *skb* succeeded the cgroup2 descendant test.
> + * 		* A negative error code, if an error occurred.
> + *
[...]
> + * int bpf_xdp_adjust_head(struct xdp_buff *xdp_md, int delta)
> + * 	Description
> + * 		Adjust (move) *xdp_md*\ **->data** by *delta* bytes. Note that
> + * 		it is possible to use a negative value for *delta*. This helper
> + * 		can be used to prepare the packet for pushing or popping
> + * 		headers.
> + *
> + * 		A call to this helper is susceptible to change data from the
> + * 		packet. Therefore, at load time, all checks on pointers
> + * 		previously done by the verifier are invalidated and must be
> + * 		performed again.
> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
> + *
LGTM. Thanks!

Acked-by: Martin KaFai Lau <kafai@fb.com>

Re: [PATCH bpf-next v3 6/8] bpf: add documentation for eBPF helpers (42-50)

[Daniel Borkmann]

On 04/17/2018 04:34 PM, Quentin Monnet wrote:
> Add documentation for eBPF helper functions to bpf.h user header file.
> This documentation can be parsed with the Python script provided in
> another commit of the patch series, in order to provide a RST document
> that can later be converted into a man page.
> 
> The objective is to make the documentation easily understandable and
> accessible to all eBPF developers, including beginners.
> 
> This patch contains descriptions for the following helper functions:
> 
> Helper from Kaixu:
> - bpf_perf_event_read()
> 
> Helpers from Martin:
> - bpf_skb_under_cgroup()
> - bpf_xdp_adjust_head()
> 
> Helpers from Sargun:
> - bpf_probe_write_user()
> - bpf_current_task_under_cgroup()
> 
> Helper from Thomas:
> - bpf_skb_change_head()
> 
> Helper from Gianluca:
> - bpf_probe_read_str()
> 
> Helpers from Chenbo:
> - bpf_get_socket_cookie()
> - bpf_get_socket_uid()
> 
> v3:
> - bpf_perf_event_read(): Fix time of selection for perf event type in
>   description. Remove occurences of "cores" to avoid confusion with
>   "CPU".
> 
> Cc: Kaixu Xia <xiakaixu@huawei.com>
> Cc: Martin KaFai Lau <kafai@fb.com>
> Cc: Sargun Dhillon <sargun@sargun.me>
> Cc: Thomas Graf <tgraf@suug.ch>
> Cc: Gianluca Borello <g.borello@gmail.com>
> Cc: Chenbo Feng <fengc@google.com>
> Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>
> ---
>  include/uapi/linux/bpf.h | 158 +++++++++++++++++++++++++++++++++++++++++++++++
>  1 file changed, 158 insertions(+)
> 
> diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
> index 3a40f5debac2..dd79a1c82adf 100644
> --- a/include/uapi/linux/bpf.h
> +++ b/include/uapi/linux/bpf.h
> @@ -753,6 +753,25 @@ union bpf_attr {
>   * 	Return
>   * 		0 on success, or a negative error in case of failure.
>   *
> + * u64 bpf_perf_event_read(struct bpf_map *map, u64 flags)
> + * 	Description
> + * 		Read the value of a perf event counter. This helper relies on a
> + * 		*map* of type **BPF_MAP_TYPE_PERF_EVENT_ARRAY**. The nature of
> + * 		the perf event counter is selected when *map* is updated with
> + * 		perf event file descriptors. The *map* is an array whose size
> + * 		is the number of available CPUs, and each cell contains a value
> + * 		relative to one CPU. The value to retrieve is indicated by
> + * 		*flags*, that contains the index of the CPU to look up, masked
> + * 		with **BPF_F_INDEX_MASK**. Alternatively, *flags* can be set to
> + * 		**BPF_F_CURRENT_CPU** to indicate that the value for the
> + * 		current CPU should be retrieved.
> + *
> + * 		Note that before Linux 4.13, only hardware perf event can be
> + * 		retrieved.
> + * 	Return
> + * 		The value of the perf event counter read from the map, or a
> + * 		negative error code in case of failure.
> + *
>   * int bpf_redirect(u32 ifindex, u64 flags)
>   * 	Description
>   * 		Redirect the packet to another net device of index *ifindex*.
> @@ -965,6 +984,17 @@ union bpf_attr {
>   * 	Return
>   * 		0 on success, or a negative error in case of failure.
>   *
> + * int bpf_skb_under_cgroup(struct sk_buff *skb, struct bpf_map *map, u32 index)
> + * 	Description
> + * 		Check whether *skb* is a descendant of the cgroup2 held by
> + * 		*map* of type **BPF_MAP_TYPE_CGROUP_ARRAY**, at *index*.
> + * 	Return
> + * 		The return value depends on the result of the test, and can be:
> + *
> + * 		* 0, if the *skb* failed the cgroup2 descendant test.
> + * 		* 1, if the *skb* succeeded the cgroup2 descendant test.
> + * 		* A negative error code, if an error occurred.
> + *
>   * u32 bpf_get_hash_recalc(struct sk_buff *skb)
>   * 	Description
>   * 		Retrieve the hash of the packet, *skb*\ **->hash**. If it is
> @@ -985,6 +1015,37 @@ union bpf_attr {
>   * 	Return
>   * 		A pointer to the current task struct.
>   *
> + * int bpf_probe_write_user(void *dst, const void *src, u32 len)
> + * 	Description
> + * 		Attempt in a safe way to write *len* bytes from the buffer
> + * 		*src* to *dst* in memory. It only works for threads that are in
> + * 		user context.

Plus the dst address must be a valid user space address.

> + * 		This helper should not be used to implement any kind of
> + * 		security mechanism because of TOC-TOU attacks, but rather to
> + * 		debug, divert, and manipulate execution of semi-cooperative
> + * 		processes.
> + *
> + * 		Keep in mind that this feature is meant for experiments, and it
> + * 		has a risk of crashing the system and running programs.

Ditto, crashing user space applications.

> + * 		Therefore, when an eBPF program using this helper is attached,
> + * 		a warning including PID and process name is printed to kernel
> + * 		logs.
> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
> + *
> + * int bpf_current_task_under_cgroup(struct bpf_map *map, u32 index)
> + * 	Description
> + * 		Check whether the probe is being run is the context of a given
> + * 		subset of the cgroup2 hierarchy. The cgroup2 to test is held by
> + * 		*map* of type **BPF_MAP_TYPE_CGROUP_ARRAY**, at *index*.
> + * 	Return
> + * 		The return value depends on the result of the test, and can be:
> + *
> + * 		* 0, if the *skb* task belongs to the cgroup2.
> + * 		* 1, if the *skb* task does not belong to the cgroup2.
> + * 		* A negative error code, if an error occurred.
> + *
>   * int bpf_skb_change_tail(struct sk_buff *skb, u32 len, u64 flags)
>   * 	Description
>   * 		Resize (trim or grow) the packet associated to *skb* to the
> @@ -1069,6 +1130,103 @@ union bpf_attr {
>   * 	Return
>   * 		The id of current NUMA node.
>   *
> + * int bpf_skb_change_head(struct sk_buff *skb, u32 len, u64 flags)
> + * 	Description
> + * 		Grows headroom of packet associated to *skb* and adjusts the
> + * 		offset of the MAC header accordingly, adding *len* bytes of
> + * 		space. It automatically extends and reallocates memory as
> + * 		required.
> + *
> + * 		This helper can be used on a layer 3 *skb* to push a MAC header
> + * 		for redirection into a layer 2 device.
> + *
> + * 		All values for *flags* are reserved for future usage, and must
> + * 		be left at zero.
> + *
> + * 		A call to this helper is susceptible to change data from the
> + * 		packet. Therefore, at load time, all checks on pointers
> + * 		previously done by the verifier are invalidated and must be
> + * 		performed again.
> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
> + *
> + * int bpf_xdp_adjust_head(struct xdp_buff *xdp_md, int delta)
> + * 	Description
> + * 		Adjust (move) *xdp_md*\ **->data** by *delta* bytes. Note that
> + * 		it is possible to use a negative value for *delta*. This helper
> + * 		can be used to prepare the packet for pushing or popping
> + * 		headers.
> + *
> + * 		A call to this helper is susceptible to change data from the
> + * 		packet. Therefore, at load time, all checks on pointers
> + * 		previously done by the verifier are invalidated and must be
> + * 		performed again.
> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
> + *
> + * int bpf_probe_read_str(void *dst, int size, const void *unsafe_ptr)
> + * 	Description
> + * 		Copy a NUL terminated string from an unsafe address
> + * 		*unsafe_ptr* to *dst*. The *size* should include the
> + * 		terminating NUL byte. In case the string length is smaller than
> + * 		*size*, the target is not padded with further NUL bytes. If the
> + * 		string length is larger than *size*, just *size*-1 bytes are
> + * 		copied and the last byte is set to NUL.
> + *
> + * 		On success, the length of the copied string is returned. This
> + * 		makes this helper useful in tracing programs for reading
> + * 		strings, and more importantly to get its length at runtime. See
> + * 		the following snippet:
> + *
> + * 		::
> + *
> + * 			SEC("kprobe/sys_open")
> + * 			void bpf_sys_open(struct pt_regs *ctx)
> + * 			{
> + * 			        char buf[PATHLEN]; // PATHLEN is defined to 256
> + * 			        int res = bpf_probe_read_str(buf, sizeof(buf),
> + * 				                             ctx->di);
> + *
> + * 				// Consume buf, for example push it to
> + * 				// userspace via bpf_perf_event_output(); we
> + * 				// can use res (the string length) as event
> + * 				// size, after checking its boundaries.
> + * 			}
> + *
> + * 		In comparison, using **bpf_probe_read()** helper here instead
> + * 		to read the string would require to estimate the length at
> + * 		compile time, and would often result in copying more memory
> + * 		than necessary.
> + *
> + * 		Another useful use case is when parsing individual process
> + * 		arguments or individual environment variables navigating
> + * 		*current*\ **->mm->arg_start** and *current*\
> + * 		**->mm->env_start**: using this helper and the return value,
> + * 		one can quickly iterate at the right offset of the memory area.
> + * 	Return
> + * 		On success, the strictly positive length of the string,
> + * 		including the trailing NUL character. On error, a negative
> + * 		value.
> + *
> + * u64 bpf_get_socket_cookie(struct sk_buff *skb)
> + * 	Description
> + * 		Retrieve the socket cookie generated by the kernel from a
> + * 		**struct sk_buff** with a known socket. If none has been set
> + * 		yet, generate a new cookie. This helper can be useful for
> + * 		monitoring per socket networking traffic statistics as it
> + * 		provides a unique socket identifier per namespace.
> + * 	Return
> + * 		A 8-byte long non-decreasing number on success, or 0 if the
> + * 		socket field is missing inside *skb*.
> + *
> + * u32 bpf_get_socket_uid(struct sk_buff *skb)
> + * 	Return
> + * 		The owner UID of the socket associated to *skb*. If the socket
> + * 		is **NULL**, or if it is not a full socket (i.e. if it is a
> + * 		time-wait or a request socket instead), **overflowuid** value
> + * 		is returned (note that **overflowuid** might also be the actual
> + * 		UID value for the socket).
> + *
>   * u32 bpf_set_hash(struct sk_buff *skb, u32 hash)
>   * 	Description
>   * 		Set the full hash for *skb* (set the field *skb*\ **->hash**)
> 

[PATCH bpf-next v3 7/8] bpf: add documentation for eBPF helpers (51-57)

[Quentin Monnet]

Add documentation for eBPF helper functions to bpf.h user header file.
This documentation can be parsed with the Python script provided in
another commit of the patch series, in order to provide a RST document
that can later be converted into a man page.

The objective is to make the documentation easily understandable and
accessible to all eBPF developers, including beginners.

This patch contains descriptions for the following helper functions:

Helpers from Lawrence:
- bpf_setsockopt()
- bpf_getsockopt()
- bpf_sock_ops_cb_flags_set()

Helpers from Yonghong:
- bpf_perf_event_read_value()
- bpf_perf_prog_read_value()

Helper from Josef:
- bpf_override_return()

Helper from Andrey:
- bpf_bind()

v3:
- bpf_perf_event_read_value(): Fix time of selection for perf event type
  in description. Remove occurences of "cores" to avoid confusion with
  "CPU".
- bpf_bind(): Remove last paragraph of description, which was off topic.

Cc: Lawrence Brakmo <brakmo@fb.com>
Cc: Yonghong Song <yhs@fb.com>
Cc: Josef Bacik <jbacik@fb.com>
Cc: Andrey Ignatov <rdna@fb.com>
Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>

fix patch 7: Yonghong and Andrey
---
 include/uapi/linux/bpf.h | 178 +++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 178 insertions(+)

diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
index dd79a1c82adf..350459c583de 100644
--- a/include/uapi/linux/bpf.h
+++ b/include/uapi/linux/bpf.h
@@ -1234,6 +1234,28 @@ union bpf_attr {
  * 	Return
  * 		0
  *
+ * int bpf_setsockopt(struct bpf_sock_ops_kern *bpf_socket, int level, int optname, char *optval, int optlen)
+ * 	Description
+ * 		Emulate a call to **setsockopt()** on the socket associated to
+ * 		*bpf_socket*, which must be a full socket. The *level* at
+ * 		which the option resides and the name *optname* of the option
+ * 		must be specified, see **setsockopt(2)** for more information.
+ * 		The option value of length *optlen* is pointed by *optval*.
+ *
+ * 		This helper actually implements a subset of **setsockopt()**.
+ * 		It supports the following *level*\ s:
+ *
+ * 		* **SOL_SOCKET**, which supports the following *optname*\ s:
+ * 		  **SO_RCVBUF**, **SO_SNDBUF**, **SO_MAX_PACING_RATE**,
+ * 		  **SO_PRIORITY**, **SO_RCVLOWAT**, **SO_MARK**.
+ * 		* **IPPROTO_TCP**, which supports the following *optname*\ s:
+ * 		  **TCP_CONGESTION**, **TCP_BPF_IW**,
+ * 		  **TCP_BPF_SNDCWND_CLAMP**.
+ * 		* **IPPROTO_IP**, which supports *optname* **IP_TOS**.
+ * 		* **IPPROTO_IPV6**, which supports *optname* **IPV6_TCLASS**.
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
+ *
  * int bpf_skb_adjust_room(struct sk_buff *skb, u32 len_diff, u32 mode, u64 flags)
  * 	Description
  * 		Grow or shrink the room for data in the packet associated to
@@ -1281,6 +1303,162 @@ union bpf_attr {
  * 		performed again.
  * 	Return
  * 		0 on success, or a negative error in case of failure.
+ *
+ * int bpf_perf_event_read_value(struct bpf_map *map, u64 flags, struct bpf_perf_event_value *buf, u32 buf_size)
+ * 	Description
+ * 		Read the value of a perf event counter, and store it into *buf*
+ * 		of size *buf_size*. This helper relies on a *map* of type
+ * 		**BPF_MAP_TYPE_PERF_EVENT_ARRAY**. The nature of the perf event
+ * 		counter is selected when *map* is updated with perf event file
+ * 		descriptors. The *map* is an array whose size is the number of
+ * 		available CPUs, and each cell contains a value relative to one
+ * 		CPU. The value to retrieve is indicated by *flags*, that
+ * 		contains the index of the CPU to look up, masked with
+ * 		**BPF_F_INDEX_MASK**. Alternatively, *flags* can be set to
+ * 		**BPF_F_CURRENT_CPU** to indicate that the value for the
+ * 		current CPU should be retrieved.
+ *
+ * 		This helper behaves in a way close to
+ * 		**bpf_perf_event_read**\ () helper, save that instead of
+ * 		just returning the value observed, it fills the *buf*
+ * 		structure. This allows for additional data to be retrieved: in
+ * 		particular, the enabled and running times (in *buf*\
+ * 		**->enabled** and *buf*\ **->running**, respectively) are
+ * 		copied.
+ *
+ * 		These values are interesting, because hardware PMU (Performance
+ * 		Monitoring Unit) counters are limited resources. When there are
+ * 		more PMU based perf events opened than available counters,
+ * 		kernel will multiplex these events so each event gets certain
+ * 		percentage (but not all) of the PMU time. In case that
+ * 		multiplexing happens, the number of samples or counter value
+ * 		will not reflect the case compared to when no multiplexing
+ * 		occurs. This makes comparison between different runs difficult.
+ * 		Typically, the counter value should be normalized before
+ * 		comparing to other experiments. The usual normalization is done
+ * 		as follows.
+ *
+ * 		::
+ *
+ * 			normalized_counter = counter * t_enabled / t_running
+ *
+ * 		Where t_enabled is the time enabled for event and t_running is
+ * 		the time running for event since last normalization. The
+ * 		enabled and running times are accumulated since the perf event
+ * 		open. To achieve scaling factor between two invocations of an
+ * 		eBPF program, users can can use CPU id as the key (which is
+ * 		typical for perf array usage model) to remember the previous
+ * 		value and do the calculation inside the eBPF program.
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
+ *
+ * int bpf_perf_prog_read_value(struct bpf_perf_event_data_kern *ctx, struct bpf_perf_event_value *buf, u32 buf_size)
+ * 	Description
+ * 		For en eBPF program attached to a perf event, retrieve the
+ * 		value of the event counter associated to *ctx* and store it in
+ * 		the structure pointed by *buf* and of size *buf_size*. Enabled
+ * 		and running times are also stored in the structure (see
+ * 		description of helper **bpf_perf_event_read_value**\ () for
+ * 		more details).
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
+ *
+ * int bpf_getsockopt(struct bpf_sock_ops_kern *bpf_socket, int level, int optname, char *optval, int optlen)
+ * 	Description
+ * 		Emulate a call to **getsockopt()** on the socket associated to
+ * 		*bpf_socket*, which must be a full socket. The *level* at
+ * 		which the option resides and the name *optname* of the option
+ * 		must be specified, see **getsockopt(2)** for more information.
+ * 		The retrieved value is stored in the structure pointed by
+ * 		*opval* and of length *optlen*.
+ *
+ * 		This helper actually implements a subset of **getsockopt()**.
+ * 		It supports the following *level*\ s:
+ *
+ * 		* **IPPROTO_TCP**, which supports *optname*
+ * 		  **TCP_CONGESTION**.
+ * 		* **IPPROTO_IP**, which supports *optname* **IP_TOS**.
+ * 		* **IPPROTO_IPV6**, which supports *optname* **IPV6_TCLASS**.
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
+ *
+ * int bpf_override_return(struct pt_reg *regs, u64 rc)
+ * 	Description
+ * 		Used for error injection, this helper uses kprobes to override
+ * 		the return value of the probed function, and to set it to *rc*.
+ * 		The first argument is the context *regs* on which the kprobe
+ * 		works.
+ *
+ * 		This helper works by setting setting the PC (program counter)
+ * 		to an override function which is run in place of the original
+ * 		probed function. This means the probed function is not run at
+ * 		all. The replacement function just returns with the required
+ * 		value.
+ *
+ * 		This helper has security implications, and thus is subject to
+ * 		restrictions. It is only available if the kernel was compiled
+ * 		with the **CONFIG_BPF_KPROBE_OVERRIDE** configuration
+ * 		option, and in this case it only works on functions tagged with
+ * 		**ALLOW_ERROR_INJECTION** in the kernel code.
+ *
+ * 		Also, the helper is only available for the architectures having
+ * 		the CONFIG_FUNCTION_ERROR_INJECTION option. As of this writing,
+ * 		x86 architecture is the only one to support this feature.
+ * 	Return
+ * 		0
+ *
+ * int bpf_sock_ops_cb_flags_set(struct bpf_sock_ops_kern *bpf_sock, int argval)
+ * 	Description
+ * 		Attempt to set the value of the **bpf_sock_ops_cb_flags** field
+ * 		for the full TCP socket associated to *bpf_sock_ops* to
+ * 		*argval*.
+ *
+ * 		The primary use of this field is to determine if there should
+ * 		be calls to eBPF programs of type
+ * 		**BPF_PROG_TYPE_SOCK_OPS** at various points in the TCP
+ * 		code. A program of the same type can change its value, per
+ * 		connection and as necessary, when the connection is
+ * 		established. This field is directly accessible for reading, but
+ * 		this helper must be used for updates in order to return an
+ * 		error if an eBPF program tries to set a callback that is not
+ * 		supported in the current kernel.
+ *
+ * 		The supported callback values that *argval* can combine are:
+ *
+ * 		* **BPF_SOCK_OPS_RTO_CB_FLAG** (retransmission time out)
+ * 		* **BPF_SOCK_OPS_RETRANS_CB_FLAG** (retransmission)
+ * 		* **BPF_SOCK_OPS_STATE_CB_FLAG** (TCP state change)
+ *
+ * 		Here are some examples of where one could call such eBPF
+ * 		program:
+ *
+ * 		* When RTO fires.
+ * 		* When a packet is retransmitted.
+ * 		* When the connection terminates.
+ * 		* When a packet is sent.
+ * 		* When a packet is received.
+ * 	Return
+ * 		Code **-EINVAL** if the socket is not a full TCP socket;
+ * 		otherwise, a positive number containing the bits that could not
+ * 		be set is returned (which comes down to 0 if all bits were set
+ * 		as required).
+ *
+ * int bpf_bind(struct bpf_sock_addr_kern *ctx, struct sockaddr *addr, int addr_len)
+ * 	Description
+ * 		Bind the socket associated to *ctx* to the address pointed by
+ * 		*addr*, of length *addr_len*. This allows for making outgoing
+ * 		connection from the desired IP address, which can be useful for
+ * 		example when all processes inside a cgroup should use one
+ * 		single IP address on a host that has multiple IP configured.
+ *
+ * 		This helper works for IPv4 and IPv6, TCP and UDP sockets. The
+ * 		domain (*addr*\ **->sa_family**) must be **AF_INET** (or
+ * 		**AF_INET6**). Looking for a free port to bind to can be
+ * 		expensive, therefore binding to port is not permitted by the
+ * 		helper: *addr*\ **->sin_port** (or **sin6_port**, respectively)
+ * 		must be set to zero.
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
  */
 #define __BPF_FUNC_MAPPER(FN)		\
 	FN(unspec),			\
-- 
2.14.1

Re: [PATCH bpf-next v3 7/8] bpf: add documentation for eBPF helpers (51-57)

[Yonghong Song]

On 4/17/18 7:34 AM, Quentin Monnet wrote:
> Add documentation for eBPF helper functions to bpf.h user header file.
> This documentation can be parsed with the Python script provided in
> another commit of the patch series, in order to provide a RST document
> that can later be converted into a man page.
> 
> The objective is to make the documentation easily understandable and
> accessible to all eBPF developers, including beginners.
> 
> This patch contains descriptions for the following helper functions:
> 
> Helpers from Lawrence:
> - bpf_setsockopt()
> - bpf_getsockopt()
> - bpf_sock_ops_cb_flags_set()
> 
> Helpers from Yonghong:
> - bpf_perf_event_read_value()
> - bpf_perf_prog_read_value()
> 
> Helper from Josef:
> - bpf_override_return()
> 
> Helper from Andrey:
> - bpf_bind()
> 
> v3:
> - bpf_perf_event_read_value(): Fix time of selection for perf event type
>    in description. Remove occurences of "cores" to avoid confusion with
>    "CPU".
> - bpf_bind(): Remove last paragraph of description, which was off topic.
> 
> Cc: Lawrence Brakmo <brakmo@fb.com>
> Cc: Yonghong Song <yhs@fb.com>
> Cc: Josef Bacik <jbacik@fb.com>
> Cc: Andrey Ignatov <rdna@fb.com>
> Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>
> 
> fix patch 7: Yonghong and Andrey
> ---
>   include/uapi/linux/bpf.h | 178 +++++++++++++++++++++++++++++++++++++++++++++++
>   1 file changed, 178 insertions(+)
> 
> diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
> index dd79a1c82adf..350459c583de 100644
> --- a/include/uapi/linux/bpf.h
> +++ b/include/uapi/linux/bpf.h
> @@ -1234,6 +1234,28 @@ union bpf_attr {
>    * 	Return
>    * 		0
>    *
> + * int bpf_setsockopt(struct bpf_sock_ops_kern *bpf_socket, int level, int optname, char *optval, int optlen)
> + * 	Description
> + * 		Emulate a call to **setsockopt()** on the socket associated to
> + * 		*bpf_socket*, which must be a full socket. The *level* at
> + * 		which the option resides and the name *optname* of the option
> + * 		must be specified, see **setsockopt(2)** for more information.
> + * 		The option value of length *optlen* is pointed by *optval*.
> + *
> + * 		This helper actually implements a subset of **setsockopt()**.
> + * 		It supports the following *level*\ s:
> + *
> + * 		* **SOL_SOCKET**, which supports the following *optname*\ s:
> + * 		  **SO_RCVBUF**, **SO_SNDBUF**, **SO_MAX_PACING_RATE**,
> + * 		  **SO_PRIORITY**, **SO_RCVLOWAT**, **SO_MARK**.
> + * 		* **IPPROTO_TCP**, which supports the following *optname*\ s:
> + * 		  **TCP_CONGESTION**, **TCP_BPF_IW**,
> + * 		  **TCP_BPF_SNDCWND_CLAMP**.
> + * 		* **IPPROTO_IP**, which supports *optname* **IP_TOS**.
> + * 		* **IPPROTO_IPV6**, which supports *optname* **IPV6_TCLASS**.
> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
> + *
>    * int bpf_skb_adjust_room(struct sk_buff *skb, u32 len_diff, u32 mode, u64 flags)
>    * 	Description
>    * 		Grow or shrink the room for data in the packet associated to
> @@ -1281,6 +1303,162 @@ union bpf_attr {
>    * 		performed again.
>    * 	Return
>    * 		0 on success, or a negative error in case of failure.
> + *
> + * int bpf_perf_event_read_value(struct bpf_map *map, u64 flags, struct bpf_perf_event_value *buf, u32 buf_size)
> + * 	Description
> + * 		Read the value of a perf event counter, and store it into *buf*
> + * 		of size *buf_size*. This helper relies on a *map* of type
> + * 		**BPF_MAP_TYPE_PERF_EVENT_ARRAY**. The nature of the perf event
> + * 		counter is selected when *map* is updated with perf event file
> + * 		descriptors. The *map* is an array whose size is the number of
> + * 		available CPUs, and each cell contains a value relative to one
> + * 		CPU. The value to retrieve is indicated by *flags*, that
> + * 		contains the index of the CPU to look up, masked with
> + * 		**BPF_F_INDEX_MASK**. Alternatively, *flags* can be set to
> + * 		**BPF_F_CURRENT_CPU** to indicate that the value for the
> + * 		current CPU should be retrieved.
> + *
> + * 		This helper behaves in a way close to
> + * 		**bpf_perf_event_read**\ () helper, save that instead of
> + * 		just returning the value observed, it fills the *buf*
> + * 		structure. This allows for additional data to be retrieved: in
> + * 		particular, the enabled and running times (in *buf*\
> + * 		**->enabled** and *buf*\ **->running**, respectively) are
> + * 		copied.
> + *
> + * 		These values are interesting, because hardware PMU (Performance
> + * 		Monitoring Unit) counters are limited resources. When there are
> + * 		more PMU based perf events opened than available counters,
> + * 		kernel will multiplex these events so each event gets certain
> + * 		percentage (but not all) of the PMU time. In case that
> + * 		multiplexing happens, the number of samples or counter value
> + * 		will not reflect the case compared to when no multiplexing
> + * 		occurs. This makes comparison between different runs difficult.
> + * 		Typically, the counter value should be normalized before
> + * 		comparing to other experiments. The usual normalization is done
> + * 		as follows.
> + *
> + * 		::
> + *
> + * 			normalized_counter = counter * t_enabled / t_running
> + *
> + * 		Where t_enabled is the time enabled for event and t_running is
> + * 		the time running for event since last normalization. The
> + * 		enabled and running times are accumulated since the perf event
> + * 		open. To achieve scaling factor between two invocations of an
> + * 		eBPF program, users can can use CPU id as the key (which is
> + * 		typical for perf array usage model) to remember the previous
> + * 		value and do the calculation inside the eBPF program.
> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
> + *
> + * int bpf_perf_prog_read_value(struct bpf_perf_event_data_kern *ctx, struct bpf_perf_event_value *buf, u32 buf_size)
> + * 	Description
> + * 		For en eBPF program attached to a perf event, retrieve the
> + * 		value of the event counter associated to *ctx* and store it in
> + * 		the structure pointed by *buf* and of size *buf_size*. Enabled
> + * 		and running times are also stored in the structure (see
> + * 		description of helper **bpf_perf_event_read_value**\ () for
> + * 		more details).
> + * 	Return
> + * 		0 on success, or a negative error in case of failure.

Acked-by: Yonghong Song <yhs@fb.com>
for the following two helpers:
   - bpf_perf_event_read_value()
   - bpf_perf_prog_read_value()

Re: [PATCH bpf-next v3 7/8] bpf: add documentation for eBPF helpers (51-57)

[Andrey Ignatov]

Quentin Monnet <quentin.monnet@netronome.com> [Tue, 2018-04-17 07:36 -0700]:
> Add documentation for eBPF helper functions to bpf.h user header file.
> This documentation can be parsed with the Python script provided in
> another commit of the patch series, in order to provide a RST document
> that can later be converted into a man page.
> 
> The objective is to make the documentation easily understandable and
> accessible to all eBPF developers, including beginners.
> 
> This patch contains descriptions for the following helper functions:
> 
> Helpers from Lawrence:
> - bpf_setsockopt()
> - bpf_getsockopt()
> - bpf_sock_ops_cb_flags_set()
> 
> Helpers from Yonghong:
> - bpf_perf_event_read_value()
> - bpf_perf_prog_read_value()
> 
> Helper from Josef:
> - bpf_override_return()
> 
> Helper from Andrey:
> - bpf_bind()
> 
> v3:
> - bpf_perf_event_read_value(): Fix time of selection for perf event type
>   in description. Remove occurences of "cores" to avoid confusion with
>   "CPU".
> - bpf_bind(): Remove last paragraph of description, which was off topic.
> 
> Cc: Lawrence Brakmo <brakmo@fb.com>
> Cc: Yonghong Song <yhs@fb.com>
> Cc: Josef Bacik <jbacik@fb.com>
> Cc: Andrey Ignatov <rdna@fb.com>
> Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>
> 
> fix patch 7: Yonghong and Andrey
> ---
>  include/uapi/linux/bpf.h | 178 +++++++++++++++++++++++++++++++++++++++++++++++
>  1 file changed, 178 insertions(+)
> 
> diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
> index dd79a1c82adf..350459c583de 100644
> --- a/include/uapi/linux/bpf.h
> +++ b/include/uapi/linux/bpf.h
> + *
> + * int bpf_bind(struct bpf_sock_addr_kern *ctx, struct sockaddr *addr, int addr_len)
> + * 	Description
> + * 		Bind the socket associated to *ctx* to the address pointed by
> + * 		*addr*, of length *addr_len*. This allows for making outgoing
> + * 		connection from the desired IP address, which can be useful for
> + * 		example when all processes inside a cgroup should use one
> + * 		single IP address on a host that has multiple IP configured.
> + *
> + * 		This helper works for IPv4 and IPv6, TCP and UDP sockets. The
> + * 		domain (*addr*\ **->sa_family**) must be **AF_INET** (or
> + * 		**AF_INET6**). Looking for a free port to bind to can be
> + * 		expensive, therefore binding to port is not permitted by the
> + * 		helper: *addr*\ **->sin_port** (or **sin6_port**, respectively)
> + * 		must be set to zero.
> + * 	Return
> + * 		0 on success, or a negative error in case of failure.

Acked-by: Andrey Ignatov <rdna@fb.com>
for bpf_bind() helper.

-- 
Andrey Ignatov

Re: [PATCH bpf-next v3 7/8] bpf: add documentation for eBPF helpers (51-57)

[Daniel Borkmann]

On 04/17/2018 04:34 PM, Quentin Monnet wrote:
> Add documentation for eBPF helper functions to bpf.h user header file.
> This documentation can be parsed with the Python script provided in
> another commit of the patch series, in order to provide a RST document
> that can later be converted into a man page.
> 
> The objective is to make the documentation easily understandable and
> accessible to all eBPF developers, including beginners.
> 
> This patch contains descriptions for the following helper functions:
> 
> Helpers from Lawrence:
> - bpf_setsockopt()
> - bpf_getsockopt()
> - bpf_sock_ops_cb_flags_set()
> 
> Helpers from Yonghong:
> - bpf_perf_event_read_value()
> - bpf_perf_prog_read_value()
> 
> Helper from Josef:
> - bpf_override_return()
> 
> Helper from Andrey:
> - bpf_bind()
> 
> v3:
> - bpf_perf_event_read_value(): Fix time of selection for perf event type
>   in description. Remove occurences of "cores" to avoid confusion with
>   "CPU".
> - bpf_bind(): Remove last paragraph of description, which was off topic.
> 
> Cc: Lawrence Brakmo <brakmo@fb.com>
> Cc: Yonghong Song <yhs@fb.com>
> Cc: Josef Bacik <jbacik@fb.com>
> Cc: Andrey Ignatov <rdna@fb.com>
> Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>
> 
> fix patch 7: Yonghong and Andrey
> ---
>  include/uapi/linux/bpf.h | 178 +++++++++++++++++++++++++++++++++++++++++++++++
>  1 file changed, 178 insertions(+)
> 
> diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
> index dd79a1c82adf..350459c583de 100644
> --- a/include/uapi/linux/bpf.h
> +++ b/include/uapi/linux/bpf.h
> @@ -1234,6 +1234,28 @@ union bpf_attr {
>   * 	Return
>   * 		0
>   *
> + * int bpf_setsockopt(struct bpf_sock_ops_kern *bpf_socket, int level, int optname, char *optval, int optlen)
> + * 	Description
> + * 		Emulate a call to **setsockopt()** on the socket associated to
> + * 		*bpf_socket*, which must be a full socket. The *level* at
> + * 		which the option resides and the name *optname* of the option
> + * 		must be specified, see **setsockopt(2)** for more information.
> + * 		The option value of length *optlen* is pointed by *optval*.
> + *
> + * 		This helper actually implements a subset of **setsockopt()**.
> + * 		It supports the following *level*\ s:
> + *
> + * 		* **SOL_SOCKET**, which supports the following *optname*\ s:
> + * 		  **SO_RCVBUF**, **SO_SNDBUF**, **SO_MAX_PACING_RATE**,
> + * 		  **SO_PRIORITY**, **SO_RCVLOWAT**, **SO_MARK**.
> + * 		* **IPPROTO_TCP**, which supports the following *optname*\ s:
> + * 		  **TCP_CONGESTION**, **TCP_BPF_IW**,
> + * 		  **TCP_BPF_SNDCWND_CLAMP**.
> + * 		* **IPPROTO_IP**, which supports *optname* **IP_TOS**.
> + * 		* **IPPROTO_IPV6**, which supports *optname* **IPV6_TCLASS**.
> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
> + *
>   * int bpf_skb_adjust_room(struct sk_buff *skb, u32 len_diff, u32 mode, u64 flags)
>   * 	Description
>   * 		Grow or shrink the room for data in the packet associated to
> @@ -1281,6 +1303,162 @@ union bpf_attr {
>   * 		performed again.
>   * 	Return
>   * 		0 on success, or a negative error in case of failure.
> + *
> + * int bpf_perf_event_read_value(struct bpf_map *map, u64 flags, struct bpf_perf_event_value *buf, u32 buf_size)
> + * 	Description
> + * 		Read the value of a perf event counter, and store it into *buf*
> + * 		of size *buf_size*. This helper relies on a *map* of type
> + * 		**BPF_MAP_TYPE_PERF_EVENT_ARRAY**. The nature of the perf event
> + * 		counter is selected when *map* is updated with perf event file
> + * 		descriptors. The *map* is an array whose size is the number of
> + * 		available CPUs, and each cell contains a value relative to one
> + * 		CPU. The value to retrieve is indicated by *flags*, that
> + * 		contains the index of the CPU to look up, masked with
> + * 		**BPF_F_INDEX_MASK**. Alternatively, *flags* can be set to
> + * 		**BPF_F_CURRENT_CPU** to indicate that the value for the
> + * 		current CPU should be retrieved.
> + *
> + * 		This helper behaves in a way close to
> + * 		**bpf_perf_event_read**\ () helper, save that instead of
> + * 		just returning the value observed, it fills the *buf*
> + * 		structure. This allows for additional data to be retrieved: in
> + * 		particular, the enabled and running times (in *buf*\
> + * 		**->enabled** and *buf*\ **->running**, respectively) are
> + * 		copied.

Since you mention bpf_perf_event_read() here, we should mention that
bpf_perf_event_read_value() is recommended over bpf_perf_event_read()
in general. The latter bpf_perf_event_read() has some ABI quirks where
error and counter value are used as a return code (which is obviously
wrong to do since ranges may overlap). bpf_perf_event_read_value()
fixed this but also provides more features at the same time over the
old interface.

> + * 		These values are interesting, because hardware PMU (Performance
> + * 		Monitoring Unit) counters are limited resources. When there are
> + * 		more PMU based perf events opened than available counters,
> + * 		kernel will multiplex these events so each event gets certain
> + * 		percentage (but not all) of the PMU time. In case that
> + * 		multiplexing happens, the number of samples or counter value
> + * 		will not reflect the case compared to when no multiplexing
> + * 		occurs. This makes comparison between different runs difficult.
> + * 		Typically, the counter value should be normalized before
> + * 		comparing to other experiments. The usual normalization is done
> + * 		as follows.
> + *
> + * 		::
> + *
> + * 			normalized_counter = counter * t_enabled / t_running
> + *
> + * 		Where t_enabled is the time enabled for event and t_running is
> + * 		the time running for event since last normalization. The
> + * 		enabled and running times are accumulated since the perf event
> + * 		open. To achieve scaling factor between two invocations of an
> + * 		eBPF program, users can can use CPU id as the key (which is
> + * 		typical for perf array usage model) to remember the previous
> + * 		value and do the calculation inside the eBPF program.
> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
> + *
> + * int bpf_perf_prog_read_value(struct bpf_perf_event_data_kern *ctx, struct bpf_perf_event_value *buf, u32 buf_size)
> + * 	Description
> + * 		For en eBPF program attached to a perf event, retrieve the
> + * 		value of the event counter associated to *ctx* and store it in
> + * 		the structure pointed by *buf* and of size *buf_size*. Enabled
> + * 		and running times are also stored in the structure (see
> + * 		description of helper **bpf_perf_event_read_value**\ () for
> + * 		more details).

Ditto, mentioning here that bpf_perf_event_read_value() should be used instead.

> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
> + *
> + * int bpf_getsockopt(struct bpf_sock_ops_kern *bpf_socket, int level, int optname, char *optval, int optlen)
> + * 	Description
> + * 		Emulate a call to **getsockopt()** on the socket associated to
> + * 		*bpf_socket*, which must be a full socket. The *level* at
> + * 		which the option resides and the name *optname* of the option
> + * 		must be specified, see **getsockopt(2)** for more information.
> + * 		The retrieved value is stored in the structure pointed by
> + * 		*opval* and of length *optlen*.
> + *
> + * 		This helper actually implements a subset of **getsockopt()**.
> + * 		It supports the following *level*\ s:
> + *
> + * 		* **IPPROTO_TCP**, which supports *optname*
> + * 		  **TCP_CONGESTION**.
> + * 		* **IPPROTO_IP**, which supports *optname* **IP_TOS**.
> + * 		* **IPPROTO_IPV6**, which supports *optname* **IPV6_TCLASS**.
> + * 	Return
> + * 		0 on success, or a negative error in case of failure.
> + *
[...]

[PATCH bpf-next v3 8/8] bpf: add documentation for eBPF helpers (58-64)

[Quentin Monnet]

Add documentation for eBPF helper functions to bpf.h user header file.
This documentation can be parsed with the Python script provided in
another commit of the patch series, in order to provide a RST document
that can later be converted into a man page.

The objective is to make the documentation easily understandable and
accessible to all eBPF developers, including beginners.

This patch contains descriptions for the following helper functions, all
written by John:

- bpf_redirect_map()
- bpf_sk_redirect_map()
- bpf_sock_map_update()
- bpf_msg_redirect_map()
- bpf_msg_apply_bytes()
- bpf_msg_cork_bytes()
- bpf_msg_pull_data()

v3:
- bpf_sk_redirect_map(): Improve description of BPF_F_INGRESS flag.
- bpf_msg_redirect_map(): Improve description of BPF_F_INGRESS flag.
- bpf_redirect_map(): Fix note on CPU redirection, not fully implemented
  for generic XDP but supported on native XDP.

Cc: Jesper Dangaard Brouer <brouer@redhat.com>
Cc: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>
---
 include/uapi/linux/bpf.h | 140 +++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 140 insertions(+)

diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
index 350459c583de..3d329538498f 100644
--- a/include/uapi/linux/bpf.h
+++ b/include/uapi/linux/bpf.h
@@ -1276,6 +1276,50 @@ union bpf_attr {
  * 	Return
  * 		0 on success, or a negative error in case of failure.
  *
+ * int bpf_redirect_map(struct bpf_map *map, u32 key, u64 flags)
+ * 	Description
+ * 		Redirect the packet to the endpoint referenced by *map* at
+ * 		index *key*. Depending on its type, his *map* can contain
+ * 		references to net devices (for forwarding packets through other
+ * 		ports), or to CPUs (for redirecting XDP frames to another CPU;
+ * 		but this is only implemented for native XDP (with driver
+ * 		support) as of this writing).
+ *
+ * 		All values for *flags* are reserved for future usage, and must
+ * 		be left at zero.
+ * 	Return
+ * 		**XDP_REDIRECT** on success, or **XDP_ABORT** on error.
+ *
+ * int bpf_sk_redirect_map(struct bpf_map *map, u32 key, u64 flags)
+ * 	Description
+ * 		Redirect the packet to the socket referenced by *map* (of type
+ * 		**BPF_MAP_TYPE_SOCKMAP**) at index *key*. Both ingress and
+ * 		egress interfaces can be used for redirection. The
+ * 		**BPF_F_INGRESS** value in *flags* is used to make the
+ * 		distinction (ingress path is selected if the flag is present,
+ * 		egress path otherwise). This is the only flag supported for now.
+ * 	Return
+ * 		**SK_PASS** on success, or **SK_DROP** on error.
+ *
+ * int bpf_sock_map_update(struct bpf_sock_ops_kern *skops, struct bpf_map *map, void *key, u64 flags)
+ * 	Description
+ * 		Add an entry to, or update a *map* referencing sockets. The
+ * 		*skops* is used as a new value for the entry associated to
+ * 		*key*. *flags* is one of:
+ *
+ * 		**BPF_NOEXIST**
+ * 			The entry for *key* must not exist in the map.
+ * 		**BPF_EXIST**
+ * 			The entry for *key* must already exist in the map.
+ * 		**BPF_ANY**
+ * 			No condition on the existence of the entry for *key*.
+ *
+ * 		If the *map* has eBPF programs (parser and verdict), those will
+ * 		be inherited by the socket being added. If the socket is
+ * 		already attached to eBPF programs, this results in an error.
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
+ *
  * int bpf_xdp_adjust_meta(struct xdp_buff *xdp_md, int delta)
  * 	Description
  * 		Adjust the address pointed by *xdp_md*\ **->data_meta** by
@@ -1443,6 +1487,102 @@ union bpf_attr {
  * 		be set is returned (which comes down to 0 if all bits were set
  * 		as required).
  *
+ * int bpf_msg_redirect_map(struct sk_msg_buff *msg, struct bpf_map *map, u32 key, u64 flags)
+ * 	Description
+ * 		This helper is used in programs implementing policies at the
+ * 		socket level. If the message *msg* is allowed to pass (i.e. if
+ * 		the verdict eBPF program returns **SK_PASS**), redirect it to
+ * 		the socket referenced by *map* (of type
+ * 		**BPF_MAP_TYPE_SOCKMAP**) at index *key*. Both ingress and
+ * 		egress interfaces can be used for redirection. The
+ * 		**BPF_F_INGRESS** value in *flags* is used to make the
+ * 		distinction (ingress path is selected if the flag is present,
+ * 		egress path otherwise). This is the only flag supported for now.
+ * 	Return
+ * 		**SK_PASS** on success, or **SK_DROP** on error.
+ *
+ * int bpf_msg_apply_bytes(struct sk_msg_buff *msg, u32 bytes)
+ * 	Description
+ * 		For socket policies, apply the verdict of the eBPF program to
+ * 		the next *bytes* (number of bytes) of message *msg*.
+ *
+ * 		For example, this helper can be used in the following cases:
+ *
+ * 		* A single **sendmsg**\ () or **sendfile**\ () system call
+ * 		  contains multiple logical messages that the eBPF program is
+ * 		  supposed to read and for which it should apply a verdict.
+ * 		* An eBPF program only cares to read the first *bytes* of a
+ * 		  *msg*. If the message has a large payload, then setting up
+ * 		  and calling the eBPF program repeatedly for all bytes, even
+ * 		  though the verdict is already known, would create unnecessary
+ * 		  overhead.
+ *
+ * 		When called from within an eBPF program, the helper sets a
+ * 		counter internal to the BPF infrastructure, that is used to
+ * 		apply the last verdict to the next *bytes*. If *bytes* is
+ * 		smaller than the current data being processed from a
+ * 		**sendmsg**\ () or **sendfile**\ () system call, the first
+ * 		*bytes* will be sent and the eBPF program will be re-run with
+ * 		the pointer for start of data pointing to byte number *bytes*
+ * 		**+ 1**. If *bytes* is larger than the current data being
+ * 		processed, then the eBPF verdict will be applied to multiple
+ * 		**sendmsg**\ () or **sendfile**\ () calls until *bytes* are
+ * 		consumed.
+ *
+ * 		Note that if a socket closes with the internal counter holding
+ * 		a non-zero value, this is not a problem because data is not
+ * 		being buffered for *bytes* and is sent as it is received.
+ * 	Return
+ * 		0
+ *
+ * int bpf_msg_cork_bytes(struct sk_msg_buff *msg, u32 bytes)
+ * 	Description
+ * 		For socket policies, prevent the execution of the verdict eBPF
+ * 		program for message *msg* until *bytes* (byte number) have been
+ * 		accumulated.
+ *
+ * 		This can be used when one needs a specific number of bytes
+ * 		before a verdict can be assigned, even if the data spans
+ * 		multiple **sendmsg**\ () or **sendfile**\ () calls. The extreme
+ * 		case would be a user calling **sendmsg**\ () repeatedly with
+ * 		1-byte long message segments. Obviously, this is bad for
+ * 		performance, but it is still valid. If the eBPF program needs
+ * 		*bytes* bytes to validate a header, this helper can be used to
+ * 		prevent the eBPF program to be called again until *bytes* have
+ * 		been accumulated.
+ * 	Return
+ * 		0
+ *
+ * int bpf_msg_pull_data(struct sk_msg_buff *msg, u32 start, u32 end, u64 flags)
+ * 	Description
+ * 		For socket policies, pull in non-linear data from user space
+ * 		for *msg* and set pointers *msg*\ **->data** and *msg*\
+ * 		**->data_end** to *start* and *end* bytes offsets into *msg*,
+ * 		respectively.
+ *
+ * 		If a program of type **BPF_PROG_TYPE_SK_MSG** is run on a
+ * 		*msg* it can only parse data that the (**data**, **data_end**)
+ * 		pointers have already consumed. For **sendmsg**\ () hooks this
+ * 		is likely the first scatterlist element. But for calls relying
+ * 		on the **sendpage** handler (e.g. **sendfile**\ ()) this will
+ * 		be the range (**0**, **0**) because the data is shared with
+ * 		user space and by default the objective is to avoid allowing
+ * 		user space to modify data while (or after) eBPF verdict is
+ * 		being decided. This helper can be used to pull in data and to
+ * 		set the start and end pointer to given values. Data will be
+ * 		copied if necessary (i.e. if data was not linear and if start
+ * 		and end pointers do not point to the same chunk).
+ *
+ * 		A call to this helper is susceptible to change data from the
+ * 		packet. Therefore, at load time, all checks on pointers
+ * 		previously done by the verifier are invalidated and must be
+ * 		performed again.
+ *
+ * 		All values for *flags* are reserved for future usage, and must
+ * 		be left at zero.
+ * 	Return
+ * 		0 on success, or a negative error in case of failure.
+ *
  * int bpf_bind(struct bpf_sock_addr_kern *ctx, struct sockaddr *addr, int addr_len)
  * 	Description
  * 		Bind the socket associated to *ctx* to the address pointed by
-- 
2.14.1

Re: [PATCH bpf-next v3 8/8] bpf: add documentation for eBPF helpers (58-64)

[Jesper Dangaard Brouer]

On Tue, 17 Apr 2018 15:34:38 +0100
Quentin Monnet <quentin.monnet@netronome.com> wrote:

> diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
> index 350459c583de..3d329538498f 100644
> --- a/include/uapi/linux/bpf.h
> +++ b/include/uapi/linux/bpf.h
> @@ -1276,6 +1276,50 @@ union bpf_attr {
>   * 	Return
>   * 		0 on success, or a negative error in case of failure.
>   *
> + * int bpf_redirect_map(struct bpf_map *map, u32 key, u64 flags)
> + * 	Description
> + * 		Redirect the packet to the endpoint referenced by *map* at
> + * 		index *key*. Depending on its type, his *map* can contain
                                                    ^^^

"his" -> "this"

> + * 		references to net devices (for forwarding packets through other
> + * 		ports), or to CPUs (for redirecting XDP frames to another CPU;
> + * 		but this is only implemented for native XDP (with driver
> + * 		support) as of this writing).
> + *
> + * 		All values for *flags* are reserved for future usage, and must
> + * 		be left at zero.
> + * 	Return
> + * 		**XDP_REDIRECT** on success, or **XDP_ABORT** on error.
> + *

"XDP_ABORT" -> "XDP_ABORTED"

I don't know if it's worth mentioning in the doc/man-page; that for XDP
using bpf_redirect_map() is a HUGE performance advantage, compared to
the bpf_redirect() call ?

-- 
Best regards,
  Jesper Dangaard Brouer
  MSc.CS, Principal Kernel Engineer at Red Hat
  LinkedIn: http://www.linkedin.com/in/brouer

Re: [PATCH bpf-next v3 8/8] bpf: add documentation for eBPF helpers (58-64)

[Quentin Monnet]

2018-04-18 15:34 UTC+0200 ~ Jesper Dangaard Brouer <brouer@redhat.com>
> On Tue, 17 Apr 2018 15:34:38 +0100
> Quentin Monnet <quentin.monnet@netronome.com> wrote:
> 
>> diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
>> index 350459c583de..3d329538498f 100644
>> --- a/include/uapi/linux/bpf.h
>> +++ b/include/uapi/linux/bpf.h
>> @@ -1276,6 +1276,50 @@ union bpf_attr {
>>   * 	Return
>>   * 		0 on success, or a negative error in case of failure.
>>   *
>> + * int bpf_redirect_map(struct bpf_map *map, u32 key, u64 flags)
>> + * 	Description
>> + * 		Redirect the packet to the endpoint referenced by *map* at
>> + * 		index *key*. Depending on its type, his *map* can contain
>                                                     ^^^
> 
> "his" -> "this"

Thanks!

>> + * 		references to net devices (for forwarding packets through other
>> + * 		ports), or to CPUs (for redirecting XDP frames to another CPU;
>> + * 		but this is only implemented for native XDP (with driver
>> + * 		support) as of this writing).
>> + *
>> + * 		All values for *flags* are reserved for future usage, and must
>> + * 		be left at zero.
>> + * 	Return
>> + * 		**XDP_REDIRECT** on success, or **XDP_ABORT** on error.
>> + *
> 
> "XDP_ABORT" -> "XDP_ABORTED"

Ouch. And I did the same for bpf_redirect(). Thanks for the catch.

> 
> I don't know if it's worth mentioning in the doc/man-page; that for XDP
> using bpf_redirect_map() is a HUGE performance advantage, compared to
> the bpf_redirect() call ?

It seems worth to me. How would you simply explain the reason for this
difference?

Quentin

Re: [PATCH bpf-next v3 8/8] bpf: add documentation for eBPF helpers (58-64)

[Jesper Dangaard Brouer]

On Wed, 18 Apr 2018 15:09:41 +0100
Quentin Monnet <quentin.monnet@netronome.com> wrote:

> 2018-04-18 15:34 UTC+0200 ~ Jesper Dangaard Brouer <brouer@redhat.com>
> > On Tue, 17 Apr 2018 15:34:38 +0100
> > Quentin Monnet <quentin.monnet@netronome.com> wrote:
> >   
> >> diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
> >> index 350459c583de..3d329538498f 100644
> >> --- a/include/uapi/linux/bpf.h
> >> +++ b/include/uapi/linux/bpf.h
> >> @@ -1276,6 +1276,50 @@ union bpf_attr {
> >>   * 	Return
> >>   * 		0 on success, or a negative error in case of failure.
> >>   *
> >> + * int bpf_redirect_map(struct bpf_map *map, u32 key, u64 flags)
> >> + * 	Description
> >> + * 		Redirect the packet to the endpoint referenced by *map* at
> >> + * 		index *key*. Depending on its type, his *map* can contain  
> >                                                     ^^^
> > 
> > "his" -> "this"  
> 
> Thanks!
> 
> >> + * 		references to net devices (for forwarding packets through other
> >> + * 		ports), or to CPUs (for redirecting XDP frames to another CPU;
> >> + * 		but this is only implemented for native XDP (with driver
> >> + * 		support) as of this writing).
> >> + *
> >> + * 		All values for *flags* are reserved for future usage, and must
> >> + * 		be left at zero.
> >> + * 	Return
> >> + * 		**XDP_REDIRECT** on success, or **XDP_ABORT** on error.
> >> + *  
> > 
> > "XDP_ABORT" -> "XDP_ABORTED"  
> 
> Ouch. And I did the same for bpf_redirect(). Thanks for the catch.
> 
> > 
> > I don't know if it's worth mentioning in the doc/man-page; that for XDP
> > using bpf_redirect_map() is a HUGE performance advantage, compared to
> > the bpf_redirect() call ?  
> 
> It seems worth to me. How would you simply explain the reason for this
> difference?

The basic reason is "bulking effect", as devmap avoids the NIC
tailptr/doorbell update on every packet... how to write that in a doc
format?

I wrote about why XDP_REDIRECT with maps are smart here:
 http://vger.kernel.org/netconf2017_files/XDP_devel_update_NetConf2017_Seoul.pdf

Using maps for redirect, hopefully makes XDP_REDIRECT the last driver
XDP action code we need.  As new types of redirect can be introduced
without driver changes. See that AF_XDP also uses a map.

It is more subtle, but maps also function as a sorting step. Imagine
your XDP program need to redirect out different interfaces (or CPUs in
cpumap case), and packets arrive intermixed.  Packets get sorted into
the different map indexes, and the xdp_do_flush_map() will trigger the
flush operation.


Happened to have an i40e NIC benchmark setup, and ran a single flow pktgen test.

Results with 'xdp_redirect_map'
 13589297 pps (13,589,297) 

Results with 'xdp_redirect' NOT using devmap:
  7567575 pps (7,567,575)

Just to point out the performance benefit of devmap...

-- 
Best regards,
  Jesper Dangaard Brouer
  MSc.CS, Principal Kernel Engineer at Red Hat
  LinkedIn: http://www.linkedin.com/in/brouer

Re: [PATCH bpf-next v3 8/8] bpf: add documentation for eBPF helpers (58-64)

[Quentin Monnet]

2018-04-18 17:43 UTC+0200 ~ Jesper Dangaard Brouer <brouer@redhat.com>
> On Wed, 18 Apr 2018 15:09:41 +0100
> Quentin Monnet <quentin.monnet@netronome.com> wrote:
> 
>> 2018-04-18 15:34 UTC+0200 ~ Jesper Dangaard Brouer <brouer@redhat.com>
>>> On Tue, 17 Apr 2018 15:34:38 +0100
>>> Quentin Monnet <quentin.monnet@netronome.com> wrote:
>>>   
>>>> diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
>>>> index 350459c583de..3d329538498f 100644
>>>> --- a/include/uapi/linux/bpf.h
>>>> +++ b/include/uapi/linux/bpf.h
>>>> @@ -1276,6 +1276,50 @@ union bpf_attr {
>>>>   * 	Return
>>>>   * 		0 on success, or a negative error in case of failure.
>>>>   *
>>>> + * int bpf_redirect_map(struct bpf_map *map, u32 key, u64 flags)
>>>> + * 	Description
>>>> + * 		Redirect the packet to the endpoint referenced by *map* at
>>>> + * 		index *key*. Depending on its type, his *map* can contain  
>>>                                                     ^^^
>>>
>>> "his" -> "this"  
>>
>> Thanks!
>>
>>>> + * 		references to net devices (for forwarding packets through other
>>>> + * 		ports), or to CPUs (for redirecting XDP frames to another CPU;
>>>> + * 		but this is only implemented for native XDP (with driver
>>>> + * 		support) as of this writing).
>>>> + *
>>>> + * 		All values for *flags* are reserved for future usage, and must
>>>> + * 		be left at zero.
>>>> + * 	Return
>>>> + * 		**XDP_REDIRECT** on success, or **XDP_ABORT** on error.
>>>> + *  
>>>
>>> "XDP_ABORT" -> "XDP_ABORTED"  
>>
>> Ouch. And I did the same for bpf_redirect(). Thanks for the catch.
>>
>>>
>>> I don't know if it's worth mentioning in the doc/man-page; that for XDP
>>> using bpf_redirect_map() is a HUGE performance advantage, compared to
>>> the bpf_redirect() call ?  
>>
>> It seems worth to me. How would you simply explain the reason for this
>> difference?
> 
> The basic reason is "bulking effect", as devmap avoids the NIC
> tailptr/doorbell update on every packet... how to write that in a doc
> format?
> 
> I wrote about why XDP_REDIRECT with maps are smart here:
>  http://vger.kernel.org/netconf2017_files/XDP_devel_update_NetConf2017_Seoul.pdf
> 
> Using maps for redirect, hopefully makes XDP_REDIRECT the last driver
> XDP action code we need.  As new types of redirect can be introduced
> without driver changes. See that AF_XDP also uses a map.
> 
> It is more subtle, but maps also function as a sorting step. Imagine
> your XDP program need to redirect out different interfaces (or CPUs in
> cpumap case), and packets arrive intermixed.  Packets get sorted into
> the different map indexes, and the xdp_do_flush_map() will trigger the
> flush operation.
> 
> 
> Happened to have an i40e NIC benchmark setup, and ran a single flow pktgen test.
> 
> Results with 'xdp_redirect_map'
>  13589297 pps (13,589,297) 
> 
> Results with 'xdp_redirect' NOT using devmap:
>   7567575 pps (7,567,575)
> 
> Just to point out the performance benefit of devmap...


Thanks for those details! This is an impressive change in performance
indeed.

I think I will just keep it simple for the documentation. I will add the
following for bpf_redirect_map():

    When used to redirect packets to net devices, this helper
    provides a high performance increase over **bpf_redirect**\ ().
    This is due to various implementation details of the underlying
    mechanisms, one of which is the fact that **bpf_redirect_map**\ ()
    tries to send packet as a "bulk" to the device.

And also append the following to bpf_redirect():

    The same effect can be attained with the more generic
    **bpf_redirect_map**\ (), which requires specific maps
    to be used but offers better performance.

Best,
Quentin

Re: [PATCH bpf-next v3 8/8] bpf: add documentation for eBPF helpers (58-64)

[Jesper Dangaard Brouer]

On Thu, 19 Apr 2018 13:44:41 +0100
Quentin Monnet <quentin.monnet@netronome.com> wrote:

> 2018-04-18 17:43 UTC+0200 ~ Jesper Dangaard Brouer <brouer@redhat.com>
> > On Wed, 18 Apr 2018 15:09:41 +0100
> > Quentin Monnet <quentin.monnet@netronome.com> wrote:
> >   
> >> 2018-04-18 15:34 UTC+0200 ~ Jesper Dangaard Brouer <brouer@redhat.com>  
> >>> On Tue, 17 Apr 2018 15:34:38 +0100
> >>> Quentin Monnet <quentin.monnet@netronome.com> wrote:
> >>>     
> >>>> diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
> >>>> index 350459c583de..3d329538498f 100644
> >>>> --- a/include/uapi/linux/bpf.h
> >>>> +++ b/include/uapi/linux/bpf.h
> >>>> @@ -1276,6 +1276,50 @@ union bpf_attr {
> >>>>   * 	Return
> >>>>   * 		0 on success, or a negative error in case of failure.
> >>>>   *
> >>>> + * int bpf_redirect_map(struct bpf_map *map, u32 key, u64 flags)
> >>>> + * 	Description
> >>>> + * 		Redirect the packet to the endpoint referenced by *map* at
> >>>> + * 		index *key*. Depending on its type, his *map* can contain    
> >>>                                                     ^^^
> >>>
> >>> "his" -> "this"    
> >>
> >> Thanks!
> >>  
> >>>> + * 		references to net devices (for forwarding packets through other
> >>>> + * 		ports), or to CPUs (for redirecting XDP frames to another CPU;
> >>>> + * 		but this is only implemented for native XDP (with driver
> >>>> + * 		support) as of this writing).
> >>>> + *
> >>>> + * 		All values for *flags* are reserved for future usage, and must
> >>>> + * 		be left at zero.
> >>>> + * 	Return
> >>>> + * 		**XDP_REDIRECT** on success, or **XDP_ABORT** on error.
> >>>> + *    
> >>>
> >>> "XDP_ABORT" -> "XDP_ABORTED"    
> >>
> >> Ouch. And I did the same for bpf_redirect(). Thanks for the catch.
> >>  
> >>>
> >>> I don't know if it's worth mentioning in the doc/man-page; that for XDP
> >>> using bpf_redirect_map() is a HUGE performance advantage, compared to
> >>> the bpf_redirect() call ?    
> >>
> >> It seems worth to me. How would you simply explain the reason for this
> >> difference?  
> > 
> > The basic reason is "bulking effect", as devmap avoids the NIC
> > tailptr/doorbell update on every packet... how to write that in a doc
> > format?
> > 
> > I wrote about why XDP_REDIRECT with maps are smart here:
> >  http://vger.kernel.org/netconf2017_files/XDP_devel_update_NetConf2017_Seoul.pdf
> > 
> > Using maps for redirect, hopefully makes XDP_REDIRECT the last driver
> > XDP action code we need.  As new types of redirect can be introduced
> > without driver changes. See that AF_XDP also uses a map.
> > 
> > It is more subtle, but maps also function as a sorting step. Imagine
> > your XDP program need to redirect out different interfaces (or CPUs in
> > cpumap case), and packets arrive intermixed.  Packets get sorted into
> > the different map indexes, and the xdp_do_flush_map() will trigger the
> > flush operation.
> > 
> > 
> > Happened to have an i40e NIC benchmark setup, and ran a single flow pktgen test.
> > 
> > Results with 'xdp_redirect_map'
> >  13589297 pps (13,589,297) 
> > 
> > Results with 'xdp_redirect' NOT using devmap:
> >   7567575 pps (7,567,575)
> > 
> > Just to point out the performance benefit of devmap...  
> 
> 
> Thanks for those details! This is an impressive change in performance
> indeed.
> 
> I think I will just keep it simple for the documentation. I will add the
> following for bpf_redirect_map():
> 
>     When used to redirect packets to net devices, this helper
>     provides a high performance increase over **bpf_redirect**\ ().
>     This is due to various implementation details of the underlying
>     mechanisms, one of which is the fact that **bpf_redirect_map**\ ()
>     tries to send packet as a "bulk" to the device.
> 
> And also append the following to bpf_redirect():
> 
>     The same effect can be attained with the more generic
>     **bpf_redirect_map**\ (), which requires specific maps
>     to be used but offers better performance.

This sounds good to me! :-)

-- 
Best regards,
  Jesper Dangaard Brouer
  MSc.CS, Principal Kernel Engineer at Red Hat
  LinkedIn: http://www.linkedin.com/in/brouer