* Re: [PATCH v2 bpf-next] tools/bpf: add missing strings.h include
From: Alexei Starovoitov @ 2019-02-08 2:21 UTC (permalink / raw)
To: Andrii Nakryiko
Cc: acme, andrii.nakryiko, yhs, songliubraving, ast, kafai, netdev,
kernel-team
In-Reply-To: <20190207192924.2280663-1-andriin@fb.com>
On Thu, Feb 07, 2019 at 11:29:24AM -0800, Andrii Nakryiko wrote:
> Few files in libbpf are using bzero() function (defined in strings.h header), but
> don't include corresponding header. When libbpf is added as a dependency to pahole,
> this undeterministically causes warnings on some machines:
>
> bpf.c:225:2: warning: implicit declaration of function ‘bzero’ [-Wimplicit-function-declaration]
> bzero(&attr, sizeof(attr));
> ^~~~~
>
> Signed-off-by: Andrii Nakryiko <andriin@fb.com>
> Reported-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Applied, Thanks
^ permalink raw reply
* Re: [PATCH bpf-next v3] tools/bpf: add log_level to bpf_load_program_attr
From: Alexei Starovoitov @ 2019-02-08 2:31 UTC (permalink / raw)
To: Yonghong Song; +Cc: netdev, Alexei Starovoitov, Daniel Borkmann, kernel-team
In-Reply-To: <20190207173451.3474545-1-yhs@fb.com>
On Thu, Feb 07, 2019 at 09:34:51AM -0800, Yonghong Song wrote:
> The kernel verifier has three levels of logs:
> 0: no logs
> 1: logs mostly useful
> > 1: verbose
>
> Current libbpf API functions bpf_load_program_xattr() and
> bpf_load_program() cannot specify log_level.
> The bcc, however, provides an interface for user to
> specify log_level 2 for verbose output.
>
> This patch added log_level into structure
> bpf_load_program_attr, so users, including bcc, can use
> bpf_load_program_xattr() to change log_level. The
> supported log_level is 0, 1, and 2.
>
> The bpf selftest test_sock.c is modified to enable log_level = 2.
> If the "verbose" in test_sock.c is changed to true,
> the test will output logs like below:
> $ ./test_sock
> func#0 @0
> 0: R1=ctx(id=0,off=0,imm=0) R10=fp0,call_-1
> 0: (bf) r6 = r1
> 1: R1=ctx(id=0,off=0,imm=0) R6_w=ctx(id=0,off=0,imm=0) R10=fp0,call_-1
> 1: (61) r7 = *(u32 *)(r6 +28)
> invalid bpf_context access off=28 size=4
>
> Test case: bind4 load with invalid access: src_ip6 .. [PASS]
> ...
> Test case: bind6 allow all .. [PASS]
> Summary: 16 PASSED, 0 FAILED
>
> Some test_sock tests are negative tests and verbose verifier
> log will be printed out as shown in the above.
>
> Signed-off-by: Yonghong Song <yhs@fb.com>
> ---
> tools/lib/bpf/bpf.c | 22 +++++++++++++++++-----
> tools/lib/bpf/bpf.h | 1 +
> tools/testing/selftests/bpf/test_sock.c | 9 ++++++++-
> 3 files changed, 26 insertions(+), 6 deletions(-)
>
> Changelog:
> v2 -> v3:
> . returning error immediately for inconsistent log_level/log_buf/log_buf_sz
> values
> . for log_level=1/2, always doing logging during bpf_prog_load.
> v1 -> v2:
> . make log_level as the last member of struct bpf_load_program_attr.
> . return -EINVAL if bpf_load_program_attr.log_level > 2.
>
> diff --git a/tools/lib/bpf/bpf.c b/tools/lib/bpf/bpf.c
> index 3defad77dc7a..3f5290ee1b47 100644
> --- a/tools/lib/bpf/bpf.c
> +++ b/tools/lib/bpf/bpf.c
> @@ -214,10 +214,15 @@ int bpf_load_program_xattr(const struct bpf_load_program_attr *load_attr,
> {
> void *finfo = NULL, *linfo = NULL;
> union bpf_attr attr;
> + __u32 log_level;
> __u32 name_len;
> int fd;
>
> - if (!load_attr)
> + if (!load_attr || !log_buf != !log_buf_sz)
> + return -EINVAL;
> +
> + log_level = load_attr->log_level;
> + if (log_level > 2 || (log_level && !log_buf))
> return -EINVAL;
>
> name_len = load_attr->name ? strlen(load_attr->name) : 0;
> @@ -228,9 +233,16 @@ int bpf_load_program_xattr(const struct bpf_load_program_attr *load_attr,
> attr.insn_cnt = (__u32)load_attr->insns_cnt;
> attr.insns = ptr_to_u64(load_attr->insns);
> attr.license = ptr_to_u64(load_attr->license);
> - attr.log_buf = ptr_to_u64(NULL);
> - attr.log_size = 0;
> - attr.log_level = 0;
> +
> + attr.log_level = log_level;
> + if (log_level) {
> + attr.log_buf = ptr_to_u64(log_buf);
> + attr.log_size = log_buf_sz;
> + } else {
> + attr.log_buf = ptr_to_u64(NULL);
> + attr.log_size = 0;
> + }
> +
> attr.kern_version = load_attr->kern_version;
> attr.prog_ifindex = load_attr->prog_ifindex;
> attr.prog_btf_fd = load_attr->prog_btf_fd;
> @@ -286,7 +298,7 @@ int bpf_load_program_xattr(const struct bpf_load_program_attr *load_attr,
> goto done;
> }
>
> - if (!log_buf || !log_buf_sz)
> + if (log_level || !log_buf)
> goto done;
>
> /* Try again with log */
> diff --git a/tools/lib/bpf/bpf.h b/tools/lib/bpf/bpf.h
> index ed09eed2dc3b..6ffdd79bea89 100644
> --- a/tools/lib/bpf/bpf.h
> +++ b/tools/lib/bpf/bpf.h
> @@ -85,6 +85,7 @@ struct bpf_load_program_attr {
> __u32 line_info_rec_size;
> const void *line_info;
> __u32 line_info_cnt;
> + __u32 log_level;
> };
>
> /* Flags to direct loading requirements */
> diff --git a/tools/testing/selftests/bpf/test_sock.c b/tools/testing/selftests/bpf/test_sock.c
> index 561ffb6d6433..fb679ac3d4b0 100644
> --- a/tools/testing/selftests/bpf/test_sock.c
> +++ b/tools/testing/selftests/bpf/test_sock.c
> @@ -20,6 +20,7 @@
> #define MAX_INSNS 512
>
> char bpf_log_buf[BPF_LOG_BUF_SIZE];
> +static bool verbose = false;
If there was a command line flag to toggle 'verbose' it would have made it
easier to test. In this form test_sock.c needs to be manually edited
to see that log_level=2 is actually working.
Applied to bpf-next anyway.
^ permalink raw reply
* [PATCH v3 bpf-next 0/4] tools/btf: extend libbpf APIs to work with btf w/o kernel
From: Andrii Nakryiko @ 2019-02-08 2:55 UTC (permalink / raw)
To: alexei.starovoitov, andrii.nakryiko, songliubraving, yhs, ast,
kafai, netdev, kernel-team, daniel
Cc: Andrii Nakryiko
This patchset introduces a set of new APIs that make it possible to work with BTF
more effectively (and without involving kernel) for applications like pahole that
need to manipulate .BTF and .BTF.ext data.
Patch #1 changes existing btf__new() API call to only load and initialize
struct btf, while exposing new btf__load() API to attempt to load and validate
BTF in kernel.
Patch #2 adds btf__get_raw_data() API allowing to get access to raw BTF data from
struct btf.
Patch #3 adds similar btf_ext__get_raw_data() API for working with struct btf_ext.
Patch #4 removes not-yet-stable btf__get_strings() API which was added to be able
to test contents of struct btf for btf__dedup(). It's now superseded by raw APIs.
v2->v3:
- const void* variants of btf__get_raw_data()
- added btf_ext__get_raw_data()
- removed btf__get_strings() and adapated test_btf.c to use btf__get_raw_data()
v1->v2:
- btf_load() returns just error, not fd
- fix ordering in libbpf.map
Andrii Nakryiko (4):
btf: separate btf creation and loading
btf: expose API to work with raw btf data
btf: expose API to work with raw btf_ext data
tools/bpf: remove btf__get_strings superseded() by raw data API
tools/lib/bpf/btf.c | 145 +++++++++++++------------
tools/lib/bpf/btf.h | 6 +-
tools/lib/bpf/libbpf.c | 2 +-
tools/lib/bpf/libbpf.map | 3 +
tools/testing/selftests/bpf/test_btf.c | 39 ++++---
5 files changed, 110 insertions(+), 85 deletions(-)
--
2.17.1
^ permalink raw reply
* [PATCH v3 bpf-next 4/4] tools/bpf: remove btf__get_strings superseded() by raw data API
From: Andrii Nakryiko @ 2019-02-08 2:55 UTC (permalink / raw)
To: alexei.starovoitov, andrii.nakryiko, songliubraving, yhs, ast,
kafai, netdev, kernel-team, daniel
Cc: Andrii Nakryiko
In-Reply-To: <20190208025555.4027769-1-andriin@fb.com>
Now that we have btf__get_raw_data() it's trivial for tests to iterate
over all strings for testing purposes, which eliminates the need for
btf__get_strings() API.
Signed-off-by: Andrii Nakryiko <andriin@fb.com>
---
tools/lib/bpf/btf.c | 7 -----
tools/lib/bpf/btf.h | 2 --
tools/testing/selftests/bpf/test_btf.c | 39 +++++++++++++++++---------
3 files changed, 26 insertions(+), 22 deletions(-)
diff --git a/tools/lib/bpf/btf.c b/tools/lib/bpf/btf.c
index c87cc3d71b9f..a986dc28f17d 100644
--- a/tools/lib/bpf/btf.c
+++ b/tools/lib/bpf/btf.c
@@ -447,13 +447,6 @@ const void *btf__get_raw_data(const struct btf *btf, __u32 *size)
return btf->data;
}
-void btf__get_strings(const struct btf *btf, const char **strings,
- __u32 *str_len)
-{
- *strings = btf->strings;
- *str_len = btf->hdr->str_len;
-}
-
const char *btf__name_by_offset(const struct btf *btf, __u32 offset)
{
if (offset < btf->hdr->str_len)
diff --git a/tools/lib/bpf/btf.h b/tools/lib/bpf/btf.h
index ad9f648260c2..6179291f2cec 100644
--- a/tools/lib/bpf/btf.h
+++ b/tools/lib/bpf/btf.h
@@ -67,8 +67,6 @@ LIBBPF_API __s64 btf__resolve_size(const struct btf *btf, __u32 type_id);
LIBBPF_API int btf__resolve_type(const struct btf *btf, __u32 type_id);
LIBBPF_API int btf__fd(const struct btf *btf);
LIBBPF_API const void *btf__get_raw_data(const struct btf *btf, __u32 *size);
-LIBBPF_API void btf__get_strings(const struct btf *btf, const char **strings,
- __u32 *str_len);
LIBBPF_API const char *btf__name_by_offset(const struct btf *btf, __u32 offset);
LIBBPF_API int btf__get_from_id(__u32 id, struct btf **btf);
LIBBPF_API int btf__get_map_kv_tids(const struct btf *btf, const char *map_name,
diff --git a/tools/testing/selftests/bpf/test_btf.c b/tools/testing/selftests/bpf/test_btf.c
index 447acc34db94..bbcacba39590 100644
--- a/tools/testing/selftests/bpf/test_btf.c
+++ b/tools/testing/selftests/bpf/test_btf.c
@@ -5882,15 +5882,17 @@ static void dump_btf_strings(const char *strs, __u32 len)
static int do_test_dedup(unsigned int test_num)
{
const struct btf_dedup_test *test = &dedup_tests[test_num - 1];
- int err = 0, i;
- __u32 test_nr_types, expect_nr_types, test_str_len, expect_str_len;
- void *raw_btf;
- unsigned int raw_btf_size;
+ __u32 test_nr_types, expect_nr_types, test_btf_size, expect_btf_size;
+ const struct btf_header *test_hdr, *expect_hdr;
struct btf *test_btf = NULL, *expect_btf = NULL;
+ const void *test_btf_data, *expect_btf_data;
const char *ret_test_next_str, *ret_expect_next_str;
const char *test_strs, *expect_strs;
const char *test_str_cur, *test_str_end;
const char *expect_str_cur, *expect_str_end;
+ unsigned int raw_btf_size;
+ void *raw_btf;
+ int err = 0, i;
fprintf(stderr, "BTF dedup test[%u] (%s):", test_num, test->descr);
@@ -5927,23 +5929,34 @@ static int do_test_dedup(unsigned int test_num)
goto done;
}
- btf__get_strings(test_btf, &test_strs, &test_str_len);
- btf__get_strings(expect_btf, &expect_strs, &expect_str_len);
- if (CHECK(test_str_len != expect_str_len,
- "test_str_len:%u != expect_str_len:%u",
- test_str_len, expect_str_len)) {
+ test_btf_data = btf__get_raw_data(test_btf, &test_btf_size);
+ expect_btf_data = btf__get_raw_data(expect_btf, &expect_btf_size);
+ if (CHECK(test_btf_size != expect_btf_size,
+ "test_btf_size:%u != expect_btf_size:%u",
+ test_btf_size, expect_btf_size)) {
+ err = -1;
+ goto done;
+ }
+
+ test_hdr = test_btf_data;
+ test_strs = test_btf_data + test_hdr->str_off;
+ expect_hdr = expect_btf_data;
+ expect_strs = expect_btf_data + expect_hdr->str_off;
+ if (CHECK(test_hdr->str_len != expect_hdr->str_len,
+ "test_hdr->str_len:%u != expect_hdr->str_len:%u",
+ test_hdr->str_len, expect_hdr->str_len)) {
fprintf(stderr, "\ntest strings:\n");
- dump_btf_strings(test_strs, test_str_len);
+ dump_btf_strings(test_strs, test_hdr->str_len);
fprintf(stderr, "\nexpected strings:\n");
- dump_btf_strings(expect_strs, expect_str_len);
+ dump_btf_strings(expect_strs, expect_hdr->str_len);
err = -1;
goto done;
}
test_str_cur = test_strs;
- test_str_end = test_strs + test_str_len;
+ test_str_end = test_strs + test_hdr->str_len;
expect_str_cur = expect_strs;
- expect_str_end = expect_strs + expect_str_len;
+ expect_str_end = expect_strs + expect_hdr->str_len;
while (test_str_cur < test_str_end && expect_str_cur < expect_str_end) {
size_t test_len, expect_len;
--
2.17.1
^ permalink raw reply related
* [PATCH v3 bpf-next 1/4] btf: separate btf creation and loading
From: Andrii Nakryiko @ 2019-02-08 2:55 UTC (permalink / raw)
To: alexei.starovoitov, andrii.nakryiko, songliubraving, yhs, ast,
kafai, netdev, kernel-team, daniel
Cc: Andrii Nakryiko
In-Reply-To: <20190208025555.4027769-1-andriin@fb.com>
This change splits out previous btf__new functionality of constructing
struct btf and loading it into kernel into two:
- btf__new() just creates and initializes struct btf
- btf__load() attempts to load existing struct btf into kernel
btf__free will still close BTF fd, if it was ever loaded successfully
into kernel.
This change allows users of libbpf to manipulate BTF using its API,
without the need to unnecessarily load it into kernel.
One of the intended use cases is pahole using libbpf to do DWARF to BTF
conversion and deduplication using libbpf, while handling ELF sections
overwrites and other concerns on its own.
Fixes: 2d3feca8c44f ("bpf: btf: print map dump and lookup with btf info")
Signed-off-by: Andrii Nakryiko <andriin@fb.com>
Acked-by: Song Liu <songliubraving@fb.com>
---
tools/lib/bpf/btf.c | 53 ++++++++++++++++++++++------------------
tools/lib/bpf/btf.h | 1 +
tools/lib/bpf/libbpf.c | 2 +-
tools/lib/bpf/libbpf.map | 1 +
4 files changed, 32 insertions(+), 25 deletions(-)
diff --git a/tools/lib/bpf/btf.c b/tools/lib/bpf/btf.c
index ab6528c935a1..24ac64492d8a 100644
--- a/tools/lib/bpf/btf.c
+++ b/tools/lib/bpf/btf.c
@@ -366,8 +366,6 @@ void btf__free(struct btf *btf)
struct btf *btf__new(__u8 *data, __u32 size)
{
- __u32 log_buf_size = 0;
- char *log_buf = NULL;
struct btf *btf;
int err;
@@ -377,15 +375,6 @@ struct btf *btf__new(__u8 *data, __u32 size)
btf->fd = -1;
- log_buf = malloc(BPF_LOG_BUF_SIZE);
- if (!log_buf) {
- err = -ENOMEM;
- goto done;
- }
-
- *log_buf = 0;
- log_buf_size = BPF_LOG_BUF_SIZE;
-
btf->data = malloc(size);
if (!btf->data) {
err = -ENOMEM;
@@ -395,17 +384,6 @@ struct btf *btf__new(__u8 *data, __u32 size)
memcpy(btf->data, data, size);
btf->data_size = size;
- btf->fd = bpf_load_btf(btf->data, btf->data_size,
- log_buf, log_buf_size, false);
-
- if (btf->fd == -1) {
- err = -errno;
- pr_warning("Error loading BTF: %s(%d)\n", strerror(errno), errno);
- if (log_buf && *log_buf)
- pr_warning("%s\n", log_buf);
- goto done;
- }
-
err = btf_parse_hdr(btf);
if (err)
goto done;
@@ -417,8 +395,6 @@ struct btf *btf__new(__u8 *data, __u32 size)
err = btf_parse_type_sec(btf);
done:
- free(log_buf);
-
if (err) {
btf__free(btf);
return ERR_PTR(err);
@@ -427,6 +403,35 @@ struct btf *btf__new(__u8 *data, __u32 size)
return btf;
}
+int btf__load(struct btf* btf) {
+ __u32 log_buf_size = BPF_LOG_BUF_SIZE;
+ char *log_buf = NULL;
+ int err = 0;
+
+ if (btf->fd >= 0)
+ return -EEXIST;
+
+ log_buf = malloc(log_buf_size);
+ if (!log_buf)
+ return -ENOMEM;
+
+ *log_buf = 0;
+
+ btf->fd = bpf_load_btf(btf->data, btf->data_size,
+ log_buf, log_buf_size, false);
+ if (btf->fd < 0) {
+ err = -errno;
+ pr_warning("Error loading BTF: %s(%d)\n", strerror(errno), errno);
+ if (*log_buf)
+ pr_warning("%s\n", log_buf);
+ goto done;
+ }
+
+done:
+ free(log_buf);
+ return err;
+}
+
int btf__fd(const struct btf *btf)
{
return btf->fd;
diff --git a/tools/lib/bpf/btf.h b/tools/lib/bpf/btf.h
index b393da90cc85..6fbbde95abc0 100644
--- a/tools/lib/bpf/btf.h
+++ b/tools/lib/bpf/btf.h
@@ -57,6 +57,7 @@ struct btf_ext_header {
LIBBPF_API void btf__free(struct btf *btf);
LIBBPF_API struct btf *btf__new(__u8 *data, __u32 size);
+LIBBPF_API int btf__load(struct btf* btf);
LIBBPF_API __s32 btf__find_by_name(const struct btf *btf,
const char *type_name);
LIBBPF_API __u32 btf__get_nr_types(const struct btf *btf);
diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c
index 47969aa0faf8..ff86a43a4336 100644
--- a/tools/lib/bpf/libbpf.c
+++ b/tools/lib/bpf/libbpf.c
@@ -835,7 +835,7 @@ static int bpf_object__elf_collect(struct bpf_object *obj, int flags)
obj->efile.maps_shndx = idx;
else if (strcmp(name, BTF_ELF_SEC) == 0) {
obj->btf = btf__new(data->d_buf, data->d_size);
- if (IS_ERR(obj->btf)) {
+ if (IS_ERR(obj->btf) || btf__load(obj->btf)) {
pr_warning("Error loading ELF section %s: %ld. Ignored and continue.\n",
BTF_ELF_SEC, PTR_ERR(obj->btf));
obj->btf = NULL;
diff --git a/tools/lib/bpf/libbpf.map b/tools/lib/bpf/libbpf.map
index 89c1149e32ee..f5372df143f4 100644
--- a/tools/lib/bpf/libbpf.map
+++ b/tools/lib/bpf/libbpf.map
@@ -137,6 +137,7 @@ LIBBPF_0.0.2 {
btf__get_map_kv_tids;
btf__get_nr_types;
btf__get_strings;
+ btf__load;
btf_ext__free;
btf_ext__func_info_rec_size;
btf_ext__line_info_rec_size;
--
2.17.1
^ permalink raw reply related
* [PATCH v3 bpf-next 3/4] btf: expose API to work with raw btf_ext data
From: Andrii Nakryiko @ 2019-02-08 2:55 UTC (permalink / raw)
To: alexei.starovoitov, andrii.nakryiko, songliubraving, yhs, ast,
kafai, netdev, kernel-team, daniel
Cc: Andrii Nakryiko
In-Reply-To: <20190208025555.4027769-1-andriin@fb.com>
This patch changes struct btf_ext to retain original data in sequential
block of memory, which makes it possible to expose
btf_ext__get_raw_data() interface, that's similar to
btf__get_raw_data(), allowing users of libbpf to get access to raw
representation of .BTF.ext section.
Signed-off-by: Andrii Nakryiko <andriin@fb.com>
---
tools/lib/bpf/btf.c | 85 +++++++++++++++++++++-------------------
tools/lib/bpf/btf.h | 2 +
tools/lib/bpf/libbpf.map | 1 +
3 files changed, 47 insertions(+), 41 deletions(-)
diff --git a/tools/lib/bpf/btf.c b/tools/lib/bpf/btf.c
index 8730f6c3be9e..c87cc3d71b9f 100644
--- a/tools/lib/bpf/btf.c
+++ b/tools/lib/bpf/btf.c
@@ -41,9 +41,8 @@ struct btf {
struct btf_ext_info {
/*
- * info points to a deep copy of the individual info section
- * (e.g. func_info and line_info) from the .BTF.ext.
- * It does not include the __u32 rec_size.
+ * info points to the individual info section (e.g. func_info and
+ * line_info) from the .BTF.ext. It does not include the __u32 rec_size.
*/
void *info;
__u32 rec_size;
@@ -51,8 +50,13 @@ struct btf_ext_info {
};
struct btf_ext {
+ union {
+ struct btf_ext_header *hdr;
+ void *data;
+ };
struct btf_ext_info func_info;
struct btf_ext_info line_info;
+ __u32 data_size;
};
struct btf_ext_info_sec {
@@ -603,19 +607,13 @@ struct btf_ext_sec_copy_param {
};
static int btf_ext_copy_info(struct btf_ext *btf_ext,
- __u8 *data, __u32 data_size,
struct btf_ext_sec_copy_param *ext_sec)
{
- const struct btf_ext_header *hdr = (struct btf_ext_header *)data;
const struct btf_ext_info_sec *sinfo;
struct btf_ext_info *ext_info;
__u32 info_left, record_size;
/* The start of the info sec (including the __u32 record_size). */
- const void *info;
-
- /* data and data_size do not include btf_ext_header from now on */
- data = data + hdr->hdr_len;
- data_size -= hdr->hdr_len;
+ void *info;
if (ext_sec->off & 0x03) {
pr_debug(".BTF.ext %s section is not aligned to 4 bytes\n",
@@ -623,16 +621,15 @@ static int btf_ext_copy_info(struct btf_ext *btf_ext,
return -EINVAL;
}
- if (data_size < ext_sec->off ||
- ext_sec->len > data_size - ext_sec->off) {
+ info = btf_ext->data + btf_ext->hdr->hdr_len + ext_sec->off;
+ info_left = ext_sec->len;
+
+ if (btf_ext->data + btf_ext->data_size < info + ext_sec->len) {
pr_debug("%s section (off:%u len:%u) is beyond the end of the ELF section .BTF.ext\n",
- ext_sec->desc, ext_sec->off, ext_sec->len);
+ ext_sec->desc, ext_sec->off, ext_sec->len);
return -EINVAL;
}
- info = data + ext_sec->off;
- info_left = ext_sec->len;
-
/* At least a record size */
if (info_left < sizeof(__u32)) {
pr_debug(".BTF.ext %s record size not found\n", ext_sec->desc);
@@ -644,7 +641,7 @@ static int btf_ext_copy_info(struct btf_ext *btf_ext,
if (record_size < ext_sec->min_rec_size ||
record_size & 0x03) {
pr_debug("%s section in .BTF.ext has invalid record size %u\n",
- ext_sec->desc, record_size);
+ ext_sec->desc, record_size);
return -EINVAL;
}
@@ -690,42 +687,35 @@ static int btf_ext_copy_info(struct btf_ext *btf_ext,
ext_info = ext_sec->ext_info;
ext_info->len = ext_sec->len - sizeof(__u32);
ext_info->rec_size = record_size;
- ext_info->info = malloc(ext_info->len);
- if (!ext_info->info)
- return -ENOMEM;
- memcpy(ext_info->info, info + sizeof(__u32), ext_info->len);
+ ext_info->info = info + sizeof(__u32);
return 0;
}
-static int btf_ext_copy_func_info(struct btf_ext *btf_ext,
- __u8 *data, __u32 data_size)
+static int btf_ext_copy_func_info(struct btf_ext *btf_ext)
{
- const struct btf_ext_header *hdr = (struct btf_ext_header *)data;
struct btf_ext_sec_copy_param param = {
- .off = hdr->func_info_off,
- .len = hdr->func_info_len,
+ .off = btf_ext->hdr->func_info_off,
+ .len = btf_ext->hdr->func_info_len,
.min_rec_size = sizeof(struct bpf_func_info_min),
.ext_info = &btf_ext->func_info,
.desc = "func_info"
};
- return btf_ext_copy_info(btf_ext, data, data_size, ¶m);
+ return btf_ext_copy_info(btf_ext, ¶m);
}
-static int btf_ext_copy_line_info(struct btf_ext *btf_ext,
- __u8 *data, __u32 data_size)
+static int btf_ext_copy_line_info(struct btf_ext *btf_ext)
{
- const struct btf_ext_header *hdr = (struct btf_ext_header *)data;
struct btf_ext_sec_copy_param param = {
- .off = hdr->line_info_off,
- .len = hdr->line_info_len,
+ .off = btf_ext->hdr->line_info_off,
+ .len = btf_ext->hdr->line_info_len,
.min_rec_size = sizeof(struct bpf_line_info_min),
.ext_info = &btf_ext->line_info,
.desc = "line_info",
};
- return btf_ext_copy_info(btf_ext, data, data_size, ¶m);
+ return btf_ext_copy_info(btf_ext, ¶m);
}
static int btf_ext_parse_hdr(__u8 *data, __u32 data_size)
@@ -765,9 +755,7 @@ void btf_ext__free(struct btf_ext *btf_ext)
{
if (!btf_ext)
return;
-
- free(btf_ext->func_info.info);
- free(btf_ext->line_info.info);
+ free(btf_ext->data);
free(btf_ext);
}
@@ -784,13 +772,23 @@ struct btf_ext *btf_ext__new(__u8 *data, __u32 size)
if (!btf_ext)
return ERR_PTR(-ENOMEM);
- err = btf_ext_copy_func_info(btf_ext, data, size);
- if (err) {
- btf_ext__free(btf_ext);
- return ERR_PTR(err);
+ btf_ext->data_size = size;
+ btf_ext->data = malloc(size);
+ if (!btf_ext->data) {
+ err = -ENOMEM;
+ goto done;
}
+ memcpy(btf_ext->data, data, size);
+
+ err = btf_ext_copy_func_info(btf_ext);
+ if (err)
+ goto done;
+
+ err = btf_ext_copy_line_info(btf_ext);
+ if (err)
+ goto done;
- err = btf_ext_copy_line_info(btf_ext, data, size);
+done:
if (err) {
btf_ext__free(btf_ext);
return ERR_PTR(err);
@@ -799,6 +797,11 @@ struct btf_ext *btf_ext__new(__u8 *data, __u32 size)
return btf_ext;
}
+const void *btf_ext__get_raw_data(const struct btf_ext* btf_ext, __u32 *size) {
+ *size = btf_ext->data_size;
+ return btf_ext->data;
+}
+
static int btf_ext_reloc_info(const struct btf *btf,
const struct btf_ext_info *ext_info,
const char *sec_name, __u32 insns_cnt,
diff --git a/tools/lib/bpf/btf.h b/tools/lib/bpf/btf.h
index d6c0a5576fdc..ad9f648260c2 100644
--- a/tools/lib/bpf/btf.h
+++ b/tools/lib/bpf/btf.h
@@ -78,6 +78,8 @@ LIBBPF_API int btf__get_map_kv_tids(const struct btf *btf, const char *map_name,
LIBBPF_API struct btf_ext *btf_ext__new(__u8 *data, __u32 size);
LIBBPF_API void btf_ext__free(struct btf_ext *btf_ext);
+LIBBPF_API const void *btf_ext__get_raw_data(const struct btf_ext* btf_ext,
+ __u32 *size);
LIBBPF_API int btf_ext__reloc_func_info(const struct btf *btf,
const struct btf_ext *btf_ext,
const char *sec_name, __u32 insns_cnt,
diff --git a/tools/lib/bpf/libbpf.map b/tools/lib/bpf/libbpf.map
index 9e10467f8cbb..eb78c7c261d9 100644
--- a/tools/lib/bpf/libbpf.map
+++ b/tools/lib/bpf/libbpf.map
@@ -141,6 +141,7 @@ LIBBPF_0.0.2 {
btf__load;
btf_ext__free;
btf_ext__func_info_rec_size;
+ btf_ext__get_raw_data;
btf_ext__line_info_rec_size;
btf_ext__new;
btf_ext__reloc_func_info;
--
2.17.1
^ permalink raw reply related
* [PATCH v3 bpf-next 2/4] btf: expose API to work with raw btf data
From: Andrii Nakryiko @ 2019-02-08 2:55 UTC (permalink / raw)
To: alexei.starovoitov, andrii.nakryiko, songliubraving, yhs, ast,
kafai, netdev, kernel-team, daniel
Cc: Andrii Nakryiko
In-Reply-To: <20190208025555.4027769-1-andriin@fb.com>
This patch exposes new API btf__get_raw_data() that allows to get a copy
of raw BTF data out of struct btf. This is useful for external programs
that need to manipulate raw data, e.g., pahole using btf__dedup() to
deduplicate BTF type info and then writing it back to file.
Signed-off-by: Andrii Nakryiko <andriin@fb.com>
Acked-by: Song Liu <songliubraving@fb.com>
---
tools/lib/bpf/btf.c | 6 ++++++
tools/lib/bpf/btf.h | 1 +
tools/lib/bpf/libbpf.map | 1 +
3 files changed, 8 insertions(+)
diff --git a/tools/lib/bpf/btf.c b/tools/lib/bpf/btf.c
index 24ac64492d8a..8730f6c3be9e 100644
--- a/tools/lib/bpf/btf.c
+++ b/tools/lib/bpf/btf.c
@@ -437,6 +437,12 @@ int btf__fd(const struct btf *btf)
return btf->fd;
}
+const void *btf__get_raw_data(const struct btf *btf, __u32 *size)
+{
+ *size = btf->data_size;
+ return btf->data;
+}
+
void btf__get_strings(const struct btf *btf, const char **strings,
__u32 *str_len)
{
diff --git a/tools/lib/bpf/btf.h b/tools/lib/bpf/btf.h
index 6fbbde95abc0..d6c0a5576fdc 100644
--- a/tools/lib/bpf/btf.h
+++ b/tools/lib/bpf/btf.h
@@ -66,6 +66,7 @@ LIBBPF_API const struct btf_type *btf__type_by_id(const struct btf *btf,
LIBBPF_API __s64 btf__resolve_size(const struct btf *btf, __u32 type_id);
LIBBPF_API int btf__resolve_type(const struct btf *btf, __u32 type_id);
LIBBPF_API int btf__fd(const struct btf *btf);
+LIBBPF_API const void *btf__get_raw_data(const struct btf *btf, __u32 *size);
LIBBPF_API void btf__get_strings(const struct btf *btf, const char **strings,
__u32 *str_len);
LIBBPF_API const char *btf__name_by_offset(const struct btf *btf, __u32 offset);
diff --git a/tools/lib/bpf/libbpf.map b/tools/lib/bpf/libbpf.map
index f5372df143f4..9e10467f8cbb 100644
--- a/tools/lib/bpf/libbpf.map
+++ b/tools/lib/bpf/libbpf.map
@@ -136,6 +136,7 @@ LIBBPF_0.0.2 {
btf__dedup;
btf__get_map_kv_tids;
btf__get_nr_types;
+ btf__get_raw_data;
btf__get_strings;
btf__load;
btf_ext__free;
--
2.17.1
^ permalink raw reply related
* [PATCH net-next] can: kvaser_usb: Use struct_size() in alloc_candev()
From: Gustavo A. R. Silva @ 2019-02-08 3:10 UTC (permalink / raw)
To: Wolfgang Grandegger, Marc Kleine-Budde, David S. Miller
Cc: linux-can, netdev, linux-kernel, Gustavo A. R. Silva
One of the more common cases of allocation size calculations is finding
the size of a structure that has a zero-sized array at the end, along
with memory for some number of elements for that array. For example:
struct foo {
int stuff;
void *entry[];
};
instance = alloc(sizeof(struct foo) + count * sizeof(void *));
Instead of leaving these open-coded and prone to type mistakes, we can
now use the new struct_size() helper:
instance = alloc(struct_size(instance, entry, count));
This code was detected with the help of Coccinelle.
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
---
drivers/net/can/usb/kvaser_usb/kvaser_usb_core.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/drivers/net/can/usb/kvaser_usb/kvaser_usb_core.c b/drivers/net/can/usb/kvaser_usb/kvaser_usb_core.c
index c89c7d4900d7..0f1d3e807d63 100644
--- a/drivers/net/can/usb/kvaser_usb/kvaser_usb_core.c
+++ b/drivers/net/can/usb/kvaser_usb/kvaser_usb_core.c
@@ -643,8 +643,7 @@ static int kvaser_usb_init_one(struct kvaser_usb *dev,
return err;
}
- netdev = alloc_candev(sizeof(*priv) +
- dev->max_tx_urbs * sizeof(*priv->tx_contexts),
+ netdev = alloc_candev(struct_size(priv, tx_contexts, dev->max_tx_urbs),
dev->max_tx_urbs);
if (!netdev) {
dev_err(&dev->intf->dev, "Cannot alloc candev\n");
--
2.20.1
^ permalink raw reply related
* [PATCH net-next] net: usb: cdc-phonet: use struct_size() in alloc_netdev()
From: Gustavo A. R. Silva @ 2019-02-08 3:13 UTC (permalink / raw)
To: David S. Miller; +Cc: linux-usb, netdev, linux-kernel, Gustavo A. R. Silva
One of the more common cases of allocation size calculations is finding
the size of a structure that has a zero-sized array at the end, along
with memory for some number of elements for that array. For example:
struct foo {
int stuff;
void *entry[];
};
instance = alloc(sizeof(struct foo) + count * sizeof(void *));
Instead of leaving these open-coded and prone to type mistakes, we can
now use the new struct_size() helper:
instance = alloc(struct_size(instance, entry, count));
This code was detected with the help of Coccinelle.
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
---
drivers/net/usb/cdc-phonet.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/net/usb/cdc-phonet.c b/drivers/net/usb/cdc-phonet.c
index 78b16eb9e58c..63aaae487995 100644
--- a/drivers/net/usb/cdc-phonet.c
+++ b/drivers/net/usb/cdc-phonet.c
@@ -361,8 +361,8 @@ static int usbpn_probe(struct usb_interface *intf, const struct usb_device_id *i
else
return -EINVAL;
- dev = alloc_netdev(sizeof(*pnd) + sizeof(pnd->urbs[0]) * rxq_size,
- ifname, NET_NAME_UNKNOWN, usbpn_setup);
+ dev = alloc_netdev(struct_size(pnd, urbs, rxq_size), ifname,
+ NET_NAME_UNKNOWN, usbpn_setup);
if (!dev)
return -ENOMEM;
--
2.20.1
^ permalink raw reply related
* [PATCH net-next] wan: wanxl: use struct_size() in kzalloc()
From: Gustavo A. R. Silva @ 2019-02-08 3:16 UTC (permalink / raw)
To: Krzysztof Halasa, David S. Miller
Cc: netdev, linux-kernel, Gustavo A. R. Silva
One of the more common cases of allocation size calculations is finding
the size of a structure that has a zero-sized array at the end, along
with memory for some number of elements for that array. For example:
struct foo {
int stuff;
struct boo entry[];
};
size = sizeof(struct foo) + count * sizeof(struct boo);
instance = alloc(size, GFP_KERNEL)
Instead of leaving these open-coded and prone to type mistakes, we can
now use the new struct_size() helper:
instance = alloc(struct_size(instance, entry, count), GFP_KERNEL)
Notice that, in this case, variable alloc_size is not necessary, hence
it is removed.
This code was detected with the help of Coccinelle.
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
---
drivers/net/wan/wanxl.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/drivers/net/wan/wanxl.c b/drivers/net/wan/wanxl.c
index d573a57bc301..459ce52a0a4d 100644
--- a/drivers/net/wan/wanxl.c
+++ b/drivers/net/wan/wanxl.c
@@ -565,7 +565,7 @@ static int wanxl_pci_init_one(struct pci_dev *pdev,
u32 plx_phy; /* PLX PCI base address */
u32 mem_phy; /* memory PCI base addr */
u8 __iomem *mem; /* memory virtual base addr */
- int i, ports, alloc_size;
+ int i, ports;
#ifndef MODULE
pr_info_once("%s\n", version);
@@ -601,8 +601,7 @@ static int wanxl_pci_init_one(struct pci_dev *pdev,
default: ports = 4;
}
- alloc_size = sizeof(struct card) + ports * sizeof(struct port);
- card = kzalloc(alloc_size, GFP_KERNEL);
+ card = kzalloc(struct_size(card, ports, ports), GFP_KERNEL);
if (card == NULL) {
pci_release_regions(pdev);
pci_disable_device(pdev);
--
2.20.1
^ permalink raw reply related
* [PATCH net-next] wimax/i2400m: use struct_size() helper
From: Gustavo A. R. Silva @ 2019-02-08 3:22 UTC (permalink / raw)
To: Inaky Perez-Gonzalez, linux-wimax, David S. Miller
Cc: netdev, linux-kernel, Gustavo A. R. Silva
One of the more common cases of allocation size calculations is finding
the size of a structure that has a zero-sized array at the end, along
with memory for some number of elements for that array. For example:
struct foo {
int stuff;
void *entry[];
};
size = sizeof(struct foo) + count * sizeof(void *);
instance = alloc(size, GFP_KERNEL);
Instead of leaving these open-coded and prone to type mistakes, we can
now use the new struct_size() helper:
size = struct_size(instance, entry, count);
This code was detected with the help of Coccinelle.
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
---
drivers/net/wimax/i2400m/rx.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/net/wimax/i2400m/rx.c b/drivers/net/wimax/i2400m/rx.c
index 0b602951ff6b..d28b96d06919 100644
--- a/drivers/net/wimax/i2400m/rx.c
+++ b/drivers/net/wimax/i2400m/rx.c
@@ -1260,8 +1260,8 @@ int i2400m_rx(struct i2400m *i2400m, struct sk_buff *skb)
goto error_msg_hdr_check;
result = -EIO;
num_pls = le16_to_cpu(msg_hdr->num_pls);
- pl_itr = sizeof(*msg_hdr) + /* Check payload descriptor(s) */
- num_pls * sizeof(msg_hdr->pld[0]);
+ /* Check payload descriptor(s) */
+ pl_itr = struct_size(msg_hdr, pld, num_pls);
pl_itr = ALIGN(pl_itr, I2400M_PL_ALIGN);
if (pl_itr > skb_len) { /* got all the payload descriptors? */
dev_err(dev, "RX: HW BUG? message too short (%u bytes) for "
--
2.20.1
^ permalink raw reply related
* [PATCH net-next] mlxsw: spectrum_router: Use struct_size() in kzalloc()
From: Gustavo A. R. Silva @ 2019-02-08 3:42 UTC (permalink / raw)
To: Jiri Pirko, Ido Schimmel, David S. Miller
Cc: netdev, linux-kernel, Gustavo A. R. Silva
One of the more common cases of allocation size calculations is finding
the size of a structure that has a zero-sized array at the end, along
with memory for some number of elements for that array. For example:
struct foo {
int stuff;
struct boo entry[];
};
size = sizeof(struct foo) + count * sizeof(struct boo);
instance = kzalloc(size, GFP_KERNEL)
Instead of leaving these open-coded and prone to type mistakes, we can
now use the new struct_size() helper:
instance = kzalloc(struct_size(instance, entry, count), GFP_KERNEL)
Notice that, in this case, variable alloc_size is not necessary, hence
it is removed.
This code was detected with the help of Coccinelle.
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
---
.../net/ethernet/mellanox/mlxsw/spectrum_router.c | 12 ++++--------
1 file changed, 4 insertions(+), 8 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
index 230e1f6e192b..e3a3535f9f23 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
@@ -3813,13 +3813,11 @@ mlxsw_sp_nexthop4_group_create(struct mlxsw_sp *mlxsw_sp, struct fib_info *fi)
struct mlxsw_sp_nexthop_group *nh_grp;
struct mlxsw_sp_nexthop *nh;
struct fib_nh *fib_nh;
- size_t alloc_size;
int i;
int err;
- alloc_size = sizeof(*nh_grp) +
- fi->fib_nhs * sizeof(struct mlxsw_sp_nexthop);
- nh_grp = kzalloc(alloc_size, GFP_KERNEL);
+ nh_grp = kzalloc(struct_size(nh_grp, nexthops, fi->fib_nhs),
+ GFP_KERNEL);
if (!nh_grp)
return ERR_PTR(-ENOMEM);
nh_grp->priv = fi;
@@ -5045,13 +5043,11 @@ mlxsw_sp_nexthop6_group_create(struct mlxsw_sp *mlxsw_sp,
struct mlxsw_sp_nexthop_group *nh_grp;
struct mlxsw_sp_rt6 *mlxsw_sp_rt6;
struct mlxsw_sp_nexthop *nh;
- size_t alloc_size;
int i = 0;
int err;
- alloc_size = sizeof(*nh_grp) +
- fib6_entry->nrt6 * sizeof(struct mlxsw_sp_nexthop);
- nh_grp = kzalloc(alloc_size, GFP_KERNEL);
+ nh_grp = kzalloc(struct_size(nh_grp, nexthops, fib6_entry->nrt6),
+ GFP_KERNEL);
if (!nh_grp)
return ERR_PTR(-ENOMEM);
INIT_LIST_HEAD(&nh_grp->fib_list);
--
2.20.1
^ permalink raw reply related
* [PATCH net-next] nfp: flower: cmsg: use struct_size() helper
From: Gustavo A. R. Silva @ 2019-02-08 3:47 UTC (permalink / raw)
To: Jakub Kicinski, David S. Miller
Cc: oss-drivers, netdev, linux-kernel, Gustavo A. R. Silva
One of the more common cases of allocation size calculations is finding
the size of a structure that has a zero-sized array at the end, along
with memory for some number of elements for that array. For example:
struct foo {
int stuff;
void *entry[];
};
size = sizeof(struct foo) + count * sizeof(void *);
instance = alloc(size, GFP_KERNEL);
Instead of leaving these open-coded and prone to type mistakes, we can
now use the new struct_size() helper:
instance = alloc(struct_size(instance, entry, count), GFP_KERNEL);
Notice that, in this case, variable size is not necessary, hence
it is removed.
This code was detected with the help of Coccinelle.
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
---
drivers/net/ethernet/netronome/nfp/flower/cmsg.c | 6 ++----
1 file changed, 2 insertions(+), 4 deletions(-)
diff --git a/drivers/net/ethernet/netronome/nfp/flower/cmsg.c b/drivers/net/ethernet/netronome/nfp/flower/cmsg.c
index 56b22ea32474..cf9e1118ee8f 100644
--- a/drivers/net/ethernet/netronome/nfp/flower/cmsg.c
+++ b/drivers/net/ethernet/netronome/nfp/flower/cmsg.c
@@ -45,11 +45,9 @@ nfp_flower_cmsg_mac_repr_start(struct nfp_app *app, unsigned int num_ports)
{
struct nfp_flower_cmsg_mac_repr *msg;
struct sk_buff *skb;
- unsigned int size;
- size = sizeof(*msg) + num_ports * sizeof(msg->ports[0]);
- skb = nfp_flower_cmsg_alloc(app, size, NFP_FLOWER_CMSG_TYPE_MAC_REPR,
- GFP_KERNEL);
+ skb = nfp_flower_cmsg_alloc(app, struct_size(msg, ports, num_ports),
+ NFP_FLOWER_CMSG_TYPE_MAC_REPR, GFP_KERNEL);
if (!skb)
return NULL;
--
2.20.1
^ permalink raw reply related
* Re: [PATCH net-next] xprtrdma: Use struct_size() in kzalloc()
From: Gustavo A. R. Silva @ 2019-02-08 3:49 UTC (permalink / raw)
To: Chuck Lever
Cc: Trond Myklebust, Anna Schumaker, Bruce Fields, Jeff Layton,
David S. Miller, Linux NFS Mailing List, netdev, linux-kernel
In-Reply-To: <07CB966E-A946-4956-8480-C0FC13E13E4E@oracle.com>
On 1/31/19 8:11 AM, Chuck Lever wrote:
>
>
>> On Jan 30, 2019, at 7:46 PM, Gustavo A. R. Silva <gustavo@embeddedor.com> wrote:
>>
>> One of the more common cases of allocation size calculations is finding
>> the size of a structure that has a zero-sized array at the end, along
>> with memory for some number of elements for that array. For example:
>>
>> struct foo {
>> int stuff;
>> struct boo entry[];
>> };
>>
>> instance = kzalloc(sizeof(struct foo) + count * sizeof(struct boo), GFP_KERNEL);
>>
>> Instead of leaving these open-coded and prone to type mistakes, we can
>> now use the new struct_size() helper:
>>
>> instance = kzalloc(struct_size(instance, entry, count), GFP_KERNEL);
>>
>> This code was detected with the help of Coccinelle.
>>
>> Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
>
> Reviewed-by: Chuck Lever <chuck.lever@oracle.com>
>
Thanks, Chuck.
--
Gustavo
^ permalink raw reply
* [PATCH net-next] bnx2x: Use struct_size() in kzalloc()
From: Gustavo A. R. Silva @ 2019-02-08 3:29 UTC (permalink / raw)
To: Ariel Elior, Sudarsana Kalluru, GR-everest-linux-l2,
David S. Miller
Cc: netdev, linux-kernel, Gustavo A. R. Silva
One of the more common cases of allocation size calculations is finding
the size of a structure that has a zero-sized array at the end, along
with memory for some number of elements for that array. For example:
struct foo {
int stuff;
struct boo entry[];
};
size = sizeof(struct foo) + count * sizeof(struct boo);
instance = kzalloc(size, GFP_KERNEL)
Instead of leaving these open-coded and prone to type mistakes, we can
now use the new struct_size() helper:
instance = kzalloc(struct_size(instance, entry, count), GFP_KERNEL)
Notice that, in this case, variable fsz is not necessary, hence
it is removed.
This code was detected with the help of Coccinelle.
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
---
drivers/net/ethernet/broadcom/bnx2x/bnx2x_vfpf.c | 8 ++------
1 file changed, 2 insertions(+), 6 deletions(-)
diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_vfpf.c b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_vfpf.c
index 8e0a317b31f7..a9bdc21873d3 100644
--- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_vfpf.c
+++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_vfpf.c
@@ -1654,13 +1654,9 @@ static int bnx2x_vf_mbx_macvlan_list(struct bnx2x *bp,
{
int i, j;
struct bnx2x_vf_mac_vlan_filters *fl = NULL;
- size_t fsz;
- fsz = tlv->n_mac_vlan_filters *
- sizeof(struct bnx2x_vf_mac_vlan_filter) +
- sizeof(struct bnx2x_vf_mac_vlan_filters);
-
- fl = kzalloc(fsz, GFP_KERNEL);
+ fl = kzalloc(struct_size(fl, filters, tlv->n_mac_vlan_filters),
+ GFP_KERNEL);
if (!fl)
return -ENOMEM;
--
2.20.1
^ permalink raw reply related
* [PATCH net-next] fm10k: use struct_size() in kzalloc()
From: Gustavo A. R. Silva @ 2019-02-08 3:55 UTC (permalink / raw)
To: Jeff Kirsher, David S. Miller
Cc: intel-wired-lan, netdev, linux-kernel, Gustavo A. R. Silva
One of the more common cases of allocation size calculations is finding
the size of a structure that has a zero-sized array at the end, along
with memory for some number of elements for that array. For example:
struct foo {
int stuff;
struct boo entry[];
};
size = sizeof(struct foo) + count * sizeof(struct boo);
instance = kzalloc(size, GFP_KERNEL);
Instead of leaving these open-coded and prone to type mistakes, we can
now use the new struct_size() helper:
instance = kzalloc(struct_size(instance, entry, count), GFP_KERNEL);
Notice that, in this case, variable size is not necessary, hence
it is removed.
This code was detected with the help of Coccinelle.
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
---
drivers/net/ethernet/intel/fm10k/fm10k_main.c | 6 ++----
1 file changed, 2 insertions(+), 4 deletions(-)
diff --git a/drivers/net/ethernet/intel/fm10k/fm10k_main.c b/drivers/net/ethernet/intel/fm10k/fm10k_main.c
index 6fd15a734324..5a0419421511 100644
--- a/drivers/net/ethernet/intel/fm10k/fm10k_main.c
+++ b/drivers/net/ethernet/intel/fm10k/fm10k_main.c
@@ -1603,14 +1603,12 @@ static int fm10k_alloc_q_vector(struct fm10k_intfc *interface,
{
struct fm10k_q_vector *q_vector;
struct fm10k_ring *ring;
- int ring_count, size;
+ int ring_count;
ring_count = txr_count + rxr_count;
- size = sizeof(struct fm10k_q_vector) +
- (sizeof(struct fm10k_ring) * ring_count);
/* allocate q_vector and rings */
- q_vector = kzalloc(size, GFP_KERNEL);
+ q_vector = kzalloc(struct_size(q_vector, ring, ring_count), GFP_KERNEL);
if (!q_vector)
return -ENOMEM;
--
2.20.1
^ permalink raw reply related
* Re: [PATCH][next] Bluetooth: a2mp: Use struct_size() helper
From: Joe Perches @ 2019-02-08 4:00 UTC (permalink / raw)
To: Gustavo A. R. Silva, Marcel Holtmann, Johan Hedberg,
David S. Miller
Cc: linux-bluetooth, netdev, linux-kernel
In-Reply-To: <20190208002817.GA15338@embeddedor>
On Thu, 2019-02-07 at 18:28 -0600, Gustavo A. R. Silva wrote:
> One of the more common cases of allocation size calculations is finding
> the size of a structure that has a zero-sized array at the end, along
> with memory for some number of elements for that array. For example:
>
> struct foo {
> int stuff;
> struct boo entry[];
> };
>
> size = sizeof(struct foo) + count * sizeof(struct boo);
> instance = alloc(size, GFP_KERNEL)
>
> Instead of leaving these open-coded and prone to type mistakes, we can
> now use the new struct_size() helper:
>
> size = struct_size(instance, entry, count);
> instance = alloc(size, GFP_KERNEL)
>
> This code was detected with the help of Coccinelle.
[]
> diff --git a/net/bluetooth/a2mp.c b/net/bluetooth/a2mp.c
[]
> @@ -174,7 +174,7 @@ static int a2mp_discover_req(struct amp_mgr *mgr, struct sk_buff *skb,
> num_ctrl++;
> }
>
> - len = num_ctrl * sizeof(struct a2mp_cl) + sizeof(*rsp);
> + len = struct_size(rsp, cl, num_ctrl);
> rsp = kmalloc(len, GFP_ATOMIC);
> if (!rsp) {
> read_unlock(&hci_dev_list_lock);
At least a weakness in this code is len is u16
and struct_size is size_t so there's a size
truncation possible.
^ permalink raw reply
* Re: [PATCH][next] Bluetooth: hci_event: Use struct_size() helper
From: Joe Perches @ 2019-02-08 4:02 UTC (permalink / raw)
To: Gustavo A. R. Silva, Marcel Holtmann, Johan Hedberg,
David S. Miller
Cc: linux-bluetooth, netdev, linux-kernel
In-Reply-To: <20190208004033.GA15609@embeddedor>
On Thu, 2019-02-07 at 18:40 -0600, Gustavo A. R. Silva wrote:
> Make use of the struct_size() helper instead of an open-coded version
> in order to avoid any potential type mistakes, in particular in the
> context in which this code is being used.
>
> So, change the following form:
>
> sizeof(*ev) + ev->num_hndl * sizeof(struct hci_comp_pkts_info)
>
> to :
>
> struct_size(ev, handles, ev->num_hndl)
>
> This code was detected with the help of Coccinelle.
[]
> diff --git a/net/bluetooth/hci_event.c b/net/bluetooth/hci_event.c
[]
> @@ -3556,8 +3556,8 @@ static void hci_num_comp_pkts_evt(struct hci_dev *hdev, struct sk_buff *skb)
> return;
> }
>
> - if (skb->len < sizeof(*ev) || skb->len < sizeof(*ev) +
> - ev->num_hndl * sizeof(struct hci_comp_pkts_info)) {
> + if (skb->len < sizeof(*ev) ||
> + skb->len < struct_size(ev, handles, ev->num_hndl)) {
> BT_DBG("%s bad parameters", hdev->name);
> return;
> }
> @@ -3644,8 +3644,8 @@ static void hci_num_comp_blocks_evt(struct hci_dev *hdev, struct sk_buff *skb)
> return;
> }
>
> - if (skb->len < sizeof(*ev) || skb->len < sizeof(*ev) +
> - ev->num_hndl * sizeof(struct hci_comp_blocks_info)) {
> + if (skb->len < sizeof(*ev) ||
> + skb->len < struct_size(ev, handles, ev->num_hndl)) {
Unless these are in a real fast path where skb->len < sizeof(*ev)
is pretty likely, it seems a bit redundant as the multiplication
is pretty cheap and num_hndl is unsigned (actually __u8)
This could/should be simply
if (skb->len < struct_size(...))
^ permalink raw reply
* [PATCH net] net: dsa: microchip: add switch offload forwarding support
From: Tristram.Ha @ 2019-02-08 4:05 UTC (permalink / raw)
To: David S. Miller, Sergio Paracuellos
Cc: Tristram Ha, Andrew Lunn, Florian Fainelli, Pavel Machek,
UNGLinuxDriver, netdev
From: Tristram Ha <Tristram.Ha@microchip.com>
The flag offload_fwd_mark is set as the switch can forward frames by
itself.
This can be considered a fix to a problem introduced in commit
c2e866911e254067 where the port membership are not set in sync. The flag
offload_fwd_mark just needs to be set in tag_ksz.c to prevent the software
bridge from forwarding duplicate multicast frames.
Fixes: c2e866911e254067 ("microchip: break KSZ9477 DSA driver into two files")
Signed-off-by: Tristram Ha <Tristram.Ha@microchip.com>
---
drivers/net/dsa/microchip/ksz9477.c | 7 ++++---
net/dsa/tag_ksz.c | 2 ++
2 files changed, 6 insertions(+), 3 deletions(-)
diff --git a/drivers/net/dsa/microchip/ksz9477.c b/drivers/net/dsa/microchip/ksz9477.c
index 89ed059..674d77e 100644
--- a/drivers/net/dsa/microchip/ksz9477.c
+++ b/drivers/net/dsa/microchip/ksz9477.c
@@ -397,6 +397,7 @@ static void ksz9477_port_stp_state_set(struct dsa_switch *ds, int port,
struct ksz_port *p = &dev->ports[port];
u8 data;
int member = -1;
+ int forward = dev->member;
ksz_pread8(dev, port, P_STP_CTRL, &data);
data &= ~(PORT_TX_ENABLE | PORT_RX_ENABLE | PORT_LEARN_DISABLE);
@@ -464,10 +465,10 @@ static void ksz9477_port_stp_state_set(struct dsa_switch *ds, int port,
}
/* When topology has changed the function ksz_update_port_member
- * should be called to modify port forwarding behavior. However
- * as the offload_fwd_mark indication cannot be reported here
- * the switch forwarding function is not enabled.
+ * should be called to modify port forwarding behavior.
*/
+ if (forward != dev->member)
+ ksz_update_port_member(dev, port);
}
static void ksz9477_flush_dyn_mac_table(struct ksz_device *dev, int port)
diff --git a/net/dsa/tag_ksz.c b/net/dsa/tag_ksz.c
index da71b9e..927e9c8 100644
--- a/net/dsa/tag_ksz.c
+++ b/net/dsa/tag_ksz.c
@@ -67,6 +67,8 @@ static struct sk_buff *ksz_common_rcv(struct sk_buff *skb,
pskb_trim_rcsum(skb, skb->len - len);
+ skb->offload_fwd_mark = true;
+
return skb;
}
--
1.9.1
^ permalink raw reply related
* [PATCH v1 net-next 0/4] net: dsa: microchip: add MIB counters support
From: Tristram.Ha @ 2019-02-08 4:07 UTC (permalink / raw)
To: Sergio Paracuellos, Andrew Lunn, Florian Fainelli, Pavel Machek
Cc: Tristram Ha, UNGLinuxDriver, netdev
From: Tristram Ha <Tristram.Ha@microchip.com>
This series of patches is to modify the KSZ9477 DSA driver to read MIB
counters periodically to avoid overflow.
v1
- Use readx_poll_time
- Do not clear MIB counters when port is enabled
- Do not advertise 1000 half-duplex mode when port is enabled
- Do not use freeze function as MIB counters may miss counts
Tristram Ha (4):
net: dsa: microchip: prepare PHY for proper advertisement
net: dsa: microchip: add MIB counter reading support
net: dsa: microchip: use readx_poll_time for polling
net: dsa: microchip: remove unnecessary include headers
drivers/net/dsa/microchip/ksz9477.c | 249 ++++++++++++++++++---------------
drivers/net/dsa/microchip/ksz_common.c | 115 ++++++++++++++-
drivers/net/dsa/microchip/ksz_common.h | 6 +-
drivers/net/dsa/microchip/ksz_priv.h | 11 +-
4 files changed, 264 insertions(+), 117 deletions(-)
--
1.9.1
^ permalink raw reply
* [PATCH v1 net-next 1/4] net: dsa: microchip: prepare PHY for proper advertisement
From: Tristram.Ha @ 2019-02-08 4:07 UTC (permalink / raw)
To: Sergio Paracuellos, Andrew Lunn, Florian Fainelli, Pavel Machek
Cc: Tristram Ha, UNGLinuxDriver, netdev
In-Reply-To: <1549598829-25970-1-git-send-email-Tristram.Ha@microchip.com>
From: Tristram Ha <Tristram.Ha@microchip.com>
Prepare PHY for proper advertisement and get link status for the port.
Signed-off-by: Tristram Ha <Tristram.Ha@microchip.com>
---
drivers/net/dsa/microchip/ksz9477.c | 17 ++++++++++++++++-
drivers/net/dsa/microchip/ksz_common.c | 19 ++++++++++++++++++-
drivers/net/dsa/microchip/ksz_common.h | 4 +++-
drivers/net/dsa/microchip/ksz_priv.h | 4 +++-
4 files changed, 40 insertions(+), 4 deletions(-)
diff --git a/drivers/net/dsa/microchip/ksz9477.c b/drivers/net/dsa/microchip/ksz9477.c
index 89ed059..0fdb22d 100644
--- a/drivers/net/dsa/microchip/ksz9477.c
+++ b/drivers/net/dsa/microchip/ksz9477.c
@@ -2,7 +2,7 @@
/*
* Microchip KSZ9477 switch driver main logic
*
- * Copyright (C) 2017-2018 Microchip Technology Inc.
+ * Copyright (C) 2017-2019 Microchip Technology Inc.
*/
#include <linux/delay.h>
@@ -965,6 +965,19 @@ static void ksz9477_port_mirror_del(struct dsa_switch *ds, int port,
PORT_MIRROR_SNIFFER, false);
}
+static void ksz9477_phy_setup(struct ksz_device *dev, int port,
+ struct phy_device *phy)
+{
+ /* ETHTOOL_LINK_MODE_Pause_BIT and ETHTOOL_LINK_MODE_Asym_Pause_BIT
+ * can be removed to disable flow control when rate limiting is used.
+ */
+ if (port < dev->phy_port_cnt) {
+ /* The MAC actually cannot run in 1000 half-duplex mode. */
+ phy_remove_link_mode(phy,
+ ETHTOOL_LINK_MODE_1000baseT_Half_BIT);
+ }
+}
+
static void ksz9477_port_setup(struct ksz_device *dev, int port, bool cpu_port)
{
u8 data8;
@@ -1151,6 +1164,7 @@ static int ksz9477_setup(struct dsa_switch *ds)
.setup = ksz9477_setup,
.phy_read = ksz9477_phy_read16,
.phy_write = ksz9477_phy_write16,
+ .adjust_link = ksz_adjust_link,
.port_enable = ksz_enable_port,
.port_disable = ksz_disable_port,
.get_strings = ksz9477_get_strings,
@@ -1298,6 +1312,7 @@ static void ksz9477_switch_exit(struct ksz_device *dev)
.get_port_addr = ksz9477_get_port_addr,
.cfg_port_member = ksz9477_cfg_port_member,
.flush_dyn_mac_table = ksz9477_flush_dyn_mac_table,
+ .phy_setup = ksz9477_phy_setup,
.port_setup = ksz9477_port_setup,
.shutdown = ksz9477_reset_switch,
.detect = ksz9477_switch_detect,
diff --git a/drivers/net/dsa/microchip/ksz_common.c b/drivers/net/dsa/microchip/ksz_common.c
index 8a5111f..a57bda7 100644
--- a/drivers/net/dsa/microchip/ksz_common.c
+++ b/drivers/net/dsa/microchip/ksz_common.c
@@ -2,7 +2,7 @@
/*
* Microchip switch driver main logic
*
- * Copyright (C) 2017-2018 Microchip Technology Inc.
+ * Copyright (C) 2017-2019 Microchip Technology Inc.
*/
#include <linux/delay.h>
@@ -61,6 +61,22 @@ int ksz_phy_write16(struct dsa_switch *ds, int addr, int reg, u16 val)
}
EXPORT_SYMBOL_GPL(ksz_phy_write16);
+void ksz_adjust_link(struct dsa_switch *ds, int port,
+ struct phy_device *phydev)
+{
+ struct ksz_device *dev = ds->priv;
+ struct ksz_port *p = &dev->ports[port];
+
+ if (phydev->link) {
+ dev->live_ports |= (1 << port) & dev->on_ports;
+ } else if (p->phydev.link) {
+ p->link_just_down = 1;
+ dev->live_ports &= ~(1 << port);
+ }
+ p->phydev = *phydev;
+}
+EXPORT_SYMBOL_GPL(ksz_adjust_link);
+
int ksz_sset_count(struct dsa_switch *ds, int port, int sset)
{
struct ksz_device *dev = ds->priv;
@@ -238,6 +254,7 @@ int ksz_enable_port(struct dsa_switch *ds, int port, struct phy_device *phy)
/* setup slave port */
dev->dev_ops->port_setup(dev, port, false);
+ dev->dev_ops->phy_setup(dev, port, phy);
/* port_stp_state_set() will be called after to enable the port so
* there is no need to do anything.
diff --git a/drivers/net/dsa/microchip/ksz_common.h b/drivers/net/dsa/microchip/ksz_common.h
index 2dd832d..6d49319 100644
--- a/drivers/net/dsa/microchip/ksz_common.h
+++ b/drivers/net/dsa/microchip/ksz_common.h
@@ -1,7 +1,7 @@
/* SPDX-License-Identifier: GPL-2.0
* Microchip switch driver common header
*
- * Copyright (C) 2017-2018 Microchip Technology Inc.
+ * Copyright (C) 2017-2019 Microchip Technology Inc.
*/
#ifndef __KSZ_COMMON_H
@@ -13,6 +13,8 @@
int ksz_phy_read16(struct dsa_switch *ds, int addr, int reg);
int ksz_phy_write16(struct dsa_switch *ds, int addr, int reg, u16 val);
+void ksz_adjust_link(struct dsa_switch *ds, int port,
+ struct phy_device *phydev);
int ksz_sset_count(struct dsa_switch *ds, int port, int sset);
int ksz_port_bridge_join(struct dsa_switch *ds, int port,
struct net_device *br);
diff --git a/drivers/net/dsa/microchip/ksz_priv.h b/drivers/net/dsa/microchip/ksz_priv.h
index 60b4901..0fdc58b 100644
--- a/drivers/net/dsa/microchip/ksz_priv.h
+++ b/drivers/net/dsa/microchip/ksz_priv.h
@@ -2,7 +2,7 @@
*
* Microchip KSZ series switch common definitions
*
- * Copyright (C) 2017-2018 Microchip Technology Inc.
+ * Copyright (C) 2017-2019 Microchip Technology Inc.
*/
#ifndef __KSZ_PRIV_H
@@ -137,6 +137,8 @@ struct ksz_dev_ops {
u32 (*get_port_addr)(int port, int offset);
void (*cfg_port_member)(struct ksz_device *dev, int port, u8 member);
void (*flush_dyn_mac_table)(struct ksz_device *dev, int port);
+ void (*phy_setup)(struct ksz_device *dev, int port,
+ struct phy_device *phy);
void (*port_setup)(struct ksz_device *dev, int port, bool cpu_port);
void (*r_phy)(struct ksz_device *dev, u16 phy, u16 reg, u16 *val);
void (*w_phy)(struct ksz_device *dev, u16 phy, u16 reg, u16 val);
--
1.9.1
^ permalink raw reply related
* Re: [PATCH][next] Bluetooth: a2mp: Use struct_size() helper
From: Gustavo A. R. Silva @ 2019-02-08 4:07 UTC (permalink / raw)
To: Joe Perches, Marcel Holtmann, Johan Hedberg, David S. Miller
Cc: linux-bluetooth, netdev, linux-kernel
In-Reply-To: <4faff31066c10285bab0afdb0c1f88f6d3d1a21b.camel@perches.com>
On 2/7/19 10:00 PM, Joe Perches wrote:
> On Thu, 2019-02-07 at 18:28 -0600, Gustavo A. R. Silva wrote:
>> One of the more common cases of allocation size calculations is finding
>> the size of a structure that has a zero-sized array at the end, along
>> with memory for some number of elements for that array. For example:
>>
>> struct foo {
>> int stuff;
>> struct boo entry[];
>> };
>>
>> size = sizeof(struct foo) + count * sizeof(struct boo);
>> instance = alloc(size, GFP_KERNEL)
>>
>> Instead of leaving these open-coded and prone to type mistakes, we can
>> now use the new struct_size() helper:
>>
>> size = struct_size(instance, entry, count);
>> instance = alloc(size, GFP_KERNEL)
>>
>> This code was detected with the help of Coccinelle.
> []
>> diff --git a/net/bluetooth/a2mp.c b/net/bluetooth/a2mp.c
> []
>> @@ -174,7 +174,7 @@ static int a2mp_discover_req(struct amp_mgr *mgr, struct sk_buff *skb,
>> num_ctrl++;
>> }
>>
>> - len = num_ctrl * sizeof(struct a2mp_cl) + sizeof(*rsp);
>> + len = struct_size(rsp, cl, num_ctrl);
>> rsp = kmalloc(len, GFP_ATOMIC);
>> if (!rsp) {
>> read_unlock(&hci_dev_list_lock);
>
> At least a weakness in this code is len is u16
> and struct_size is size_t so there's a size
> truncation possible.
>
>
That's true. I didn't change the type to size_t because of the call
to le16_to_cpu():
u16 len = le16_to_cpu(hdr->len);
I've been changing the type of the variable in other cases.
--
Gustavo
^ permalink raw reply
* [PATCH v1 net-next 3/4] net: dsa: microchip: use readx_poll_time for polling
From: Tristram.Ha @ 2019-02-08 4:07 UTC (permalink / raw)
To: Sergio Paracuellos, Andrew Lunn, Florian Fainelli, Pavel Machek
Cc: Tristram Ha, UNGLinuxDriver, netdev
In-Reply-To: <1549598829-25970-1-git-send-email-Tristram.Ha@microchip.com>
From: Tristram Ha <Tristram.Ha@microchip.com>
Replace register polling functions using timeout with readx_poll_time call.
Signed-off-by: Tristram Ha <Tristram.Ha@microchip.com>
---
drivers/net/dsa/microchip/ksz9477.c | 91 +++++++++++--------------------------
1 file changed, 27 insertions(+), 64 deletions(-)
diff --git a/drivers/net/dsa/microchip/ksz9477.c b/drivers/net/dsa/microchip/ksz9477.c
index 4502e13..8391b9e 100644
--- a/drivers/net/dsa/microchip/ksz9477.c
+++ b/drivers/net/dsa/microchip/ksz9477.c
@@ -114,28 +114,11 @@ static void ksz9477_port_cfg32(struct ksz_device *dev, int port, int offset,
data; \
})
-static int ksz9477_wait_vlan_ctrl_ready(struct ksz_device *dev, u32 waiton,
- int timeout)
-{
- u8 data;
-
- do {
- ksz_read8(dev, REG_SW_VLAN_CTRL, &data);
- if (!(data & waiton))
- break;
- usleep_range(1, 10);
- } while (timeout-- > 0);
-
- if (timeout <= 0)
- return -ETIMEDOUT;
-
- return 0;
-}
-
static int ksz9477_get_vlan_table(struct ksz_device *dev, u16 vid,
u32 *vlan_table)
{
int ret;
+ u8 data;
mutex_lock(&dev->vlan_mutex);
@@ -143,7 +126,8 @@ static int ksz9477_get_vlan_table(struct ksz_device *dev, u16 vid,
ksz_write8(dev, REG_SW_VLAN_CTRL, VLAN_READ | VLAN_START);
/* wait to be cleared */
- ret = ksz9477_wait_vlan_ctrl_ready(dev, VLAN_START, 1000);
+ ret = readx_poll_timeout(read8_op, REG_SW_VLAN_CTRL, data,
+ !(data & VLAN_START), 10, 1000);
if (ret < 0) {
dev_dbg(dev->dev, "Failed to read vlan table\n");
goto exit;
@@ -165,6 +149,7 @@ static int ksz9477_set_vlan_table(struct ksz_device *dev, u16 vid,
u32 *vlan_table)
{
int ret;
+ u8 data;
mutex_lock(&dev->vlan_mutex);
@@ -176,7 +161,8 @@ static int ksz9477_set_vlan_table(struct ksz_device *dev, u16 vid,
ksz_write8(dev, REG_SW_VLAN_CTRL, VLAN_START | VLAN_WRITE);
/* wait to be cleared */
- ret = ksz9477_wait_vlan_ctrl_ready(dev, VLAN_START, 1000);
+ ret = readx_poll_timeout(read8_op, REG_SW_VLAN_CTRL, data,
+ !(data & VLAN_START), 10, 1000);
if (ret < 0) {
dev_dbg(dev->dev, "Failed to write vlan table\n");
goto exit;
@@ -211,42 +197,6 @@ static void ksz9477_write_table(struct ksz_device *dev, u32 *table)
ksz_write32(dev, REG_SW_ALU_VAL_D, table[3]);
}
-static int ksz9477_wait_alu_ready(struct ksz_device *dev, u32 waiton,
- int timeout)
-{
- u32 data;
-
- do {
- ksz_read32(dev, REG_SW_ALU_CTRL__4, &data);
- if (!(data & waiton))
- break;
- usleep_range(1, 10);
- } while (timeout-- > 0);
-
- if (timeout <= 0)
- return -ETIMEDOUT;
-
- return 0;
-}
-
-static int ksz9477_wait_alu_sta_ready(struct ksz_device *dev, u32 waiton,
- int timeout)
-{
- u32 data;
-
- do {
- ksz_read32(dev, REG_SW_ALU_STAT_CTRL__4, &data);
- if (!(data & waiton))
- break;
- usleep_range(1, 10);
- } while (timeout-- > 0);
-
- if (timeout <= 0)
- return -ETIMEDOUT;
-
- return 0;
-}
-
static int ksz9477_reset_switch(struct ksz_device *dev)
{
u8 data8;
@@ -649,7 +599,8 @@ static int ksz9477_port_fdb_add(struct dsa_switch *ds, int port,
ksz_write32(dev, REG_SW_ALU_CTRL__4, ALU_READ | ALU_START);
/* wait to be finished */
- ret = ksz9477_wait_alu_ready(dev, ALU_START, 1000);
+ ret = readx_poll_timeout(read32_op, REG_SW_ALU_CTRL__4, data,
+ !(data & ALU_START), 10, 1000);
if (ret < 0) {
dev_dbg(dev->dev, "Failed to read ALU\n");
goto exit;
@@ -673,7 +624,8 @@ static int ksz9477_port_fdb_add(struct dsa_switch *ds, int port,
ksz_write32(dev, REG_SW_ALU_CTRL__4, ALU_WRITE | ALU_START);
/* wait to be finished */
- ret = ksz9477_wait_alu_ready(dev, ALU_START, 1000);
+ ret = readx_poll_timeout(read32_op, REG_SW_ALU_CTRL__4, data,
+ !(data & ALU_START), 10, 1000);
if (ret < 0)
dev_dbg(dev->dev, "Failed to write ALU\n");
@@ -706,7 +658,8 @@ static int ksz9477_port_fdb_del(struct dsa_switch *ds, int port,
ksz_write32(dev, REG_SW_ALU_CTRL__4, ALU_READ | ALU_START);
/* wait to be finished */
- ret = ksz9477_wait_alu_ready(dev, ALU_START, 1000);
+ ret = readx_poll_timeout(read32_op, REG_SW_ALU_CTRL__4, data,
+ !(data & ALU_START), 10, 1000);
if (ret < 0) {
dev_dbg(dev->dev, "Failed to read ALU\n");
goto exit;
@@ -740,7 +693,8 @@ static int ksz9477_port_fdb_del(struct dsa_switch *ds, int port,
ksz_write32(dev, REG_SW_ALU_CTRL__4, ALU_WRITE | ALU_START);
/* wait to be finished */
- ret = ksz9477_wait_alu_ready(dev, ALU_START, 1000);
+ ret = readx_poll_timeout(read32_op, REG_SW_ALU_CTRL__4, data,
+ !(data & ALU_START), 10, 1000);
if (ret < 0)
dev_dbg(dev->dev, "Failed to write ALU\n");
@@ -832,6 +786,7 @@ static void ksz9477_port_mdb_add(struct dsa_switch *ds, int port,
u32 static_table[4];
u32 data;
int index;
+ int ret;
u32 mac_hi, mac_lo;
mac_hi = ((mdb->addr[0] << 8) | mdb->addr[1]);
@@ -847,7 +802,10 @@ static void ksz9477_port_mdb_add(struct dsa_switch *ds, int port,
ksz_write32(dev, REG_SW_ALU_STAT_CTRL__4, data);
/* wait to be finished */
- if (ksz9477_wait_alu_sta_ready(dev, ALU_STAT_START, 1000) < 0) {
+ ret = readx_poll_timeout(read32_op, REG_SW_ALU_STAT_CTRL__4,
+ data, !(data & ALU_STAT_START),
+ 10, 1000);
+ if (ret < 0) {
dev_dbg(dev->dev, "Failed to read ALU STATIC\n");
goto exit;
}
@@ -888,7 +846,9 @@ static void ksz9477_port_mdb_add(struct dsa_switch *ds, int port,
ksz_write32(dev, REG_SW_ALU_STAT_CTRL__4, data);
/* wait to be finished */
- if (ksz9477_wait_alu_sta_ready(dev, ALU_STAT_START, 1000) < 0)
+ ret = readx_poll_timeout(read32_op, REG_SW_ALU_STAT_CTRL__4, data,
+ !(data & ALU_STAT_START), 10, 1000);
+ if (ret < 0)
dev_dbg(dev->dev, "Failed to read ALU STATIC\n");
exit:
@@ -918,7 +878,9 @@ static int ksz9477_port_mdb_del(struct dsa_switch *ds, int port,
ksz_write32(dev, REG_SW_ALU_STAT_CTRL__4, data);
/* wait to be finished */
- ret = ksz9477_wait_alu_sta_ready(dev, ALU_STAT_START, 1000);
+ ret = readx_poll_timeout(read32_op, REG_SW_ALU_STAT_CTRL__4,
+ data, !(data & ALU_STAT_START),
+ 10, 1000);
if (ret < 0) {
dev_dbg(dev->dev, "Failed to read ALU STATIC\n");
goto exit;
@@ -960,7 +922,8 @@ static int ksz9477_port_mdb_del(struct dsa_switch *ds, int port,
ksz_write32(dev, REG_SW_ALU_STAT_CTRL__4, data);
/* wait to be finished */
- ret = ksz9477_wait_alu_sta_ready(dev, ALU_STAT_START, 1000);
+ ret = readx_poll_timeout(read32_op, REG_SW_ALU_STAT_CTRL__4, data,
+ !(data & ALU_STAT_START), 10, 1000);
if (ret < 0)
dev_dbg(dev->dev, "Failed to read ALU STATIC\n");
--
1.9.1
^ permalink raw reply related
* [PATCH v1 net-next 2/4] net: dsa: microchip: add MIB counter reading support
From: Tristram.Ha @ 2019-02-08 4:07 UTC (permalink / raw)
To: Sergio Paracuellos, Andrew Lunn, Florian Fainelli, Pavel Machek
Cc: Tristram Ha, UNGLinuxDriver, netdev
In-Reply-To: <1549598829-25970-1-git-send-email-Tristram.Ha@microchip.com>
From: Tristram Ha <Tristram.Ha@microchip.com>
Add MIB counter reading support.
Signed-off-by: Tristram Ha <Tristram.Ha@microchip.com>
---
drivers/net/dsa/microchip/ksz9477.c | 139 +++++++++++++++++++++++----------
drivers/net/dsa/microchip/ksz_common.c | 96 +++++++++++++++++++++++
drivers/net/dsa/microchip/ksz_common.h | 2 +
drivers/net/dsa/microchip/ksz_priv.h | 7 +-
4 files changed, 198 insertions(+), 46 deletions(-)
diff --git a/drivers/net/dsa/microchip/ksz9477.c b/drivers/net/dsa/microchip/ksz9477.c
index 0fdb22d..4502e13 100644
--- a/drivers/net/dsa/microchip/ksz9477.c
+++ b/drivers/net/dsa/microchip/ksz9477.c
@@ -10,6 +10,7 @@
#include <linux/gpio.h>
#include <linux/kernel.h>
#include <linux/module.h>
+#include <linux/iopoll.h>
#include <linux/platform_data/microchip-ksz.h>
#include <linux/phy.h>
#include <linux/etherdevice.h>
@@ -18,8 +19,8 @@
#include <net/switchdev.h>
#include "ksz_priv.h"
-#include "ksz_common.h"
#include "ksz9477_reg.h"
+#include "ksz_common.h"
static const struct {
int index;
@@ -92,6 +93,27 @@ static void ksz9477_port_cfg32(struct ksz_device *dev, int port, int offset,
ksz_write32(dev, addr, data);
}
+#define read8_op(addr) \
+({ \
+ u8 data; \
+ ksz_read8(dev, addr, &data); \
+ data; \
+})
+
+#define read32_op(addr) \
+({ \
+ u32 data; \
+ ksz_read32(dev, addr, &data); \
+ data; \
+})
+
+#define pread32_op(addr) \
+({ \
+ u32 data; \
+ ksz_pread32(dev, port, addr, &data); \
+ data; \
+})
+
static int ksz9477_wait_vlan_ctrl_ready(struct ksz_device *dev, u32 waiton,
int timeout)
{
@@ -259,6 +281,70 @@ static int ksz9477_reset_switch(struct ksz_device *dev)
return 0;
}
+static void ksz9477_r_mib_cnt(struct ksz_device *dev, int port, u16 addr,
+ u64 *cnt)
+{
+ u32 data;
+ int ret;
+ struct ksz_port *p = &dev->ports[port];
+
+ /* retain the flush/freeze bit */
+ data = p->freeze ? MIB_COUNTER_FLUSH_FREEZE : 0;
+ data |= MIB_COUNTER_READ;
+ data |= (addr << MIB_COUNTER_INDEX_S);
+ ksz_pwrite32(dev, port, REG_PORT_MIB_CTRL_STAT__4, data);
+
+ ret = readx_poll_timeout(pread32_op, REG_PORT_MIB_CTRL_STAT__4, data,
+ !(data & MIB_COUNTER_READ), 10, 1000);
+
+ /* failed to read MIB. get out of loop */
+ if (ret < 0) {
+ dev_dbg(dev->dev, "Failed to get MIB\n");
+ return;
+ }
+
+ /* count resets upon read */
+ ksz_pread32(dev, port, REG_PORT_MIB_DATA, &data);
+ *cnt += data;
+}
+
+static void ksz9477_r_mib_pkt(struct ksz_device *dev, int port, u16 addr,
+ u64 *dropped, u64 *cnt)
+{
+ addr = ksz9477_mib_names[addr].index;
+ ksz9477_r_mib_cnt(dev, port, addr, cnt);
+}
+
+static void ksz9477_freeze_mib(struct ksz_device *dev, int port, bool freeze)
+{
+ struct ksz_port *p = &dev->ports[port];
+ u32 val = freeze ? MIB_COUNTER_FLUSH_FREEZE : 0;
+
+ /* enable/disable the port for flush/freeze function */
+ mutex_lock(&p->mib.cnt_mutex);
+ ksz_pwrite32(dev, port, REG_PORT_MIB_CTRL_STAT__4, val);
+
+ /* used by MIB counter reading code to know freeze is enabled */
+ p->freeze = freeze;
+ mutex_unlock(&p->mib.cnt_mutex);
+}
+
+static void ksz9477_port_init_cnt(struct ksz_device *dev, int port)
+{
+ struct ksz_port_mib *mib = &dev->ports[port].mib;
+
+ /* flush all enabled port MIB counters */
+ mutex_lock(&mib->cnt_mutex);
+ ksz_pwrite32(dev, port, REG_PORT_MIB_CTRL_STAT__4,
+ MIB_COUNTER_FLUSH_FREEZE);
+ ksz_write8(dev, REG_SW_MAC_CTRL_6, SW_MIB_COUNTER_FLUSH);
+ ksz_pwrite32(dev, port, REG_PORT_MIB_CTRL_STAT__4, 0);
+ mutex_unlock(&mib->cnt_mutex);
+
+ mib->cnt_ptr = 0;
+ memset(mib->counters, 0, dev->mib_cnt * sizeof(u64));
+}
+
static enum dsa_tag_protocol ksz9477_get_tag_protocol(struct dsa_switch *ds,
int port)
{
@@ -342,47 +428,6 @@ static void ksz9477_get_strings(struct dsa_switch *ds, int port,
}
}
-static void ksz_get_ethtool_stats(struct dsa_switch *ds, int port,
- uint64_t *buf)
-{
- struct ksz_device *dev = ds->priv;
- int i;
- u32 data;
- int timeout;
-
- mutex_lock(&dev->stats_mutex);
-
- for (i = 0; i < TOTAL_SWITCH_COUNTER_NUM; i++) {
- data = MIB_COUNTER_READ;
- data |= ((ksz9477_mib_names[i].index & 0xFF) <<
- MIB_COUNTER_INDEX_S);
- ksz_pwrite32(dev, port, REG_PORT_MIB_CTRL_STAT__4, data);
-
- timeout = 1000;
- do {
- ksz_pread32(dev, port, REG_PORT_MIB_CTRL_STAT__4,
- &data);
- usleep_range(1, 10);
- if (!(data & MIB_COUNTER_READ))
- break;
- } while (timeout-- > 0);
-
- /* failed to read MIB. get out of loop */
- if (!timeout) {
- dev_dbg(dev->dev, "Failed to get MIB\n");
- break;
- }
-
- /* count resets upon read */
- ksz_pread32(dev, port, REG_PORT_MIB_DATA, &data);
-
- dev->mib_value[i] += (uint64_t)data;
- buf[i] = dev->mib_value[i];
- }
-
- mutex_unlock(&dev->stats_mutex);
-}
-
static void ksz9477_cfg_port_member(struct ksz_device *dev, int port,
u8 member)
{
@@ -1153,9 +1198,14 @@ static int ksz9477_setup(struct dsa_switch *ds)
/* queue based egress rate limit */
ksz_cfg(dev, REG_SW_MAC_CTRL_5, SW_OUT_RATE_LIMIT_QUEUE_BASED, true);
+ /* enable global MIB counter freeze function */
+ ksz_cfg(dev, REG_SW_MAC_CTRL_6, SW_MIB_COUNTER_FREEZE, true);
+
/* start switch */
ksz_cfg(dev, REG_SW_OPERATION, SW_START, true);
+ ksz_init_mib_timer(dev);
+
return 0;
}
@@ -1290,6 +1340,7 @@ static int ksz9477_switch_init(struct ksz_device *dev)
if (!dev->ports)
return -ENOMEM;
for (i = 0; i < dev->mib_port_cnt; i++) {
+ mutex_init(&dev->ports[i].mib.cnt_mutex);
dev->ports[i].mib.counters =
devm_kzalloc(dev->dev,
sizeof(u64) *
@@ -1314,6 +1365,10 @@ static void ksz9477_switch_exit(struct ksz_device *dev)
.flush_dyn_mac_table = ksz9477_flush_dyn_mac_table,
.phy_setup = ksz9477_phy_setup,
.port_setup = ksz9477_port_setup,
+ .r_mib_cnt = ksz9477_r_mib_cnt,
+ .r_mib_pkt = ksz9477_r_mib_pkt,
+ .freeze_mib = ksz9477_freeze_mib,
+ .port_init_cnt = ksz9477_port_init_cnt,
.shutdown = ksz9477_reset_switch,
.detect = ksz9477_switch_detect,
.init = ksz9477_switch_init,
diff --git a/drivers/net/dsa/microchip/ksz_common.c b/drivers/net/dsa/microchip/ksz_common.c
index a57bda7..62d4344 100644
--- a/drivers/net/dsa/microchip/ksz_common.c
+++ b/drivers/net/dsa/microchip/ksz_common.c
@@ -40,6 +40,82 @@ void ksz_update_port_member(struct ksz_device *dev, int port)
}
EXPORT_SYMBOL_GPL(ksz_update_port_member);
+static void port_r_cnt(struct ksz_device *dev, int port)
+{
+ struct ksz_port_mib *mib = &dev->ports[port].mib;
+ u64 *dropped;
+
+ /* Some ports may not have MIB counters before SWITCH_COUNTER_NUM. */
+ while (mib->cnt_ptr < dev->reg_mib_cnt) {
+ dev->dev_ops->r_mib_cnt(dev, port, mib->cnt_ptr,
+ &mib->counters[mib->cnt_ptr]);
+ ++mib->cnt_ptr;
+ }
+
+ /* last one in storage */
+ dropped = &mib->counters[dev->mib_cnt];
+
+ /* Some ports may not have MIB counters after SWITCH_COUNTER_NUM. */
+ while (mib->cnt_ptr < dev->mib_cnt) {
+ dev->dev_ops->r_mib_pkt(dev, port, mib->cnt_ptr,
+ dropped, &mib->counters[mib->cnt_ptr]);
+ ++mib->cnt_ptr;
+ }
+ mib->cnt_ptr = 0;
+}
+
+static void ksz_mib_read_work(struct work_struct *work)
+{
+ struct ksz_device *dev =
+ container_of(work, struct ksz_device, mib_read);
+ struct ksz_port *p;
+ struct ksz_port_mib *mib;
+ int i;
+
+ for (i = 0; i < dev->mib_port_cnt; i++) {
+ p = &dev->ports[i];
+ if (!p->on)
+ continue;
+ mib = &p->mib;
+ mutex_lock(&mib->cnt_mutex);
+
+ /* read only dropped counters when link is not up */
+ if (p->link_just_down)
+ p->link_just_down = 0;
+ else if (!p->phydev.link)
+ mib->cnt_ptr = dev->reg_mib_cnt;
+ port_r_cnt(dev, i);
+ mutex_unlock(&mib->cnt_mutex);
+ }
+}
+
+static void mib_monitor(struct timer_list *t)
+{
+ struct ksz_device *dev = from_timer(dev, t, mib_read_timer);
+
+ mod_timer(&dev->mib_read_timer, jiffies + dev->mib_read_interval);
+ schedule_work(&dev->mib_read);
+}
+
+void ksz_init_mib_timer(struct ksz_device *dev)
+{
+ int i;
+
+ /* Read MIB counters every 30 seconds to avoid overflow. */
+ dev->mib_read_interval = msecs_to_jiffies(30000);
+
+ INIT_WORK(&dev->mib_read, ksz_mib_read_work);
+ timer_setup(&dev->mib_read_timer, mib_monitor, 0);
+
+ for (i = 0; i < dev->mib_port_cnt; i++)
+ dev->dev_ops->port_init_cnt(dev, i);
+
+ /* Start the timer 2 seconds later. */
+ dev->mib_read_timer.expires = jiffies + msecs_to_jiffies(2000);
+ add_timer(&dev->mib_read_timer);
+}
+EXPORT_SYMBOL_GPL(ksz_init_mib_timer);
+
int ksz_phy_read16(struct dsa_switch *ds, int addr, int reg)
{
struct ksz_device *dev = ds->priv;
@@ -88,6 +164,20 @@ int ksz_sset_count(struct dsa_switch *ds, int port, int sset)
}
EXPORT_SYMBOL_GPL(ksz_sset_count);
+void ksz_get_ethtool_stats(struct dsa_switch *ds, int port, uint64_t *buf)
+{
+ struct ksz_device *dev = ds->priv;
+ struct ksz_port_mib *mib;
+
+ mib = &dev->ports[port].mib;
+
+ mutex_lock(&mib->cnt_mutex);
+ port_r_cnt(dev, port);
+ memcpy(buf, mib->counters, dev->mib_cnt * sizeof(u64));
+ mutex_unlock(&mib->cnt_mutex);
+}
+EXPORT_SYMBOL_GPL(ksz_get_ethtool_stats);
+
int ksz_port_bridge_join(struct dsa_switch *ds, int port,
struct net_device *br)
{
@@ -355,6 +445,12 @@ int ksz_switch_register(struct ksz_device *dev,
void ksz_switch_remove(struct ksz_device *dev)
{
+ /* timer started */
+ if (dev->mib_read_timer.expires) {
+ del_timer_sync(&dev->mib_read_timer);
+ flush_work(&dev->mib_read);
+ }
+
dev->dev_ops->exit(dev);
dsa_unregister_switch(dev->ds);
diff --git a/drivers/net/dsa/microchip/ksz_common.h b/drivers/net/dsa/microchip/ksz_common.h
index 6d49319..f5f5fd8 100644
--- a/drivers/net/dsa/microchip/ksz_common.h
+++ b/drivers/net/dsa/microchip/ksz_common.h
@@ -8,6 +8,7 @@
#define __KSZ_COMMON_H
void ksz_update_port_member(struct ksz_device *dev, int port);
+void ksz_init_mib_timer(struct ksz_device *dev);
/* Common DSA access functions */
@@ -16,6 +17,7 @@
void ksz_adjust_link(struct dsa_switch *ds, int port,
struct phy_device *phydev);
int ksz_sset_count(struct dsa_switch *ds, int port, int sset);
+void ksz_get_ethtool_stats(struct dsa_switch *ds, int port, uint64_t *buf);
int ksz_port_bridge_join(struct dsa_switch *ds, int port,
struct net_device *br);
void ksz_port_bridge_leave(struct dsa_switch *ds, int port,
diff --git a/drivers/net/dsa/microchip/ksz_priv.h b/drivers/net/dsa/microchip/ksz_priv.h
index 0fdc58b..a35bb34 100644
--- a/drivers/net/dsa/microchip/ksz_priv.h
+++ b/drivers/net/dsa/microchip/ksz_priv.h
@@ -14,8 +14,6 @@
#include <linux/etherdevice.h>
#include <net/dsa.h>
-#include "ksz9477_reg.h"
-
struct ksz_io_ops;
struct vlan_table {
@@ -23,6 +21,7 @@ struct vlan_table {
};
struct ksz_port_mib {
+ struct mutex cnt_mutex; /* structure access */
u8 cnt_ptr;
u64 *counters;
};
@@ -39,6 +38,7 @@ struct ksz_port {
u32 sgmii:1; /* port is SGMII */
u32 force:1;
u32 link_just_down:1; /* link just goes down */
+ u32 freeze:1; /* MIB counter freeze is enabled */
struct ksz_port_mib mib;
};
@@ -79,8 +79,6 @@ struct ksz_device {
struct vlan_table *vlan_cache;
- u64 mib_value[TOTAL_SWITCH_COUNTER_NUM];
-
u8 *txbuf;
struct ksz_port *ports;
@@ -153,6 +151,7 @@ struct ksz_dev_ops {
u64 *cnt);
void (*r_mib_pkt)(struct ksz_device *dev, int port, u16 addr,
u64 *dropped, u64 *cnt);
+ void (*freeze_mib)(struct ksz_device *dev, int port, bool freeze);
void (*port_init_cnt)(struct ksz_device *dev, int port);
int (*shutdown)(struct ksz_device *dev);
int (*detect)(struct ksz_device *dev);
--
1.9.1
^ permalink raw reply related
* [PATCH v1 net-next 4/4] net: dsa: microchip: remove unnecessary include headers
From: Tristram.Ha @ 2019-02-08 4:07 UTC (permalink / raw)
To: Sergio Paracuellos, Andrew Lunn, Florian Fainelli, Pavel Machek
Cc: Tristram Ha, UNGLinuxDriver, netdev
In-Reply-To: <1549598829-25970-1-git-send-email-Tristram.Ha@microchip.com>
From: Tristram Ha <Tristram.Ha@microchip.com>
Remove unnecessary header include lines.
Signed-off-by: Tristram Ha <Tristram.Ha@microchip.com>
---
drivers/net/dsa/microchip/ksz9477.c | 4 ----
1 file changed, 4 deletions(-)
diff --git a/drivers/net/dsa/microchip/ksz9477.c b/drivers/net/dsa/microchip/ksz9477.c
index 8391b9e..7c51edd 100644
--- a/drivers/net/dsa/microchip/ksz9477.c
+++ b/drivers/net/dsa/microchip/ksz9477.c
@@ -5,15 +5,11 @@
* Copyright (C) 2017-2019 Microchip Technology Inc.
*/
-#include <linux/delay.h>
-#include <linux/export.h>
-#include <linux/gpio.h>
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/iopoll.h>
#include <linux/platform_data/microchip-ksz.h>
#include <linux/phy.h>
-#include <linux/etherdevice.h>
#include <linux/if_bridge.h>
#include <net/dsa.h>
#include <net/switchdev.h>
--
1.9.1
^ permalink raw reply related
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox