- Linux-kselftest-mirror - lists.linaro.org

[PATCH bpf-next] bpf: ringbuf: Support consuming BPF_MAP_TYPE_RINGBUF from prog

by Daniel Xu

Right now there exists prog produce / userspace consume and userspace produce / prog consume support. But it is also useful to have prog produce / prog consume. For example, we want to track the latency overhead of cpumap in production. Since we need to store enqueue timestamps somewhere and cpumap is MPSC, we need an MPSC data structure to shadow cpumap. BPF ringbuf is such a data structure. Rather than reimplement (possibly poorly) a custom ringbuffer in BPF, extend the existing interface to allow the final quadrant of ringbuf usecases to be filled. Note we ignore userspace to userspace use case - there is no need to involve kernel for that. Signed-off-by: Daniel Xu <dxu(a)dxuuu.xyz> --- kernel/bpf/verifier.c | 6 +- tools/testing/selftests/bpf/Makefile | 3 +- .../selftests/bpf/prog_tests/ringbuf.c | 50 +++++++++++++++ .../bpf/progs/test_ringbuf_bpf_to_bpf.c | 64 +++++++++++++++++++ 4 files changed, 120 insertions(+), 3 deletions(-) create mode 100644 tools/testing/selftests/bpf/progs/test_ringbuf_bpf_to_bpf.c diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 53d0556fbbf3..56bfe559f228 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -9142,7 +9142,8 @@ static int check_map_func_compatibility(struct bpf_verifier_env *env, func_id != BPF_FUNC_ringbuf_query && func_id != BPF_FUNC_ringbuf_reserve_dynptr && func_id != BPF_FUNC_ringbuf_submit_dynptr && - func_id != BPF_FUNC_ringbuf_discard_dynptr) + func_id != BPF_FUNC_ringbuf_discard_dynptr && + func_id != BPF_FUNC_user_ringbuf_drain) goto error; break; case BPF_MAP_TYPE_USER_RINGBUF: @@ -9276,7 +9277,8 @@ static int check_map_func_compatibility(struct bpf_verifier_env *env, goto error; break; case BPF_FUNC_user_ringbuf_drain: - if (map->map_type != BPF_MAP_TYPE_USER_RINGBUF) + if (map->map_type != BPF_MAP_TYPE_USER_RINGBUF && + map->map_type != BPF_MAP_TYPE_RINGBUF) goto error; break; case BPF_FUNC_get_stackid: diff --git a/tools/testing/selftests/bpf/Makefile b/tools/testing/selftests/bpf/Makefile index 9905e3739dd0..233109843d4d 100644 --- a/tools/testing/selftests/bpf/Makefile +++ b/tools/testing/selftests/bpf/Makefile @@ -503,7 +503,8 @@ LINKED_SKELS := test_static_linked.skel.h linked_funcs.skel.h \ LSKELS := fentry_test.c fexit_test.c fexit_sleep.c atomics.c \ trace_printk.c trace_vprintk.c map_ptr_kern.c \ core_kern.c core_kern_overflow.c test_ringbuf.c \ - test_ringbuf_n.c test_ringbuf_map_key.c test_ringbuf_write.c + test_ringbuf_n.c test_ringbuf_map_key.c test_ringbuf_write.c \ + test_ringbuf_bpf_to_bpf.c # Generate both light skeleton and libbpf skeleton for these LSKELS_EXTRA := test_ksyms_module.c test_ksyms_weak.c kfunc_call_test.c \ diff --git a/tools/testing/selftests/bpf/prog_tests/ringbuf.c b/tools/testing/selftests/bpf/prog_tests/ringbuf.c index da430df45aa4..3e7955573ac5 100644 --- a/tools/testing/selftests/bpf/prog_tests/ringbuf.c +++ b/tools/testing/selftests/bpf/prog_tests/ringbuf.c @@ -17,6 +17,7 @@ #include "test_ringbuf_n.lskel.h" #include "test_ringbuf_map_key.lskel.h" #include "test_ringbuf_write.lskel.h" +#include "test_ringbuf_bpf_to_bpf.lskel.h" #define EDONE 7777 @@ -497,6 +498,53 @@ static void ringbuf_map_key_subtest(void) test_ringbuf_map_key_lskel__destroy(skel_map_key); } +static void ringbuf_bpf_to_bpf_subtest(void) +{ + struct test_ringbuf_bpf_to_bpf_lskel *skel; + int err, i; + + skel = test_ringbuf_bpf_to_bpf_lskel__open(); + if (!ASSERT_OK_PTR(skel, "test_ringbuf_bpf_to_bpf_lskel__open")) + return; + + skel->maps.ringbuf.max_entries = getpagesize(); + skel->bss->pid = getpid(); + + err = test_ringbuf_bpf_to_bpf_lskel__load(skel); + if (!ASSERT_OK(err, "test_ringbuf_bpf_to_bpf_lskel__load")) + goto cleanup; + + ringbuf = ring_buffer__new(skel->maps.ringbuf.map_fd, NULL, NULL, NULL); + if (!ASSERT_OK_PTR(ringbuf, "ring_buffer__new")) + goto cleanup; + + err = test_ringbuf_bpf_to_bpf_lskel__attach(skel); + if (!ASSERT_OK(err, "test_ringbuf_bpf_to_bpf_lskel__attach")) + goto cleanup_ringbuf; + + /* Produce N_SAMPLES samples in the ring buffer by calling getpid() */ + skel->bss->value = SAMPLE_VALUE; + for (i = 0; i < N_SAMPLES; i++) + syscall(__NR_getpgid); + + /* Trigger bpf-side consumption */ + syscall(__NR_prctl); + + /* Check correct number samples were consumed */ + if (!ASSERT_EQ(skel->bss->round_tripped, N_SAMPLES, "round_tripped")) + goto cleanup_ringbuf; + + /* Check all samples were consumed */ + err = ring_buffer__consume(ringbuf); + if (!ASSERT_EQ(err, 0, "rb_consume")) + goto cleanup_ringbuf; + +cleanup_ringbuf: + ring_buffer__free(ringbuf); +cleanup: + test_ringbuf_bpf_to_bpf_lskel__destroy(skel); +} + void test_ringbuf(void) { if (test__start_subtest("ringbuf")) @@ -507,4 +555,6 @@ void test_ringbuf(void) ringbuf_map_key_subtest(); if (test__start_subtest("ringbuf_write")) ringbuf_write_subtest(); + if (test__start_subtest("ringbuf_bpf_to_bpf")) + ringbuf_bpf_to_bpf_subtest(); } diff --git a/tools/testing/selftests/bpf/progs/test_ringbuf_bpf_to_bpf.c b/tools/testing/selftests/bpf/progs/test_ringbuf_bpf_to_bpf.c new file mode 100644 index 000000000000..378154922024 --- /dev/null +++ b/tools/testing/selftests/bpf/progs/test_ringbuf_bpf_to_bpf.c @@ -0,0 +1,64 @@ +// SPDX-License-Identifier: GPL-2.0 + +#include <linux/bpf.h> +#include <unistd.h> +#include <bpf/bpf_helpers.h> +#include "bpf_misc.h" + +struct sample { + long value; +}; + +struct { + __uint(type, BPF_MAP_TYPE_RINGBUF); +} ringbuf SEC(".maps"); + +int pid = 0; +long value = 0; +int round_tripped = 0; + +SEC("fentry/" SYS_PREFIX "sys_getpgid") +int test_ringbuf_bpf_to_bpf_produce(void *ctx) +{ + int cur_pid = bpf_get_current_pid_tgid() >> 32; + struct sample *sample; + + if (cur_pid != pid) + return 0; + + sample = bpf_ringbuf_reserve(&ringbuf, sizeof(*sample), 0); + if (!sample) + return 0; + sample->value = value; + + bpf_ringbuf_submit(sample, 0); + return 0; +} + +static long consume_cb(struct bpf_dynptr *dynptr, void *context) +{ + struct sample *sample = NULL; + + sample = bpf_dynptr_data(dynptr, 0, sizeof(*sample)); + if (!sample) + return 0; + + if (sample->value == value) + round_tripped++; + + return 0; +} + +SEC("fentry/" SYS_PREFIX "sys_prctl") +int test_ringbuf_bpf_to_bpf_consume(void *ctx) +{ + int cur_pid = bpf_get_current_pid_tgid() >> 32; + + if (cur_pid != pid) + return 0; + + bpf_user_ringbuf_drain(&ringbuf, consume_cb, NULL, 0); + return 0; +} + +char _license[] SEC("license") = "GPL"; -- 2.46.0

1 year, 2 months

4
16
0 0

[PATCH] selftest/mm: Do not use hint for riscv mmap

by Chunyan Zhang

When the virtual address range selftest is run on RISC-V platforms, it is observed that using the hint address when calling mmap cannot get the address in the range of that validate_addr() checks, also that will cause '/proc/self/maps' have gaps larger than MAP_CHUNK_SIZE. Signed-off-by: Chunyan Zhang <zhangchunyan(a)iscas.ac.cn> --- tools/testing/selftests/mm/virtual_address_range.c | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/tools/testing/selftests/mm/virtual_address_range.c b/tools/testing/selftests/mm/virtual_address_range.c index 4e4c1e311247..25f3eb304999 100644 --- a/tools/testing/selftests/mm/virtual_address_range.c +++ b/tools/testing/selftests/mm/virtual_address_range.c @@ -64,6 +64,14 @@ #define NR_CHUNKS_HIGH NR_CHUNKS_384TB #endif +#if defined(__riscv) && (__riscv_xlen == 64) +static char *hind_addr(void) +{ + return NULL; +} + +static void validate_addr(char *ptr, int high_addr) { } +#else static char *hind_addr(void) { int bits = HIGH_ADDR_SHIFT + rand() % (63 - HIGH_ADDR_SHIFT); @@ -81,6 +89,7 @@ static void validate_addr(char *ptr, int high_addr) if (addr > HIGH_ADDR_MARK) ksft_exit_fail_msg("Bad address %lx\n", addr); } +#endif static int validate_lower_address_hint(void) { -- 2.34.1

1 year, 2 months

3
2
0 0

[PATCH] kunit: tool: Build compile_commands.json

by Brendan Jackman

compile_commands.json is used by clangd[1] to provide code navigation and completion functionality to editors. See [2] for an example configuration that includes this functionality for VSCode. It can currently be built manually when using kunit.py, by running: ./scripts/clang-tools/gen_compile_commands.py -d .kunit With this change however, it's built automatically so you don't need to manually keep it up to date. Unlike the manual approach, having make build the compile_commands.json means that it appears in the build output tree instead of at the root of the source tree, so you'll need to add --compile-commands-dir= to your clangd args for it to be found. [1] https://clangd.llvm.org/ [2] https://github.com/FlorentRevest/linux-kernel-vscode Signed-off-by: Brendan Jackman <jackmanb(a)google.com> --- tools/testing/kunit/kunit_kernel.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/tools/testing/kunit/kunit_kernel.py b/tools/testing/kunit/kunit_kernel.py index 7254c110ff23..61931c4926fd 100644 --- a/tools/testing/kunit/kunit_kernel.py +++ b/tools/testing/kunit/kunit_kernel.py @@ -72,7 +72,8 @@ class LinuxSourceTreeOperations: raise ConfigError(e.output.decode()) def make(self, jobs: int, build_dir: str, make_options: Optional[List[str]]) -> None: - command = ['make', 'ARCH=' + self._linux_arch, 'O=' + build_dir, '--jobs=' + str(jobs)] + command = ['make', 'all', 'compile_commands.json', 'ARCH=' + self._linux_arch, + 'O=' + build_dir, '--jobs=' + str(jobs)] if make_options: command.extend(make_options) if self._cross_compile: --- base-commit: 3c999d1ae3c75991902a1a7dad0cb62c2a3008b4 change-id: 20240516-kunit-compile-commands-d994074fc2be Best regards, -- Brendan Jackman <jackmanb(a)google.com>

1 year, 2 months

3
2
0 0

[PATCH v5] lib/math: Add int_pow test suite

by Luis Felipe Hernandez

Adds test suite for integer based power function. Signed-off-by: Luis Felipe Hernandez <luis.hernandez093(a)gmail.com> --- Changes in v5: - Fix kernel test bot warning - Rebase with latest Changes in v4: - Address checkpatch warning and make kconfig description longer - Use GPL-2.0-only for consistency - Spelling fix fith -> fifth Changes in v3: - Fix compiler warning: explicitly define constant as unsigned int - Add changes in patch revisions Changes in v2: - Address review feedback - Add kconfig entry - Use correct dir and file convention for KUnit - Fix typo - Remove unused static_stub header - Refactor test suite to use paramerterized test cases - Add close to max allowable value to in large_result test case - Add test case with non-power of two exponent - Fix module license --- lib/Kconfig.debug | 16 +++++++++++ lib/math/Makefile | 1 + lib/math/tests/Makefile | 3 ++ lib/math/tests/int_pow_kunit.c | 52 ++++++++++++++++++++++++++++++++++ 4 files changed, 72 insertions(+) create mode 100644 lib/math/tests/Makefile create mode 100644 lib/math/tests/int_pow_kunit.c diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index a30c03a66172..c415f8ca43a7 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -3051,3 +3051,19 @@ config RUST_KERNEL_DOCTESTS endmenu # "Rust" endmenu # Kernel hacking + +config INT_POW_TEST + tristate "Integer exponentiation (int_pow) test" if !KUNIT_ALL_TESTS + depends on KUNIT + default KUNIT_ALL_TESTS + help + This option enables the KUnit test suite for the int_pow function, + which performs integer exponentiation. The test suite is designed to + verify that the implementation of int_pow correctly computes the power + of a given base raised to a given exponent. + + Enabling this option will include tests that check various scenarios + and edge cases to ensure the accuracy and reliability of the exponentiation + function. + + If unsure, say N diff --git a/lib/math/Makefile b/lib/math/Makefile index 91fcdb0c9efe..3c1f92a7459d 100644 --- a/lib/math/Makefile +++ b/lib/math/Makefile @@ -5,5 +5,6 @@ obj-$(CONFIG_CORDIC) += cordic.o obj-$(CONFIG_PRIME_NUMBERS) += prime_numbers.o obj-$(CONFIG_RATIONAL) += rational.o +obj-$(CONFIG_INT_POW_TEST) += tests/int_pow_kunit.o obj-$(CONFIG_TEST_DIV64) += test_div64.o obj-$(CONFIG_RATIONAL_KUNIT_TEST) += rational-test.o diff --git a/lib/math/tests/Makefile b/lib/math/tests/Makefile new file mode 100644 index 000000000000..6a169123320a --- /dev/null +++ b/lib/math/tests/Makefile @@ -0,0 +1,3 @@ +# SPDX-License-Identifier: GPL-2.0-only + +obj-$(CONFIG_INT_POW_TEST) += int_pow_kunit.o diff --git a/lib/math/tests/int_pow_kunit.c b/lib/math/tests/int_pow_kunit.c new file mode 100644 index 000000000000..34b33677d458 --- /dev/null +++ b/lib/math/tests/int_pow_kunit.c @@ -0,0 +1,52 @@ +// SPDX-License-Identifier: GPL-2.0-only + +#include <kunit/test.h> +#include <linux/math.h> + +struct test_case_params { + u64 base; + unsigned int exponent; + u64 expected_result; + const char *name; +}; + +static const struct test_case_params params[] = { + { 64, 0, 1, "Power of zero" }, + { 64, 1, 64, "Power of one"}, + { 0, 5, 0, "Base zero" }, + { 1, 64, 1, "Base one" }, + { 2, 2, 4, "Two squared"}, + { 2, 3, 8, "Two cubed"}, + { 5, 5, 3125, "Five raised to the fifth power" }, + { U64_MAX, 1, U64_MAX, "Max base" }, + { 2, 63, 9223372036854775808ULL, "Large result"}, +}; + +static void get_desc(const struct test_case_params *tc, char *desc) +{ + strscpy(desc, tc->name, KUNIT_PARAM_DESC_SIZE); +} + +KUNIT_ARRAY_PARAM(int_pow, params, get_desc); + +static void int_pow_test(struct kunit *test) +{ + const struct test_case_params *tc = (const struct test_case_params *)test->param_value; + + KUNIT_EXPECT_EQ(test, tc->expected_result, int_pow(tc->base, tc->exponent)); +} + +static struct kunit_case math_int_pow_test_cases[] = { + KUNIT_CASE_PARAM(int_pow_test, int_pow_gen_params), + {} +}; + +static struct kunit_suite int_pow_test_suite = { + .name = "math-int_pow", + .test_cases = math_int_pow_test_cases, +}; + +kunit_test_suites(&int_pow_test_suite); + +MODULE_DESCRIPTION("math.int_pow KUnit test suite"); +MODULE_LICENSE("GPL"); -- 2.46.0

1 year, 2 months

3
2
0 0

[PATCH v2 0/3] selftests: kvm: s390: Add ucontrol memory selftests

by Christoph Schlameuss

This patch series adds a some not yet picked selftests to the kvm s390x selftest suite. The additional test cases are covering: * Assert KVM_EXIT_S390_UCONTROL exit on not mapped memory access * Assert functionality of storage keys in ucontrol VM * Assert that memory region operations are rejected for ucontrol VMs Running the test cases requires sys_admin capabilities to start the ucontrol VM. This can be achieved by running as root or with a command like: sudo setpriv --reuid nobody --inh-caps -all,+sys_admin \ --ambient-caps -all,+sys_admin --bounding-set -all,+sys_admin \ ./ucontrol_test --- The patches in this series have been part of the previous patch series. The test cases added here do depend on the fixture added in the earlier patches. From v5 PATCH 7-9 the segment and page table generation has been removed and DAT has been disabled. Since DAT is not necessary to validate the KVM code. https://lore.kernel.org/kvm/20240807154512.316936-1-schlameuss@linux.ibm.co… v2: - Reenable KSS intercept and handle it within skey test. - Modify the checked register between storing (sske) and reading (iske) it within the test program to make sure the. - Add an additional state assertion in the end of uc_skey - Fix some typos and white spaces. v1: - Remove segment and page table generation and disable DAT. This is not necessary to validate the KVM code. Christoph Schlameuss (3): selftests: kvm: s390: Add uc_map_unmap VM test case selftests: kvm: s390: Add uc_skey VM test case selftests: kvm: s390: Verify reject memory region operations for ucontrol VMs .../selftests/kvm/s390x/ucontrol_test.c | 232 +++++++++++++++++- 1 file changed, 230 insertions(+), 2 deletions(-) -- 2.46.0

1 year, 2 months

2
6
0 0

[PATCH v2] selftest/powerpc/benchmark: remove requirement libc-dev

by Madhavan Srinivasan

Currently exec-target.c file is linked as static and this post a requirement to install libc dev package to build. Without it, build-break when compiling selftest/powerpc/benchmark. CC exec_target /usr/bin/ld: cannot find -lc: No such file or directory collect2: error: ld returned 1 exit status exec_target.c is using "syscall" library function which could be replaced with a inline assembly and the same is proposed as a fix here. Suggested-by: Michael Ellerman <mpe(a)ellerman.id.au> Signed-off-by: Madhavan Srinivasan <maddy(a)linux.ibm.com> --- Chnagelog v1: - Add comment for clobber register and proper list of clobber registers as suggested by Michael Ellerman and Christophe Leroy .../selftests/powerpc/benchmarks/Makefile | 2 +- .../selftests/powerpc/benchmarks/exec_target.c | 16 ++++++++++++++-- 2 files changed, 15 insertions(+), 3 deletions(-) diff --git a/tools/testing/selftests/powerpc/benchmarks/Makefile b/tools/testing/selftests/powerpc/benchmarks/Makefile index 1321922038d0..ca4483c238b9 100644 --- a/tools/testing/selftests/powerpc/benchmarks/Makefile +++ b/tools/testing/selftests/powerpc/benchmarks/Makefile @@ -18,4 +18,4 @@ $(OUTPUT)/context_switch: LDLIBS += -lpthread $(OUTPUT)/fork: LDLIBS += -lpthread -$(OUTPUT)/exec_target: CFLAGS += -static -nostartfiles +$(OUTPUT)/exec_target: CFLAGS += -nostartfiles diff --git a/tools/testing/selftests/powerpc/benchmarks/exec_target.c b/tools/testing/selftests/powerpc/benchmarks/exec_target.c index c14b0fc1edde..a6408d3f26cd 100644 --- a/tools/testing/selftests/powerpc/benchmarks/exec_target.c +++ b/tools/testing/selftests/powerpc/benchmarks/exec_target.c @@ -7,10 +7,22 @@ */ #define _GNU_SOURCE -#include <unistd.h> #include <sys/syscall.h> void _start(void) { - syscall(SYS_exit, 0); + asm volatile ( + "li %%r0, %[sys_exit];" + "li %%r3, 0;" + "sc;" + : + : [sys_exit] "i" (SYS_exit) + /* + * "sc" will clobber r0, r3-r13, cr0, ctr, xer and memory. + * Even though sys_exit never returns, handle clobber + * registers. + */ + : "r0", "r3", "r4", "r5", "r6", "r7", "r8", "r9", "r10", + "r11", "r12", "r13", "cr0", "ctr", "xer", "memory" + ); } -- 2.45.2

1 year, 2 months

3
2
0 0

[PATCH RESEND v2 00/19] random: Resolve circular include dependency and include <linux/percpu.h>

by Uros Bizjak

Resent due to missing linux-kernel@ mailing list inclusion. There were several attempts to resolve circular include dependency after the addition of percpu.h: 1c9df907da83 ("random: fix circular include dependency on arm64 after addition of percpu.h"), c0842fbc1b18 ("random32: move the pseudo-random 32-bit definitions to prandom.h") and finally d9f29deb7fe8 ("prandom: Remove unused include") that completely removes the inclusion of <linux/percpu.h>. Due to legacy reasons, <linux/random.h> includes <linux/prandom.h>, but with the commit entry remark: --quote-- A further cleanup step would be to remove this from <linux/random.h> entirely, and make people who use the prandom infrastructure include just the new header file. That's a bit of a churn patch, but grepping for "prandom_" and "next_pseudo_random32" "struct rnd_state" should catch most users. But it turns out that that nice cleanup step is fairly painful, because a _lot_ of code currently seems to depend on the implicit include of <linux/random.h>, which can currently come in a lot of ways, including such fairly core headfers as <linux/net.h>. So the "nice cleanup" part may or may never happen. --/quote-- We would like to include <linux/percpu.h> in <linux/prandom.h>. In [1] we would like to repurpose __percpu tag as a named address space qualifier, where __percpu macro uses defines from <linux/percpu.h>. The major roadblock to inclusion of <linux/percpu.h> is the above mentioned legacy inclusion of <linux/prandom.h> in <linux/random.h> that causes circular include dependency that prevents <linux/percpu.h> inclusion. This patch series is the "nice cleanup" part that: a) Substitutes the inclusion of <linux/random.h> with the inclusion of <linux/prandom.h> where needed (patches 1 - 17). b) Removes legacy inclusion of <linux/prandom.h> from <linux/random.h> (patch 18). c) Includes <linux/percpu.h> in <linux/prandom.h> (patch 19). The whole series was tested by compiling the kernel for x86_64 allconfig and some popular architectures, namely arm64 defconfig, powerpc defconfig and loongarch defconfig. [1] https://lore.kernel.org/lkml/20240812115945.484051-4-ubizjak@gmail.com/ Cc: Dave Hansen <dave.hansen(a)linux.intel.com> Cc: Andy Lutomirski <luto(a)kernel.org> Cc: Peter Zijlstra <peterz(a)infradead.org> Cc: Thomas Gleixner <tglx(a)linutronix.de> Cc: Ingo Molnar <mingo(a)redhat.com> Cc: Borislav Petkov <bp(a)alien8.de> Cc: x86(a)kernel.org Cc: "H. Peter Anvin" <hpa(a)zytor.com> Cc: Jani Nikula <jani.nikula(a)linux.intel.com> Cc: Joonas Lahtinen <joonas.lahtinen(a)linux.intel.com> Cc: Rodrigo Vivi <rodrigo.vivi(a)intel.com> Cc: Tvrtko Ursulin <tursulin(a)ursulin.net> Cc: David Airlie <airlied(a)gmail.com> Cc: Daniel Vetter <daniel(a)ffwll.ch> Cc: Maarten Lankhorst <maarten.lankhorst(a)linux.intel.com> Cc: Maxime Ripard <mripard(a)kernel.org> Cc: Thomas Zimmermann <tzimmermann(a)suse.de> Cc: Hans Verkuil <hverkuil(a)xs4all.nl> Cc: Mauro Carvalho Chehab <mchehab(a)kernel.org> Cc: Miquel Raynal <miquel.raynal(a)bootlin.com> Cc: Richard Weinberger <richard(a)nod.at> Cc: Vignesh Raghavendra <vigneshr(a)ti.com> Cc: Eric Biggers <ebiggers(a)kernel.org> Cc: "Theodore Y. Ts'o" <tytso(a)mit.edu> Cc: Jaegeuk Kim <jaegeuk(a)kernel.org> Cc: "Jason A. Donenfeld" <Jason(a)zx2c4.com> Cc: Linus Torvalds <torvalds(a)linux-foundation.org> Cc: Hannes Reinecke <hare(a)suse.de> Cc: "James E.J. Bottomley" <James.Bottomley(a)HansenPartnership.com> Cc: "Martin K. Petersen" <martin.petersen(a)oracle.com> Cc: Alexei Starovoitov <ast(a)kernel.org> Cc: Daniel Borkmann <daniel(a)iogearbox.net> Cc: John Fastabend <john.fastabend(a)gmail.com> Cc: Andrii Nakryiko <andrii(a)kernel.org> Cc: Martin KaFai Lau <martin.lau(a)linux.dev> Cc: Eduard Zingerman <eddyz87(a)gmail.com> Cc: Song Liu <song(a)kernel.org> Cc: Yonghong Song <yonghong.song(a)linux.dev> Cc: KP Singh <kpsingh(a)kernel.org> Cc: Stanislav Fomichev <sdf(a)fomichev.me> Cc: Hao Luo <haoluo(a)google.com> Cc: Jiri Olsa <jolsa(a)kernel.org> Cc: Andrew Morton <akpm(a)linux-foundation.org> Cc: Brendan Higgins <brendan.higgins(a)linux.dev> Cc: David Gow <davidgow(a)google.com> Cc: Rae Moar <rmoar(a)google.com> Cc: "David S. Miller" <davem(a)davemloft.net> Cc: Eric Dumazet <edumazet(a)google.com> Cc: Jakub Kicinski <kuba(a)kernel.org> Cc: Paolo Abeni <pabeni(a)redhat.com> Cc: Jiri Pirko <jiri(a)resnulli.us> Cc: Petr Mladek <pmladek(a)suse.com> Cc: Steven Rostedt <rostedt(a)goodmis.org> Cc: Andy Shevchenko <andriy.shevchenko(a)linux.intel.com> Cc: Rasmus Villemoes <linux(a)rasmusvillemoes.dk> Cc: Sergey Senozhatsky <senozhatsky(a)chromium.org> Cc: Stephen Hemminger <stephen(a)networkplumber.org> Cc: Jamal Hadi Salim <jhs(a)mojatatu.com> Cc: Cong Wang <xiyou.wangcong(a)gmail.com> Cc: Kent Overstreet <kent.overstreet(a)linux.dev> --- v2: - Reword commit messages to mention the removal of legacy inclusion of <linux/prandom.h> from <linux/random.h> - Add missing substitution in crypto/testmgr.c (reported by kernel test robot) - Add Acked-by:. Uros Bizjak (19): x86/kaslr: Include <linux/prandom.h> instead of <linux/random.h> crypto: testmgr: Include <linux/prandom.h> instead of <linux/random.h> drm/i915/selftests: Include <linux/prandom.h> instead of <linux/random.h> drm/lib: Include <linux/prandom.h> instead of <linux/random.h> media: vivid: Include <linux/prandom.h> in vivid-vid-cap.c mtd: tests: Include <linux/prandom.h> instead of <linux/random.h> fscrypt: Include <linux/once.h> in fs/crypto/keyring.c scsi: libfcoe: Include <linux/prandom.h> instead of <linux/random.h> bpf: Include <linux/prandom.h> instead of <linux/random.h> lib/interval_tree_test.c: Include <linux/prandom.h> instead of <linux/random.h> kunit: string-stream-test: Include <linux/prandom.h> instead of <linux/random.h> random32: Include <linux/prandom.h> instead of <linux/random.h> lib/rbtree-test: Include <linux/prandom.h> instead of <linux/random.h> bpf/tests: Include <linux/prandom.h> instead of <linux/random.h> lib/test_parman: Include <linux/prandom.h> instead of <linux/random.h> lib/test_scanf: Include <linux/prandom.h> instead of <linux/random.h> netem: Include <linux/prandom.h> in sch_netem.c random: Do not include <linux/prandom.h> in <linux/random.h> prandom: Include <linux/percpu.h> in <linux/prandom.h> arch/x86/mm/kaslr.c | 2 +- crypto/testmgr.c | 2 +- drivers/gpu/drm/i915/selftests/i915_gem.c | 2 +- drivers/gpu/drm/i915/selftests/i915_random.h | 2 +- drivers/gpu/drm/i915/selftests/scatterlist.c | 2 +- drivers/gpu/drm/lib/drm_random.h | 2 +- drivers/media/test-drivers/vivid/vivid-vid-cap.c | 1 + drivers/mtd/tests/oobtest.c | 2 +- drivers/mtd/tests/pagetest.c | 2 +- drivers/mtd/tests/subpagetest.c | 2 +- fs/crypto/keyring.c | 1 + include/linux/prandom.h | 1 + include/linux/random.h | 7 ------- include/scsi/libfcoe.h | 2 +- kernel/bpf/core.c | 2 +- lib/interval_tree_test.c | 2 +- lib/kunit/string-stream-test.c | 1 + lib/random32.c | 2 +- lib/rbtree_test.c | 2 +- lib/test_bpf.c | 2 +- lib/test_parman.c | 2 +- lib/test_scanf.c | 2 +- net/sched/sch_netem.c | 1 + 23 files changed, 22 insertions(+), 24 deletions(-) -- 2.46.0

1 year, 2 months

8
28
0 0

[PATCH net-next v26 00/13] Device Memory TCP

by Mina Almasry

v26: https://patchwork.kernel.org/project/netdevbpf/list/?series=888227&state=* ==== No major changes. Only applied Reviewed-by tags from Jakub and addressed reported nits. v25: https://patchwork.kernel.org/project/netdevbpf/list/?series=885396&state=* === Major changes: - Moved devmem.h and mp_dmabuf_devmem.h to internal header files. - Changed the page_pool_params to take in a queue_idx rather than a struct netdev_rx_queue. - Added WARN_ON_ONCE around __skb_checksum readability check and added check to skb_checksum_help(). Other more minor feedback addressed as well. v24: https://patchwork.kernel.org/project/netdevbpf/list/?series=884556&state=* ==== No major changes. Mostly addressing issues in the error paths of dmabuf binding, and code cleanups/improvements from reviewers: Changes: - Fix failing ynl regen error. - Error path fixes & extack error messages in dmabuf binding. - Code cleanup in introspection. - gitignore ynl.d generated file. Full devmem TCP changes including the full GVE driver implementation is here: https://github.com/mina/linux/commits/tcpdevmem-v24/ v23: https://patchwork.kernel.org/project/netdevbpf/list/?series=882978&state=* ==== Fixing relatively minor issues called out in v22. (thanks again!) Mostly code cleanups, extack error messages, and minor reworks. Nothing major really changed, so the exact changes per commit is called in the commit messages. Full devmem TCP changes including the full GVE driver implementation is here: https://github.com/mina/linux/commits/tcpdevmem-v23/ v22: https://patchwork.kernel.org/project/netdevbpf/list/?series=881158&state=* ==== v22 aims to resolve the pending issue pointed to in v21, which is the interaction with xdp. In this series I rebase on top of the minor refactor which refactors propagating xdp configuration to slave devices: https://patchwork.kernel.org/project/netdevbpf/list/?series=881994&state=* I then disable setting xdp on devices using memory providers, and propagating xdp configuration to devices using memory providers. Full devmem TCP changes including the full GVE driver implementation is here: https://github.com/mina/linux/commits/tcpdevmem-v22/ v21: https://patchwork.kernel.org/project/netdevbpf/list/?series=880735&state=* ==== v20 addressed some comments and resolved a test failure, but introduced an unfortunate build error with a config edge case I wasn't testing. v21 simply resolves that error. Major Changes: - Resolve build error with CONFIG_PAGE_POOL=n && CONFIG_NET=y Full devmem TCP changes including the full GVE driver implementation is here: https://github.com/mina/linux/commits/tcpdevmem-v21/ v20: https://patchwork.kernel.org/project/netdevbpf/list/?series=879373&state=* ==== v20 aims to resolve a couple of bug reports against v19, and addresses some review comments around the page_pool_check_memory_provider mechanism. Major changes: - Test edge cases such as header split disabled in selftest. - Change `offset = 0` back to `offset = offset - start` to resolve issue found in RX path by Taehee (thanks!) - Address a few comments around page_pool_check_memory_provider() from Pavel & Jakub. - Removed some unnecessary includes across various patches in the series. - Removed unnecessary EXPORT_SYMBOL(page_pool_mem_providers) (Jakub). - Fix regression caused by incorrect dev_get_max_mp_channel check, along with rename (Jakub). Full devmem TCP changes including the full GVE driver implementation is here: https://github.com/mina/linux/commits/tcpdevmem-v20/ v19: https://patchwork.kernel.org/project/netdevbpf/list/?series=876852&state=* ==== v18 got a thorough review (thanks!), and this iteration addresses the feedback. Major changes: - Prevent deactivating mp bound queues. - Prevent installing xdp on mp bound netdevs, or installing mps on xdp installed netdevs. - Fix corner cases in netlink API vis-a-vis missing attributes. - Iron out the unreadable netmem driver support story. To be honest, the conversation with Jakub & Pavel got a bit confusing for me. I've implemented an approach in this set that makes sense to me, and AFAICT, addresses the requirements. It may be good as-is, or it may be a conversation starter/continuer. To be honest IMO there are many ways to skin this cat and I don't see an extremely strong reason to go for one approach over another. Here is one approach you may like. - Don't reset niov dma_addr on allocation & free. - Add some tests to the selftest that catches some of the issues around missing netlink attributes or deactivating mp-bound queues. Full devmem TCP changes including the full GVE driver implementation is here: https://github.com/mina/linux/commits/tcpdevmem-v19/ v18: https://patchwork.kernel.org/project/netdevbpf/list/?series=874848&state=* ==== v17 got minor feedback: (a) to beef up the description on patch 1 and (b) to remove the leading underscores in the header definition. I applied (a). (b) seems to be against current conventions so I did not apply before further discussion. Full devmem TCP changes including the full GVE driver implementation is here: https://github.com/mina/linux/commits/tcpdevmem-v17/ v17: https://patchwork.kernel.org/project/netdevbpf/list/?series=869900&state=* ==== v16 also got a very thorough review and some testing (thanks again!). Thes version addresses all the concerns reported on v15, in terms of feedback and issues reported. Major changes: - Use ASSERT_RTNL. - Moved around some of the page_pool helpers definitions so I can hide some netmem helpers in private files as Jakub suggested. - Don't make every net_iov hold a ref on the binding as Jakub suggested. - Fix issue reported by Taehee where we access queues after they have been freed. Full devmem TCP changes including the full GVE driver implementation is here: https://github.com/mina/linux/commits/tcpdevmem-v17/ v16: https://patchwork.kernel.org/project/netdevbpf/list/?series=866353&state=* ==== v15 got a thorough review and some testing, and this version addresses almost all the feedback. Some more minor comments where the authors said it could be done later, I left out. Major changes: - Addition of dma-buf introspection to page-pool-get and queue-get. - Fixes to selftests suggested by Taehee. - Fixes to documentation suggested by Donald. - A couple of suggestions and fixes to TCP patches by Eric and David. - Fixes to number assignements suggested by Arnd. - Use rtnl_lock()ing to guard against queue reconfiguration while the page_pool initialization is happening. (Jakub). - Fixes to a few warnings reproduced by Taehee. - Fixes to dma-buf binding suggested by Taehee and Jakub. - Fixes to netlink UAPI suggested by Jakub - Applied a number of Reviewed-bys and Acked-bys (including ones I lost from v13+). Full devmem TCP changes including the full GVE driver implementation is here: https://github.com/mina/linux/commits/tcpdevmem-v16/ One caveat: Taehee reproduced a KASAN warning and reported it here: https://lore.kernel.org/netdev/CAMArcTUdCxOBYGF3vpbq=eBvqZfnc44KBaQTN7H-wqd… I estimate the issue to be minor and easily fixable: https://lore.kernel.org/netdev/CAHS8izNgaqC--GGE2xd85QB=utUnOHmioCsDd1TNxJW… I hope to be able to follow up with a fix to net tree as net-next closes imminently, but if this iteration doesn't make it in, I will repost with a fix squashed after net-next reopens, no problem. v15: https://patchwork.kernel.org/project/netdevbpf/list/?series=865481&state=* ==== No material changes in this version, only a fix to linking against libynl.a from the last version. Per Jakub's instructions I've pulled one of his patches into this series, and now use the new libynl.a correctly, I hope. As usual, the full devmem TCP changes including the full GVE driver implementation is here: https://github.com/mina/linux/commits/tcpdevmem-v15/ v14: https://patchwork.kernel.org/project/netdevbpf/list/?series=865135&archive=… ==== No material changes in this version. Only rebase and re-verification on top of net-next. v13, I think, raced with commit ebad6d0334793 ("net/ipv4: Use nested-BH locking for ipv4_tcp_sk.") being merged to net-next that caused a patchwork failure to apply. This series should apply cleanly on commit c4532232fa2a4 ("selftests: net: remove unneeded IP_GRE config"). I did not wait the customary 24hr as Jakub said it's OK to repost as soon as I build test the rebased version: https://lore.kernel.org/netdev/20240625075926.146d769d@kernel.org/ v13: https://patchwork.kernel.org/project/netdevbpf/list/?series=861406&archive=… ==== Major changes: -------------- This iteration addresses Pavel's review comments, applies his reviewed-by's, and seeks to fix the patchwork build error (sorry!). As usual, the full devmem TCP changes including the full GVE driver implementation is here: https://github.com/mina/linux/commits/tcpdevmem-v13/ v12: https://patchwork.kernel.org/project/netdevbpf/list/?series=859747&state=* ==== Major changes: -------------- This iteration only addresses one minor comment from Pavel with regards to the trace printing of netmem, and the patchwork build error introduced in v11 because I missed doing an allmodconfig build, sorry. Other than that v11, AFAICT, received no feedback. There is one discussion about how the specifics of plugging io uring memory through the page pool, but not relevant to content in this particular patchset, AFAICT. As usual, the full devmem TCP changes including the full GVE driver implementation is here: https://github.com/mina/linux/commits/tcpdevmem-v12/ v11: https://patchwork.kernel.org/project/netdevbpf/list/?series=857457&state=* ==== Major Changes: -------------- v11 addresses feedback received in v10. The major change is the removal of the memory provider ops as requested by Christoph. We still accomplish the same thing, but utilizing direct function calls with if statements rather than generic ops. Additionally address sparse warnings, bugs and review comments from folks that reviewed. As usual, the full devmem TCP changes including the full GVE driver implementation is here: https://github.com/mina/linux/commits/tcpdevmem-v11/ Detailed changelog: ------------------- - Fixes in netdev_rx_queue_restart() from Pavel & David. - Remove commit e650e8c3a36f5 ("net: page_pool: create hooks for custom page providers") from the series to address Christoph's feedback and rebased other patches on the series on this change. - Fixed build errors with CONFIG_DMA_SHARED_BUFFER && !CONFIG_GENERIC_ALLOCATOR build. - Fixed sparse warnings pointed out by Paolo. - Drop unnecessary gro_pull_from_frag0 checks. - Added Bagas reviewed-by to docs. v10: https://patchwork.kernel.org/project/netdevbpf/list/?series=852422&state=* ==== Major Changes: -------------- v9 was sent right before the merge window closed (sorry!). v10 is almost a re-send of the series now that the merge window re-opened. Only rebased to latest net-next and addressed some minor iterative comments received on v9. As usual, the full devmem TCP changes including the full GVE driver implementation is here: https://github.com/mina/linux/commits/tcpdevmem-v10/ Detailed changelog: ------------------- - Fixed tokens leaking in DONTNEED setsockopt (Nikolay). - Moved net_iov_dma_addr() to devmem.c and made it a devmem specific helpers (David). - Rename hook alloc_pages to alloc_netmems as alloc_pages is now preprocessor macro defined and causes a build error. v9: === Major Changes: -------------- GVE queue API has been merged. Submitting this version as non-RFC after rebasing on top of the merged API, and dropped the out of tree queue API I was carrying on github. Addressed the little feedback v8 has received. Detailed changelog: ------------------ - Added new patch from David Wei to this series for netdev_rx_queue_restart() - Fixed sparse error. - Removed CONFIG_ checks in netmem_is_net_iov() - Flipped skb->readable to skb->unreadable - Minor fixes to selftests & docs. RFC v8: ======= Major Changes: -------------- - Fixed build error generated by patch-by-patch build. - Applied docs suggestions from Randy. RFC v7: ======= Major Changes: -------------- This revision largely rebases on top of net-next and addresses the feedback RFCv6 received from folks, namely Jakub, Yunsheng, Arnd, David, & Pavel. The series remains in RFC because the queue-API ndos defined in this series are not yet implemented. I have a GVE implementation I carry out of tree for my testing. A upstreamable GVE implementation is in the works. Aside from that, in my estimation all the patches are ready for review/merge. Please do take a look. As usual the full devmem TCP changes including the full GVE driver implementation is here: https://github.com/mina/linux/commits/tcpdevmem-v7/ Detailed changelog: - Use admin-perm in netlink API. - Addressed feedback from Jakub with regards to netlink API implementation. - Renamed devmem.c functions to something more appropriate for that file. - Improve the performance seen through the page_pool benchmark. - Fix the value definition of all the SO_DEVMEM_* uapi. - Various fixes to documentation. Perf - page-pool benchmark: --------------------------- Improved performance of bench_page_pool_simple.ko tests compared to v6: https://pastebin.com/raw/v5dYRg8L net-next base: 8 cycle fast path. RFC v6: 10 cycle fast path. RFC v7: 9 cycle fast path. RFC v7 with CONFIG_DMA_SHARED_BUFFER disabled: 8 cycle fast path, same as baseline. Perf - Devmem TCP benchmark: --------------------- Perf is about the same regardless of the changes in v7, namely the removal of the static_branch_unlikely to improve the page_pool benchmark performance: 189/200gbps bi-directional throughput with RX devmem TCP and regular TCP TX i.e. ~95% line rate. RFC v6: ======= Major Changes: -------------- This revision largely rebases on top of net-next and addresses the little feedback RFCv5 received. The series remains in RFC because the queue-API ndos defined in this series are not yet implemented. I have a GVE implementation I carry out of tree for my testing. A upstreamable GVE implementation is in the works. Aside from that, in my estimation all the patches are ready for review/merge. Please do take a look. As usual the full devmem TCP changes including the full GVE driver implementation is here: https://github.com/mina/linux/commits/tcpdevmem-v6/ This version also comes with some performance data recorded in the cover letter (see below changelog). Detailed changelog: - Rebased on top of the merged netmem_ref changes. - Converted skb->dmabuf to skb->readable (Pavel). Pavel's original suggestion was to remove the skb->dmabuf flag entirely, but when I looked into it closely, I found the issue that if we remove the flag we have to dereference the shinfo(skb) pointer to obtain the first frag to tell whether an skb is readable or not. This can cause a performance regression if it dirties the cache line when the shinfo(skb) was not really needed. Instead, I converted the skb->dmabuf flag into a generic skb->readable flag which can be re-used by io_uring 0-copy RX. - Squashed a few locking optimizations from Eric Dumazet in the RX path and the DEVMEM_DONTNEED setsockopt. - Expanded the tests a bit. Added validation for invalid scenarios and added some more coverage. Perf - page-pool benchmark: --------------------------- bench_page_pool_simple.ko tests with and without these changes: https://pastebin.com/raw/ncHDwAbn AFAIK the number that really matters in the perf tests is the 'tasklet_page_pool01_fast_path Per elem'. This one measures at about 8 cycles without the changes but there is some 1 cycle noise in some results. With the patches this regresses to 9 cycles with the changes but there is 1 cycle noise occasionally running this test repeatedly. Lastly I tried disable the static_branch_unlikely() in netmem_is_net_iov() check. To my surprise disabling the static_branch_unlikely() check reduces the fast path back to 8 cycles, but the 1 cycle noise remains. Perf - Devmem TCP benchmark: --------------------- 189/200gbps bi-directional throughput with RX devmem TCP and regular TCP TX i.e. ~95% line rate. Major changes in RFC v5: ======================== 1. Rebased on top of 'Abstract page from net stack' series and used the new netmem type to refer to LSB set pointers instead of re-using struct page. 2. Downgraded this series back to RFC and called it RFC v5. This is because this series is now dependent on 'Abstract page from net stack'[1] and the queue API. Both are removed from the series to reduce the patch # and those bits are fairly independent or pre-requisite work. 3. Reworked the page_pool devmem support to use netmem and for some more unified handling. 4. Reworked the reference counting of net_iov (renamed from page_pool_iov) to use pp_ref_count for refcounting. The full changes including the dependent series and GVE page pool support is here: https://github.com/mina/linux/commits/tcpdevmem-rfcv5/ [1] https://patchwork.kernel.org/project/netdevbpf/list/?series=810774 Major changes in v1: ==================== 1. Implemented MVP queue API ndos to remove the userspace-visible driver reset. 2. Fixed issues in the napi_pp_put_page() devmem frag unref path. 3. Removed RFC tag. Many smaller addressed comments across all the patches (patches have individual change log). Full tree including the rest of the GVE driver changes: https://github.com/mina/linux/commits/tcpdevmem-v1 Changes in RFC v3: ================== 1. Pulled in the memory-provider dependency from Jakub's RFC[1] to make the series reviewable and mergeable. 2. Implemented multi-rx-queue binding which was a todo in v2. 3. Fix to cmsg handling. The sticking point in RFC v2[2] was the device reset required to refill the device rx-queues after the dmabuf bind/unbind. The solution suggested as I understand is a subset of the per-queue management ops Jakub suggested or similar: https://lore.kernel.org/netdev/20230815171638.4c057dcd@kernel.org/ This is not addressed in this revision, because: 1. This point was discussed at netconf & netdev and there is openness to using the current approach of requiring a device reset. 2. Implementing individual queue resetting seems to be difficult for my test bed with GVE. My prototype to test this ran into issues with the rx-queues not coming back up properly if reset individually. At the moment I'm unsure if it's a mistake in the POC or a genuine issue in the virtualization stack behind GVE, which currently doesn't test individual rx-queue restart. 3. Our usecases are not bothered by requiring a device reset to refill the buffer queues, and we'd like to support NICs that run into this limitation with resetting individual queues. My thought is that drivers that have trouble with per-queue configs can use the support in this series, while drivers that support new netdev ops to reset individual queues can automatically reset the queue as part of the dma-buf bind/unbind. The same approach with device resets is presented again for consideration with other sticking points addressed. This proposal includes the rx devmem path only proposed for merge. For a snapshot of my entire tree which includes the GVE POC page pool support & device memory support: https://github.com/torvalds/linux/compare/master...mina:linux:tcpdevmem-v3 [1] https://lore.kernel.org/netdev/f8270765-a27b-6ccf-33ea-cda097168d79@redhat.… [2] https://lore.kernel.org/netdev/CAHS8izOVJGJH5WF68OsRWFKJid1_huzzUK+hpKbLcL4… Changes in RFC v2: ================== The sticking point in RFC v1[1] was the dma-buf pages approach we used to deliver the device memory to the TCP stack. RFC v2 is a proof-of-concept that attempts to resolve this by implementing scatterlist support in the networking stack, such that we can import the dma-buf scatterlist directly. This is the approach proposed at a high level here[2]. Detailed changes: 1. Replaced dma-buf pages approach with importing scatterlist into the page pool. 2. Replace the dma-buf pages centric API with a netlink API. 3. Removed the TX path implementation - there is no issue with implementing the TX path with scatterlist approach, but leaving out the TX path makes it easier to review. 4. Functionality is tested with this proposal, but I have not conducted perf testing yet. I'm not sure there are regressions, but I removed perf claims from the cover letter until they can be re-confirmed. 5. Added Signed-off-by: contributors to the implementation. 6. Fixed some bugs with the RX path since RFC v1. Any feedback welcome, but specifically the biggest pending questions needing feedback IMO are: 1. Feedback on the scatterlist-based approach in general. 2. Netlink API (Patch 1 & 2). 3. Approach to handle all the drivers that expect to receive pages from the page pool (Patch 6). [1] https://lore.kernel.org/netdev/dfe4bae7-13a0-3c5d-d671-f61b375cb0b4@gmail.c… [2] https://lore.kernel.org/netdev/CAHS8izPm6XRS54LdCDZVd0C75tA1zHSu6jLVO8nzTLX… ================== * TL;DR: Device memory TCP (devmem TCP) is a proposal for transferring data to and/or from device memory efficiently, without bouncing the data to a host memory buffer. * Problem: A large amount of data transfers have device memory as the source and/or destination. Accelerators drastically increased the volume of such transfers. Some examples include: - ML accelerators transferring large amounts of training data from storage into GPU/TPU memory. In some cases ML training setup time can be as long as 50% of TPU compute time, improving data transfer throughput & efficiency can help improving GPU/TPU utilization. - Distributed training, where ML accelerators, such as GPUs on different hosts, exchange data among them. - Distributed raw block storage applications transfer large amounts of data with remote SSDs, much of this data does not require host processing. Today, the majority of the Device-to-Device data transfers the network are implemented as the following low level operations: Device-to-Host copy, Host-to-Host network transfer, and Host-to-Device copy. The implementation is suboptimal, especially for bulk data transfers, and can put significant strains on system resources, such as host memory bandwidth, PCIe bandwidth, etc. One important reason behind the current state is the kernel’s lack of semantics to express device to network transfers. * Proposal: In this patch series we attempt to optimize this use case by implementing socket APIs that enable the user to: 1. send device memory across the network directly, and 2. receive incoming network packets directly into device memory. Packet _payloads_ go directly from the NIC to device memory for receive and from device memory to NIC for transmit. Packet _headers_ go to/from host memory and are processed by the TCP/IP stack normally. The NIC _must_ support header split to achieve this. Advantages: - Alleviate host memory bandwidth pressure, compared to existing network-transfer + device-copy semantics. - Alleviate PCIe BW pressure, by limiting data transfer to the lowest level of the PCIe tree, compared to traditional path which sends data through the root complex. * Patch overview: ** Part 1: netlink API Gives user ability to bind dma-buf to an RX queue. ** Part 2: scatterlist support Currently the standard for device memory sharing is DMABUF, which doesn't generate struct pages. On the other hand, networking stack (skbs, drivers, and page pool) operate on pages. We have 2 options: 1. Generate struct pages for dmabuf device memory, or, 2. Modify the networking stack to process scatterlist. Approach #1 was attempted in RFC v1. RFC v2 implements approach #2. ** part 3: page pool support We piggy back on page pool memory providers proposal: https://github.com/kuba-moo/linux/tree/pp-providers It allows the page pool to define a memory provider that provides the page allocation and freeing. It helps abstract most of the device memory TCP changes from the driver. ** part 4: support for unreadable skb frags Page pool iovs are not accessible by the host; we implement changes throughput the networking stack to correctly handle skbs with unreadable frags. ** Part 5: recvmsg() APIs We define user APIs for the user to send and receive device memory. Not included with this series is the GVE devmem TCP support, just to simplify the review. Code available here if desired: https://github.com/mina/linux/tree/tcpdevmem This series is built on top of net-next with Jakub's pp-providers changes cherry-picked. * NIC dependencies: 1. (strict) Devmem TCP require the NIC to support header split, i.e. the capability to split incoming packets into a header + payload and to put each into a separate buffer. Devmem TCP works by using device memory for the packet payload, and host memory for the packet headers. 2. (optional) Devmem TCP works better with flow steering support & RSS support, i.e. the NIC's ability to steer flows into certain rx queues. This allows the sysadmin to enable devmem TCP on a subset of the rx queues, and steer devmem TCP traffic onto these queues and non devmem TCP elsewhere. The NIC I have access to with these properties is the GVE with DQO support running in Google Cloud, but any NIC that supports these features would suffice. I may be able to help reviewers bring up devmem TCP on their NICs. * Testing: The series includes a udmabuf kselftest that show a simple use case of devmem TCP and validates the entire data path end to end without a dependency on a specific dmabuf provider. ** Test Setup Kernel: net-next with this series and memory provider API cherry-picked locally. Hardware: Google Cloud A3 VMs. NIC: GVE with header split & RSS & flow steering support. Cc: Pavel Begunkov <asml.silence(a)gmail.com> Cc: David Wei <dw(a)davidwei.uk> Cc: Jason Gunthorpe <jgg(a)ziepe.ca> Cc: Yunsheng Lin <linyunsheng(a)huawei.com> Cc: Shailend Chand <shailend(a)google.com> Cc: Harshitha Ramamurthy <hramamurthy(a)google.com> Cc: Shakeel Butt <shakeel.butt(a)linux.dev> Cc: Jeroen de Borst <jeroendb(a)google.com> Cc: Praveen Kaligineedi <pkaligineedi(a)google.com> Cc: Bagas Sanjaya <bagasdotme(a)gmail.com> Cc: Steven Rostedt <rostedt(a)goodmis.org> Cc: Christoph Hellwig <hch(a)infradead.org> Cc: Nikolay Aleksandrov <razor(a)blackwall.org> Cc: Taehee Yoo <ap420073(a)gmail.com> Cc: Donald Hunter <donald.hunter(a)gmail.com> Mina Almasry (13): netdev: add netdev_rx_queue_restart() net: netdev netlink api to bind dma-buf to a net device netdev: support binding dma-buf to netdevice netdev: netdevice devmem allocator page_pool: devmem support memory-provider: dmabuf devmem memory provider net: support non paged skb frags net: add support for skbs with unreadable frags tcp: RX path for devmem TCP net: add SO_DEVMEM_DONTNEED setsockopt to release RX frags net: add devmem TCP documentation selftests: add ncdevmem, netcat for devmem TCP netdev: add dmabuf introspection Documentation/netlink/specs/netdev.yaml | 61 +++ Documentation/networking/devmem.rst | 269 +++++++++++ Documentation/networking/index.rst | 1 + arch/alpha/include/uapi/asm/socket.h | 6 + arch/mips/include/uapi/asm/socket.h | 6 + arch/parisc/include/uapi/asm/socket.h | 6 + arch/sparc/include/uapi/asm/socket.h | 6 + include/linux/netdevice.h | 2 + include/linux/skbuff.h | 61 ++- include/linux/skbuff_ref.h | 9 +- include/linux/socket.h | 1 + include/net/netdev_rx_queue.h | 5 + include/net/netmem.h | 132 +++++- include/net/page_pool/helpers.h | 39 +- include/net/page_pool/types.h | 23 +- include/net/sock.h | 2 + include/net/tcp.h | 3 +- include/trace/events/page_pool.h | 12 +- include/uapi/asm-generic/socket.h | 6 + include/uapi/linux/netdev.h | 13 + include/uapi/linux/uio.h | 18 + net/Kconfig | 5 + net/core/Makefile | 2 + net/core/datagram.c | 6 + net/core/dev.c | 33 +- net/core/devmem.c | 389 ++++++++++++++++ net/core/devmem.h | 180 ++++++++ net/core/gro.c | 3 +- net/core/mp_dmabuf_devmem.h | 44 ++ net/core/netdev-genl-gen.c | 23 + net/core/netdev-genl-gen.h | 6 + net/core/netdev-genl.c | 139 +++++- net/core/netdev_rx_queue.c | 81 ++++ net/core/netmem_priv.h | 31 ++ net/core/page_pool.c | 119 +++-- net/core/page_pool_priv.h | 46 ++ net/core/page_pool_user.c | 32 +- net/core/skbuff.c | 77 +++- net/core/sock.c | 68 +++ net/ethtool/common.c | 8 + net/ipv4/esp4.c | 3 +- net/ipv4/tcp.c | 263 ++++++++++- net/ipv4/tcp_input.c | 13 +- net/ipv4/tcp_ipv4.c | 16 + net/ipv4/tcp_minisocks.c | 2 + net/ipv4/tcp_output.c | 5 +- net/ipv6/esp6.c | 3 +- net/packet/af_packet.c | 4 +- net/xdp/xsk_buff_pool.c | 5 + tools/include/uapi/linux/netdev.h | 13 + tools/net/ynl/lib/.gitignore | 1 + tools/testing/selftests/net/.gitignore | 1 + tools/testing/selftests/net/Makefile | 9 + tools/testing/selftests/net/ncdevmem.c | 570 ++++++++++++++++++++++++ 54 files changed, 2757 insertions(+), 124 deletions(-) create mode 100644 Documentation/networking/devmem.rst create mode 100644 net/core/devmem.c create mode 100644 net/core/devmem.h create mode 100644 net/core/mp_dmabuf_devmem.h create mode 100644 net/core/netdev_rx_queue.c create mode 100644 net/core/netmem_priv.h create mode 100644 tools/testing/selftests/net/ncdevmem.c -- 2.46.0.598.g6f2099f65c-goog

1 year, 2 months

2
14
0 0

[PATCH net 0/3] selftests: mptcp: misc. small fixes

by Matthieu Baerts (NGI0)

Here are some various fixes for the MPTCP selftests. Patch 1 fixes a recently modified test to continue to work as expected on older kernels. This is a fix for a recent fix that can be backported up to v5.15. Patch 2 and 3 include dependences when exporting or installing the tests. Two fixes for v6.11-rc1. Signed-off-by: Matthieu Baerts (NGI0) <matttbe(a)kernel.org> --- Matthieu Baerts (NGI0) (3): selftests: mptcp: join: restrict fullmesh endp on 1st sf selftests: mptcp: include lib.sh file selftests: mptcp: include net_helper.sh file tools/testing/selftests/net/mptcp/Makefile | 2 ++ tools/testing/selftests/net/mptcp/mptcp_join.sh | 4 +++- 2 files changed, 5 insertions(+), 1 deletion(-) --- base-commit: 48aa361c5db0b380c2b75c24984c0d3e7c1e8c09 change-id: 20240910-net-selftests-mptcp-fix-install-2b82ae5a99c8 Best regards, -- Matthieu Baerts (NGI0) <matttbe(a)kernel.org>

1 year, 2 months

2
4
0 0

[PATCH] selftests: kselftest: Use strerror() on nolibc

by zhangjiao2

From: zhang jiao <zhangjiao2(a)cmss.chinamobile.com> Nolibc gained an implementation of strerror() recently. Use it and drop the ifndef. Signed-off-by: zhang jiao <zhangjiao2(a)cmss.chinamobile.com> --- tools/testing/selftests/kselftest.h | 8 -------- 1 file changed, 8 deletions(-) diff --git a/tools/testing/selftests/kselftest.h b/tools/testing/selftests/kselftest.h index e195ec156859..29fedf609611 100644 --- a/tools/testing/selftests/kselftest.h +++ b/tools/testing/selftests/kselftest.h @@ -373,15 +373,7 @@ static inline __noreturn __printf(1, 2) void ksft_exit_fail_msg(const char *msg, static inline __noreturn void ksft_exit_fail_perror(const char *msg) { -#ifndef NOLIBC ksft_exit_fail_msg("%s: %s (%d)\n", msg, strerror(errno), errno); -#else - /* - * nolibc doesn't provide strerror() and it seems - * inappropriate to add one, just print the errno. - */ - ksft_exit_fail_msg("%s: %d)\n", msg, errno); -#endif } static inline __noreturn void ksft_exit_xfail(void) -- 2.33.0

1 year, 2 months

3
3
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-kselftest-mirror