Re: [PATCH net v3 1/2] skmsg: prevent empty ingress skb from enqueuing

Hi John, On Wed, 2024-07-03 at 09:54 +0800, Geliang Tang wrote: > On Tue, 2024-07-02 at 18:03 -0700, John Fastabend wrote: > > Geliang Tang wrote: > > > From: Geliang Tang > > > > > > Run this BPF selftests (./test_progs -t sockmap_basic) on a > > > Loongarch > > > platform, a Kernel panic occur

[PATCH bpf-next v8 9/9] selftests/bpf: Drop make_socket in sk_lookup

From: Geliang Tang Use local helper make_client() instead of using make_socket() and connect(). Then make_socket() and inetaddr_len() can be dropped now. Signed-off-by: Geliang Tang --- .../selftests/bpf/prog_tests/sk_lookup.c | 45 +-- 1 file changed, 2 insertions(+), 43

[PATCH bpf-next v8 8/9] selftests/bpf: Use connect_to_addr in sk_lookup

From: Geliang Tang Use public network helpers make_sockaddr() and connect_to_addr() instead of using the local defined function make_socket() and connect(). This make_socket() can be dropped latter. Signed-off-by: Geliang Tang --- .../selftests/bpf/prog_tests/sk_lookup.c | 22 ---

[PATCH bpf-next v8 7/9] selftests/bpf: Use connect_to_fd_opts in sk_lookup

From: Geliang Tang This patch uses public helper connect_to_fd_opts() exported in network_helpers.h instead of using make_socket() and connect() in prog_tests/sk_lookup.c. This can simplify the code. Signed-off-by: Geliang Tang --- .../selftests/bpf/prog_tests/sk_lookup.c | 25 +++

[PATCH bpf-next v8 6/9] selftests/bpf: Use start_server_str in sk_lookup

From: Geliang Tang This patch uses public helper start_server_addr() instead of local defined function make_server() in prog_tests/sk_lookup.c to avoid duplicate code. Add a helper setsockopts() to set SOL_CUSTOM sockopt looply, set it to setsockopt pointer of struct network_helper_opts, and pas

[PATCH bpf-next v8 4/9] selftests/bpf: Use make_sockaddr in sockmap_ktls

From: Geliang Tang This patch uses public helper make_sockaddr() exported in network_helpers.h instead of open-coding in sockmap_ktls.c. This can avoid duplicate code. Acked-by: Eduard Zingerman Signed-off-by: Geliang Tang --- .../selftests/bpf/prog_tests/sockmap_ktls.c | 16 ++--

[PATCH bpf-next v8 5/9] selftests/bpf: Close fd in error path in drop_on_reuseport

From: Geliang Tang Server 1 fd should be closed in the error path when update_lookup_map() fails. This patch fixes it by goto "close_srv1" instead of "detach" lable in that case. Acked-by: Eduard Zingerman Signed-off-by: Geliang Tang --- tools/testing/selftests/bpf/prog_tests/sk_lookup.c | 2

[PATCH bpf-next v8 3/9] selftests/bpf: Use connect_to_fd_opts in sockmap_ktls

From: Geliang Tang Use public network helper connect_to_fd_opts() instead of open-coding it in prog_tests/sockmap_ktls.c. This can avoid duplicate code. Acked-by: Eduard Zingerman Signed-off-by: Geliang Tang --- .../selftests/bpf/prog_tests/sockmap_ktls.c| 14 ++ 1 file ch

[PATCH bpf-next v8 2/9] selftests/bpf: Use start_server_str in sockmap_ktls

From: Geliang Tang Include network_helpers.h in prog_tests/sockmap_ktls.c, use public network helper start_server_str() instead of local defined function tcp_server(). This can avoid duplicate code. Technically, this is not a one-for-one replacement, as start_server_str() also does bind(). But t

[PATCH bpf-next v8 1/9] selftests/bpf: Add backlog for network_helper_opts

From: Geliang Tang Some callers expect __start_server() helper to pass their own "backlog" value to listen() instead of the default of 1. So this patch adds struct member "backlog" for network_helper_opts to allow callers to set "backlog" value via start_server_str() helper. listen(fd, 0 /* back

[PATCH bpf-next v8 0/9] use network helpers, part 8

From: Geliang Tang v8: - only patch 8 updated, to fix errors reported by CI. v7: - address Martin's comments in v6. (thanks) - use MAX(opts->backlog, 0) instead of opts->backlog. - use connect_to_fd_opts instead connect_to_fd. - more ASSERT_* to check errors. v6: - update patch 6 as Danie

Re: [PATCH 1/1] KVM: selftests: pmu_counters_test: increase robustness of LLC cache misses

On Thu, 2024-06-27 at 10:42 -0700, Sean Christopherson wrote: > On Fri, Jun 21, 2024, Maxim Levitsky wrote: > > Currently this test does a single CLFLUSH on its memory location > > but due to speculative execution this might not cause LLC misses. > > > > Instead, do a cache flush on each loop iter

Re: [PATCH 0/6] skip ENOTSUPP for BPF selftests

On Thu, 2024-07-04 at 14:48 +0800, Geliang Tang wrote: > From: Geliang Tang > > BPF selftests seem to have not been fully tested on Loongarch > platforms. > There are so many "ENOTSUPP" (-524) errors when running BPF selftests > on > them since lacking BPF trampoline on Loongarch. > > For these

[PATCH bpf-next v2 6/6] selftests/bpf: Skip ENOTSUPP in ASSERT_GE

From: Geliang Tang There are still some "ENOTSUPP" (-524) errors left when running BPF selftests on a Loongarch platform since ASSERT_GE() are used there to check the return values, not ASSERT_OK(): ''' test_bpf_cookie:PASS:skel_open 0 nsec #17/1bpf_cookie/kprobe:OK #17/2bpf_cookie/mu

[PATCH bpf-next v2 5/6] selftests/bpf: Skip ENOTSUPP in ASSERT_OK_PTR

From: Geliang Tang Although the "Segmentation fault" errors are fixed in the last commit, run bpf_tcp_ca selftests (./test_progs -t bpf_tcp_ca) on a Loongarch platform, still some other "ENOTSUPP" (-524) errors left since lacking BPF trampoline on Loongarch: ''' test_dctcp:PASS:bpf_dctcp__open_

[PATCH bpf-next v2 4/6] selftests/bpf: Null checks for link in bpf_tcp_ca

From: Geliang Tang Run bpf_tcp_ca selftests (./test_progs -t bpf_tcp_ca) on a Loongarch platform, some "Segmentation fault" errors occur: ''' test_dctcp:PASS:bpf_dctcp__open_and_load 0 nsec test_dctcp:FAIL:bpf_map__attach_struct_ops unexpected error: -524 #29/1bpf_tcp_ca/dctcp:FAIL test_

[PATCH bpf-next v2 3/6] selftests/bpf: Use ASSERT_OK to skip ENOTSUPP

From: Geliang Tang There are so many "ENOTSUPP" (-524) errors when running BPF selftests on a Loongarch platform since lacking BPF trampoline on Loongarch: ''' test_d_path_basic:PASS:setup 0 nsec libbpf: prog 'prog_stat': failed to attach: unknown error (-524) libbpf: prog 'prog_stat': failed

[PATCH bpf-next v2 2/6] selftests/bpf: Skip ENOTSUPP in ASSERT_OK

From: Geliang Tang Just like handling ENOTSUPP in test_lsm_cgroup_functional(), this patch adds a new helper test_progs_get_error() to check whether the input error is ENOTSUPP (524) or ENOTSUP (95). If it is, invoke test__skip() to skip the test instead of using test__fail(). Use this helper in

[PATCH bpf-next v2 1/6] selftests/bpf: Define ENOTSUPP in testing_helpers.h

From: Geliang Tang ENOTSUPP are defined in so many places in bpf selftests, this patch moves this definition into testing_helpers.h, which is almost included by each tests. And drop all other duplicate definitions. Signed-off-by: Geliang Tang --- tools/testing/selftests/bpf/prog_tests/bpf_tcp_

[PATCH bpf-next v2 0/6] skip ENOTSUPP BPF selftests

From: Geliang Tang v2: - Although all CI tests passed on x86_64 "bpf/vmtest-bpf-next-VM_Test-22 Logs for x86_64-gcc / test (test_progs, false, 360) / test_progs on x86_64 with gcc", some unexpect "SKIP"s are showed in the log: #29/1bpf_tcp_ca/dctcp:SKIP #29/2bpf_tcp_ca/cubic:OK #2

Re: [PATCH bpf-next v7 8/9] selftests/bpf: Use connect_to_addr in sk_lookup

On Thu, 2024-07-04 at 16:59 +0800, Geliang Tang wrote: > From: Geliang Tang > > Use public network helpers make_sockaddr() and connect_to_addr() > instead > of using the local defined function make_socket() and connect(). > > This make_socket() can be dropped latter. > > Signed-off-by: Geliang

Re: [GIT PULL] Kselftest fixes for v6.10

The pull request you sent on Thu, 4 Jul 2024 14:38:16 +0200: > https://git.kernel.org/pub/scm/linux/kernel/git/mic/linux.git > tags/kselftest-fix-2024-07-04 has been merged into torvalds/linux.git: https://git.kernel.org/torvalds/c/4d85acef10252c59e3b6c197c61d9252ff950431 Thank you! -- Deet-

Re: [RFC PATCH] selftests: introduce and use SELFTESTS_CC_IS_CLANG instead of LLVM

On Wed, Jul 03, 2024 at 08:04:52PM -0700, John Hubbard wrote: > Current practice in the selftests Makefiles is to use $(LLVM) as a way > to decide if clang is being used as the compiler (and/or the linker > front end). Unfortunately, this does not cover all of the use cases: > > 1) CC could have b

Re: [PATCH bpf-next] arm64, bpf: Add 12-argument support for bpf trampoline

On Thu, Jul 4, 2024 at 7:33 PM Puranjay Mohan wrote: > > The arm64 bpf JIT currently supports attaching the trampoline to > functions with <= 8 arguments. This is because up to 8 arguments can be > passed in registers r0-r7. If there are more than 8 arguments then the > 9th and later arguments are

Re: [PATCH net-next v15 03/14] netdev: support binding dma-buf to netdevice

On Fri, Jun 28, 2024 at 9:43 AM Mina Almasry wrote: > Hi Mina, > Add a netdev_dmabuf_binding struct which represents the > dma-buf-to-netdevice binding. The netlink API will bind the dma-buf to > rx queues on the netdevice. On the binding, the dma_buf_attach > & dma_buf_map_attachment will occur

[PATCH bpf-next] arm64, bpf: Add 12-argument support for bpf trampoline

The arm64 bpf JIT currently supports attaching the trampoline to functions with <= 8 arguments. This is because up to 8 arguments can be passed in registers r0-r7. If there are more than 8 arguments then the 9th and later arguments are passed on the stack, with SP pointing to the first stacked argu

Re: [PATCH net-next v15 03/14] netdev: support binding dma-buf to netdevice

On Fri, Jun 28, 2024 at 9:43 AM Mina Almasry wrote: > Hi Mina, Thanks a lot for this work! > Add a netdev_dmabuf_binding struct which represents the > dma-buf-to-netdevice binding. The netlink API will bind the dma-buf to > rx queues on the netdevice. On the binding, the dma_buf_attach > & dma_b

Re: [PATCH] riscv: selftests: Fix vsetivli args for clang

Hello: This patch was applied to riscv/linux.git (fixes) by Palmer Dabbelt : On Tue, 02 Jul 2024 18:54:48 -0700 you wrote: > Clang does not support implicit LMUL in the vset* instruction sequences. > Introduce an explicit LMUL in the vsetivli instruction. > > Signed-off-by: Charlie Jenkins > Fi

[GIT PULL] Kselftest fixes for v6.10

Hi Linus, This PR fixes a few kselftests [1]. This has been in linux-next for a week and rebased to add Mark Brown's Tested-by. The race condition found while writing this fix is not new and seems specific to UML's hostfs (I also tested against ext4 and btrfs without being able to trigger this i

Re: [PATCH RFC 0/3] add support for mm-local memory allocations

On Fri, 2024-06-21 at 22:14 +0200, Roman Kagan wrote: > > Compared to the approach used in the orignal series, where a dedicated kernel > address range and thus a dedicated PGD was used for mm-local allocations, the > one proposed here may have certain drawbacks, in particular > > - using user ad

Re: [PATCH bpf-next v3 2/3] selftests/bpf: Add mptcp pm_nl_ctl link

Hello, @BPF people: this new tool doesn't compile on the BPF CI [1]. Can I have a hand to find the best way to fix this please? (see below) [1] https://github.com/kernel-patches/bpf/pull/7296 On 03/07/2024 20:57, Matthieu Baerts (NGI0) wrote: > From: Geliang Tang > > This patch adds a symlink

Re: [RFC PATCH] selftests: introduce and use SELFTESTS_CC_IS_CLANG instead of LLVM

On 04/07/2024 04:04, John Hubbard wrote: > Current practice in the selftests Makefiles is to use $(LLVM) as a way > to decide if clang is being used as the compiler (and/or the linker > front end). Unfortunately, this does not cover all of the use cases: > > 1) CC could have been set within selfte

[PATCH bpf-next v7 9/9] selftests/bpf: Drop make_socket in sk_lookup

From: Geliang Tang Use local helper make_client() instead of using make_socket() and connect(). Then make_socket() and inetaddr_len() can be dropped now. Signed-off-by: Geliang Tang --- .../selftests/bpf/prog_tests/sk_lookup.c | 45 +-- 1 file changed, 2 insertions(+), 43

[PATCH bpf-next v7 8/9] selftests/bpf: Use connect_to_addr in sk_lookup

From: Geliang Tang Use public network helpers make_sockaddr() and connect_to_addr() instead of using the local defined function make_socket() and connect(). This make_socket() can be dropped latter. Signed-off-by: Geliang Tang --- .../selftests/bpf/prog_tests/sk_lookup.c | 20 +--

[PATCH bpf-next v7 7/9] selftests/bpf: Use connect_to_fd_opts in sk_lookup

From: Geliang Tang This patch uses public helper connect_to_fd_opts() exported in network_helpers.h instead of using make_socket() and connect() in prog_tests/sk_lookup.c. This can simplify the code. Signed-off-by: Geliang Tang --- .../selftests/bpf/prog_tests/sk_lookup.c | 25 +++

[PATCH bpf-next v7 6/9] selftests/bpf: Use start_server_str in sk_lookup

From: Geliang Tang This patch uses public helper start_server_addr() instead of local defined function make_server() in prog_tests/sk_lookup.c to avoid duplicate code. Add a helper setsockopts() to set SOL_CUSTOM sockopt looply, set it to setsockopt pointer of struct network_helper_opts, and pas

[PATCH bpf-next v7 5/9] selftests/bpf: Close fd in error path in drop_on_reuseport

From: Geliang Tang Server 1 fd should be closed in the error path when update_lookup_map() fails. This patch fixes it by goto "close_srv1" instead of "detach" lable in that case. Acked-by: Eduard Zingerman Signed-off-by: Geliang Tang --- tools/testing/selftests/bpf/prog_tests/sk_lookup.c | 2

[PATCH bpf-next v7 4/9] selftests/bpf: Use make_sockaddr in sockmap_ktls

From: Geliang Tang This patch uses public helper make_sockaddr() exported in network_helpers.h instead of open-coding in sockmap_ktls.c. This can avoid duplicate code. Acked-by: Eduard Zingerman Signed-off-by: Geliang Tang --- .../selftests/bpf/prog_tests/sockmap_ktls.c | 16 ++--

[PATCH bpf-next v7 3/9] selftests/bpf: Use connect_to_fd_opts in sockmap_ktls

From: Geliang Tang Use public network helper connect_to_fd_opts() instead of open-coding it in prog_tests/sockmap_ktls.c. This can avoid duplicate code. Acked-by: Eduard Zingerman Signed-off-by: Geliang Tang --- .../selftests/bpf/prog_tests/sockmap_ktls.c| 14 ++ 1 file ch

[PATCH bpf-next v7 2/9] selftests/bpf: Use start_server_str in sockmap_ktls

From: Geliang Tang Include network_helpers.h in prog_tests/sockmap_ktls.c, use public network helper start_server_str() instead of local defined function tcp_server(). This can avoid duplicate code. Technically, this is not a one-for-one replacement, as start_server_str() also does bind(). But t

[PATCH bpf-next v7 1/9] selftests/bpf: Add backlog for network_helper_opts

From: Geliang Tang Some callers expect __start_server() helper to pass their own "backlog" value to listen() instead of the default of 1. So this patch adds struct member "backlog" for network_helper_opts to allow callers to set "backlog" value via start_server_str() helper. listen(fd, 0 /* back

[PATCH bpf-next v7 0/9] use network helpers, part 8

From: Geliang Tang v7: - address Martin's comments in v6. (thanks) - use MAX(opts->backlog, 0) instead of opts->backlog. - use connect_to_fd_opts instead connect_to_fd. - more ASSERT_* to check errors. v6: - update patch 6 as Daniel suggested. (thanks) v5: - keep make_server and make_clie

[PATCH net-next v9 10/10] selftests: openvswitch: add psample test

Add a test to verify sampling packets via psample works. In order to do that, create a subcommand in ovs-dpctl.py to listen to on the psample multicast group and print samples. Reviewed-by: Aaron Conole Tested-by: Ilya Maximets Signed-off-by: Adrian Moreno --- .../selftests/net/openvswitch/op

[PATCH net-next v9 09/10] selftests: openvswitch: parse trunc action

The trunc action was supported decode-able but not parse-able. Add support for parsing the action string. Reviewed-by: Aaron Conole Signed-off-by: Adrian Moreno --- .../testing/selftests/net/openvswitch/ovs-dpctl.py | 13 + 1 file changed, 13 insertions(+) diff --git a/tools/testi

[PATCH net-next v9 08/10] selftests: openvswitch: add userspace parsing

The userspace action lacks parsing support plus it contains a bug in the name of one of its attributes. This patch makes userspace action work. Reviewed-by: Aaron Conole Signed-off-by: Adrian Moreno --- .../selftests/net/openvswitch/ovs-dpctl.py| 24 +-- 1 file changed, 22

[PATCH net-next v9 07/10] selftests: openvswitch: add psample action

Add sample and psample action support to ovs-dpctl.py. Refactor common attribute parsing logic into an external function. Reviewed-by: Aaron Conole Signed-off-by: Adrian Moreno --- .../selftests/net/openvswitch/ovs-dpctl.py| 162 +- 1 file changed, 161 insertions(+), 1 dele

[PATCH net-next v9 00/10] net: openvswitch: Add sample multicasting.

** Background ** Currently, OVS supports several packet sampling mechanisms (sFlow, per-bridge IPFIX, per-flow IPFIX). These end up being translated into a userspace action that needs to be handled by ovs-vswitchd's handler threads only to be forwarded to some third party application that will some

Re: [PATCH net-next v15 14/14] selftests: add ncdevmem, netcat for devmem TCP

On Fri, Jun 28, 2024 at 9:38 AM Mina Almasry wrote: > Hi Mina, Thank you so much for this work! > ncdevmem is a devmem TCP netcat. It works similarly to netcat, but it > sends and receives data using the devmem TCP APIs. It uses udmabuf as > the dmabuf provider. It is compatible with a regular n

Re: [PATCH net-next v8 00/10] net: openvswitch: Add sample multicasting.

On Wed, Jul 03, 2024 at 07:35:19PM GMT, Jakub Kicinski wrote: > On Tue, 2 Jul 2024 11:53:17 +0200 Adrian Moreno wrote: > > selftests: openvswitch: add psample action > > selftests: openvswitch: add userspace parsing > > selftests: openvswitch: parse trunc action > > selftests: openvswitch:

[PATCH v4 7/7] selftests/x86: fix printk warnings reported by clang

These warnings are all of the form, "the format specified a short (signed or unsigned) int, but the value is a full length int". Acked-by: Muhammad Usama Anjum Signed-off-by: John Hubbard --- tools/testing/selftests/x86/sigreturn.c | 2 +- tools/testing/selftests/x86/test_vsyscall.c | 10 +

[PATCH v4 6/7] selftests/x86: remove (or use) unused variables and functions

When building with clang, via: make LLVM=1 -C tools/testing/selftests ...quite a few functions are variables are generating "unused" warnings. Fix the warnings by deleting the unused items. One item, the "nerrs" variable in vsdo_restorer.c's main(), is unused but probably wants to be returne

[PATCH v4 5/7] selftests/x86: avoid -no-pie warnings from clang during compilation

When building with clang, via: make LLVM=1 -C tools/testing/selftests ...clang warns that -no-pie is "unused during compilation". This occurs because clang only wants to see -no-pie during linking. Here, we don't have a separate linking stage, so a compiler warning is unavoidable without (wa

[PATCH v4 1/7] selftests/x86: fix Makefile dependencies to work with clang

When building with clang, via: make LLVM=1 -C tools/testing/selftests ...the following build failure occurs in selftests/x86: clang: error: cannot specify -o when generating multiple output files This happens because, although gcc doesn't complain if you invoke it like this: gcc fil

[PATCH v4 4/7] selftests/x86: build sysret_rip.c with clang

When building with clang, via: make LLVM=1 -C tools/testing/selftests ...the build fails because clang's inline asm doesn't support all of the features that are used in the asm() snippet in sysret_rip.c. Fix this by moving the asm code into the clang_helpers_64.S file, where it can be built

[PATCH v4 3/7] selftests/x86: build fsgsbase_restore.c with clang

When building with clang, via: make LLVM=1 -C tools/testing/selftests Fix this by moving the inline asm to "pure" assembly, in two new files: clang_helpers_32.S, clang_helpers_64.S. As a bonus, the pure asm avoids the need for ifdefs, and is now very simple and easy on the eyes. Acked-by: M

[PATCH v4 2/7] selftests: x86: test_FISTTP: use fisttps instead of ambiguous fisttp

From: Muhammad Usama Anjum Use fisttps instead of fisttp to specify correctly that the output variable is of size short. test_FISTTP.c:28:3: error: ambiguous instructions require an explicit suffix (could be 'fisttps', or 'fisttpl') 28 | " fisttp res16""\n" | ^

[PATCH v4 0/7] selftests/x86: fix build errors and warnings found via clang

Changes since v3: 1) Rebased onto Linux 6.10-rc6+. 2) Added Muhammad's acks for the series. Cover letter for v3: Hi, Dave Hansen, Muhammad Usama Anjum, here is the combined series that we discussed yesterday [1]. As I mentioned then, this is a bit intrusive--but no more than necessary, IMHO.

Re: [PATCH v3 2/2] selftests/resctrl: Adjust SNC support messages

Hi Reinette, On 2024-07-03 at 13:51:03 -0700, Reinette Chatre wrote: >Hi Maciej, > >On 7/3/24 12:43 AM, Maciej Wieczór-Retman wrote: >> On 3.07.2024 00:21, Reinette Chatre wrote: >> > On 7/1/24 7:18 AM, Maciej Wieczor-Retman wrote: > >> > > diff --git a/tools/testing/selftests/resctrl/cache.c >>