November 2024 - Linux-kselftest-mirror

[PATCH 0/6] kselftest/arm64: Test floating point signal context restore in fp-stress

by Mark Brown

Currently we test signal delivery to the programs run by fp-stress but our signal handlers simply count the number of signals seen and don't do anything with the floating point state. The original fpsimd-test and sve-test programs had signal handlers called irritators which modify the live register state, verifying that we restore the signal context on return, but a combination of misleading comments and code resulted in them never being used and the equivalent handlers in the other tests being stubbed or omitted. Clarify the code, implement effective irritator handlers for the test programs that can have them and then switch the signals generated by the fp-stress program over to use the irritators, ensuring that we validate that we restore the saved signal context properly. Signed-off-by: Mark Brown <broonie(a)kernel.org> --- Mark Brown (6): kselftest/arm64: Correct misleading comments on fp-stress irritators kselftest/arm64: Remove unused ADRs from irritator handlers kselftest/arm64: Corrupt P15 in the irritator when testing SSVE kselftest/arm64: Implement irritators for ZA and ZT kselftest/arm64: Provide a SIGUSR1 handler in the kernel mode FP stress test kselftest/arm64: Test signal handler state modification in fp-stress tools/testing/selftests/arm64/fp/fp-stress.c | 2 +- tools/testing/selftests/arm64/fp/fpsimd-test.S | 4 +--- tools/testing/selftests/arm64/fp/kernel-test.c | 4 ++++ tools/testing/selftests/arm64/fp/sve-test.S | 6 ++---- tools/testing/selftests/arm64/fp/za-test.S | 13 ++++--------- tools/testing/selftests/arm64/fp/zt-test.S | 13 ++++--------- 6 files changed, 16 insertions(+), 26 deletions(-) --- base-commit: 8e929cb546ee42c9a61d24fae60605e9e3192354 change-id: 20241023-arm64-fp-stress-irritator-5415fe92adef Best regards, -- Mark Brown <broonie(a)kernel.org>

1 month, 2 weeks

2
15
0 0

Re: [PATCH v2 2/9] initramfs_test: kunit tests for initramfs unpacking

by David Disseldorp

[adding kunit lists to cc, see lore link below for context] https://lore.kernel.org/linux-fsdevel/20241104141750.16119-3-ddiss@suse.de/ On Wed, 6 Nov 2024 07:04:21 +0800, kernel test robot wrote: ... > All warnings (new ones prefixed by >>, old ones prefixed by <<): > > >> WARNING: modpost: vmlinux: section mismatch in reference: initramfs_test_cases+0x0 (section: .data) -> initramfs_test_extract (section: .init.text) > >> WARNING: modpost: vmlinux: section mismatch in reference: initramfs_test_cases+0x30 (section: .data) -> initramfs_test_fname_overrun (section: .init.text) > >> WARNING: modpost: vmlinux: section mismatch in reference: initramfs_test_cases+0x60 (section: .data) -> initramfs_test_data (section: .init.text) > >> WARNING: modpost: vmlinux: section mismatch in reference: initramfs_test_cases+0x90 (section: .data) -> initramfs_test_csum (section: .init.text) > >> WARNING: modpost: vmlinux: section mismatch in reference: initramfs_test_cases+0xc0 (section: .data) -> initramfs_test_hardlink (section: .init.text) > >> WARNING: modpost: vmlinux: section mismatch in reference: initramfs_test_cases+0xf0 (section: .data) -> initramfs_test_many (section: .init.text) If I understand correctly, the kunit_case initramfs_test_cases[] members can't be flagged __initdata, as they need to be present post-init for results queries via debugfs. The remaining -Wimplicit-function-declaration reports are all valid. I'll fix them in v3. Thanks, David

1 month, 2 weeks

1
0
0 0

[PATCH] selftests/ftrace: update kprobe syntax error test for ppc64le

by Hari Bathini

For ppc64le, depending on the kernel configuration used, offset 16 from function start address can also be considered function entry. Update the test case to accommodate such configurations. Signed-off-by: Hari Bathini <hbathini(a)linux.ibm.com> --- .../selftests/ftrace/test.d/kprobe/kprobe_syntax_errors.tc | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/tools/testing/selftests/ftrace/test.d/kprobe/kprobe_syntax_errors.tc b/tools/testing/selftests/ftrace/test.d/kprobe/kprobe_syntax_errors.tc index a16c6a6f6055..c03b94cc5784 100644 --- a/tools/testing/selftests/ftrace/test.d/kprobe/kprobe_syntax_errors.tc +++ b/tools/testing/selftests/ftrace/test.d/kprobe/kprobe_syntax_errors.tc @@ -111,7 +111,11 @@ check_error 'p vfs_read $arg* ^$arg*' # DOUBLE_ARGS if !grep -q 'kernel return probes support:' README; then check_error 'r vfs_read ^$arg*' # NOFENTRY_ARGS fi +if [ "$(uname -m)" = "ppc64le" ]; then +check_error 'p vfs_read+20 ^$arg*' # NOFENTRY_ARGS +else check_error 'p vfs_read+8 ^$arg*' # NOFENTRY_ARGS +fi check_error 'p vfs_read ^hoge' # NO_BTFARG check_error 'p kfree ^$arg10' # NO_BTFARG (exceed the number of parameters) check_error 'r kfree ^$retval' # NO_RETVAL -- 2.47.0

1 month, 2 weeks

4
14
0 0

[PATCH net-next] selftests: net: include lib/sh/*.sh with lib.sh

by Matthieu Baerts (NGI0)

Recently, the net/lib.sh file has been modified to include defer.sh from net/lib/sh/ directory. The Makefile from net/lib has been modified accordingly, but not the ones from the sub-targets using net/lib.sh. Because of that, the new file is not installed as expected when installing the Forwarding, MPTCP, and Netfilter targets, e.g. # make -C tools/testing/selftests TARGETS=net/mptcp install \ INSTALL_PATH=/tmp/kself # cd /tmp/kself/ # ./run_kselftest.sh -c net/mptcp TAP version 13 1..7 # timeout set to 1800 # selftests: net/mptcp: mptcp_connect.sh # ./../lib.sh: line 5: /tmp/kself/net/lib/sh/defer.sh: No such file or directory # (...) This can be fixed simply by adding all the .sh files from net/lib/sh directory to the TEST_INCLUDES variable in the different Makefile's. Fixes: a6e263f125cd ("selftests: net: lib: Introduce deferred commands") Signed-off-by: Matthieu Baerts (NGI0) <matttbe(a)kernel.org> --- tools/testing/selftests/net/forwarding/Makefile | 3 ++- tools/testing/selftests/net/mptcp/Makefile | 2 +- tools/testing/selftests/net/netfilter/Makefile | 3 ++- 3 files changed, 5 insertions(+), 3 deletions(-) diff --git a/tools/testing/selftests/net/forwarding/Makefile b/tools/testing/selftests/net/forwarding/Makefile index 224346426ef2220470bf2dbef66198eae9331f56..7d885cff8d79bc6882d2341d0a1a59891fc1cb2e 100644 --- a/tools/testing/selftests/net/forwarding/Makefile +++ b/tools/testing/selftests/net/forwarding/Makefile @@ -126,6 +126,7 @@ TEST_FILES := devlink_lib.sh \ tc_common.sh TEST_INCLUDES := \ - ../lib.sh + ../lib.sh \ + $(wildcard ../lib/sh/*.sh) include ../../lib.mk diff --git a/tools/testing/selftests/net/mptcp/Makefile b/tools/testing/selftests/net/mptcp/Makefile index 5d796622e73099d0a0331c1bc8a41f09e1d36b4d..8e3fc05a539797c5e0feab72be69db623ef3fa96 100644 --- a/tools/testing/selftests/net/mptcp/Makefile +++ b/tools/testing/selftests/net/mptcp/Makefile @@ -11,7 +11,7 @@ TEST_GEN_FILES = mptcp_connect pm_nl_ctl mptcp_sockopt mptcp_inq TEST_FILES := mptcp_lib.sh settings -TEST_INCLUDES := ../lib.sh ../net_helper.sh +TEST_INCLUDES := ../lib.sh $(wildcard ../lib/sh/*.sh) ../net_helper.sh EXTRA_CLEAN := *.pcap diff --git a/tools/testing/selftests/net/netfilter/Makefile b/tools/testing/selftests/net/netfilter/Makefile index 542f7886a0bc2ac016f41d8b70357a8e0c1d271b..9d009f74cfc2da66428b1e23d5884e2c3bc4a85d 100644 --- a/tools/testing/selftests/net/netfilter/Makefile +++ b/tools/testing/selftests/net/netfilter/Makefile @@ -55,4 +55,5 @@ TEST_FILES := lib.sh TEST_FILES += packetdrill TEST_INCLUDES := \ - ../lib.sh + ../lib.sh \ + $(wildcard ../lib/sh/*.sh) --- base-commit: ecf99864ea6b1843773589a935bb026951bf12dd change-id: 20241104-net-next-selftests-lib-sh-deps-cc359ca5602f Best regards, -- Matthieu Baerts (NGI0) <matttbe(a)kernel.org>

1 month, 2 weeks

3
2
0 0

[PATCH bpf-next/net v3] selftests/bpf: Drop netns helpers in mptcp

by Geliang Tang

From: Geliang Tang <tanggeliang(a)kylinos.cn> New netns selftest helpers netns_new() and netns_free() has been added in network_helpers.c, let's use them in mptcp selftests too instead of using MPTCP's own helpers create_netns() and cleanup_netns(). Signed-off-by: Geliang Tang <tanggeliang(a)kylinos.cn> Reviewed-by: Matthieu Baerts (NGI0) <matttbe(a)kernel.org> --- v3: - No code has changed. - Change subject prefix as "bpf-next/net" since CI complains that this patch cannot be merged. - Add Matt's Reviewed-by tag. v2: - Use netns_new/netns_free instead of create_netns/cleanup_netns as Martin suggested. v1: selftests/bpf: Use make/remove netns helpers in mptcp --- .../testing/selftests/bpf/prog_tests/mptcp.c | 42 ++++++------------- 1 file changed, 12 insertions(+), 30 deletions(-) diff --git a/tools/testing/selftests/bpf/prog_tests/mptcp.c b/tools/testing/selftests/bpf/prog_tests/mptcp.c index be3cad2aff77..f8eb7f9d4fd2 100644 --- a/tools/testing/selftests/bpf/prog_tests/mptcp.c +++ b/tools/testing/selftests/bpf/prog_tests/mptcp.c @@ -69,24 +69,6 @@ struct mptcp_storage { char ca_name[TCP_CA_NAME_MAX]; }; -static struct nstoken *create_netns(void) -{ - SYS(fail, "ip netns add %s", NS_TEST); - SYS(fail, "ip -net %s link set dev lo up", NS_TEST); - - return open_netns(NS_TEST); -fail: - return NULL; -} - -static void cleanup_netns(struct nstoken *nstoken) -{ - if (nstoken) - close_netns(nstoken); - - SYS_NOFAIL("ip netns del %s", NS_TEST); -} - static int start_mptcp_server(int family, const char *addr_str, __u16 port, int timeout_ms) { @@ -206,15 +188,15 @@ static int run_test(int cgroup_fd, int server_fd, bool is_mptcp) static void test_base(void) { - struct nstoken *nstoken = NULL; + struct netns_obj *netns = NULL; int server_fd, cgroup_fd; cgroup_fd = test__join_cgroup("/mptcp"); if (!ASSERT_GE(cgroup_fd, 0, "test__join_cgroup")) return; - nstoken = create_netns(); - if (!ASSERT_OK_PTR(nstoken, "create_netns")) + netns = netns_new(NS_TEST, true); + if (!ASSERT_OK_PTR(netns, "netns_new")) goto fail; /* without MPTCP */ @@ -237,7 +219,7 @@ static void test_base(void) close(server_fd); fail: - cleanup_netns(nstoken); + netns_free(netns); close(cgroup_fd); } @@ -322,21 +304,21 @@ static int run_mptcpify(int cgroup_fd) static void test_mptcpify(void) { - struct nstoken *nstoken = NULL; + struct netns_obj *netns = NULL; int cgroup_fd; cgroup_fd = test__join_cgroup("/mptcpify"); if (!ASSERT_GE(cgroup_fd, 0, "test__join_cgroup")) return; - nstoken = create_netns(); - if (!ASSERT_OK_PTR(nstoken, "create_netns")) + netns = netns_new(NS_TEST, true); + if (!ASSERT_OK_PTR(netns, "netns_new")) goto fail; ASSERT_OK(run_mptcpify(cgroup_fd), "run_mptcpify"); fail: - cleanup_netns(nstoken); + netns_free(netns); close(cgroup_fd); } @@ -414,7 +396,7 @@ static void run_subflow(void) static void test_subflow(void) { struct mptcp_subflow *skel; - struct nstoken *nstoken; + struct netns_obj *netns; int cgroup_fd; cgroup_fd = test__join_cgroup("/mptcp_subflow"); @@ -437,8 +419,8 @@ static void test_subflow(void) if (!ASSERT_OK_PTR(skel->links._getsockopt_subflow, "attach _getsockopt_subflow")) goto skel_destroy; - nstoken = create_netns(); - if (!ASSERT_OK_PTR(nstoken, "create_netns: mptcp_subflow")) + netns = netns_new(NS_TEST, true); + if (!ASSERT_OK_PTR(netns, "netns_new: mptcp_subflow")) goto skel_destroy; if (endpoint_init("subflow") < 0) @@ -447,7 +429,7 @@ static void test_subflow(void) run_subflow(); close_netns: - cleanup_netns(nstoken); + netns_free(netns); skel_destroy: mptcp_subflow__destroy(skel); close_cgroup: -- 2.45.2

1 month, 2 weeks

2
1
0 0

[PATCH net-next v1 0/7] devmem TCP fixes

by Mina Almasry

A few unrelated devmem TCP fixes bundled in a series for some convenience (if that's ok). Patch 1-2: fix naming and provide page_pool_alloc_netmem for fragged netmem. Patch 3-4: fix issues with dma-buf dma addresses being potentially passed to dma_sync_for_* helpers. Patch 5-6: fix syzbot SO_DEVMEM_DONTNEED issue and add test for this case. Mina Almasry (6): net: page_pool: rename page_pool_alloc_netmem to *_netmems net: page_pool: create page_pool_alloc_netmem page_pool: disable sync for cpu for dmabuf memory provider netmem: add netmem_prefetch net: fix SO_DEVMEM_DONTNEED looping too long ncdevmem: add test for too many token_count Samiullah Khawaja (1): page_pool: Set `dma_sync` to false for devmem memory provider include/net/netmem.h | 7 ++++ include/net/page_pool/helpers.h | 50 ++++++++++++++++++-------- include/net/page_pool/types.h | 2 +- net/core/devmem.c | 9 +++-- net/core/page_pool.c | 11 +++--- net/core/sock.c | 46 ++++++++++++++---------- tools/testing/selftests/net/ncdevmem.c | 11 ++++++ 7 files changed, 93 insertions(+), 43 deletions(-) -- 2.47.0.163.g1226f6d8fa-goog

1 month, 2 weeks

3
12
0 0

[PATCH v5 0/2] selftests/resctrl: SNC kernel support discovery

by Maciej Wieczor-Retman

Changes v5: - Tests are skipped if snc_unreliable was set. - Moved resctrlfs.c changes from patch 2/2 to 1/2. - Removed CAT changes since it's not impacted by SNC in the selftest. - Updated various comments. - Fixed a bunch of minor issues pointed out in the review. Changes v4: - Printing SNC warnings at the start of every test. - Printing SNC warnings at the end of every relevant test. - Remove global snc_mode variable, consolidate snc detection functions into one. - Correct minor mistakes. Changes v3: - Reworked patch 2. - Changed minor things in patch 1 like function name and made corrections to the patch message. Changes v2: - Removed patches 2 and 3 since now this part will be supported by the kernel. Sub-Numa Clustering (SNC) allows splitting CPU cores, caches and memory into multiple NUMA nodes. When enabled, NUMA-aware applications can achieve better performance on bigger server platforms. SNC support in the kernel was merged into x86/cache [1]. With SNC enabled and kernel support in place all the tests will function normally (aside from effective cache size). There might be a problem when SNC is enabled but the system is still using an older kernel version without SNC support. Currently the only message displayed in that situation is a guess that SNC might be enabled and is causing issues. That message also is displayed whenever the test fails on an Intel platform. Add a mechanism to discover kernel support for SNC which will add more meaning and certainty to the error message. Add runtime SNC mode detection and verify how reliable that information is. Series was tested on Ice Lake server platforms with SNC disabled, SNC-2 and SNC-4. The tests were also ran with and without kernel support for SNC. Series applies cleanly on kselftest/next. [1] https://lore.kernel.org/all/20240628215619.76401-1-tony.luck@intel.com/ Previous versions of this series: [v1] https://lore.kernel.org/all/cover.1709721159.git.maciej.wieczor-retman@inte… [v2] https://lore.kernel.org/all/cover.1715769576.git.maciej.wieczor-retman@inte… [v3] https://lore.kernel.org/all/cover.1719842207.git.maciej.wieczor-retman@inte… [v4] https://lore.kernel.org/all/cover.1720774981.git.maciej.wieczor-retman@inte… Maciej Wieczor-Retman (2): selftests/resctrl: Adjust effective L3 cache size with SNC enabled selftests/resctrl: Adjust SNC support messages tools/testing/selftests/resctrl/cmt_test.c | 8 +- tools/testing/selftests/resctrl/mba_test.c | 8 +- tools/testing/selftests/resctrl/mbm_test.c | 10 +- tools/testing/selftests/resctrl/resctrl.h | 7 + .../testing/selftests/resctrl/resctrl_tests.c | 8 +- tools/testing/selftests/resctrl/resctrlfs.c | 132 ++++++++++++++++++ 6 files changed, 166 insertions(+), 7 deletions(-) -- 2.46.2

1 month, 2 weeks

2
5
0 0

[PATCH net-next v6 0/7] Suspend IRQs during application busy periods

by Joe Damato

Greetings: Welcome to v6, see changelog below. This revision includes only documentation changes: patch 7 has been updated with Bagas' suggestions and the htmldoc looks better as a result. In addition, this cover letter has been updated with a full re-run of the test data. We've included a new test case highlighting a case Sridhar asked about in our v3. See below. This series introduces a new mechanism, IRQ suspension, which allows network applications using epoll to mask IRQs during periods of high traffic while also reducing tail latency (compared to existing mechanisms, see below) during periods of low traffic. In doing so, this balances CPU consumption with network processing efficiency. Martin Karsten (CC'd) and I have been collaborating on this series for several months and have appreciated the feedback from the community on our RFC [1]. We've updated the cover letter and kernel documentation in an attempt to more clearly explain how this mechanism works, how applications can use it, and how it compares to existing mechanisms in the kernel. We've added an additional test case, 'fullbusy', achieved by modifying libevent for comparison. See below for a detailed description, link to the patch, and test results. I briefly mentioned this idea at netdev conf 2024 (for those who were there) and Martin described this idea in an earlier paper presented at Sigmetrics 2024 [2]. ~ The short explanation (TL;DR) We propose adding a new napi config parameter: irq_suspend_timeout to help balance CPU usage and network processing efficiency when using IRQ deferral and napi busy poll. If this parameter is set to a non-zero value *and* a user application has enabled preferred busy poll on a busy poll context (via the EPIOCSPARAMS ioctl introduced in commit 18e2bf0edf4d ("eventpoll: Add epoll ioctl for epoll_params")), then application calls to epoll_wait for that context will cause device IRQs and softirq processing to be suspended as long as epoll_wait successfully retrieves data from the NAPI. Each time data is retrieved, the irq_suspend_timeout is deferred. If/when network traffic subsides and epoll_wait returns no data, IRQ suspension is immediately reverted back to the existing napi_defer_hard_irqs and gro_flush_timeout mechanism which was introduced in commit 6f8b12d661d0 ("net: napi: add hard irqs deferral feature")). The irq_suspend_timeout serves as a safety mechanism. If userland takes a long time processing data, irq_suspend_timeout will fire and restart normal NAPI processing. For a more in depth explanation, please continue reading. ~ Comparison with existing mechanisms Interrupt mitigation can be accomplished in napi software, by setting napi_defer_hard_irqs and gro_flush_timeout, or via interrupt coalescing in the NIC. This can be quite efficient, but in both cases, a fixed timeout (or packet count) needs to be configured. However, a fixed timeout cannot effectively support both low- and high-load situations: At low load, an application typically processes a few requests and then waits to receive more input data. In this scenario, a large timeout will cause unnecessary latency. At high load, an application typically processes many requests before being ready to receive more input data. In this case, a small timeout will likely fire prematurely and trigger irq/softirq processing, which interferes with the application's execution. This causes overhead, most likely due to cache contention. While NICs attempt to provide adaptive interrupt coalescing schemes, these cannot properly take into account application-level processing. An alternative packet delivery mechanism is busy-polling, which results in perfect alignment of application processing and network polling. It delivers optimal performance (throughput and latency), but results in 100% cpu utilization and is thus inefficient for below-capacity workloads. We propose to add a new packet delivery mode that properly alternates between busy polling and interrupt-based delivery depending on busy and idle periods of the application. During a busy period, the system operates in busy-polling mode, which avoids interference. During an idle period, the system falls back to interrupt deferral, but with a small timeout to avoid excessive latencies. This delivery mode can also be viewed as an extension of basic interrupt deferral, but alternating between a small and a very large timeout. This delivery mode is efficient, because it avoids softirq execution interfering with application processing during busy periods. It can be used with blocking epoll_wait to conserve cpu cycles during idle periods. The effect of alternating between busy and idle periods is that performance (throughput and latency) is very close to full busy polling, while cpu utilization is lower and very close to interrupt mitigation. ~ Usage details IRQ suspension is introduced via a per-NAPI configuration parameter that controls the maximum time that IRQs can be suspended. Here's how it is intended to work: - The user application (or system administrator) uses the netdev-genl netlink interface to set the pre-existing napi_defer_hard_irqs and gro_flush_timeout NAPI config parameters to enable IRQ deferral. - The user application (or system administrator) sets the proposed irq_suspend_timeout parameter via the netdev-genl netlink interface to a larger value than gro_flush_timeout to enable IRQ suspension. - The user application issues the existing epoll ioctl to set the prefer_busy_poll flag on the epoll context. - The user application then calls epoll_wait to busy poll for network events, as it normally would. - If epoll_wait returns events to userland, IRQs are suspended for the duration of irq_suspend_timeout. - If epoll_wait finds no events and the thread is about to go to sleep, IRQ handling using napi_defer_hard_irqs and gro_flush_timeout is resumed. As long as epoll_wait is retrieving events, IRQs (and softirq processing) for the NAPI being polled remain disabled. When network traffic reduces, eventually a busy poll loop in the kernel will retrieve no data. When this occurs, regular IRQ deferral using gro_flush_timeout for the polled NAPI is re-enabled. Unless IRQ suspension is continued by subsequent calls to epoll_wait, it automatically times out after the irq_suspend_timeout timer expires. Regular deferral is also immediately re-enabled when the epoll context is destroyed. ~ Usage scenario The target scenario for IRQ suspension as packet delivery mode is a system that runs a dominant application with substantial network I/O. The target application can be configured to receive input data up to a certain batch size (via epoll_wait maxevents parameter) and this batch size determines the worst-case latency that application requests might experience. Because packet delivery is suspended during the target application's processing, the batch size also determines the worst-case latency of concurrent applications using the same RX queue(s). gro_flush_timeout should be set as small as possible, but large enough to make sure that a single request is likely not being interfered with. irq_suspend_timeout is largely a safety mechanism against misbehaving applications. It should be set large enough to cover the processing of an entire application batch, i.e., the factor between gro_flush_timeout and irq_suspend_timeout should roughly correspond to the maximum batch size that the target application would process in one go. ~ Design rationale The implementation of the IRQ suspension mechanism very nicely dovetails with the existing mechanism for IRQ deferral when preferred busy poll is enabled (introduced in commit 7fd3253a7de6 ("net: Introduce preferred busy-polling"), see that commit message for more details). While it would be possible to inject the suspend timeout via the existing epoll ioctl, it is more natural to avoid this path for one main reason: An epoll context is linked to NAPI IDs as file descriptors are added; this means any epoll context might suddenly be associated with a different net_device if the application were to replace all existing fds with fds from a different device. In this case, the scope of the suspend timeout becomes unclear and many edge cases for both the user application and the kernel are introduced Only a single iteration through napi busy polling is needed for this mechanism to work effectively. Since an important objective for this mechanism is preserving cpu cycles, exactly one iteration of the napi busy loop is invoked when busy_poll_usecs is set to 0. ~ Important call out in the implementation - Enabling per epoll-context preferred busy poll will now effectively lead to a nonblocking iteration through napi_busy_loop, even when busy_poll_usecs is 0. See patch 4. ~ Benchmark configs & descriptions The changes were benchmarked with memcached [3] using the benchmarking tool mutilate [4]. To facilitate benchmarking, a small patch [5] was applied to memcached 1.6.29 to allow setting per-epoll context preferred busy poll and other settings via environment variables. Another small patch [6] was applied to libevent to enable full busy-polling. Multiple scenarios were benchmarked as described below and the scripts used for producing these results can be found on github [7] (note: all scenarios use NAPI-based traffic splitting via SO_INCOMING_ID by passing -N to memcached): - base: - no other options enabled - deferX: - set defer_hard_irqs to 100 - set gro_flush_timeout to X,000 - napibusy: - set defer_hard_irqs to 100 - set gro_flush_timeout to 200,000 - enable busy poll via the existing ioctl (busy_poll_usecs = 64, busy_poll_budget = 64, prefer_busy_poll = true) - fullbusy: - set defer_hard_irqs to 100 - set gro_flush_timeout to 5,000,000 - enable busy poll via the existing ioctl (busy_poll_usecs = 1000, busy_poll_budget = 64, prefer_busy_poll = true) - change memcached's nonblocking epoll_wait invocation (via libevent) to using a 1 ms timeout - suspend0: - set defer_hard_irqs to 0 - set gro_flush_timeout to 0 - set irq_suspend_timeout to 20,000,000 - enable busy poll via the existing ioctl (busy_poll_usecs = 0, busy_poll_budget = 64, prefer_busy_poll = true) - suspendX: - set defer_hard_irqs to 100 - set gro_flush_timeout to X,000 - set irq_suspend_timeout to 20,000,000 - enable busy poll via the existing ioctl (busy_poll_usecs = 0, busy_poll_budget = 64, prefer_busy_poll = true) ~ Benchmark results Tested on: Single socket AMD EPYC 7662 64-Core Processor Hyperthreading disabled 4 NUMA Zones (NPS=4) 16 CPUs per NUMA zone (64 cores total) 2 x Dual port 100gbps Mellanox Technologies ConnectX-5 Ex EN NIC The test machine is configured such that a single interface has 8 RX queues. The queues' IRQs and memcached are pinned to CPUs that are NUMA-local to the interface which is under test. The NIC's interrupt coalescing configuration is left at boot-time defaults. Results: Results are shown below. The mechanism added by this series is represented by the 'suspend' cases. Data presented shows a summary over at least 15 runs of each test case [8] using the scripts on github [7]. For latency, the median is shown. For throughput and CPU utilization, the average is shown. The results also include cycles-per-query (cpq) and instruction-per-query (ipq) metrics, following the methodology proposed in [2], to augment the CPU utilization numbers, which could be skewed due to frequency scaling. We find that this does not appear to be the case as CPU utilization and low-level metrics show similar trends. These results were captured using the scripts on github [7] to illustrate how this approach compares with other pre-existing mechanisms. This data is not to be interpreted as scientific data captured in a fully isolated lab setting, but instead as best effort, illustrative information comparing and contrasting tradeoffs. The absolute QPS results are higher than our previous submission, but the relative differences between variants are equivalent. Because the patches have been rebased on 6.12, several factors have likely influenced the overall performance. Most importantly, we had to switch to a new set of basic kernel options, which has likely altered the baseline performance. Because the overall comparison of variants still holds, we have not attempted to recreate the exact set of kernel options from the previous submission. Compare: - Throughput (MAX) and latencies of base vs suspend. - CPU usage of napibusy and fullbusy during lower load (200K, 400K for example) vs suspend. - Latency of the defer variants vs suspend as timeout and load increases. - suspend0, which sets defer_hard_irqs and gro_flush_timeout to 0, has nearly the same performance as the base case (this is FAQ item #1). The overall takeaway is that the suspend variants provide a superior combination of high throughput, low latency, and low cpu utilization compared to all other variants. Each of the suspend variants works very well, but some fine-tuning between latency and cpu utilization is still possible by tuning the small timeout (gro_flush_timeout). Note: we've reorganized the results to make comparison among testcases with the same load easier. testcase load qps avglat 95%lat 99%lat cpu cpq ipq base 200K 199954 112 237 415 26 13040 11336 defer10 200K 200002 54 123 142 28 19033 16508 defer20 200K 199985 60 130 153 26 15737 14247 defer50 200K 199968 78 142 181 23 12113 11609 defer200 200K 199997 163 252 304 18 8449 9155 fullbusy 200K 199993 46 117 132 100 43959 23320 napibusy 200K 200006 100 237 275 56 25016 24866 suspend0 200K 200012 105 249 432 29 14369 11844 suspend10 200K 200004 53 123 141 32 19432 16752 suspend20 200K 200014 58 126 151 30 16356 14670 suspend50 200K 200018 73 134 176 26 13245 12416 suspend200 200K 200027 149 250 302 20 9508 9781 testcase load qps avglat 95%lat 99%lat cpu cpq ipq base 400K 399984 139 268 715 40 9437 9299 defer10 400K 400002 59 133 165 53 14089 12908 defer20 400K 400016 66 140 171 47 12085 11682 defer50 400K 400037 87 161 198 39 9528 9879 defer200 400K 399954 181 273 329 32 7326 8438 fullbusy 400K 399951 50 123 155 100 21990 16097 napibusy 400K 399997 76 221 271 83 18260 16511 suspend0 400K 399991 125 337 768 48 11051 9629 suspend10 400K 399990 57 129 161 54 13629 12841 suspend20 400K 399922 61 135 167 49 12055 11715 suspend50 400K 400024 75 148 186 42 10049 10243 suspend200 400K 399936 154 267 325 34 7770 8677 testcase load qps avglat 95%lat 99%lat cpu cpq ipq base 600K 600064 148 265 576 61 9276 8757 defer10 600K 599985 71 147 204 76 12048 10863 defer20 600K 600024 75 151 199 66 10572 10328 defer50 600K 600054 94 172 217 55 8584 9144 defer200 600K 600030 200 299 355 45 6874 8189 fullbusy 600K 599956 55 127 176 100 14650 13968 napibusy 600K 599956 64 163 252 96 14022 14153 suspend0 600K 600029 126 306 724 70 10393 8977 suspend10 600K 599997 63 137 194 70 10991 11005 suspend20 600K 600012 67 141 194 65 10108 10359 suspend50 600K 600045 80 157 203 57 8747 9320 suspend200 600K 599940 158 277 344 48 7221 8354 testcase load qps avglat 95%lat 99%lat cpu cpq ipq base 800K 800025 179 298 555 86 9572 8297 defer10 800K 799275 224 633 1271 96 10679 8904 defer20 800K 800041 114 226 328 90 10122 8917 defer50 800K 799936 118 207 288 77 8820 8607 defer200 800K 799994 228 341 403 65 7424 8130 fullbusy 800K 799964 62 136 192 100 10992 12518 napibusy 800K 799971 65 142 216 99 10911 12529 suspend0 800K 799965 126 250 533 86 9489 8496 suspend10 800K 799995 69 145 201 83 9475 9764 suspend20 800K 799931 74 151 209 79 8976 9336 suspend50 800K 799946 87 168 224 71 7993 8794 suspend200 800K 799993 160 292 357 62 6967 8184 testcase load qps avglat 95%lat 99%lat cpu cpq ipq base 1000K 915792 3498 5740 6239 97 9388 7930 defer10 1000K 876285 3896 6095 6418 99 9960 8542 defer20 1000K 914909 3107 5771 6283 97 9407 8284 defer50 1000K 928426 2977 5591 5931 97 9214 8012 defer200 1000K 959989 3097 5306 5929 96 8816 7908 fullbusy 1000K 1000102 74 155 213 100 8796 10559 napibusy 1000K 1000006 74 154 216 100 8787 10654 suspend0 1000K 960757 2223 5715 7029 98 8964 7993 suspend10 1000K 999926 80 162 222 92 8246 8922 suspend20 1000K 1000095 85 166 226 89 7966 8719 suspend50 1000K 1000067 96 180 238 84 7476 8419 suspend200 1000K 999968 163 298 363 76 6798 8061 testcase load qps avglat 95%lat 99%lat cpu cpq ipq base MAX 1054805 4152 5298 5743 100 8332 7890 defer10 MAX 937098 4598 6010 6347 100 9378 8407 defer20 MAX 988905 4389 5637 5990 100 8886 8106 defer50 MAX 1067194 3960 5216 5544 100 8235 7911 defer200 MAX 1054967 4084 5496 5821 100 8323 7871 fullbusy MAX 1248006 3472 3918 3979 100 7050 7919 napibusy MAX 1128384 3742 7958 10753 100 7776 7872 suspend0 MAX 1034456 4242 5668 6042 100 8497 7912 suspend10 MAX 1229229 3513 3926 3986 100 7156 7937 suspend20 MAX 1226845 3514 3939 3985 100 7171 7937 suspend50 MAX 1230757 3513 3935 3983 100 7140 7935 suspend200 MAX 1230424 3503 3934 3984 100 7142 7927 ~ FAQ - Why is a new parameter needed? Does irq_suspend_timeout override gro_flush_timeout? Using the suspend mechanism causes the system to alternate between polling mode and irq-driven packet delivery. During busy periods, irq_suspend_timeout overrides gro_flush_timeout and keeps the system busy polling, but when epoll finds no events, the setting of gro_flush_timeout and napi_defer_hard_irqs determine the next step. There are essentially three possible loops for network processing and packet delivery: 1) hardirq -> softirq -> napi poll; basic interrupt delivery 2) timer -> softirq -> napi poll; deferred irq processing 3) epoll -> busy-poll -> napi poll; busy looping Loop 2 can take control from Loop 1, if gro_flush_timeout and napi_defer_hard_irqs are set. If gro_flush_timeout and napi_defer_hard_irqs are set, Loops 2 and 3 "wrestle" with each other for control. During busy periods, irq_suspend_timeout is used as timer in Loop 2, which essentially tilts this in favour of Loop 3. If gro_flush_timeout and napi_defer_hard_irqs are not set, Loop 3 cannot take control from Loop 1. Therefore, setting gro_flush_timeout and napi_defer_hard_irqs is the recommended usage, because otherwise setting irq_suspend_timeout might not have any discernible effect. This is shown in the results above: compare suspend0 with the base case. Note that the lack of napi_defer_hard_irqs and gro_flush_timeout produce similar results for both, which encourages the use of napi_defer_hard_irqs and gro_flush_timeout in addition to irq_suspend_timeout. - Can the new timeout value be threaded through the new epoll ioctl ? Only with difficulty. The epoll ioctl sets options on an epoll context and the NAPI ID associated with an epoll context can change based on what file descriptors a user app adds to the epoll context. This would introduce complexity in the API from the user perspective and also complexity in the kernel. - Can irq suspend be built by combining NIC coalescing and gro_flush_timeout ? No. The problem is that the long timeout must engage if and only if prefer-busy is active. When using NIC coalescing for the short timeout (without napi_defer_hard_irqs/gro_flush_timeout), an interrupt after an idle period will trigger softirq, which will run napi polling. At this point, prefer-busy is not active, so NIC interrupts would be re-enabled. Then it is not possible for the longer timeout to interject to switch control back to polling. In other words, only by using the software timer for the short timeout, it is possible to extend the timeout without having to reprogram the NIC timer or reach down directly and disable interrupts. Using gro_flush_timeout for the long timeout also has problems, for the same underlying reason. In the current napi implementation, gro_flush_timeout is not tied to prefer-busy. We'd either have to change that and in the process modify the existing deferral mechanism, or introduce a state variable to determine whether gro_flush_timeout is used as long timeout for irq suspend or whether it is used for its default purpose. In an earlier version, we did try something similar to the latter and made it work, but it ends up being a lot more convoluted than our current proposal. - Isn't it already possible to combine busy looping with irq deferral? Yes, in fact enabling irq deferral via napi_defer_hard_irqs and gro_flush_timeout is a precondition for prefer_busy_poll to have an effect. If the application also uses a tight busy loop with essentially nonblocking epoll_wait (accomplished with a very short timeout parameter), this is the fullbusy case shown in the results. An application using blocking epoll_wait is shown as the napibusy case in the results. It's a hybrid approach that provides limited latency benefits compared to the base case and plain irq deferral, but not as good as fullbusy or suspend. ~ Special thanks Several people were involved in earlier stages of the development of this mechanism whom we'd like to thank: - Peter Cai (CC'd), for the initial kernel patch and his contributions to the paper. - Mohammadamin Shafie (CC'd), for testing various versions of the kernel patch and providing helpful feedback. Thanks, Martin and Joe [1]: https://lore.kernel.org/netdev/20240812125717.413108-1-jdamato@fastly.com/ [2]: https://doi.org/10.1145/3626780 [3]: https://github.com/memcached/memcached/blob/master/doc/napi_ids.txt [4]: https://github.com/leverich/mutilate [5]: https://raw.githubusercontent.com/martinkarsten/irqsuspend/main/patches/mem… [6]: https://raw.githubusercontent.com/martinkarsten/irqsuspend/main/patches/lib… [7]: https://github.com/martinkarsten/irqsuspend [8]: https://github.com/martinkarsten/irqsuspend/tree/main/results v6: - Updated the cover letter with a full re-run of all test cases, including a new case suspend0, as requested by Sridhar previously. - Updated the kernel documentation in patch 7 as suggested by Bagas Sanjaya, which improved the htmldoc output. v5: https://lore.kernel.org/netdev/20241103052421.518856-1-jdamato@fastly.com/ - Adjusted patch 5 to only suspend IRQs when ep_send_events returns a positive return value. This issue was pointed out by Hillf Danton. - Updated the commit message of patch 6 which still mentioned netcat, despite the code being updated in v4 to replace it with socat and fixed misspelling of netdevsim. - Fixed a minor typo in patch 7 and removed an unnecessary paragraph. - Added Sridhar Samudrala's Reviewed-by to patch 1-5 and 7. v4: https://lore.kernel.org/netdev/20241102005214.32443-1-jdamato@fastly.com/ - Added a new FAQ item to cover letter. - Updated patch 6 to use socat instead of nc in busy_poll_test.sh and updated busy_poller.c to use netlink directly to configure napi params. - Updated the kernel documentation in patch 7 to include more details. - Dropped Stanislav's Acked-by and Bagas' Reviewed-by from patch 7 since the documentation was updated. v3: https://lore.kernel.org/netdev/20241101004846.32532-1-jdamato@fastly.com/ - Added Stanislav Fomichev's Acked-by to every patch except the newly added selftest. - Added Bagas Sanjaya's Reviewed-by to the documentation patch. - Fixed the commit message of patch 2 to remove a reference to the now non-existent sysfs setting. - Added a self test which tests both "regular" busy poll and busy poll with suspend enabled. This was added as patch 6 as requested by Paolo. netdevsim was chosen instead of veth due to netdevsim's pre-existing support for netdev-genl. See the commit message of patch 6 for more details. v2: https://lore.kernel.org/bpf/20241021015311.95468-1-jdamato@fastly.com/ - Cover letter updated, including a re-run of test data. - Patch 1 rewritten to use netdev-genl instead of sysfs. - Patch 3 updated with a comment added to napi_resume_irqs. - Patch 4 rebased to apply now that commit b9ca079dd6b0 ("eventpoll: Annotate data-race of busy_poll_usecs") has been picked up from VFS. - Patch 6 updated the kernel documentation. rfc -> v1: - Cover letter updated to include more details. - Patch 1 updated to remove the documentation added. This was moved to patch 6 with the rest of the docs (see below). - Patch 5 updated to fix an error uncovered by the kernel build robot. See patch 5's changelog for more details. - Patch 6 added which updates kernel documentation. Joe Damato (2): selftests: net: Add busy_poll_test docs: networking: Describe irq suspension Martin Karsten (5): net: Add napi_struct parameter irq_suspend_timeout net: Suspend softirq when prefer_busy_poll is set net: Add control functions for irq suspension eventpoll: Trigger napi_busy_loop, if prefer_busy_poll is set eventpoll: Control irq suspension for prefer_busy_poll Documentation/netlink/specs/netdev.yaml | 7 + Documentation/networking/napi.rst | 170 ++++++++- fs/eventpoll.c | 36 +- include/linux/netdevice.h | 2 + include/net/busy_poll.h | 3 + include/uapi/linux/netdev.h | 1 + net/core/dev.c | 58 +++- net/core/dev.h | 25 ++ net/core/netdev-genl-gen.c | 5 +- net/core/netdev-genl.c | 12 + tools/include/uapi/linux/netdev.h | 1 + tools/testing/selftests/net/.gitignore | 1 + tools/testing/selftests/net/Makefile | 3 +- tools/testing/selftests/net/busy_poll_test.sh | 164 +++++++++ tools/testing/selftests/net/busy_poller.c | 328 ++++++++++++++++++ 15 files changed, 805 insertions(+), 11 deletions(-) create mode 100755 tools/testing/selftests/net/busy_poll_test.sh create mode 100644 tools/testing/selftests/net/busy_poller.c base-commit: dbb9a7ef347828870df3e5e6ddf19469a3277fc9 -- 2.25.1

1 month, 2 weeks

2
3
0 0

[PATCH net-next v7 00/12] selftests: ncdevmem: Add ncdevmem to ksft

by Stanislav Fomichev

The goal of the series is to simplify and make it possible to use ncdevmem in an automated way from the ksft python wrapper. ncdevmem is slowly mutated into a state where it uses stdout to print the payload and the python wrapper is added to make sure the arrived payload matches the expected one. v7: - fix validation (Mina) - add support for working with non ::ffff-prefixed addresses (Mina) v6: - fix compilation issue in 'Unify error handling' patch (Jakub) v5: - properly handle errors from inet_pton() and socket() (Paolo) - remove unneeded import from python selftest (Paolo) v4: - keep usage example with validation (Mina) - fix compilation issue in one patch (s/start_queues/start_queue/) v3: - keep and refine the comment about ncdevmem invocation (Mina) - add the comment about not enforcing exit status for ntuple reset (Mina) - make configure_headersplit more robust (Mina) - use num_queues/2 in selftest and let the users override it (Mina) - remove memory_provider.memcpy_to_device (Mina) - keep ksft as is (don't use -v validate flags): we are gonna need a --debug-disable flag to make it less chatty; otherwise it times out when sending too much data; so leaving it as a separate follow up v2: - don't remove validation (Mina) - keep 5-tuple flow steering but use it only when -c is provided (Mina) - remove separate flag for probing (Mina) - move ncdevmem under drivers/net/hw, not drivers/net (Jakub) Cc: Mina Almasry <almasrymina(a)google.com> Stanislav Fomichev (12): selftests: ncdevmem: Redirect all non-payload output to stderr selftests: ncdevmem: Separate out dmabuf provider selftests: ncdevmem: Unify error handling selftests: ncdevmem: Make client_ip optional selftests: ncdevmem: Remove default arguments selftests: ncdevmem: Switch to AF_INET6 selftests: ncdevmem: Properly reset flow steering selftests: ncdevmem: Use YNL to enable TCP header split selftests: ncdevmem: Remove hard-coded queue numbers selftests: ncdevmem: Run selftest when none of the -s or -c has been provided selftests: ncdevmem: Move ncdevmem under drivers/net/hw selftests: ncdevmem: Add automated test .../selftests/drivers/net/hw/.gitignore | 1 + .../testing/selftests/drivers/net/hw/Makefile | 9 + .../selftests/drivers/net/hw/devmem.py | 45 + .../selftests/drivers/net/hw/ncdevmem.c | 786 ++++++++++++++++++ tools/testing/selftests/net/.gitignore | 1 - tools/testing/selftests/net/Makefile | 8 - tools/testing/selftests/net/ncdevmem.c | 570 ------------- 7 files changed, 841 insertions(+), 579 deletions(-) create mode 100644 tools/testing/selftests/drivers/net/hw/.gitignore create mode 100755 tools/testing/selftests/drivers/net/hw/devmem.py create mode 100644 tools/testing/selftests/drivers/net/hw/ncdevmem.c delete mode 100644 tools/testing/selftests/net/ncdevmem.c -- 2.47.0

1 month, 2 weeks

4
30
0 0

[PATCH v3 0/4] Add support for the Bus Lock Threshold

by Manali Shukla

Misbehaving guests can cause bus locks to degrade the performance of a system. Non-WB (write-back) and misaligned locked RMW (read-modify-write) instructions are referred to as "bus locks" and require system wide synchronization among all processors to guarantee the atomicity. The bus locks can impose notable performance penalties for all processors within the system. Support for the Bus Lock Threshold is indicated by CPUID Fn8000_000A_EDX[29] BusLockThreshold=1, the VMCB provides a Bus Lock Threshold enable bit and an unsigned 16-bit Bus Lock Threshold count. VMCB intercept bit VMCB Offset Bits Function 14h 5 Intercept bus lock operations Bus lock threshold count VMCB Offset Bits Function 120h 15:0 Bus lock counter During VMRUN, the bus lock threshold count is fetched and stored in an internal count register. Prior to executing a bus lock within the guest, the processor verifies the count in the bus lock register. If the count is greater than zero, the processor executes the bus lock, reducing the count. However, if the count is zero, the bus lock operation is not performed, and instead, a Bus Lock Threshold #VMEXIT is triggered to transfer control to the Virtual Machine Monitor (VMM). A Bus Lock Threshold #VMEXIT is reported to the VMM with VMEXIT code 0xA5h, VMEXIT_BUSLOCK. EXITINFO1 and EXITINFO2 are set to 0 on a VMEXIT_BUSLOCK. On a #VMEXIT, the processor writes the current value of the Bus Lock Threshold Counter to the VMCB. More details about the Bus Lock Threshold feature can be found in AMD APM [1]. v2 -> v3 - Drop parch to add virt tag in /proc/cpuinfo. - Incorporated Tom's review comments. v1 -> v2 - Incorporated misc review comments from Sean. - Removed bus_lock_counter module parameter. - Set the value of bus_lock_counter to zero by default and reload the value by 1 in bus lock exit handler. - Add documentation for the behavioral difference for KVM_EXIT_BUS_LOCK. - Improved selftest for buslock to work on SVM and VMX. - Rewrite the commit messages. Patches are prepared on kvm-next/next (efbc6bd090f4). Testing done: - Added a selftest for the Bus Lock Threshold functionality. - The bus lock threshold selftest has been tested on both Intel and AMD platforms. - Tested the Bus Lock Threshold functionality on SEV, SEV-ES, SEV-SNP guests. - Tested the Bus Lock Threshold functionality on nested guests. v1: https://lore.kernel.org/kvm/20240709175145.9986-4-manali.shukla@amd.com/T/ v2: https://lore.kernel.org/kvm/20241001063413.687787-4-manali.shukla@amd.com/T/ [1]: AMD64 Architecture Programmer's Manual Pub. 24593, April 2024, Vol 2, 15.14.5 Bus Lock Threshold. https://bugzilla.kernel.org/attachment.cgi?id=306250 Manali Shukla (2): x86/cpufeatures: Add CPUID feature bit for the Bus Lock Threshold KVM: X86: Add documentation about behavioral difference for KVM_EXIT_BUS_LOCK Nikunj A Dadhania (2): KVM: SVM: Enable Bus lock threshold exit KVM: selftests: Add bus lock exit test Documentation/virt/kvm/api.rst | 4 + arch/x86/include/asm/cpufeatures.h | 1 + arch/x86/include/asm/svm.h | 5 +- arch/x86/include/uapi/asm/svm.h | 2 + arch/x86/kvm/svm/nested.c | 10 ++ arch/x86/kvm/svm/svm.c | 27 ++++ tools/testing/selftests/kvm/Makefile | 1 + .../selftests/kvm/x86_64/kvm_buslock_test.c | 130 ++++++++++++++++++ 8 files changed, 179 insertions(+), 1 deletion(-) create mode 100644 tools/testing/selftests/kvm/x86_64/kvm_buslock_test.c base-commit: efbc6bd090f48ccf64f7a8dd5daea775821d57ec -- 2.34.1

1 month, 2 weeks

4
13
0 0

2024

2023

2022

2021

2020

2019

2018

2017

Linux-kselftest-mirror November 2024