September 2024 - Linux-kselftest-mirror

[PATCH 0/5] riscv: add support for Zaamo and Zalrsc extensions

by Clément Léger

Since commit e87412e621f1 ("integrate Zaamo and Zalrsc text (#1304)"), the A extension has been described as a set of instructions provided by Zaamo and Zalrsc. Add these two extensions. This series is based on the Zc one [1]. Link: https://lore.kernel.org/linux-riscv/20240619113529.676940-1-cleger@rivosinc… --- Clément Léger (5): dt-bindings: riscv: add Zaamo and Zalrsc ISA extension description riscv: add parsing for Zaamo and Zalrsc extensions riscv: hwprobe: export Zaamo and Zalrsc extensions RISC-V: KVM: Allow Zaamo/Zalrsc extensions for Guest/VM KVM: riscv: selftests: Add Zaamo/Zalrsc extensions to get-reg-list test Documentation/arch/riscv/hwprobe.rst | 8 ++++++++ .../devicetree/bindings/riscv/extensions.yaml | 19 +++++++++++++++++++ arch/riscv/include/asm/hwcap.h | 2 ++ arch/riscv/include/uapi/asm/hwprobe.h | 2 ++ arch/riscv/include/uapi/asm/kvm.h | 2 ++ arch/riscv/kernel/cpufeature.c | 9 ++++++++- arch/riscv/kernel/sys_hwprobe.c | 2 ++ arch/riscv/kvm/vcpu_onereg.c | 4 ++++ .../selftests/kvm/riscv/get-reg-list.c | 8 ++++++++ 9 files changed, 55 insertions(+), 1 deletion(-) -- 2.45.2

2 days, 4 hours

4
11
0 0

[PATCH 1/2] selftest/mm: Fix typo in virtual_address_range

by Chunyan Zhang

The function name should be *hint* address, so correct it. Signed-off-by: Chunyan Zhang <zhangchunyan(a)iscas.ac.cn> --- tools/testing/selftests/mm/virtual_address_range.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/testing/selftests/mm/virtual_address_range.c b/tools/testing/selftests/mm/virtual_address_range.c index 4e4c1e311247..2a2b69e91950 100644 --- a/tools/testing/selftests/mm/virtual_address_range.c +++ b/tools/testing/selftests/mm/virtual_address_range.c @@ -64,7 +64,7 @@ #define NR_CHUNKS_HIGH NR_CHUNKS_384TB #endif -static char *hind_addr(void) +static char *hint_addr(void) { int bits = HIGH_ADDR_SHIFT + rand() % (63 - HIGH_ADDR_SHIFT); @@ -185,7 +185,7 @@ int main(int argc, char *argv[]) } for (i = 0; i < NR_CHUNKS_HIGH; i++) { - hint = hind_addr(); + hint = hint_addr(); hptr[i] = mmap(hint, MAP_CHUNK_SIZE, PROT_READ | PROT_WRITE, MAP_PRIVATE | MAP_ANONYMOUS, -1, 0); -- 2.34.1

1 week, 2 days

4
5
0 0

[PATCH 00/15] KVM: x86: Introduce new ioctl KVM_TRANSLATE2

by Nikolas Wipper

This series introduces a new ioctl KVM_TRANSLATE2, which expands on KVM_TRANSLATE. It is required to implement Hyper-V's HvTranslateVirtualAddress hyper-call as part of the ongoing effort to emulate HyperV's Virtual Secure Mode (VSM) within KVM and QEMU. The hyper- call requires several new KVM APIs, one of which is KVM_TRANSLATE2, which implements the core functionality of the hyper-call. The rest of the required functionality will be implemented in subsequent series. Other than translating guest virtual addresses, the ioctl allows the caller to control whether the access and dirty bits are set during the page walk. It also allows specifying an access mode instead of returning viable access modes, which enables setting the bits up to the level that caused a failure. Additionally, the ioctl provides more information about why the page walk failed, and which page table is responsible. This functionality is not available within KVM_TRANSLATE, and can't be added without breaking backwards compatiblity, thus a new ioctl is required. The ioctl was designed to facilitate as many other use cases as possible apart from VSM. The error codes were intentionally chosen to be broad enough to avoid exposing architecture specific details. Even though HvTranslateVirtualAddress only really needs one flag to set the accessed and dirty bits whenever possible, that was split into several flags so that future users can chose more gradually when these bits should be set. Furthermore, as much information as possible is provided to the caller. The patch series includes selftests for the ioctl, as well as fuzzy testing on random garbage guest page table entries. All previously passing KVM selftests and KVM unit tests still pass. Series overview: - 1: Document the new ioctl - 2-11: Update the page walker in preparation - 12-14: Implement the ioctl - 15: Implement testing This series, alongside the series by Nicolas Saenz Julienne [1] introducing the core building blocks for VSM and the accompanying QEMU implementation [2], is capable of booting Windows Server 2019. Both series are also available on GitHub [3]. [1] https://lore.kernel.org/linux-hyperv/20240609154945.55332-1-nsaenz@amazon.c… [2] https://github.com/vianpl/qemu/tree/vsm/next [3] https://github.com/vianpl/linux/tree/vsm/next Best, Nikolas Nikolas Wipper (15): KVM: Add API documentation for KVM_TRANSLATE2 KVM: x86/mmu: Abort page walk if permission checks fail KVM: x86/mmu: Introduce exception flag for unmapped GPAs KVM: x86/mmu: Store GPA in exception if applicable KVM: x86/mmu: Introduce flags parameter to page walker KVM: x86/mmu: Implement PWALK_SET_ACCESSED in page walker KVM: x86/mmu: Implement PWALK_SET_DIRTY in page walker KVM: x86/mmu: Implement PWALK_FORCE_SET_ACCESSED in page walker KVM: x86/mmu: Introduce status parameter to page walker KVM: x86/mmu: Implement PWALK_STATUS_READ_ONLY_PTE_GPA in page walker KVM: x86: Introduce generic gva to gpa translation function KVM: Introduce KVM_TRANSLATE2 KVM: Add KVM_TRANSLATE2 stub KVM: x86: Implement KVM_TRANSLATE2 KVM: selftests: Add test for KVM_TRANSLATE2 Documentation/virt/kvm/api.rst | 131 ++++++++ arch/x86/include/asm/kvm_host.h | 18 +- arch/x86/kvm/hyperv.c | 3 +- arch/x86/kvm/kvm_emulate.h | 8 + arch/x86/kvm/mmu.h | 10 +- arch/x86/kvm/mmu/mmu.c | 7 +- arch/x86/kvm/mmu/paging_tmpl.h | 80 +++-- arch/x86/kvm/x86.c | 123 ++++++- include/linux/kvm_host.h | 6 + include/uapi/linux/kvm.h | 33 ++ tools/testing/selftests/kvm/Makefile | 1 + .../selftests/kvm/x86_64/kvm_translate2.c | 310 ++++++++++++++++++ virt/kvm/kvm_main.c | 41 +++ 13 files changed, 724 insertions(+), 47 deletions(-) create mode 100644 tools/testing/selftests/kvm/x86_64/kvm_translate2.c -- 2.40.1 Amazon Web Services Development Center Germany GmbH Krausenstr. 38 10117 Berlin Geschaeftsfuehrung: Christian Schlaeger, Jonathan Weiss Eingetragen am Amtsgericht Charlottenburg unter HRB 257764 B Sitz: Berlin Ust-ID: DE 365 538 597

1 week, 2 days

2
18
0 0

selftests: core: unshare_test: WARNING: at mm/util.c:671 __kvmalloc_node_noprof

by Naresh Kamboju

The following kernel warning is noticed on all arch and all devices while running selftests: core: unshare_test on Linux next-20240823 and next-20240826. First seen on next-20240823. Good: next-20240822 BAD: next-20240823 and next-20240826 Reported-by: Linux Kernel Functional Testing <lkft(a)linaro.org> Crash log: -------- # selftests: core: unshare_test <4>[ 61.084149] ------------[ cut here ]------------ <4>[ 61.085175] WARNING: CPU: 0 PID: 477 at mm/util.c:671 __kvmalloc_node_noprof (mm/util.c:671 (discriminator 1)) <4>[ 61.088958] Modules linked in: crct10dif_ce sm3_ce sm3 sha3_ce sha512_ce sha512_arm64 drm fuse backlight dm_mod ip_tables x_tables <4>[ 61.093141] CPU: 0 UID: 0 PID: 477 Comm: unshare_test Not tainted 6.11.0-rc5-next-20240826 #1 <4>[ 61.094558] Hardware name: linux,dummy-virt (DT) <4>[ 61.096763] pstate: 23400009 (nzCv daif +PAN -UAO +TCO +DIT -SSBS BTYPE=--) <4>[ 61.097841] pc : __kvmalloc_node_noprof (mm/util.c:671 (discriminator 1)) <4>[ 61.099701] lr : __kvmalloc_node_noprof (mm/util.c:661) <4>[ 61.100448] sp : ffff800080abbce0 <4>[ 61.100819] x29: ffff800080abbcf0 x28: fff0000004549280 x27: 0000000000000000 <4>[ 61.101744] x26: 0000000000000000 x25: 0000000000000000 x24: fff0000003615e40 <4>[ 61.102512] x23: fff0000003615ec0 x22: bfafa45863b285c8 x21: 0000000200002000 <4>[ 61.103232] x20: 00000000ffffffff x19: 0000000000400cc0 x18: 0000000000000000 <4>[ 61.104053] x17: 0000000000000000 x16: 0000000000000000 x15: 0000000000000000 <4>[ 61.104927] x14: 0000000000000000 x13: 0000000000000000 x12: 0000000000000000 <4>[ 61.105752] x11: 0000000000000000 x10: 0000000000000000 x9 : 0000000000000000 <4>[ 61.106606] x8 : 0000000000000001 x7 : 0000000000000001 x6 : 0000000000000005 <4>[ 61.107377] x5 : 0000000000000000 x4 : fff0000004549280 x3 : 0000000000000000 <4>[ 61.108207] x2 : 0000000000000000 x1 : 000000007fffffff x0 : 0000000000000000 <4>[ 61.109262] Call trace: <4>[ 61.109619] __kvmalloc_node_noprof (mm/util.c:671 (discriminator 1)) <4>[ 61.110248] alloc_fdtable (fs/file.c:133) <4>[ 61.110751] expand_files (include/linux/atomic/atomic-arch-fallback.h:457 include/linux/atomic/atomic-instrumented.h:33 fs/file.c:177 fs/file.c:238) <4>[ 61.111171] ksys_dup3 (fs/file.c:1337) <4>[ 61.111596] __arm64_sys_dup3 (fs/file.c:1355) <4>[ 61.112006] invoke_syscall (arch/arm64/include/asm/current.h:19 arch/arm64/kernel/syscall.c:54) <4>[ 61.112480] el0_svc_common.constprop.0 (include/linux/thread_info.h:127 (discriminator 2) arch/arm64/kernel/syscall.c:140 (discriminator 2)) <4>[ 61.112955] do_el0_svc (arch/arm64/kernel/syscall.c:152) <4>[ 61.113384] el0_svc (arch/arm64/include/asm/irqflags.h:55 arch/arm64/include/asm/irqflags.h:76 arch/arm64/kernel/entry-common.c:165 arch/arm64/kernel/entry-common.c:178 arch/arm64/kernel/entry-common.c:713) <4>[ 61.113742] el0t_64_sync_handler (arch/arm64/kernel/entry-common.c:731) <4>[ 61.115181] el0t_64_sync (arch/arm64/kernel/entry.S:598) <4>[ 61.115709] ---[ end trace 0000000000000000 ]--- Crash Log links, -------- - https://qa-reports.linaro.org/lkft/linux-next-master/build/next-20240826/te… Crash failed comparison: ---------- - https://qa-reports.linaro.org/lkft/linux-next-master/build/next-20240826/te… metadata: ---- git describe: next-20240823 and next-20240826 git repo: https://gitlab.com/Linaro/lkft/mirrors/next/linux-next git sha: c79c85875f1af04040fe4492ed94ce37ad729c4d kernel config: https://storage.tuxsuite.com/public/linaro/lkft/builds/2l2pZRzhgRkPgXIKLJCI… artifact location: https://storage.tuxsuite.com/public/linaro/lkft/builds/2l2pZRzhgRkPgXIKLJCI… build url: https://storage.tuxsuite.com/public/linaro/lkft/builds/2l2pZRzhgRkPgXIKLJCI… toolchain: clang-18 and gcc-13 Steps to reproduce: --------- - https://tuxapi.tuxsuite.com/v1/groups/linaro/projects/lkft/tests/2l2paZVYTl… - https://tuxapi.tuxsuite.com/v1/groups/linaro/projects/lkft/tests/2l2paZVYTl… Please let me know if you need more information. -- Linaro LKFT https://lkft.linaro.org

1 week, 3 days

1
1
0 0

[PATCH v2] selftests/vDSO: support DT_GNU_HASH

by Fangrui Song

glibc added support for DT_GNU_HASH in 2006 and DT_HASH has been obsoleted for more than one decade in many Linux distributions. Many vDSOs support DT_GNU_HASH. This patch adds selftests support. Signed-off-by: Fangrui Song <maskray(a)google.com> Tested-by: Xi Ruoyao <xry111(a)xry111.site> -- Changes from v1: * fix style of a multi-line comment. ignore false positive suggestions from checkpath.pl: `ELF(Word) *` --- tools/testing/selftests/vDSO/parse_vdso.c | 105 ++++++++++++++++------ 1 file changed, 79 insertions(+), 26 deletions(-) diff --git a/tools/testing/selftests/vDSO/parse_vdso.c b/tools/testing/selftests/vDSO/parse_vdso.c index 4ae417372e9e..dbc946dee4b1 100644 --- a/tools/testing/selftests/vDSO/parse_vdso.c +++ b/tools/testing/selftests/vDSO/parse_vdso.c @@ -47,6 +47,7 @@ static struct vdso_info /* Symbol table */ ELF(Sym) *symtab; const char *symstrings; + ELF(Word) *gnu_hash; ELF(Word) *bucket, *chain; ELF(Word) nbucket, nchain; @@ -75,6 +76,16 @@ static unsigned long elf_hash(const char *name) return h; } +static uint32_t gnu_hash(const char *name) +{ + const unsigned char *s = (void *)name; + uint32_t h = 5381; + + for (; *s; s++) + h += h * 32 + *s; + return h; +} + void vdso_init_from_sysinfo_ehdr(uintptr_t base) { size_t i; @@ -117,6 +128,7 @@ void vdso_init_from_sysinfo_ehdr(uintptr_t base) */ ELF(Word) *hash = 0; vdso_info.symstrings = 0; + vdso_info.gnu_hash = 0; vdso_info.symtab = 0; vdso_info.versym = 0; vdso_info.verdef = 0; @@ -137,6 +149,11 @@ void vdso_init_from_sysinfo_ehdr(uintptr_t base) ((uintptr_t)dyn[i].d_un.d_ptr + vdso_info.load_offset); break; + case DT_GNU_HASH: + vdso_info.gnu_hash = + (ELF(Word) *)((uintptr_t)dyn[i].d_un.d_ptr + + vdso_info.load_offset); + break; case DT_VERSYM: vdso_info.versym = (ELF(Versym) *) ((uintptr_t)dyn[i].d_un.d_ptr @@ -149,17 +166,26 @@ void vdso_init_from_sysinfo_ehdr(uintptr_t base) break; } } - if (!vdso_info.symstrings || !vdso_info.symtab || !hash) + if (!vdso_info.symstrings || !vdso_info.symtab || + (!hash && !vdso_info.gnu_hash)) return; /* Failed */ if (!vdso_info.verdef) vdso_info.versym = 0; /* Parse the hash table header. */ - vdso_info.nbucket = hash[0]; - vdso_info.nchain = hash[1]; - vdso_info.bucket = &hash[2]; - vdso_info.chain = &hash[vdso_info.nbucket + 2]; + if (vdso_info.gnu_hash) { + vdso_info.nbucket = vdso_info.gnu_hash[0]; + /* The bucket array is located after the header (4 uint32) and the bloom + * filter (size_t array of gnu_hash[2] elements). */ + vdso_info.bucket = vdso_info.gnu_hash + 4 + + sizeof(size_t) / 4 * vdso_info.gnu_hash[2]; + } else { + vdso_info.nbucket = hash[0]; + vdso_info.nchain = hash[1]; + vdso_info.bucket = &hash[2]; + vdso_info.chain = &hash[vdso_info.nbucket + 2]; + } /* That's all we need. */ vdso_info.valid = true; @@ -203,6 +229,26 @@ static bool vdso_match_version(ELF(Versym) ver, && !strcmp(name, vdso_info.symstrings + aux->vda_name); } +static bool check_sym(ELF(Sym) *sym, ELF(Word) i, const char *name, + const char *version, unsigned long ver_hash) +{ + /* Check for a defined global or weak function w/ right name. */ + if (ELF64_ST_TYPE(sym->st_info) != STT_FUNC) + return false; + if (ELF64_ST_BIND(sym->st_info) != STB_GLOBAL && + ELF64_ST_BIND(sym->st_info) != STB_WEAK) + return false; + if (strcmp(name, vdso_info.symstrings + sym->st_name)) + return false; + + /* Check symbol version. */ + if (vdso_info.versym && + !vdso_match_version(vdso_info.versym[i], version, ver_hash)) + return false; + + return true; +} + void *vdso_sym(const char *version, const char *name) { unsigned long ver_hash; @@ -210,29 +256,36 @@ void *vdso_sym(const char *version, const char *name) return 0; ver_hash = elf_hash(version); - ELF(Word) chain = vdso_info.bucket[elf_hash(name) % vdso_info.nbucket]; + ELF(Word) i; - for (; chain != STN_UNDEF; chain = vdso_info.chain[chain]) { - ELF(Sym) *sym = &vdso_info.symtab[chain]; + if (vdso_info.gnu_hash) { + uint32_t h1 = gnu_hash(name), h2, *hashval; - /* Check for a defined global or weak function w/ right name. */ - if (ELF64_ST_TYPE(sym->st_info) != STT_FUNC) - continue; - if (ELF64_ST_BIND(sym->st_info) != STB_GLOBAL && - ELF64_ST_BIND(sym->st_info) != STB_WEAK) - continue; - if (sym->st_shndx == SHN_UNDEF) - continue; - if (strcmp(name, vdso_info.symstrings + sym->st_name)) - continue; - - /* Check symbol version. */ - if (vdso_info.versym - && !vdso_match_version(vdso_info.versym[chain], - version, ver_hash)) - continue; - - return (void *)(vdso_info.load_offset + sym->st_value); + i = vdso_info.bucket[h1 % vdso_info.nbucket]; + if (i == 0) + return 0; + h1 |= 1; + hashval = vdso_info.bucket + vdso_info.nbucket + + (i - vdso_info.gnu_hash[1]); + for (;; i++) { + ELF(Sym) *sym = &vdso_info.symtab[i]; + h2 = *hashval++; + if (h1 == (h2 | 1) && + check_sym(sym, i, name, version, ver_hash)) + return (void *)(vdso_info.load_offset + + sym->st_value); + if (h2 & 1) + break; + } + } else { + i = vdso_info.bucket[elf_hash(name) % vdso_info.nbucket]; + for (; i; i = vdso_info.chain[i]) { + ELF(Sym) *sym = &vdso_info.symtab[i]; + if (sym->st_shndx != SHN_UNDEF && + check_sym(sym, i, name, version, ver_hash)) + return (void *)(vdso_info.load_offset + + sym->st_value); + } } return 0; -- 2.46.0.662.g92d0881bb0-goog

1 week, 4 days

4
6
0 0

[PATCH v4] memfd: `MFD_NOEXEC_SEAL` should not imply `MFD_ALLOW_SEALING`

by Barnabás Pőcze

`MFD_NOEXEC_SEAL` should remove the executable bits and set `F_SEAL_EXEC` to prevent further modifications to the executable bits as per the comment in the uapi header file: not executable and sealed to prevent changing to executable However, commit 105ff5339f498a ("mm/memfd: add MFD_NOEXEC_SEAL and MFD_EXEC") that introduced this feature made it so that `MFD_NOEXEC_SEAL` unsets `F_SEAL_SEAL`, essentially acting as a superset of `MFD_ALLOW_SEALING`. Nothing implies that it should be so, and indeed up until the second version of the of the patchset[0] that introduced `MFD_EXEC` and `MFD_NOEXEC_SEAL`, `F_SEAL_SEAL` was not removed, however, it was changed in the third revision of the patchset[1] without a clear explanation. This behaviour is surprising for application developers, there is no documentation that would reveal that `MFD_NOEXEC_SEAL` has the additional effect of `MFD_ALLOW_SEALING`. Additionally, combined with `vm.memfd_noexec=2` it has the effect of making all memfds initially sealable. So do not remove `F_SEAL_SEAL` when `MFD_NOEXEC_SEAL` is requested, thereby returning to the pre-Linux 6.3 behaviour of only allowing sealing when `MFD_ALLOW_SEALING` is specified. Now, this is technically a uapi break. However, the damage is expected to be minimal. To trigger user visible change, a program has to do the following steps: - create memfd: - with `MFD_NOEXEC_SEAL`, - without `MFD_ALLOW_SEALING`; - try to add seals / check the seals. But that seems unlikely to happen intentionally since this change essentially reverts the kernel's behaviour to that of Linux <6.3, so if a program worked correctly on those older kernels, it will likely work correctly after this change. I have used Debian Code Search and GitHub to try to find potential breakages, and I could only find a single one. dbus-broker's memfd_create() wrapper is aware of this implicit `MFD_ALLOW_SEALING` behaviour, and tries to work around it[2]. This workaround will break. Luckily, this only affects the test suite, it does not affect the normal operations of dbus-broker. There is a PR with a fix[3]. I also carried out a smoke test by building a kernel with this change and booting an Arch Linux system into GNOME and Plasma sessions. There was also a previous attempt to address this peculiarity by introducing a new flag[4]. [0]: https://lore.kernel.org/lkml/20220805222126.142525-3-jeffxu@google.com/ [1]: https://lore.kernel.org/lkml/20221202013404.163143-3-jeffxu@google.com/ [2]: https://github.com/bus1/dbus-broker/blob/9eb0b7e5826fc76cad7b025bc46f267d4a… [3]: https://github.com/bus1/dbus-broker/pull/366 [4]: https://lore.kernel.org/lkml/20230714114753.170814-1-david@readahead.eu/ Cc: stable(a)vger.kernel.org Signed-off-by: Barnabás Pőcze <pobrn(a)protonmail.com> --- * v3: https://lore.kernel.org/linux-mm/20240611231409.3899809-1-jeffxu@chromium.o… * v2: https://lore.kernel.org/linux-mm/20240524033933.135049-1-jeffxu@google.com/ * v1: https://lore.kernel.org/linux-mm/20240513191544.94754-1-pobrn@protonmail.co… This fourth version returns to removing the inconsistency as opposed to documenting its existence, with the same code change as v1 but with a somewhat extended commit message. This is sent because I believe it is worth at least a try; it can be easily reverted if bigger application breakages are discovered than initially imagined. --- mm/memfd.c | 9 ++++----- tools/testing/selftests/memfd/memfd_test.c | 2 +- 2 files changed, 5 insertions(+), 6 deletions(-) diff --git a/mm/memfd.c b/mm/memfd.c index 7d8d3ab3fa37..8b7f6afee21d 100644 --- a/mm/memfd.c +++ b/mm/memfd.c @@ -356,12 +356,11 @@ SYSCALL_DEFINE2(memfd_create, inode->i_mode &= ~0111; file_seals = memfd_file_seals_ptr(file); - if (file_seals) { - *file_seals &= ~F_SEAL_SEAL; + if (file_seals) *file_seals |= F_SEAL_EXEC; - } - } else if (flags & MFD_ALLOW_SEALING) { - /* MFD_EXEC and MFD_ALLOW_SEALING are set */ + } + + if (flags & MFD_ALLOW_SEALING) { file_seals = memfd_file_seals_ptr(file); if (file_seals) *file_seals &= ~F_SEAL_SEAL; diff --git a/tools/testing/selftests/memfd/memfd_test.c b/tools/testing/selftests/memfd/memfd_test.c index 95af2d78fd31..7b78329f65b6 100644 --- a/tools/testing/selftests/memfd/memfd_test.c +++ b/tools/testing/selftests/memfd/memfd_test.c @@ -1151,7 +1151,7 @@ static void test_noexec_seal(void) mfd_def_size, MFD_CLOEXEC | MFD_NOEXEC_SEAL); mfd_assert_mode(fd, 0666); - mfd_assert_has_seals(fd, F_SEAL_EXEC); + mfd_assert_has_seals(fd, F_SEAL_SEAL | F_SEAL_EXEC); mfd_fail_chmod(fd, 0777); close(fd); } -- 2.45.2

2 weeks, 2 days

4
5
0 0

[PATCH v2 net-next] wireguard: allowedips: Add WGALLOWEDIP_F_REMOVE_ME flag

by Jordan Rife

With the current API the only way to remove an allowed IP is to completely rebuild the allowed IPs set for a peer using WGPEER_F_REPLACE_ALLOWEDIPS. In other words, if my current configuration is such that a peer has allowed IP IPs 192.168.0.2 and 192.168.0.3 and I want to remove 192.168.0.2 the actual transition looks like this. [192.168.0.2, 192.168.0.3] <-- Initial state [] <-- Step 1: Allowed IPs removed for peer [192.168.0.3] <-- Step 2: Allowed IPs added back for peer This is true even if the allowed IP list is small and the update does not need to be batched into multiple WG_CMD_SET_DEVICE requests, as the removal and subsequent addition of IPs is non-atomic within a single request. Consequently, wg_allowedips_lookup_dst and wg_allowedips_lookup_src may return NULL while reconfiguring a peer even for packets bound for IPs a user did not intend to remove leading to unintended interruptions in connectivity. This presents in userspace as failed calls to sendto and sendmsg for UDP sockets. In my case, I ran netperf while repeatedly reconfiguring the allowed IPs for a peer with wg. /usr/local/bin/netperf -H 10.102.73.72 -l 10m -t UDP_STREAM -- -R 1 -m 1024 send_data: data send error: No route to host (errno 113) netperf: send_omni: send_data failed: No route to host While this may not be of particular concern for environments where peers and allowed IPs are mostly static, Cilium manages peers and allowed IPs in a dynamic environment where peers (i.e. Kubernetes nodes) and allowed IPs (i.e. Pods running on those nodes) can frequently change. Cilium must continually keep its WireGuard device's configuration in sync with its cluster state leading to unnecessary churn and packet drops. This patch introduces a new flag called WGALLOWEDIP_F_REMOVE_ME which in the same way that WGPEER_F_REMOVE_ME allows a user to remove a single peer from a WireGuard device's configuration allows a user to remove an IP from a peer's set of allowed IPs. This has two benefits. First, it allows systems such as Cilium to avoid introducing connectivity blips while reconfiguring a WireGuard device. Second, it allows us to more efficiently keep the device's configuration in sync with the cluster state, as we no longer need to do frequent rebuilds of the allowed IPs list for each peer. Instead, the device's configuration can be incrementally updated. This patch also bumps WG_GENL_VERSION which can be used by clients to detect whether or not their system supports the WGALLOWEDIP_F_REMOVE_ME flag. ======= Changes ======= v1->v2 ------ * Fixed some Sparse warnings Signed-off-by: Jordan Rife <jrife(a)google.com> --- drivers/net/wireguard/allowedips.c | 103 ++++++++++---- drivers/net/wireguard/allowedips.h | 4 + drivers/net/wireguard/netlink.c | 45 +++++-- drivers/net/wireguard/selftest/allowedips.c | 30 +++++ include/uapi/linux/wireguard.h | 11 +- tools/testing/selftests/wireguard/Makefile | 18 +++ tools/testing/selftests/wireguard/netns.sh | 38 ++++++ tools/testing/selftests/wireguard/remove-ip.c | 126 ++++++++++++++++++ 8 files changed, 333 insertions(+), 42 deletions(-) create mode 100644 tools/testing/selftests/wireguard/Makefile create mode 100644 tools/testing/selftests/wireguard/remove-ip.c diff --git a/drivers/net/wireguard/allowedips.c b/drivers/net/wireguard/allowedips.c index 4b8528206cc8a..ff52259dd8d81 100644 --- a/drivers/net/wireguard/allowedips.c +++ b/drivers/net/wireguard/allowedips.c @@ -249,6 +249,56 @@ static int add(struct allowedips_node __rcu **trie, u8 bits, const u8 *key, return 0; } +static void _remove(struct allowedips_node *node, struct mutex *lock) +{ + struct allowedips_node *child, **parent_bit, *parent; + bool free_parent; + + list_del_init(&node->peer_list); + RCU_INIT_POINTER(node->peer, NULL); + if (node->bit[0] && node->bit[1]) + return; + child = rcu_dereference_protected(node->bit[!rcu_access_pointer(node->bit[0])], + lockdep_is_held(lock)); + if (child) + child->parent_bit_packed = node->parent_bit_packed; + parent_bit = (struct allowedips_node **)(node->parent_bit_packed & ~3UL); + *parent_bit = child; + parent = (void *)parent_bit - + offsetof(struct allowedips_node, bit[node->parent_bit_packed & 1]); + free_parent = !rcu_access_pointer(node->bit[0]) && + !rcu_access_pointer(node->bit[1]) && + (node->parent_bit_packed & 3) <= 1 && + !rcu_access_pointer(parent->peer); + if (free_parent) + child = rcu_dereference_protected(parent->bit[!(node->parent_bit_packed & 1)], + lockdep_is_held(lock)); + call_rcu(&node->rcu, node_free_rcu); + if (!free_parent) + return; + if (child) + child->parent_bit_packed = parent->parent_bit_packed; + *(struct allowedips_node **)(parent->parent_bit_packed & ~3UL) = child; + call_rcu(&parent->rcu, node_free_rcu); +} + +static int remove(struct allowedips_node __rcu **trie, u8 bits, const u8 *key, + u8 cidr, struct wg_peer *peer, struct mutex *lock) +{ + struct allowedips_node *node; + + if (unlikely(cidr > bits || !peer)) + return -EINVAL; + if (!rcu_access_pointer(*trie) || + !node_placement(*trie, key, cidr, bits, &node, lock) || + peer != rcu_access_pointer(node->peer)) + return 0; + + _remove(node, lock); + + return 0; +} + void wg_allowedips_init(struct allowedips *table) { table->root4 = table->root6 = NULL; @@ -300,43 +350,38 @@ int wg_allowedips_insert_v6(struct allowedips *table, const struct in6_addr *ip, return add(&table->root6, 128, key, cidr, peer, lock); } +int wg_allowedips_remove_v4(struct allowedips *table, const struct in_addr *ip, + u8 cidr, struct wg_peer *peer, struct mutex *lock) +{ + /* Aligned so it can be passed to fls */ + u8 key[4] __aligned(__alignof(u32)); + + ++table->seq; + swap_endian(key, (const u8 *)ip, 32); + return remove(&table->root4, 32, key, cidr, peer, lock); +} + +int wg_allowedips_remove_v6(struct allowedips *table, const struct in6_addr *ip, + u8 cidr, struct wg_peer *peer, struct mutex *lock) +{ + /* Aligned so it can be passed to fls64 */ + u8 key[16] __aligned(__alignof(u64)); + + ++table->seq; + swap_endian(key, (const u8 *)ip, 128); + return remove(&table->root6, 128, key, cidr, peer, lock); +} + void wg_allowedips_remove_by_peer(struct allowedips *table, struct wg_peer *peer, struct mutex *lock) { - struct allowedips_node *node, *child, **parent_bit, *parent, *tmp; - bool free_parent; + struct allowedips_node *node, *tmp; if (list_empty(&peer->allowedips_list)) return; ++table->seq; list_for_each_entry_safe(node, tmp, &peer->allowedips_list, peer_list) { - list_del_init(&node->peer_list); - RCU_INIT_POINTER(node->peer, NULL); - if (node->bit[0] && node->bit[1]) - continue; - child = rcu_dereference_protected(node->bit[!rcu_access_pointer(node->bit[0])], - lockdep_is_held(lock)); - if (child) - child->parent_bit_packed = node->parent_bit_packed; - parent_bit = (struct allowedips_node **)(node->parent_bit_packed & ~3UL); - *parent_bit = child; - parent = (void *)parent_bit - - offsetof(struct allowedips_node, bit[node->parent_bit_packed & 1]); - free_parent = !rcu_access_pointer(node->bit[0]) && - !rcu_access_pointer(node->bit[1]) && - (node->parent_bit_packed & 3) <= 1 && - !rcu_access_pointer(parent->peer); - if (free_parent) - child = rcu_dereference_protected( - parent->bit[!(node->parent_bit_packed & 1)], - lockdep_is_held(lock)); - call_rcu(&node->rcu, node_free_rcu); - if (!free_parent) - continue; - if (child) - child->parent_bit_packed = parent->parent_bit_packed; - *(struct allowedips_node **)(parent->parent_bit_packed & ~3UL) = child; - call_rcu(&parent->rcu, node_free_rcu); + _remove(node, lock); } } diff --git a/drivers/net/wireguard/allowedips.h b/drivers/net/wireguard/allowedips.h index 2346c797eb4d8..931958cb6e100 100644 --- a/drivers/net/wireguard/allowedips.h +++ b/drivers/net/wireguard/allowedips.h @@ -38,6 +38,10 @@ int wg_allowedips_insert_v4(struct allowedips *table, const struct in_addr *ip, u8 cidr, struct wg_peer *peer, struct mutex *lock); int wg_allowedips_insert_v6(struct allowedips *table, const struct in6_addr *ip, u8 cidr, struct wg_peer *peer, struct mutex *lock); +int wg_allowedips_remove_v4(struct allowedips *table, const struct in_addr *ip, + u8 cidr, struct wg_peer *peer, struct mutex *lock); +int wg_allowedips_remove_v6(struct allowedips *table, const struct in6_addr *ip, + u8 cidr, struct wg_peer *peer, struct mutex *lock); void wg_allowedips_remove_by_peer(struct allowedips *table, struct wg_peer *peer, struct mutex *lock); /* The ip input pointer should be __aligned(__alignof(u64))) */ diff --git a/drivers/net/wireguard/netlink.c b/drivers/net/wireguard/netlink.c index f7055180ba4aa..5f2a8553ab43d 100644 --- a/drivers/net/wireguard/netlink.c +++ b/drivers/net/wireguard/netlink.c @@ -46,7 +46,8 @@ static const struct nla_policy peer_policy[WGPEER_A_MAX + 1] = { static const struct nla_policy allowedip_policy[WGALLOWEDIP_A_MAX + 1] = { [WGALLOWEDIP_A_FAMILY] = { .type = NLA_U16 }, [WGALLOWEDIP_A_IPADDR] = NLA_POLICY_MIN_LEN(sizeof(struct in_addr)), - [WGALLOWEDIP_A_CIDR_MASK] = { .type = NLA_U8 } + [WGALLOWEDIP_A_CIDR_MASK] = { .type = NLA_U8 }, + [WGALLOWEDIP_A_FLAGS] = { .type = NLA_U32 } }; static struct wg_device *lookup_interface(struct nlattr **attrs, @@ -329,6 +330,7 @@ static int set_port(struct wg_device *wg, u16 port) static int set_allowedip(struct wg_peer *peer, struct nlattr **attrs) { int ret = -EINVAL; + u32 flags = 0; u16 family; u8 cidr; @@ -337,19 +339,38 @@ static int set_allowedip(struct wg_peer *peer, struct nlattr **attrs) return ret; family = nla_get_u16(attrs[WGALLOWEDIP_A_FAMILY]); cidr = nla_get_u8(attrs[WGALLOWEDIP_A_CIDR_MASK]); + if (attrs[WGALLOWEDIP_A_FLAGS]) + flags = nla_get_u32(attrs[WGALLOWEDIP_A_FLAGS]); if (family == AF_INET && cidr <= 32 && - nla_len(attrs[WGALLOWEDIP_A_IPADDR]) == sizeof(struct in_addr)) - ret = wg_allowedips_insert_v4( - &peer->device->peer_allowedips, - nla_data(attrs[WGALLOWEDIP_A_IPADDR]), cidr, peer, - &peer->device->device_update_lock); - else if (family == AF_INET6 && cidr <= 128 && - nla_len(attrs[WGALLOWEDIP_A_IPADDR]) == sizeof(struct in6_addr)) - ret = wg_allowedips_insert_v6( - &peer->device->peer_allowedips, - nla_data(attrs[WGALLOWEDIP_A_IPADDR]), cidr, peer, - &peer->device->device_update_lock); + nla_len(attrs[WGALLOWEDIP_A_IPADDR]) == sizeof(struct in_addr)) { + if (flags & WGALLOWEDIP_F_REMOVE_ME) + ret = wg_allowedips_remove_v4(&peer->device->peer_allowedips, + nla_data(attrs[WGALLOWEDIP_A_IPADDR]), + cidr, + peer, + &peer->device->device_update_lock); + else + ret = wg_allowedips_insert_v4(&peer->device->peer_allowedips, + nla_data(attrs[WGALLOWEDIP_A_IPADDR]), + cidr, + peer, + &peer->device->device_update_lock); + } else if (family == AF_INET6 && cidr <= 128 && + nla_len(attrs[WGALLOWEDIP_A_IPADDR]) == sizeof(struct in6_addr)) { + if (flags & WGALLOWEDIP_F_REMOVE_ME) + ret = wg_allowedips_remove_v6(&peer->device->peer_allowedips, + nla_data(attrs[WGALLOWEDIP_A_IPADDR]), + cidr, + peer, + &peer->device->device_update_lock); + else + ret = wg_allowedips_insert_v6(&peer->device->peer_allowedips, + nla_data(attrs[WGALLOWEDIP_A_IPADDR]), + cidr, + peer, + &peer->device->device_update_lock); + } return ret; } diff --git a/drivers/net/wireguard/selftest/allowedips.c b/drivers/net/wireguard/selftest/allowedips.c index 3d1f64ff2e122..9f6458a889e96 100644 --- a/drivers/net/wireguard/selftest/allowedips.c +++ b/drivers/net/wireguard/selftest/allowedips.c @@ -461,6 +461,10 @@ static __init struct wg_peer *init_peer(void) wg_allowedips_insert_v##version(&t, ip##version(ipa, ipb, ipc, ipd), \ cidr, mem, &mutex) +#define remove(version, mem, ipa, ipb, ipc, ipd, cidr) \ + wg_allowedips_remove_v##version(&t, ip##version(ipa, ipb, ipc, ipd), \ + cidr, mem, &mutex) + #define maybe_fail() do { \ ++i; \ if (!_s) { \ @@ -586,6 +590,32 @@ bool __init wg_allowedips_selftest(void) test_negative(4, a, 192, 0, 0, 0); test_negative(4, a, 255, 0, 0, 0); + insert(4, a, 1, 0, 0, 0, 32); + insert(4, a, 192, 0, 0, 0, 24); + insert(6, a, 0x24446801, 0x40e40800, 0xdeaebeef, 0xdefbeef, 128); + insert(6, a, 0x24446800, 0xf0e40800, 0xeeaebeef, 0, 98); + test(4, a, 1, 0, 0, 0); + test(4, a, 192, 0, 0, 1); + test(6, a, 0x24446801, 0x40e40800, 0xdeaebeef, 0xdefbeef); + test(6, a, 0x24446800, 0xf0e40800, 0xeeaebeef, 0x10101010); + /* Must be an exact match to remove */ + remove(4, a, 192, 0, 0, 0, 32); + test(4, a, 192, 0, 0, 1); + remove(4, a, 192, 0, 0, 0, 24); + test_negative(4, a, 192, 0, 0, 1); + remove(4, a, 1, 0, 0, 0, 32); + test_negative(4, a, 1, 0, 0, 0); + /* Must be an exact match to remove */ + remove(6, a, 0x24446801, 0x40e40800, 0xdeaebeef, 0xdefbeef, 96); + test(6, a, 0x24446801, 0x40e40800, 0xdeaebeef, 0xdefbeef); + remove(6, a, 0x24446801, 0x40e40800, 0xdeaebeef, 0xdefbeef, 128); + test_negative(6, a, 0x24446801, 0x40e40800, 0xdeaebeef, 0xdefbeef); + /* Must match the peer to remove */ + remove(6, b, 0x24446800, 0xf0e40800, 0xeeaebeef, 0, 98); + test(6, a, 0x24446800, 0xf0e40800, 0xeeaebeef, 0x10101010); + remove(6, a, 0x24446800, 0xf0e40800, 0xeeaebeef, 0, 98); + test_negative(6, a, 0x24446800, 0xf0e40800, 0xeeaebeef, 0x10101010); + wg_allowedips_free(&t, &mutex); wg_allowedips_init(&t); insert(4, a, 192, 168, 0, 0, 16); diff --git a/include/uapi/linux/wireguard.h b/include/uapi/linux/wireguard.h index ae88be14c9478..e219194cb9f5a 100644 --- a/include/uapi/linux/wireguard.h +++ b/include/uapi/linux/wireguard.h @@ -101,6 +101,10 @@ * WGALLOWEDIP_A_FAMILY: NLA_U16 * WGALLOWEDIP_A_IPADDR: struct in_addr or struct in6_addr * WGALLOWEDIP_A_CIDR_MASK: NLA_U8 + * WGALLOWEDIP_A_FLAGS: NLA_U32, WGALLOWEDIP_F_REMOVE_ME if + * the specified IP should be removed, + * otherwise this IP will be added if + * it is not already present. * 0: NLA_NESTED * ... * 0: NLA_NESTED @@ -132,7 +136,7 @@ #define _WG_UAPI_WIREGUARD_H #define WG_GENL_NAME "wireguard" -#define WG_GENL_VERSION 1 +#define WG_GENL_VERSION 2 #define WG_KEY_LEN 32 @@ -184,11 +188,16 @@ enum wgpeer_attribute { }; #define WGPEER_A_MAX (__WGPEER_A_LAST - 1) +enum wgallowedip_flag { + WGALLOWEDIP_F_REMOVE_ME = 1U << 0, + __WGALLOWEDIP_F_ALL = WGALLOWEDIP_F_REMOVE_ME +}; enum wgallowedip_attribute { WGALLOWEDIP_A_UNSPEC, WGALLOWEDIP_A_FAMILY, WGALLOWEDIP_A_IPADDR, WGALLOWEDIP_A_CIDR_MASK, + WGALLOWEDIP_A_FLAGS, __WGALLOWEDIP_A_LAST }; #define WGALLOWEDIP_A_MAX (__WGALLOWEDIP_A_LAST - 1) diff --git a/tools/testing/selftests/wireguard/Makefile b/tools/testing/selftests/wireguard/Makefile new file mode 100644 index 0000000000000..4f4db54f89cb3 --- /dev/null +++ b/tools/testing/selftests/wireguard/Makefile @@ -0,0 +1,18 @@ +# SPDX-License-Identifier: GPL-2.0 +# +# Note: To build this you must install libnl-3 and libnl-genl-3 development +# packages. +remove-ip: + gcc -I/usr/include/libnl3 \ + -I../../../../usr/include \ + remove-ip.c \ + -o remove-ip \ + -lnl-genl-3 \ + -lnl-3 + +.PHONY: all +all: remove-ip + +.PHONY: clean +clean: + rm remove-ip diff --git a/tools/testing/selftests/wireguard/netns.sh b/tools/testing/selftests/wireguard/netns.sh index 405ff262ca93d..70058d6ebbe85 100755 --- a/tools/testing/selftests/wireguard/netns.sh +++ b/tools/testing/selftests/wireguard/netns.sh @@ -28,6 +28,7 @@ exec 3>&1 export LANG=C export WG_HIDE_KEYS=never NPROC=( /sys/devices/system/cpu/cpu+([0-9]) ); NPROC=${#NPROC[@]} +SCRIPT_DIR=$( cd -- "$( dirname -- "${BASH_SOURCE[0]}" )" &> /dev/null && pwd ) netns0="wg-test-$$-0" netns1="wg-test-$$-1" netns2="wg-test-$$-2" @@ -610,6 +611,43 @@ n0 wg set wg0 peer "$pub2" allowed-ips "$allowedips" } < <(n0 wg show wg0 allowed-ips) ip0 link del wg0 +# Test IP removal +allowedips=( ) +for i in {1..197}; do + allowedips+=( 192.168.0.$i ) + allowedips+=( abcd::$i ) +done +saved_ifs="$IFS" +IFS=, +allowedips="${allowedips[*]}" +IFS="$saved_ifs" +ip0 link add wg0 type wireguard +n0 wg set wg0 peer "$pub1" allowed-ips "$allowedips" +pub1_hex=$(echo "$pub1" | base64 -d | xxd -p -c 50) +n0 $SCRIPT_DIR/remove-ip wg0 "$pub1_hex" 4 192.168.0.1 +n0 $SCRIPT_DIR/remove-ip wg0 "$pub1_hex" 4 192.168.0.20 +n0 $SCRIPT_DIR/remove-ip wg0 "$pub1_hex" 4 192.168.0.100 +n0 $SCRIPT_DIR/remove-ip wg0 "$pub1_hex" 6 abcd::1 +n0 $SCRIPT_DIR/remove-ip wg0 "$pub1_hex" 6 abcd::20 +n0 $SCRIPT_DIR/remove-ip wg0 "$pub1_hex" 6 abcd::100 +n0 wg show wg0 allowed-ips +{ + read -r pub allowedips + [[ $pub == "$pub1" ]] + i=0 + for ip in $allowedips; do + [[ "$ip" != "192.168.0.1" ]] + [[ "$ip" != "192.168.0.20" ]] + [[ "$ip" != "192.168.0.100" ]] + [[ "$ip" != "abcd::1" ]] + [[ "$ip" != "abcd::20" ]] + [[ "$ip" != "abcd::100" ]] + ((++i)) + done + ((i == 388)) +} < <(n0 wg show wg0 allowed-ips) +ip0 link del wg0 + ! n0 wg show doesnotexist || false ip0 link add wg0 type wireguard diff --git a/tools/testing/selftests/wireguard/remove-ip.c b/tools/testing/selftests/wireguard/remove-ip.c new file mode 100644 index 0000000000000..242f922d99b56 --- /dev/null +++ b/tools/testing/selftests/wireguard/remove-ip.c @@ -0,0 +1,126 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <linux/wireguard.h> +#include <sys/socket.h> +#include <netinet/in.h> +#include <arpa/inet.h> +#include <netlink/socket.h> +#include <netlink/netlink.h> +#include <netlink/genl/ctrl.h> +#include <netlink/genl/genl.h> +#include <netlink/genl/family.h> + +#define CURVE25519_KEY_SIZE 32 + +const char *usage = "Usage: remove-ip INTERFACE_NAME PEER_PUBLIC_KEY_HEX IP_VERSION IP"; + +char h2b(char c) +{ + if ('0' <= c && c <= '9') + return c - '0'; + else if ('a' <= c && c <= 'f') + return 10 + (c - 'a'); + + return -1; +} + +int parse_key(const char *raw, unsigned char key[CURVE25519_KEY_SIZE]) +{ + int ret = 0; + int i; + + for (i = 0; i < CURVE25519_KEY_SIZE; i++) { + char h, l; + + h = h2b(raw[0]); + if (h < 0) + return -1; + + l = h2b(raw[1]); + if (l < 0) + return -1; + + key[i] = (h << 4) | l; + raw += 2; + } + + return 0; +} + +int main(int argc, char **argv) +{ + unsigned char addr[sizeof(struct in6_addr)]; + unsigned char pub_key[CURVE25519_KEY_SIZE]; + struct nl_sock *sock; + struct nl_msg *msg; + int addr_len; + int family; + int cidr; + int af; + + if (argc < 5) { + printf("Not enough arguments.\n\n%s\n", usage); + return -1; + } + + if (parse_key(argv[2], pub_key)) { + printf("Could not parse public key\n"); + return -1; + } + + switch (argv[3][0]) { + case '4': + af = AF_INET; + addr_len = sizeof(struct in_addr); + cidr = 32; + break; + case '6': + af = AF_INET6; + addr_len = sizeof(struct in6_addr); + cidr = 128; + break; + default: + printf("Invalid IP version\n"); + return -1; + } + + if (inet_pton(af, argv[4], &addr) <= 0) { + printf("Could not parse IP address\n"); + return -1; + } + + sock = nl_socket_alloc(); + genl_connect(sock); + family = genl_ctrl_resolve(sock, WG_GENL_NAME); + msg = nlmsg_alloc(); + genlmsg_put(msg, NL_AUTO_PID, NL_AUTO_SEQ, family, 0, NLM_F_ECHO, + WG_CMD_SET_DEVICE, WG_GENL_VERSION); + nla_put_string(msg, WGDEVICE_A_IFNAME, argv[1]); + + struct nlattr *peers = nla_nest_start(msg, WGDEVICE_A_PEERS); + struct nlattr *peer0 = nla_nest_start(msg, 0); + + nla_put(msg, WGPEER_A_PUBLIC_KEY, CURVE25519_KEY_SIZE, pub_key); + + struct nlattr *allowed_ips = nla_nest_start(msg, WGPEER_A_ALLOWEDIPS); + struct nlattr *allowed_ip0 = nla_nest_start(msg, 0); + + nla_put_u16(msg, WGALLOWEDIP_A_FAMILY, af); + nla_put(msg, WGALLOWEDIP_A_IPADDR, addr_len, &addr); + nla_put_u8(msg, WGALLOWEDIP_A_CIDR_MASK, cidr); + nla_put_u32(msg, WGALLOWEDIP_A_FLAGS, WGALLOWEDIP_F_REMOVE_ME); + nla_nest_end(msg, allowed_ip0); + nla_nest_end(msg, allowed_ips); + nla_nest_end(msg, peer0); + nla_nest_end(msg, peers); + + int err = nl_send_sync(sock, msg); + + if (err < 0) { + char message[256]; + + nl_perror(err, message); + printf("An error occurred: %d - %s\n", err, message); + } + + return err; +} -- 2.46.0.598.g6f2099f65c-goog

2 weeks, 3 days

3
7
0 0

[RFC PATCH 00/39] 1G page support for guest_memfd

by Ackerley Tng

Hello, This patchset is our exploration of how to support 1G pages in guest_memfd, and how the pages will be used in Confidential VMs. The patchset covers: + How to get 1G pages + Allowing mmap() of guest_memfd to userspace so that both private and shared memory can use the same physical pages + Splitting and reconstructing pages to support conversions and mmap() + How the VM, userspace and guest_memfd interact to support conversions + Selftests to test all the above + Selftests also demonstrate the conversion flow between VM, userspace and guest_memfd. Why 1G pages in guest memfd? Bring guest_memfd to performance and memory savings parity with VMs that are backed by HugeTLBfs. + Performance is improved with 1G pages by more TLB hits and faster page walks on TLB misses. + Memory savings from 1G pages comes from HugeTLB Vmemmap Optimization (HVO). Options for 1G page support: 1. HugeTLB 2. Contiguous Memory Allocator (CMA) 3. Other suggestions are welcome! Comparison between options: 1. HugeTLB + Refactor HugeTLB to separate allocator from the rest of HugeTLB + Pro: Graceful transition for VMs backed with HugeTLB to guest_memfd + Near term: Allows co-tenancy of HugeTLB and guest_memfd backed VMs + Pro: Can provide iterative steps toward new future allocator + Unexplored: Managing userspace-visible changes + e.g. HugeTLB's free_hugepages will decrease if HugeTLB is used, but not when future allocator is used 2. CMA + Port some HugeTLB features to be applied on CMA + Pro: Clean slate What would refactoring HugeTLB involve? (Some refactoring was done in this RFC, more can be done.) 1. Broadly involves separating the HugeTLB allocator from the rest of HugeTLB + Brings more modularity to HugeTLB + No functionality change intended + Likely step towards HugeTLB's integration into core-mm 2. guest_memfd will use just the allocator component of HugeTLB, not including the complex parts of HugeTLB like + Userspace reservations (resv_map) + Shared PMD mappings + Special page walkers What features would need to be ported to CMA? + Improved allocation guarantees + Per NUMA node pool of huge pages + Subpools per guest_memfd + Memory savings + Something like HugeTLB Vmemmap Optimization + Configuration/reporting features + Configuration of number of pages available (and per NUMA node) at and after host boot + Reporting of memory usage/availability statistics at runtime HugeTLB was picked as the source of 1G pages for this RFC because it allows a graceful transition, and retains memory savings from HVO. To illustrate this, if a host machine uses HugeTLBfs to back VMs, and a confidential VM were to be scheduled on that host, some HugeTLBfs pages would have to be given up and returned to CMA for guest_memfd pages to be rebuilt from that memory. This requires memory to be reserved for HVO to be removed and reapplied on the new guest_memfd memory. This not only slows down memory allocation but also trims the benefits of HVO. Memory would have to be reserved on the host to facilitate these transitions. Improving how guest_memfd uses the allocator in a future revision of this RFC: To provide an easier transition away from HugeTLB, guest_memfd's use of HugeTLB should be limited to these allocator functions: + reserve(node, page_size, num_pages) => opaque handle + Used when a guest_memfd inode is created to reserve memory from backend allocator + allocate(handle, mempolicy, page_size) => folio + To allocate a folio from guest_memfd's reservation + split(handle, folio, target_page_size) => void + To take a huge folio, and split it to smaller folios, restore to filemap + reconstruct(handle, first_folio, nr_pages) => void + To take a folio, and reconstruct a huge folio out of nr_pages from the first_folio + free(handle, folio) => void + To return folio to guest_memfd's reservation + error(handle, folio) => void + To handle memory errors + unreserve(handle) => void + To return guest_memfd's reservation to allocator backend Userspace should only provide a page size when creating a guest_memfd and should not have to specify HugeTLB. Overview of patches: + Patches 01-12 + Many small changes to HugeTLB, mostly to separate HugeTLBfs concepts from HugeTLB, and to expose HugeTLB functions. + Patches 13-16 + Letting guest_memfd use HugeTLB + Creation of each guest_memfd reserves pages from HugeTLB's global hstate and puts it into the guest_memfd inode's subpool + Each folio allocation takes a page from the guest_memfd inode's subpool + Patches 17-21 + Selftests for new HugeTLB features in guest_memfd + Patches 22-24 + More small changes on the HugeTLB side to expose functions needed by guest_memfd + Patch 25: + Uses the newly available functions from patches 22-24 to split HugeTLB pages. In this patch, HugeTLB folios are always split to 4K before any usage, private or shared. + Patches 26-28 + Allow mmap() in guest_memfd and faulting in shared pages + Patch 29 + Enables conversion between private/shared pages + Patch 30 + Required to zero folios after conversions to avoid leaking initialized kernel memory + Patch 31-38 + Add selftests to test mapping pages to userspace, guest/host memory sharing and update conversions tests + Patch 33 illustrates the conversion flow between VM/userspace/guest_memfd + Patch 39 + Dynamically split and reconstruct HugeTLB pages instead of always splitting before use. All earlier selftests are expected to still pass. TODOs: + Add logic to wait for safe_refcount [1] + Look into lazy splitting/reconstruction of pages + Currently, when the KVM_SET_MEMORY_ATTRIBUTES is invoked, not only is the mem_attr_array and faultability updated, the pages in the requested range are also split/reconstructed as necessary. We want to look into delaying splitting/reconstruction to fault time. + Solve race between folios being faulted in and being truncated + When running private_mem_conversions_test with more than 1 vCPU, a folio getting truncated may get faulted in by another process, causing elevated mapcounts when the folio is freed (VM_BUG_ON_FOLIO). + Add intermediate splits (1G should first split to 2M and not split directly to 4K) + Use guest's lock instead of hugetlb_lock + Use multi-index xarray/replace xarray with some other data struct for faultability flag + Refactor HugeTLB better, present generic allocator interface Please let us know your thoughts on: + HugeTLB as the choice of transitional allocator backend + Refactoring HugeTLB to provide generic allocator interface + Shared/private conversion flow + Requiring user to request kernel to unmap pages from userspace using madvise(MADV_DONTNEED) + Failing conversion on elevated mapcounts/pincounts/refcounts + Process of splitting/reconstructing page + Anything else! [1] https://lore.kernel.org/all/20240829-guest-memfd-lib-v2-0-b9afc1ff3656@quic… Ackerley Tng (37): mm: hugetlb: Simplify logic in dequeue_hugetlb_folio_vma() mm: hugetlb: Refactor vma_has_reserves() to should_use_hstate_resv() mm: hugetlb: Remove unnecessary check for avoid_reserve mm: mempolicy: Refactor out policy_node_nodemask() mm: hugetlb: Refactor alloc_buddy_hugetlb_folio_with_mpol() to interpret mempolicy instead of vma mm: hugetlb: Refactor dequeue_hugetlb_folio_vma() to use mpol mm: hugetlb: Refactor out hugetlb_alloc_folio mm: truncate: Expose preparation steps for truncate_inode_pages_final mm: hugetlb: Expose hugetlb_subpool_{get,put}_pages() mm: hugetlb: Add option to create new subpool without using surplus mm: hugetlb: Expose hugetlb_acct_memory() mm: hugetlb: Move and expose hugetlb_zero_partial_page() KVM: guest_memfd: Make guest mem use guest mem inodes instead of anonymous inodes KVM: guest_memfd: hugetlb: initialization and cleanup KVM: guest_memfd: hugetlb: allocate and truncate from hugetlb KVM: guest_memfd: Add page alignment check for hugetlb guest_memfd KVM: selftests: Add basic selftests for hugetlb-backed guest_memfd KVM: selftests: Support various types of backing sources for private memory KVM: selftests: Update test for various private memory backing source types KVM: selftests: Add private_mem_conversions_test.sh KVM: selftests: Test that guest_memfd usage is reported via hugetlb mm: hugetlb: Expose vmemmap optimization functions mm: hugetlb: Expose HugeTLB functions for promoting/demoting pages mm: hugetlb: Add functions to add/move/remove from hugetlb lists KVM: guest_memfd: Track faultability within a struct kvm_gmem_private KVM: guest_memfd: Allow mmapping guest_memfd files KVM: guest_memfd: Use vm_type to determine default faultability KVM: Handle conversions in the SET_MEMORY_ATTRIBUTES ioctl KVM: guest_memfd: Handle folio preparation for guest_memfd mmap KVM: selftests: Allow vm_set_memory_attributes to be used without asserting return value of 0 KVM: selftests: Test using guest_memfd memory from userspace KVM: selftests: Test guest_memfd memory sharing between guest and host KVM: selftests: Add notes in private_mem_kvm_exits_test for mmap-able guest_memfd KVM: selftests: Test that pinned pages block KVM from setting memory attributes to PRIVATE KVM: selftests: Refactor vm_mem_add to be more flexible KVM: selftests: Add helper to perform madvise by memslots KVM: selftests: Update private_mem_conversions_test for mmap()able guest_memfd Vishal Annapurve (2): KVM: guest_memfd: Split HugeTLB pages for guest_memfd use KVM: guest_memfd: Dynamically split/reconstruct HugeTLB page fs/hugetlbfs/inode.c | 35 +- include/linux/hugetlb.h | 54 +- include/linux/kvm_host.h | 1 + include/linux/mempolicy.h | 2 + include/linux/mm.h | 1 + include/uapi/linux/kvm.h | 26 + include/uapi/linux/magic.h | 1 + mm/hugetlb.c | 346 ++-- mm/hugetlb_vmemmap.h | 11 - mm/mempolicy.c | 36 +- mm/truncate.c | 26 +- tools/include/linux/kernel.h | 4 +- tools/testing/selftests/kvm/Makefile | 3 + .../kvm/guest_memfd_hugetlb_reporting_test.c | 222 +++ .../selftests/kvm/guest_memfd_pin_test.c | 104 ++ .../selftests/kvm/guest_memfd_sharing_test.c | 160 ++ .../testing/selftests/kvm/guest_memfd_test.c | 238 ++- .../testing/selftests/kvm/include/kvm_util.h | 45 +- .../testing/selftests/kvm/include/test_util.h | 18 + tools/testing/selftests/kvm/lib/kvm_util.c | 443 +++-- tools/testing/selftests/kvm/lib/test_util.c | 99 ++ .../kvm/x86_64/private_mem_conversions_test.c | 158 +- .../x86_64/private_mem_conversions_test.sh | 91 + .../kvm/x86_64/private_mem_kvm_exits_test.c | 11 +- virt/kvm/guest_memfd.c | 1563 ++++++++++++++++- virt/kvm/kvm_main.c | 17 + virt/kvm/kvm_mm.h | 16 + 27 files changed, 3288 insertions(+), 443 deletions(-) create mode 100644 tools/testing/selftests/kvm/guest_memfd_hugetlb_reporting_test.c create mode 100644 tools/testing/selftests/kvm/guest_memfd_pin_test.c create mode 100644 tools/testing/selftests/kvm/guest_memfd_sharing_test.c create mode 100755 tools/testing/selftests/kvm/x86_64/private_mem_conversions_test.sh -- 2.46.0.598.g6f2099f65c-goog

2 weeks, 5 days

14
92
0 0

[PATCH] selftest: drivers: Add support to check duplicate hwirq

by Joseph Jang

Validate there are no duplicate hwirq from the irq debug file system /sys/kernel/debug/irq/irqs/* per chip name. One example log show 2 duplicated hwirq in the irq debug file system. $ sudo cat /sys/kernel/debug/irq/irqs/163 handler: handle_fasteoi_irq device: 0019:00:00.0 <SNIP> node: 1 affinity: 72-143 effectiv: 76 domain: irqchip@0x0000100022040000-3 hwirq: 0xc8000000 chip: ITS-MSI flags: 0x20 $ sudo cat /sys/kernel/debug/irq/irqs/174 handler: handle_fasteoi_irq device: 0039:00:00.0 <SNIP> node: 3 affinity: 216-287 effectiv: 221 domain: irqchip@0x0000300022040000-3 hwirq: 0xc8000000 chip: ITS-MSI flags: 0x20 The irq-check.sh can help to collect hwirq and chip name from /sys/kernel/debug/irq/irqs/* and print error log when find duplicate hwirq per chip name. Kernel patch ("PCI/MSI: Fix MSI hwirq truncation") [1] fix above issue. [1]: https://lore.kernel.org/all/20240115135649.708536-1-vidyas@nvidia.com/ Signed-off-by: Joseph Jang <jjang(a)nvidia.com> Reviewed-by: Matthew R. Ochs <mochs(a)nvidia.com> --- tools/testing/selftests/drivers/irq/Makefile | 5 +++ tools/testing/selftests/drivers/irq/config | 2 + .../selftests/drivers/irq/irq-check.sh | 39 +++++++++++++++++++ 3 files changed, 46 insertions(+) create mode 100644 tools/testing/selftests/drivers/irq/Makefile create mode 100644 tools/testing/selftests/drivers/irq/config create mode 100755 tools/testing/selftests/drivers/irq/irq-check.sh diff --git a/tools/testing/selftests/drivers/irq/Makefile b/tools/testing/selftests/drivers/irq/Makefile new file mode 100644 index 000000000000..d6998017c861 --- /dev/null +++ b/tools/testing/selftests/drivers/irq/Makefile @@ -0,0 +1,5 @@ +# SPDX-License-Identifier: GPL-2.0 + +TEST_PROGS := irq-check.sh + +include ../../lib.mk diff --git a/tools/testing/selftests/drivers/irq/config b/tools/testing/selftests/drivers/irq/config new file mode 100644 index 000000000000..a53d3b713728 --- /dev/null +++ b/tools/testing/selftests/drivers/irq/config @@ -0,0 +1,2 @@ +CONFIG_GENERIC_IRQ_DEBUGFS=y +CONFIG_GENERIC_IRQ_INJECTION=y diff --git a/tools/testing/selftests/drivers/irq/irq-check.sh b/tools/testing/selftests/drivers/irq/irq-check.sh new file mode 100755 index 000000000000..e784777043a1 --- /dev/null +++ b/tools/testing/selftests/drivers/irq/irq-check.sh @@ -0,0 +1,39 @@ +#!/bin/bash +# SPDX-License-Identifier: GPL-2.0 + +# This script need root permission +uid=$(id -u) +if [ $uid -ne 0 ]; then + echo "SKIP: Must be run as root" + exit 4 +fi + +# Ensure debugfs is mounted +mount -t debugfs nodev /sys/kernel/debug 2>/dev/null +if [ ! -d "/sys/kernel/debug/irq/irqs" ]; then + echo "SKIP: irq debugfs not found" + exit 4 +fi + +# Traverse the irq debug file system directory to collect chip_name and hwirq +hwirq_list=$(for irq_file in /sys/kernel/debug/irq/irqs/*; do + # Read chip name and hwirq from the irq_file + chip_name=$(cat "$irq_file" | grep -m 1 'chip:' | awk '{print $2}') + hwirq=$(cat "$irq_file" | grep -m 1 'hwirq:' | awk '{print $2}' ) + + if [ -z "$chip_name" ] || [ -z "$hwirq" ]; then + continue + fi + + echo "$chip_name $hwirq" +done) + +dup_hwirq_list=$(echo "$hwirq_list" | sort | uniq -cd) + +if [ -n "$dup_hwirq_list" ]; then + echo "ERROR: Found duplicate hwirq" + echo "$dup_hwirq_list" + exit 1 +fi + +exit 0 -- 2.34.1

2 weeks, 6 days

4
6
0 0

[PATCH net] selftests: rds: move include.sh to TEST_FILES

by Hangbin Liu

The include.sh file is generated for inclusion and should not be executable. Otherwise, it will be added to kselftest-list.txt. Additionally, add the executable bit for test.py at the same time to ensure proper functionality. Fixes: 3ade6ce1255e ("selftests: rds: add testing infrastructure") Signed-off-by: Hangbin Liu <liuhangbin(a)gmail.com> --- tools/testing/selftests/net/rds/Makefile | 3 ++- tools/testing/selftests/net/rds/test.py | 0 2 files changed, 2 insertions(+), 1 deletion(-) mode change 100644 => 100755 tools/testing/selftests/net/rds/test.py diff --git a/tools/testing/selftests/net/rds/Makefile b/tools/testing/selftests/net/rds/Makefile index da9714bc7aad..cf30307a829b 100644 --- a/tools/testing/selftests/net/rds/Makefile +++ b/tools/testing/selftests/net/rds/Makefile @@ -4,9 +4,10 @@ all: @echo mk_build_dir="$(shell pwd)" > include.sh TEST_PROGS := run.sh \ - include.sh \ test.py +TEST_FILES := include.sh + EXTRA_CLEAN := /tmp/rds_logs include ../../lib.mk diff --git a/tools/testing/selftests/net/rds/test.py b/tools/testing/selftests/net/rds/test.py old mode 100644 new mode 100755 -- 2.39.3 (Apple Git-146)

4 weeks

3
3
0 0

2024

2023

2022

2021

2020

2019

2018

2017

Linux-kselftest-mirror September 2024