| Message ID | 20250212-arm-generic-entry-v4-0-a457ff0a61d6@linaro.org (mailing list archive) |
|---|---|
| Headers | show |
| Series | ARM: Switch to generic entry | expand |
On Wed, Feb 12, 2025 at 12:22:54PM +0100, Linus Walleij wrote: [...] > - Tested some ptrace/strace obviously, such as issuing > several instances of "ptrace find /" and let this scroll > by in the terminal over some 10 minutes or so. Could you also run the strace test suite, please? Given that it used to catch quite a few regressions in the past, it could be useful in this case as well.
On Fri, Feb 14, 2025 at 2:03 AM Dmitry V. Levin <ldv@strace.io> wrote: > On Wed, Feb 12, 2025 at 12:22:54PM +0100, Linus Walleij wrote: > [...] > > - Tested some ptrace/strace obviously, such as issuing > > several instances of "ptrace find /" and let this scroll > > by in the terminal over some 10 minutes or so. > > Could you also run the strace test suite, please? Given that it used to > catch quite a few regressions in the past, it could be useful in this case > as well. Sure, where can I find this test suite? Yours, Linus Walleij
On Fri, Feb 14, 2025 at 11:47:30AM +0100, Linus Walleij wrote: > On Fri, Feb 14, 2025 at 2:03 AM Dmitry V. Levin <ldv@strace.io> wrote: > > On Wed, Feb 12, 2025 at 12:22:54PM +0100, Linus Walleij wrote: > > [...] > > > - Tested some ptrace/strace obviously, such as issuing > > > several instances of "ptrace find /" and let this scroll > > > by in the terminal over some 10 minutes or so. > > > > Could you also run the strace test suite, please? Given that it used to > > catch quite a few regressions in the past, it could be useful in this case > > as well. > > Sure, where can I find this test suite? It's a part of strace, you can find it e.g. at https://github.com/strace/strace To build and run it one can roughly do ./bootstrap && ./configure && make -j`nproc` && make -j`nproc check
On Fri, Feb 14, 2025 at 12:05 PM Dmitry V. Levin <ldv@strace.io> wrote: > > Sure, where can I find this test suite? > > It's a part of strace, you can find it e.g. at > https://github.com/strace/strace > > To build and run it one can roughly do > ./bootstrap && ./configure && make -j`nproc` && make -j`nproc check make check produces some test failures on v6.14-rc1 on ARM even before I apply the generic entry: FAIL: filtering_syscall-syntax.test FAIL: qual_fault-syscall.test FAIL: qual_fault.test FAIL: strace--tips-full.test FAIL: strace-r.test ============================================================================ Testsuite summary for strace 6.13.0.27.bbda4 ============================================================================ # TOTAL: 1409 # PASS: 1106 # SKIP: 298 # XFAIL: 0 # FAIL: 5 # XPASS: 0 # ERROR: 0 ============================================================================ But I create more fails after my patch set ... Some have to do with fast syscall restart (I need to look into this). Then there is the fact that I had to add the PTRACE_SYSEMU and PTRACE_SYSEMU_SINGLESTEP as stubs returning -EIO (modeled after UM) to use generic entry. Do you think this affects the results? Is there a way to run a single test? I tried to check the docs but nothing obvious to me ... I guess I'm not used to this unit-tester. Yours, Linus Walleij
On Thu, Feb 20, 2025 at 03:04:22PM +0100, Linus Walleij wrote: > On Fri, Feb 14, 2025 at 12:05 PM Dmitry V. Levin <ldv@strace.io> wrote: > > > > Sure, where can I find this test suite? > > > > It's a part of strace, you can find it e.g. at > > https://github.com/strace/strace > > > > To build and run it one can roughly do > > ./bootstrap && ./configure && make -j`nproc` && make -j`nproc check > > make check produces some test failures on v6.14-rc1 on ARM > even before I apply the generic entry: > > FAIL: filtering_syscall-syntax.test > FAIL: qual_fault-syscall.test > FAIL: qual_fault.test > FAIL: strace--tips-full.test > FAIL: strace-r.test This is surprising, no tests are currently known to fail on arm. > Then there is the fact that I had to add the PTRACE_SYSEMU and > PTRACE_SYSEMU_SINGLESTEP as stubs returning -EIO > (modeled after UM) to use generic entry. Do you think this affects > the results? No, strace doesn't use PTRACE_SYSEMU* and doesn't test it either. > Is there a way to run a single test? I tried to check the docs but > nothing obvious to me ... I guess I'm not used to this unit-tester. Sure, it's a regular automake-based test suite, so you can do something like $ make -s -k check VERBOSE=1 TESTS='filtering_syscall-syntax.test qual_fault-syscall.test qual_fault.test'
First non-RFC version. This patch series converts a slew of ARM assembly into the corresponding C code, step by step moving the codebase closer to the expectations of the generic entry code, and as a last step switches ARM over to the generic entry code, and an RFC patch fixes a bunch of warnings from lockdep and the context tracker. This was inspired by Jinjie Ruans similar work for ARM64. The low-level assembly calls into arch/arm/kernel/syscall.c to invoke syscalls from userspace, and to the functions listed in arch/arm/kernel/entry.c for any other transitions to and from userspace. Looking at these functions and the call sites in the assembly on the final result should give a pretty good idea about how this works, and what the generic entry expects from an architecture. This was successfully booted on ARMv7m as well: the v7m avoids the interrupt path in the generic entry, because it never called the context tracker to begin with. It uses the common path for syscalls however and this works just fine. Adding proper context tracking to the ARMv7m IRQs is probably a good idea but a separate issue altogether. There is a git branch you can pull in and test (v6.14-rc1 based): https://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-integrator.git/log/?h=b4/arm-generic-entry-v6.14-rc1 Upsides: - Same code paths as x86, S390, RISCV, Loongarch and probably soon ARM64 is used for the ARM systems. This includes some instrumentation stubs helping out with things we haven't even started to look at such as kmsan and live patching (!). - By introducing the new callbacks to C, we can move away from the deprecated (and I think partly unmaintained) context tracking mechanism for RCU (user_exit_callable(), user_enter_callable()) in favor of what everyone else is using, i.e. calling rcu_irq_enter_check_tick() on IRQ entry. If we do not go with this patch set we can perhaps look into a separate patch just switching ARM32 to the new context tracking, as tests show the performance impact appears negligible for this. - I think also lockdep is now behaving more according to expectations (the lockdep calls in ARM64 and generic entry seems different and more fine-granular from the ARM32 code) and the three warnings I see on Vexpress boots with mainline goes away after this patch set, but I am no expert in lockdep so I cannot really tell if this is a real improvement. The patches does make ARM lockdep clean. Downsides: - I had to remove the "fast syscall restart" from Al Viro. I don't know how much it will affect performance, but if this is something we must have, let's try to make the solution generic, i.e. add fast syscall restart in the generic entry code. - The "superfast return to userspace" using just very small assembly snippets to get back to userspace on e.g. IRQs if and only if no instrumentation was compiled in, is no longer possible, since we unconditionally call into code written in C. I *think* this accounts for the majority of the ~3-4% performance impact (see measurements below). Both downsides are more or less unavoidable side effects if you just want to use the non-deprecated context tracking, as that involves calling into C from every exception, without exceptions. Testing: - Booted into Versatile Express QEMU (ARMv7), Ux500 full graphic UI (PostmarketOS Phosh, ARMv7 on hardware, Gemini ARMv4 on hardware. No special issues. - Tested some ptrace/strace obviously, such as issuing several instances of "ptrace find /" and let this scroll by in the terminal over some 10 minutes or so. - Turned on RCU torture tests and ran for a while. Seems stable and the test outputs look normal. - Ran stress-ng, works fine. - Booted with "lockdep" (CONFIG_PROVE_LOCKING). The ARM32 mainline produce 3 warnings at boot and those go away after these patches. I haven't looked closer at what it was that I inadvertedly fixed here, but I suspect the current context tracking has the same issues as what I fix in the RFC patch. Performance impact: The changes were tested using the standard syscall overhead testing oneliner: perf bench syscall all This executes 10,000,000 getppid() in sequence and measures the time taken for this to complete. The numbers vary a bit but they are consistent. In QEMU I tested with Vexpress and two CPU cores (-M vexpress-a15 -m 2G -smp cpus=2). DRM graphics and framebuffer was activated to give a bit of background IRQ activity (vsync interrupts). I ran the perf command three times on each configuration, and picked the one iteration where the original code performed the best, and the one where the patches kernel performed the worst, to get a worst-case comparison. v6.14-rc1 vexpress_defconfig, best invocation: Total time: 146.546 [sec] 14.654698 usecs/op 68,237 ops/sec v6.14-rc1 vexpress_defconfig, and this patch set, worst invocation: Total time: 156.263 [sec] 15.626398 usecs/op 63,994 ops/sec Here we see a performance degradation of around 6-7% operations/sec for a vexpress dualcore defconfig in the best vs worst case. (This isn't statistically correct, the effect is likely smaller.) Debians stock kernel was noticably faster, so I investigated what causes this. It turns out that the big performance hog for syscalls is actually PAN which cause an order of magnitude syscall performance decrease, and I think Debian armhf simply turns this off. Consistent tests with PAN disabled also see around 6-7% on that performance figure. To conclude if any of this was due to the new context tracking, at one point I tested to patch back the old context tracking on top of generic entry. This is hardly something that can be recommended, and anyway showed no noticeable overhead difference. Open questions: - I need to test with an OABI rootfs. Signed-off-by: Linus Walleij <linus.walleij@linaro.org> --- Changes in v4: - Rebased on v6.14-rc1, marked non-RFC. - Tested on ARMv7m, it works. - Fixed a bug where I missed to handle syscall "-1" which when tracing means "skip syscall". This took some time to find, taking up much of my debug time despite being so obvious :/ - Added stubs for PTRACE_SYSEMU and PTRACE_SYSEMU_SINGLESTEP. - No feedback on the "fast syscall restart" so I conclude that this is some ARM oddity, if it is needed for performance (i.e. a workload constantly restarting syscalls) we should look at recreating it inside the generic entry code. - After discussing with Ard about the IRQ stacks, altered the irqstack handling to just assume IRQ stack or overflow stack is in use if we are not on the main thread stack. - Unmark the patch to block IRQs in early IRQ context as "RFC": when doing proper context tracking this is likely plain necessary. Block IRQs in the early assembly entry directly in CPSR instead of later in the exception handler. - New cleanup patch in the tail of the patch series. - Link to v3: https://lore.kernel.org/r/20250107-arm-generic-entry-v3-0-4e5f3c15db2d@linaro.org Changes in v3: - Rewrote the code in entry.c so the IRQ handler saves pt_regs calls IRQ handler (including switching to IRQ stack!) and restores pt_regs in one function instead of one entry and one exit function. This is what every other arch using generic entry is doing, and we should do it too. - The rewrite solved the caveat warnings from the previous patch set which was blatantly not SMP safe :/ - Rewrite the data abort and prefetch abort handlers in a separate patch which we may squash in the end, but this makes the patch set easier to review. - Drop a pointless patch rewriting the NMI handlers in C, it's better to just patch into the end result in the last patch, as we're replacing handle_fiq_as_nmi(). - Syscall C invocations have to be tagged __ADDRESSABLE() in order not to upset KCFI: the file is only referenced in both ends by assembly so we need to point this out to the compiler. - Link to v2: https://lore.kernel.org/r/20241029-arm-generic-entry-v2-0-573519abef38@linaro.org Changes in v2: - Performance impact measurements have been provided. - Link to v1: https://lore.kernel.org/r/20241010-arm-generic-entry-v1-0-b94f451d087b@linaro.org --- Linus Walleij (31): ARM: Prepare includes for generic entry ARM: ptrace: Split report_syscall() ARM: entry: Skip ret_slow_syscall label ARM: process: Rewrite ret_from_fork i C ARM: process: Remove local restart ARM: entry: Invoke syscalls using C ARM: entry: Rewrite two asm calls in C ARM: entry: Move trace entry to C function ARM: entry: save the syscall sp in thread_info ARM: entry: move all tracing invocation to C ARM: entry: Merge the common and trace entry code ARM: entry: Rename syscall invocation ARM: entry: Create user_mode_enter/exit ARM: entry: Drop trace argument from usr_entry macro ARM: entry: Separate call path for syscall SWI entry ARM: entry: Drop argument to asm_irqentry macros ARM: entry: Implement syscall_exit_to_user_mode() ARM: entry: Drop the superfast ret_fast_syscall ARM: entry: Remove fast and offset register restore ARM: entry: Untangle ret_fast_syscall/to_user ARM: entry: Do not double-call exit functions ARM: entry: Move work processing to C ARM: entry: Stop exiting syscalls like IRQs ARM: entry: Complete syscall and IRQ transition to C ARM: entry: Create irqentry calls from kernel mode ARM: entry: Move in-kernel hardirq tracing to C ARM: irq: Add irqstack helper ARM: entry: Convert to generic entry ARM: entry: Handle dabt, pabt, and und as interrupts ARM: entry: Block IRQs in early IRQ context ARM: entry: Straighten syscall returns arch/arm/Kconfig | 1 + arch/arm/include/asm/entry-common.h | 66 +++++++++++ arch/arm/include/asm/entry.h | 14 +++ arch/arm/include/asm/ptrace.h | 8 +- arch/arm/include/asm/signal.h | 4 - arch/arm/include/asm/stacktrace.h | 2 +- arch/arm/include/asm/switch_to.h | 4 + arch/arm/include/asm/syscall.h | 7 ++ arch/arm/include/asm/thread_info.h | 18 +-- arch/arm/include/asm/traps.h | 5 +- arch/arm/include/uapi/asm/ptrace.h | 2 + arch/arm/kernel/Makefile | 5 +- arch/arm/kernel/asm-offsets.c | 1 + arch/arm/kernel/entry-armv.S | 82 ++++---------- arch/arm/kernel/entry-common.S | 218 +++++++++++++----------------------- arch/arm/kernel/entry-header.S | 100 ++--------------- arch/arm/kernel/entry.c | 120 ++++++++++++++++++++ arch/arm/kernel/irq.c | 6 + arch/arm/kernel/irq.h | 2 + arch/arm/kernel/process.c | 25 ++++- arch/arm/kernel/ptrace.c | 81 +------------- arch/arm/kernel/signal.c | 57 +--------- arch/arm/kernel/syscall.c | 37 ++++++ arch/arm/kernel/traps.c | 30 +---- arch/arm/mm/abort-ev4.S | 2 +- arch/arm/mm/abort-ev4t.S | 2 +- arch/arm/mm/abort-ev5t.S | 4 +- arch/arm/mm/abort-ev5tj.S | 6 +- arch/arm/mm/abort-ev6.S | 2 +- arch/arm/mm/abort-ev7.S | 2 +- arch/arm/mm/abort-lv4t.S | 36 +++--- arch/arm/mm/abort-macro.S | 2 +- arch/arm/mm/abort-nommu.S | 2 +- arch/arm/mm/fault.c | 4 +- arch/arm/mm/fault.h | 8 +- arch/arm/mm/pabort-legacy.S | 2 +- arch/arm/mm/pabort-v6.S | 2 +- arch/arm/mm/pabort-v7.S | 2 +- 38 files changed, 456 insertions(+), 515 deletions(-) --- base-commit: 2014c95afecee3e76ca4a56956a936e23283f05b change-id: 20240903-arm-generic-entry-ada145378bbe Best regards,