summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2023-04-14init: Mark start_kernel() __noreturnJosh Poimboeuf
Now that arch_call_rest_init() is __noreturn, mark its caller start_kernel() __noreturn. Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lore.kernel.org/r/7069acf026a195f26a88061227fba5a3b0337b9a.1681342859.git.jpoimboe@kernel.org
2023-04-14init: Mark [arch_call_]rest_init() __noreturnJosh Poimboeuf
In preparation for improving objtool's handling of weak noreturn functions, mark start_kernel(), arch_call_rest_init(), and rest_init() __noreturn. Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Nick Desaulniers <ndesaulniers@google.com> Link: https://lore.kernel.org/r/7194ed8a989a85b98d92e62df660f4a90435a723.1681342859.git.jpoimboe@kernel.org
2023-04-14objtool: Generate ORC data for __pfx codeJosh Poimboeuf
Allow unwinding from prefix code by copying the CFI from the starting instruction of the corresponding function. Even when the NOPs are replaced, they're still stack-invariant instructions so the same ORC entry can be reused everywhere. Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lkml.kernel.org/r/bc3344e51f3e87102f1301a0be0f72a7689ea4a4.1681331135.git.jpoimboe@kernel.org
2023-04-14x86/linkage: Fix padding for typed functionsJosh Poimboeuf
CFI typed functions are failing to get padded properly for CONFIG_CALL_PADDING. Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lkml.kernel.org/r/721f0da48d2a49fe907225711b8b76a2b787f9a8.1681331135.git.jpoimboe@kernel.org
2023-04-14objtool: Separate prefix code from stack validation codeJosh Poimboeuf
Simplify the prefix code by moving it after validate_reachable_instructions(). Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lkml.kernel.org/r/d7f31ac2de462d0cd7b1db01b7ecb525c057c8f6.1681331135.git.jpoimboe@kernel.org
2023-04-14objtool: Remove superfluous dead_end_function() checkJosh Poimboeuf
annotate_call_site() already sets 'insn->dead_end' for calls to dead end functions. Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lkml.kernel.org/r/5d603a301e9a8b1036b61503385907e154867ace.1681325924.git.jpoimboe@kernel.org
2023-04-14objtool: Add symbol iteration helpersJosh Poimboeuf
Add [sec_]for_each_sym() and use them. Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lkml.kernel.org/r/59023e5886ab125aa30702e633be7732b1acaa7e.1681325924.git.jpoimboe@kernel.org
2023-04-14objtool: Add WARN_INSN()Josh Poimboeuf
It's easier to use and also gives easy access to the instruction's containing function, which is useful for printing that function's symbol. It will also be useful in the future for rate-limiting and disassembly of warned functions. Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lkml.kernel.org/r/2eaa3155c90fba683d8723599f279c46025b75f3.1681325924.git.jpoimboe@kernel.org
2023-04-14scripts/objdump-func: Support multiple functionsJosh Poimboeuf
Allow specifying multiple functions on the cmdline. Note this removes the secret EXTRA_ARGS feature. While at it, spread out the awk to make it more readable. Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lkml.kernel.org/r/0bf5f4f5978660985037b24c6db49b114374eb4d.1681325924.git.jpoimboe@kernel.org
2023-04-14context_tracking: Fix KCSAN noinstr violationJosh Poimboeuf
With KCSAN enabled, even empty inline stubs can be out-of-lined. Force the context_tracking_guest_exit() stub inline. Fixes the following warnings: vmlinux.o: warning: objtool: vmx_vcpu_enter_exit+0x1be: call to context_tracking_guest_exit() leaves .noinstr.text section vmlinux.o: warning: objtool: svm_vcpu_enter_exit+0x85: call to context_tracking_guest_exit() leaves .noinstr.text section Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lore.kernel.org/r/dc93f45abdec90c171108b4b590b7fff5790963c.1681320026.git.jpoimboe@kernel.org
2023-04-14objtool: Add stackleak instrumentation to uaccess safe listJosh Poimboeuf
If a function has a large stack frame, the stackleak plugin adds a call to stackleak_track_stack() after the prologue. This function may be called in uaccess-enabled code. Add it to the uaccess safe list. Fixes the following warning: vmlinux.o: warning: objtool: kasan_report+0x12: call to stackleak_track_stack() with UACCESS enabled Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lkml.kernel.org/r/42e9b487ef89e9b237fd5220ad1c7cf1a2ad7eb8.1681320562.git.jpoimboe@kernel.org
2023-04-14Revert "objtool: Support addition to set CFA base"Josh Poimboeuf
Commit 468af56a7bba ("objtool: Support addition to set CFA base") was added as a preparatory patch for arm64 support, but that support never came. It triggers a false positive warning on x86, so just revert it for now. Fixes the following warning: vmlinux.o: warning: objtool: cdce925_regmap_i2c_write+0xdb: stack state mismatch: cfa1=4+120 cfa2=5+40 Fixes: 468af56a7bba ("objtool: Support addition to set CFA base") Reported-by: kernel test robot <lkp@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lore.kernel.org/oe-kbuild-all/202304080538.j5G6h1AB-lkp@intel.com/
2023-04-14lkdtm/stackleak: Fix noinstr violationJosh Poimboeuf
Fixes the following warning: vmlinux.o: warning: objtool: check_stackleak_irqoff+0x2b6: call to _printk() leaves .noinstr.text section Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lore.kernel.org/r/ee5209f53aa0a62aea58be18f2b78b17606779a6.1681320026.git.jpoimboe@kernel.org
2023-04-14sched: Fix KCSAN noinstr violationJosh Poimboeuf
With KCSAN enabled, end_of_stack() can get out-of-lined. Force it inline. Fixes the following warnings: vmlinux.o: warning: objtool: check_stackleak_irqoff+0x2b: call to end_of_stack() leaves .noinstr.text section Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lore.kernel.org/r/cc1b4d73d3a428a00d206242a68fdf99a934ca7b.1681320026.git.jpoimboe@kernel.org
2023-03-23x86,objtool: Split UNWIND_HINT_EMPTY in twoJosh Poimboeuf
Mark reported that the ORC unwinder incorrectly marks an unwind as reliable when the unwind terminates prematurely in the dark corners of return_to_handler() due to lack of information about the next frame. The problem is UNWIND_HINT_EMPTY is used in two different situations: 1) The end of the kernel stack unwind before hitting user entry, boot code, or fork entry 2) A blind spot in ORC coverage where the unwinder has to bail due to lack of information about the next frame The ORC unwinder has no way to tell the difference between the two. When it encounters an undefined stack state with 'end=1', it blindly marks the stack reliable, which can break the livepatch consistency model. Fix it by splitting UNWIND_HINT_EMPTY into UNWIND_HINT_UNDEFINED and UNWIND_HINT_END_OF_STACK. Reported-by: Mark Rutland <mark.rutland@arm.com> Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Acked-by: Steven Rostedt (Google) <rostedt@goodmis.org> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lore.kernel.org/r/fd6212c8b450d3564b855e1cb48404d6277b4d9f.1677683419.git.jpoimboe@kernel.org
2023-03-23x86,objtool: Separate unret validation from unwind hintsJosh Poimboeuf
The ENTRY unwind hint type is serving double duty as both an empty unwind hint and an unret validation annotation. Unret validation is unrelated to unwinding. Separate it out into its own annotation. Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lore.kernel.org/r/ff7448d492ea21b86d8a90264b105fbd0d751077.1677683419.git.jpoimboe@kernel.org
2023-03-23x86,objtool: Introduce ORC_TYPE_*Josh Poimboeuf
Unwind hints and ORC entry types are two distinct things. Separate them out more explicitly. Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lore.kernel.org/r/cc879d38fff8a43f8f7beb2fd56e35a5a384d7cd.1677683419.git.jpoimboe@kernel.org
2023-03-23objtool: Change UNWIND_HINT() argument orderJosh Poimboeuf
The most important argument is 'type', make that one first. Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lore.kernel.org/r/d994f8c29376c5618c75698df28fc03b52d3a868.1677683419.git.jpoimboe@kernel.org
2023-03-23objtool: Use relative pointers for annotationsJosh Poimboeuf
They produce the needed relocations while using half the space. Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lore.kernel.org/r/bed05c64e28200220c9b1754a2f3ce71f73076ea.1677683419.git.jpoimboe@kernel.org
2023-03-23objtool: Add objtool_types.hJosh Poimboeuf
Reduce the amount of header sync churn by splitting the shared objtool.h types into a new file. Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lore.kernel.org/r/dec622720851210ceafa12d4f4c5f9e73c832152.1677683419.git.jpoimboe@kernel.org
2023-03-08x86/cpu: Expose arch_cpu_idle_dead()'s prototype definitionPhilippe Mathieu-Daudé
Include <linux/cpu.h> to make sure arch_cpu_idle_dead() matches its prototype going forward. Inspired-by: Josh Poimboeuf <jpoimboe@kernel.org> Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Link: https://lore.kernel.org/r/20230214083857.50163-1-philmd@linaro.org Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
2023-03-08sched/idle: Mark arch_cpu_idle_dead() __noreturnJosh Poimboeuf
Before commit 076cbf5d2163 ("x86/xen: don't let xen_pv_play_dead() return"), in Xen, when a previously offlined CPU was brought back online, it unexpectedly resumed execution where it left off in the middle of the idle loop. There were some hacks to make that work, but the behavior was surprising as do_idle() doesn't expect an offlined CPU to return from the dead (in arch_cpu_idle_dead()). Now that Xen has been fixed, and the arch-specific implementations of arch_cpu_idle_dead() also don't return, give it a __noreturn attribute. This will cause the compiler to complain if an arch-specific implementation might return. It also improves code generation for both caller and callee. Also fixes the following warning: vmlinux.o: warning: objtool: do_idle+0x25f: unreachable instruction Reported-by: Paul E. McKenney <paulmck@kernel.org> Tested-by: Paul E. McKenney <paulmck@kernel.org> Link: https://lore.kernel.org/r/60d527353da8c99d4cf13b6473131d46719ed16d.1676358308.git.jpoimboe@kernel.org Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
2023-03-08sched/idle: Make sure weak version of arch_cpu_idle_dead() doesn't returnJosh Poimboeuf
arch_cpu_idle_dead() should never return. Make it so. Link: https://lore.kernel.org/r/cf5ad95eef50f7704bb30e7770c59bfe23372af7.1676358308.git.jpoimboe@kernel.org Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
2023-03-08xtensa/cpu: Mark cpu_die() __noreturnJosh Poimboeuf
cpu_die() doesn't return. Annotate it as such. By extension this also makes arch_cpu_idle_dead() noreturn. Acked-by: Max Filippov <jcmvbkbc@gmail.com> Link: https://lore.kernel.org/r/ad801544cab7c26a0f3bbf7cfefb67303f4cd866.1676358308.git.jpoimboe@kernel.org Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
2023-03-08xtensa/cpu: Make sure cpu_die() doesn't returnJosh Poimboeuf
cpu_die() doesn't return. Make that more explicit with a BUG(). BUG() is preferable to unreachable() because BUG() is a more explicit failure mode and avoids undefined behavior like falling off the edge of the function into whatever code happens to be next. Link: https://lore.kernel.org/r/cca346b5c87693499e630291d78fb0bf12c24290.1676358308.git.jpoimboe@kernel.org Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
2023-03-08x86/cpu: Mark play_dead() __noreturnJosh Poimboeuf
play_dead() doesn't return. Annotate it as such. By extension this also makes arch_cpu_idle_dead() noreturn. Link: https://lore.kernel.org/r/f3a069e6869c51ccfdda656b76882363bc9fcfa4.1676358308.git.jpoimboe@kernel.org Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
2023-03-08x86/cpu: Make sure play_dead() doesn't returnJosh Poimboeuf
After commit 076cbf5d2163 ("x86/xen: don't let xen_pv_play_dead() return"), play_dead() never returns. Make that more explicit with a BUG(). BUG() is preferable to unreachable() because BUG() is a more explicit failure mode and avoids undefined behavior like falling off the edge of the function into whatever code happens to be next. Link: https://lore.kernel.org/r/11e6ac1cf10f92967882926e3ac16287b50642f2.1676358308.git.jpoimboe@kernel.org Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
2023-03-08sparc/cpu: Mark cpu_play_dead() __noreturnJosh Poimboeuf
cpu_play_dead() doesn't return. Annotate it as such. By extension this also makes arch_cpu_idle_dead() noreturn. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Link: https://lore.kernel.org/r/847fdb53cc7124bb7c94e3e104e443a29be85184.1676358308.git.jpoimboe@kernel.org Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
2023-03-08sh/cpu: Expose arch_cpu_idle_dead()'s prototype definitionJosh Poimboeuf
Include <linux/cpu.h> to make sure arch_cpu_idle_dead() matches its prototype going forward. Link: https://lore.kernel.org/r/3d9661e97828fb464a48d4becf18f12604831903.1676358308.git.jpoimboe@kernel.org Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
2023-03-08sh/cpu: Mark play_dead() __noreturnJosh Poimboeuf
play_dead() doesn't return. Annotate it as such. By extension this also makes arch_cpu_idle_dead() noreturn. Link: https://lore.kernel.org/r/03549a74fad9f73576d57e6fc0b5102322f9cff4.1676358308.git.jpoimboe@kernel.org Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
2023-03-08sh/cpu: Make sure play_dead() doesn't returnJosh Poimboeuf
play_dead() doesn't return. Make that more explicit with a BUG(). BUG() is preferable to unreachable() because BUG() is a more explicit failure mode and avoids undefined behavior like falling off the edge of the function into whatever code happens to be next. Link: https://lore.kernel.org/r/d0c3ff5349adfe8fd227acc236ae2c278a05eb4c.1676358308.git.jpoimboe@kernel.org Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
2023-03-08powerpc/cpu: Mark start_secondary_resume() __noreturnJosh Poimboeuf
start_secondary_resume() doesn't return. Annotate it as such. By extension this also makes arch_cpu_idle_dead() noreturn. Acked-by: Michael Ellerman <mpe@ellerman.id.au> (powerpc) Reviewed-by: Christophe Leroy <christophe.leroy@csgroup.eu> Link: https://lore.kernel.org/r/b6b2141f832d8cd8ade65f190d04b011cda5f9bb.1676358308.git.jpoimboe@kernel.org Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
2023-03-08mips/cpu: Mark play_dead() __noreturnJosh Poimboeuf
play_dead() doesn't return. Annotate it as such. By extension this also makes arch_cpu_idle_dead() noreturn. Acked-by: Florian Fainelli <f.fainelli@gmail.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Link: https://lore.kernel.org/r/2897b51a9b8beb5b594fe66fb1d3a479ddd2a0e2.1676358308.git.jpoimboe@kernel.org Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
2023-03-08mips/cpu: Make sure play_dead() doesn't returnJosh Poimboeuf
play_dead() doesn't return. Make that more explicit with a BUG(). BUG() is preferable to unreachable() because BUG() is a more explicit failure mode and avoids undefined behavior like falling off the edge of the function into whatever code happens to be next. Acked-by: Florian Fainelli <f.fainelli@gmail.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Link: https://lore.kernel.org/r/b195e4da190bb06b7d4af15d66ce6129e2347630.1676358308.git.jpoimboe@kernel.org Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
2023-03-08mips/cpu: Expose play_dead()'s prototype definitionJosh Poimboeuf
Include <asm/smp.h> to make sure play_dead() matches its prototype going forward. Acked-by: Florian Fainelli <f.fainelli@gmail.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Link: https://lkml.kernel.org/r/20230216184249.ogaqsaykottpxtcb@treble Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
2023-03-06loongarch/cpu: Mark play_dead() __noreturnJosh Poimboeuf
play_dead() doesn't return. Annotate it as such. By extension this also makes arch_cpu_idle_dead() noreturn. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Link: https://lore.kernel.org/r/4da55acfdec8a9132c4e21ffb7edb1f846841193.1676358308.git.jpoimboe@kernel.org Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
2023-03-06loongarch/cpu: Make sure play_dead() doesn't returnJosh Poimboeuf
play_dead() doesn't return. Make that more explicit with a BUG(). BUG() is preferable to unreachable() because BUG() is a more explicit failure mode and avoids undefined behavior like falling off the edge of the function into whatever code happens to be next. Link: https://lore.kernel.org/r/21245d687ffeda34dbcf04961a2df3724f04f7c8.1676358308.git.jpoimboe@kernel.org Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
2023-03-06ia64/cpu: Mark play_dead() __noreturnJosh Poimboeuf
play_dead() doesn't return. Annotate it as such. By extension this also makes arch_cpu_idle_dead() noreturn. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Link: https://lore.kernel.org/r/7575bb38417bd8bcb5be980443f99cab29319342.1676358308.git.jpoimboe@kernel.org Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
2023-03-06csky/cpu: Make sure arch_cpu_idle_dead() doesn't returnJosh Poimboeuf
arch_cpu_idle_dead() doesn't return. Make that more explicit with a BUG(). BUG() is preferable to unreachable() because BUG() is a more explicit failure mode and avoids undefined behavior like falling off the edge of the function into whatever code happens to be next. Acked-by: Guo Ren <guoren@kernel.org> Link: https://lore.kernel.org/r/1e9ecc3d248e82973e80bc336fc9f97e3ba2708d.1676358308.git.jpoimboe@kernel.org Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
2023-03-06arm64/cpu: Mark cpu_die() __noreturnJosh Poimboeuf
cpu_die() doesn't return. Annotate it as such. By extension this also makes arch_cpu_idle_dead() noreturn. Acked-by: Mark Rutland <mark.rutland@arm.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Link: https://lkml.kernel.org/r/20230216184157.4hup6y6mmspr2kll@treble Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
2023-03-06arm/cpu: Add unreachable() to arch_cpu_idle_dead()Josh Poimboeuf
arch_cpu_idle_dead() doesn't return. Make that visible to the compiler with an unreachable() code annotation. Link: https://lkml.kernel.org/r/20230216183851.s5bnvniomq44rytu@treble Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
2023-03-06alpha/cpu: Make sure arch_cpu_idle_dead() doesn't returnJosh Poimboeuf
arch_cpu_idle_dead() doesn't return. Make that more explicit with a BUG(). BUG() is preferable to unreachable() because BUG() is a more explicit failure mode and avoids undefined behavior like falling off the edge of the function into whatever code happens to be next. Link: https://lore.kernel.org/r/19ffef09a175fecb783abcd44d6bcfeade2857eb.1676358308.git.jpoimboe@kernel.org Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
2023-03-06alpha/cpu: Expose arch_cpu_idle_dead()'s prototype declarationJosh Poimboeuf
Include <linux/cpu.h> to make sure arch_cpu_idle_dead() matches its prototype going forward. Link: https://lore.kernel.org/r/b0405c2ac5686303b6026e1ac27cfd769b21a7d0.1676358308.git.jpoimboe@kernel.org Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
2023-03-05Linux 6.3-rc1v6.3-rc1Linus Torvalds
2023-03-05cpumask: re-introduce constant-sized cpumask optimizationsLinus Torvalds
Commit aa47a7c215e7 ("lib/cpumask: deprecate nr_cpumask_bits") resulted in the cpumask operations potentially becoming hugely less efficient, because suddenly the cpumask was always considered to be variable-sized. The optimization was then later added back in a limited form by commit 6f9c07be9d02 ("lib/cpumask: add FORCE_NR_CPUS config option"), but that FORCE_NR_CPUS option is not useful in a generic kernel and more of a special case for embedded situations with fixed hardware. Instead, just re-introduce the optimization, with some changes. Instead of depending on CPUMASK_OFFSTACK being false, and then always using the full constant cpumask width, this introduces three different cpumask "sizes": - the exact size (nr_cpumask_bits) remains identical to nr_cpu_ids. This is used for situations where we should use the exact size. - the "small" size (small_cpumask_bits) is the NR_CPUS constant if it fits in a single word and the bitmap operations thus end up able to trigger the "small_const_nbits()" optimizations. This is used for the operations that have optimized single-word cases that get inlined, notably the bit find and scanning functions. - the "large" size (large_cpumask_bits) is the NR_CPUS constant if it is an sufficiently small constant that makes simple "copy" and "clear" operations more efficient. This is arbitrarily set at four words or less. As a an example of this situation, without this fixed size optimization, cpumask_clear() will generate code like movl nr_cpu_ids(%rip), %edx addq $63, %rdx shrq $3, %rdx andl $-8, %edx callq memset@PLT on x86-64, because it would calculate the "exact" number of longwords that need to be cleared. In contrast, with this patch, using a MAX_CPU of 64 (which is quite a reasonable value to use), the above becomes a single movq $0,cpumask instruction instead, because instead of caring to figure out exactly how many CPU's the system has, it just knows that the cpumask will be a single word and can just clear it all. Note that this does end up tightening the rules a bit from the original version in another way: operations that set bits in the cpumask are now limited to the actual nr_cpu_ids limit, whereas we used to do the nr_cpumask_bits thing almost everywhere in the cpumask code. But if you just clear bits, or scan for bits, we can use the simpler compile-time constants. In the process, remove 'cpumask_complement()' and 'for_each_cpu_not()' which were not useful, and which fundamentally have to be limited to 'nr_cpu_ids'. Better remove them now than have somebody introduce use of them later. Of course, on x86-64 with MAXSMP there is no sane small compile-time constant for the cpumask sizes, and we end up using the actual CPU bits, and will generate the above kind of horrors regardless. Please don't use MAXSMP unless you really expect to have machines with thousands of cores. Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2023-03-05Merge tag 'v6.3-p2' of ↵Linus Torvalds
git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6 Pull crypto fix from Herbert Xu: "Fix a regression in the caam driver" * tag 'v6.3-p2' of git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6: crypto: caam - Fix edesc/iv ordering mixup
2023-03-05Merge tag 'x86-urgent-2023-03-05' of ↵Linus Torvalds
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip Pull x86 updates from Thomas Gleixner: "A small set of updates for x86: - Return -EIO instead of success when the certificate buffer for SEV guests is not large enough - Allow STIPB to be enabled with legacy IBSR. Legacy IBRS is cleared on return to userspace for performance reasons, but the leaves user space vulnerable to cross-thread attacks which STIBP prevents. Update the documentation accordingly" * tag 'x86-urgent-2023-03-05' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: virt/sev-guest: Return -EIO if certificate buffer is not large enough Documentation/hw-vuln: Document the interaction between IBRS and STIBP x86/speculation: Allow enabling STIBP with legacy IBRS
2023-03-05Merge tag 'irq-urgent-2023-03-05' of ↵Linus Torvalds
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip Pull irq updates from Thomas Gleixner: "A set of updates for the interrupt susbsystem: - Prevent possible NULL pointer derefences in irq_data_get_affinity_mask() and irq_domain_create_hierarchy() - Take the per device MSI lock before invoking code which relies on it being hold - Make sure that MSI descriptors are unreferenced before freeing them. This was overlooked when the platform MSI code was converted to use core infrastructure and results in a fals positive warning - Remove dead code in the MSI subsystem - Clarify the documentation for pci_msix_free_irq() - More kobj_type constification" * tag 'irq-urgent-2023-03-05' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: genirq/msi, platform-msi: Ensure that MSI descriptors are unreferenced genirq/msi: Drop dead domain name assignment irqdomain: Add missing NULL pointer check in irq_domain_create_hierarchy() genirq/irqdesc: Make kobj_type structures constant PCI/MSI: Clarify usage of pci_msix_free_irq() genirq/msi: Take the per-device MSI lock before validating the control structure genirq/ipi: Fix NULL pointer deref in irq_data_get_affinity_mask()
2023-03-05Merge tag 'pull-misc' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfsLinus Torvalds
Pull vfs update from Al Viro: "Adding Christian Brauner as VFS co-maintainer" * tag 'pull-misc' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs: Adding VFS co-maintainer
2023-03-05Merge tag 'pull-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfsLinus Torvalds
Pull VM_FAULT_RETRY fixes from Al Viro: "Some of the page fault handlers do not deal with the following case correctly: - handle_mm_fault() has returned VM_FAULT_RETRY - there is a pending fatal signal - fault had happened in kernel mode Correct action in such case is not "return unconditionally" - fatal signals are handled only upon return to userland and something like copy_to_user() would end up retrying the faulting instruction and triggering the same fault again and again. What we need to do in such case is to make the caller to treat that as failed uaccess attempt - handle exception if there is an exception handler for faulting instruction or oops if there isn't one. Over the years some architectures had been fixed and now are handling that case properly; some still do not. This series should fix the remaining ones. Status: - m68k, riscv, hexagon, parisc: tested/acked by maintainers. - alpha, sparc32, sparc64: tested locally - bug has been reproduced on the unpatched kernel and verified to be fixed by this series. - ia64, microblaze, nios2, openrisc: build, but otherwise completely untested" * tag 'pull-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs: openrisc: fix livelock in uaccess nios2: fix livelock in uaccess microblaze: fix livelock in uaccess ia64: fix livelock in uaccess sparc: fix livelock in uaccess alpha: fix livelock in uaccess parisc: fix livelock in uaccess hexagon: fix livelock in uaccess riscv: fix livelock in uaccess m68k: fix livelock in uaccess
generated by cgit (git 2.34.1) at 2025-10-02 02:17:44 +0000