On Fri, Apr 05, 2024 at 03:54:45PM +0200, Klara Modin wrote:
> Hi,
>
> On 2024-03-21 17:36, Suren Baghdasaryan wrote:
> > Introduce CONFIG_MEM_ALLOC_PROFILING which provides definitions to easily
> > instrument memory allocators. It registers an "alloc_tags" codetag type
> > with /proc/allocinfo interface to output allocation tag information when
> > the feature is enabled.
> > CONFIG_MEM_ALLOC_PROFILING_DEBUG is provided for debugging the memory
> > allocation profiling instrumentation.
> > Memory allocation profiling can be enabled or disabled at runtime using
> > /proc/sys/vm/mem_profiling sysctl when CONFIG_MEM_ALLOC_PROFILING_DEBUG=n.
> > CONFIG_MEM_ALLOC_PROFILING_ENABLED_BY_DEFAULT enables memory allocation
> > profiling by default.
> >
> > Signed-off-by: Suren Baghdasaryan <surenb@xxxxxxxxxx>
> > Co-developed-by: Kent Overstreet <kent.overstreet@xxxxxxxxx>
> > Signed-off-by: Kent Overstreet <kent.overstreet@xxxxxxxxx>
>
> With this commit (9e2dcefa791e9d14006b360fba3455510fd3325d in
> next-20240404), randconfig with KCONFIG_SEED=0xE6264236 fails to build
> with the attached error. The following patch fixes the build error for me,
> but I don't know if it's correct.
Looks good - if you sound out an official patch I'll ack it.
>
> Kind regards,
> Klara Modin
>
> diff --git a/include/linux/alloc_tag.h b/include/linux/alloc_tag.h
> index 100ddf66eb8e..1c765d80298b 100644
> --- a/include/linux/alloc_tag.h
> +++ b/include/linux/alloc_tag.h
> @@ -12,6 +12,7 @@
> #include <asm/percpu.h>
> #include <linux/cpumask.h>
> #include <linux/static_key.h>
> +#include <linux/irqflags.h>
>
> struct alloc_tag_counters {
> u64 bytes;
>
> > diff --git a/include/linux/alloc_tag.h b/include/linux/alloc_tag.h
> > new file mode 100644
> > index 000000000000..b970ff1c80dc
> > --- /dev/null
> > +++ b/include/linux/alloc_tag.h
> > @@ -0,0 +1,145 @@
> > +/* SPDX-License-Identifier: GPL-2.0 */
> > +/*
> > + * allocation tagging
> > + */
> > +#ifndef _LINUX_ALLOC_TAG_H
> > +#define _LINUX_ALLOC_TAG_H
> > +
> > +#include <linux/bug.h>
> > +#include <linux/codetag.h>
> > +#include <linux/container_of.h>
> > +#include <linux/preempt.h>
> > +#include <asm/percpu.h>
> > +#include <linux/cpumask.h>
> > +#include <linux/static_key.h>
> > +
> > +struct alloc_tag_counters {
> > + u64 bytes;
> > + u64 calls;
> > +};
> > +
> > +/*
> > + * An instance of this structure is created in a special ELF section at every
> > + * allocation callsite. At runtime, the special section is treated as
> > + * an array of these. Embedded codetag utilizes codetag framework.
> > + */
> > +struct alloc_tag {
> > + struct codetag ct;
> > + struct alloc_tag_counters __percpu *counters;
> > +} __aligned(8);
> > +
> > +#ifdef CONFIG_MEM_ALLOC_PROFILING
> > +
> > +static inline struct alloc_tag *ct_to_alloc_tag(struct codetag *ct)
> > +{
> > + return container_of(ct, struct alloc_tag, ct);
> > +}
> > +
> > +#ifdef ARCH_NEEDS_WEAK_PER_CPU
> > +/*
> > + * When percpu variables are required to be defined as weak, static percpu
> > + * variables can't be used inside a function (see comments for DECLARE_PER_CPU_SECTION).
> > + */
> > +#error "Memory allocation profiling is incompatible with ARCH_NEEDS_WEAK_PER_CPU"
> > +#endif
> > +
> > +#define DEFINE_ALLOC_TAG(_alloc_tag) \
> > + static DEFINE_PER_CPU(struct alloc_tag_counters, _alloc_tag_cntr); \
> > + static struct alloc_tag _alloc_tag __used __aligned(8) \
> > + __section("alloc_tags") = { \
> > + .ct = CODE_TAG_INIT, \
> > + .counters = &_alloc_tag_cntr };
> > +
> > +DECLARE_STATIC_KEY_MAYBE(CONFIG_MEM_ALLOC_PROFILING_ENABLED_BY_DEFAULT,
> > + mem_alloc_profiling_key);
> > +
> > +static inline bool mem_alloc_profiling_enabled(void)
> > +{
> > + return static_branch_maybe(CONFIG_MEM_ALLOC_PROFILING_ENABLED_BY_DEFAULT,
> > + &mem_alloc_profiling_key);
> > +}
> > +
> > +static inline struct alloc_tag_counters alloc_tag_read(struct alloc_tag *tag)
> > +{
> > + struct alloc_tag_counters v = { 0, 0 };
> > + struct alloc_tag_counters *counter;
> > + int cpu;
> > +
> > + for_each_possible_cpu(cpu) {
> > + counter = per_cpu_ptr(tag->counters, cpu);
> > + v.bytes += counter->bytes;
> > + v.calls += counter->calls;
> > + }
> > +
> > + return v;
> > +}
> > +
> > +#ifdef CONFIG_MEM_ALLOC_PROFILING_DEBUG
> > +static inline void alloc_tag_add_check(union codetag_ref *ref, struct alloc_tag *tag)
> > +{
> > + WARN_ONCE(ref && ref->ct,
> > + "alloc_tag was not cleared (got tag for %s:%u)\n",
> > + ref->ct->filename, ref->ct->lineno);
> > +
> > + WARN_ONCE(!tag, "current->alloc_tag not set");
> > +}
> > +
> > +static inline void alloc_tag_sub_check(union codetag_ref *ref)
> > +{
> > + WARN_ONCE(ref && !ref->ct, "alloc_tag was not set\n");
> > +}
> > +#else
> > +static inline void alloc_tag_add_check(union codetag_ref *ref, struct alloc_tag *tag) {}
> > +static inline void alloc_tag_sub_check(union codetag_ref *ref) {}
> > +#endif
> > +
> > +/* Caller should verify both ref and tag to be valid */
> > +static inline void __alloc_tag_ref_set(union codetag_ref *ref, struct alloc_tag *tag)
> > +{
> > + ref->ct = &tag->ct;
> > + /*
> > + * We need in increment the call counter every time we have a new
> > + * allocation or when we split a large allocation into smaller ones.
> > + * Each new reference for every sub-allocation needs to increment call
> > + * counter because when we free each part the counter will be decremented.
> > + */
> > + this_cpu_inc(tag->counters->calls);
> > +}
> > +
> > +static inline void alloc_tag_add(union codetag_ref *ref, struct alloc_tag *tag, size_t bytes)
> > +{
> > + alloc_tag_add_check(ref, tag);
> > + if (!ref || !tag)
> > + return;
> > +
> > + __alloc_tag_ref_set(ref, tag);
> > + this_cpu_add(tag->counters->bytes, bytes);
> > +}
> > +
> > +static inline void alloc_tag_sub(union codetag_ref *ref, size_t bytes)
> > +{
> > + struct alloc_tag *tag;
> > +
> > + alloc_tag_sub_check(ref);
> > + if (!ref || !ref->ct)
> > + return;
> > +
> > + tag = ct_to_alloc_tag(ref->ct);
> > +
> > + this_cpu_sub(tag->counters->bytes, bytes);
> > + this_cpu_dec(tag->counters->calls);
> > +
> > + ref->ct = NULL;
> > +}
> > +
> > +#else /* CONFIG_MEM_ALLOC_PROFILING */
> > +
> > +#define DEFINE_ALLOC_TAG(_alloc_tag)
> > +static inline bool mem_alloc_profiling_enabled(void) { return false; }
> > +static inline void alloc_tag_add(union codetag_ref *ref, struct alloc_tag *tag,
> > + size_t bytes) {}
> > +static inline void alloc_tag_sub(union codetag_ref *ref, size_t bytes) {}
> > +
> > +#endif /* CONFIG_MEM_ALLOC_PROFILING */
> > +
> > +#endif /* _LINUX_ALLOC_TAG_H */
> In file included from ./arch/x86/include/asm/percpu.h:615,
> from ./arch/x86/include/asm/preempt.h:6,
> from ./include/linux/preempt.h:79,
> from ./include/linux/alloc_tag.h:11,
> from lib/alloc_tag.c:2:
> ./include/linux/alloc_tag.h: In function ‘__alloc_tag_ref_set’:
> ./include/asm-generic/percpu.h:155:9: error: implicit declaration of function ‘raw_local_irq_save’ [-Werror=implicit-function-declaration]
> 155 | raw_local_irq_save(__flags); \
> | ^~~~~~~~~~~~~~~~~~
> ./include/asm-generic/percpu.h:410:41: note: in expansion of macro ‘this_cpu_generic_to_op’
> 410 | #define this_cpu_add_8(pcp, val) this_cpu_generic_to_op(pcp, val, +=)
> | ^~~~~~~~~~~~~~~~~~~~~~
> ./include/linux/percpu-defs.h:368:25: note: in expansion of macro ‘this_cpu_add_8’
> 368 | case 8: stem##8(variable, __VA_ARGS__);break; \
> | ^~~~
> ./include/linux/percpu-defs.h:491:41: note: in expansion of macro ‘__pcpu_size_call’
> 491 | #define this_cpu_add(pcp, val) __pcpu_size_call(this_cpu_add_, pcp, val)
> | ^~~~~~~~~~~~~~~~
> ./include/linux/percpu-defs.h:501:41: note: in expansion of macro ‘this_cpu_add’
> 501 | #define this_cpu_inc(pcp) this_cpu_add(pcp, 1)
> | ^~~~~~~~~~~~
> ./include/linux/alloc_tag.h:106:9: note: in expansion of macro ‘this_cpu_inc’
> 106 | this_cpu_inc(tag->counters->calls);
> | ^~~~~~~~~~~~
> ./include/asm-generic/percpu.h:157:9: error: implicit declaration of function ‘raw_local_irq_restore’ [-Werror=implicit-function-declaration]
> 157 | raw_local_irq_restore(__flags); \
> | ^~~~~~~~~~~~~~~~~~~~~
> ./include/asm-generic/percpu.h:410:41: note: in expansion of macro ‘this_cpu_generic_to_op’
> 410 | #define this_cpu_add_8(pcp, val) this_cpu_generic_to_op(pcp, val, +=)
> | ^~~~~~~~~~~~~~~~~~~~~~
> ./include/linux/percpu-defs.h:368:25: note: in expansion of macro ‘this_cpu_add_8’
> 368 | case 8: stem##8(variable, __VA_ARGS__);break; \
> | ^~~~
> ./include/linux/percpu-defs.h:491:41: note: in expansion of macro ‘__pcpu_size_call’
> 491 | #define this_cpu_add(pcp, val) __pcpu_size_call(this_cpu_add_, pcp, val)
> | ^~~~~~~~~~~~~~~~
> ./include/linux/percpu-defs.h:501:41: note: in expansion of macro ‘this_cpu_add’
> 501 | #define this_cpu_inc(pcp) this_cpu_add(pcp, 1)
> | ^~~~~~~~~~~~
> ./include/linux/alloc_tag.h:106:9: note: in expansion of macro ‘this_cpu_inc’
> 106 | this_cpu_inc(tag->counters->calls);
> | ^~~~~~~~~~~~
> cc1: some warnings being treated as errors
> make[3]: *** [scripts/Makefile.build:244: lib/alloc_tag.o] Error 1
> make[2]: *** [scripts/Makefile.build:485: lib] Error 2
> make[1]: *** [/home/klara/git/linux/Makefile:1919: .] Error 2
> make: *** [Makefile:240: __sub-make] Error 2
> # bad: [2b3d5988ae2cb5cd945ddbc653f0a71706231fdd] Add linux-next specific files for 20240404
> git bisect start 'next/master'
> # status: waiting for good commit(s), bad commit known
> # good: [39cd87c4eb2b893354f3b850f916353f2658ae6f] Linux 6.9-rc2
> git bisect good 39cd87c4eb2b893354f3b850f916353f2658ae6f
> # bad: [cc7b62666779616ff52d389a344ffe2c041e36e2] Merge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/bluetooth/bluetooth-next.git
> git bisect bad cc7b62666779616ff52d389a344ffe2c041e36e2
> # bad: [d6b7dd0f8d84f9fdf2af65fceb608e3206276e81] Merge branch 'for-next' of git://git.kernel.org/pub/scm/linux/kernel/git/qcom/linux.git
> git bisect bad d6b7dd0f8d84f9fdf2af65fceb608e3206276e81
> # bad: [ad6a31687713a8f12165e730e0eb6e0de3beae56] Merge branch 'mm-everything' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
> git bisect bad ad6a31687713a8f12165e730e0eb6e0de3beae56
> # good: [59266d9886adb5c9e240129ccc606727fd3a881d] Merge branch 'fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/riscv/linux.git
> git bisect good 59266d9886adb5c9e240129ccc606727fd3a881d
> # bad: [085e5fe7388cf36ab5c02d91022229e5fade5b30] mm: merge folio_is_secretmem() and folio_fast_pin_allowed() into gup_fast_folio_allowed()
> git bisect bad 085e5fe7388cf36ab5c02d91022229e5fade5b30
> # bad: [f6a61baa9139d174170acdae8667b3246ce44db6] lib: add memory allocations report in show_mem()
> git bisect bad f6a61baa9139d174170acdae8667b3246ce44db6
> # good: [302519d9e80a7fbf2cf8d0b8961d491af648759f] asm-generic/io.h: kill vmalloc.h dependency
> git bisect good 302519d9e80a7fbf2cf8d0b8961d491af648759f
> # bad: [e6942003e682e3883847459c3d07e23c796a2782] mm: create new codetag references during page splitting
> git bisect bad e6942003e682e3883847459c3d07e23c796a2782
> # good: [ed97151dec736c1541bfac2b801108d54ebee5bc] lib: code tagging module support
> git bisect good ed97151dec736c1541bfac2b801108d54ebee5bc
> # bad: [95767bde5020afefef4205b60e71f4ebf96da74e] lib: introduce early boot parameter to avoid page_ext memory overhead
> git bisect bad 95767bde5020afefef4205b60e71f4ebf96da74e
> # bad: [9e2dcefa791e9d14006b360fba3455510fd3325d] lib: add allocation tagging support for memory allocation profiling
> git bisect bad 9e2dcefa791e9d14006b360fba3455510fd3325d
> # good: [0eccd42fbf9d7c4ae0cbec48cce637da89813c2c] lib: prevent module unloading if memory is not freed
> git bisect good 0eccd42fbf9d7c4ae0cbec48cce637da89813c2c
> # first bad commit: [9e2dcefa791e9d14006b360fba3455510fd3325d] lib: add allocation tagging support for memory allocation profiling
