linux-perf-users.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Sean Christopherson <seanjc@google.com>
To: Yury Norov <yury.norov@gmail.com>,
	Peter Zijlstra <peterz@infradead.org>,
	Ingo Molnar <mingo@redhat.com>,
	Arnaldo Carvalho de Melo <acme@kernel.org>,
	Marc Zyngier <maz@kernel.org>,
	Paolo Bonzini <pbonzini@redhat.com>
Cc: Andy Shevchenko <andriy.shevchenko@linux.intel.com>,
	Rasmus Villemoes <linux@rasmusvillemoes.dk>,
	Mark Rutland <mark.rutland@arm.com>,
	Alexander Shishkin <alexander.shishkin@linux.intel.com>,
	Jiri Olsa <jolsa@kernel.org>, Namhyung Kim <namhyung@kernel.org>,
	James Morse <james.morse@arm.com>,
	Alexandru Elisei <alexandru.elisei@arm.com>,
	Suzuki K Poulose <suzuki.poulose@arm.com>,
	Oliver Upton <oliver.upton@linux.dev>,
	linux-perf-users@vger.kernel.org, linux-kernel@vger.kernel.org,
	linux-arm-kernel@lists.infradead.org, kvmarm@lists.linux.dev,
	kvmarm@lists.cs.columbia.edu, kvm@vger.kernel.org,
	Sean Christopherson <seanjc@google.com>
Subject: [PATCH 4/9] tools: Take @bit as an "unsigned long" in {clear,set}_bit() helpers
Date: Sat, 19 Nov 2022 01:34:45 +0000	[thread overview]
Message-ID: <20221119013450.2643007-5-seanjc@google.com> (raw)
In-Reply-To: <20221119013450.2643007-1-seanjc@google.com>

Take @bit as an unsigned long instead of a signed int in clear_bit() and
set_bit() so that they match the double-underscore versions, __clear_bit()
and __set_bit().  This will allow converting users that really don't want
atomic operations to the double-underscores without introducing a
functional change, which will in turn allow making {clear,set}_bit()
atomic (as advertised).

Practically speaking, this _should_ have no functional impact.  KVM's
selftests usage is either hardcoded (Hyper-V tests) or is artificially
limited (arch_timer test and dirty_log test).  In KVM, dirty_log test is
the only mildly interesting case as it's use indirectly restricted to
unsigned 32-bit values, but in theory it could generate a negative value
when cast to a signed int.  But in that case, taking an "unsigned long"
is actually a bug fix.

Perf's usage is more difficult to audit, but any code that is affected
by the switch is likely already broken.  perf_header__{set,clear}_feat()
and perf_file_header__read() effectively use only hardcoded enums with
small, positive values, atom_new() passes an unsigned long, but its value
is capped at 128 via NR_ATOM_PER_PAGE, etc...

The only real potential for breakage is in the perf flows that take a
"cpu", but it's unlikely perf is subtly relying on a negative index into
bitmaps, e.g. "cpu" can be "-1", but only as "not valid" placeholder.

Note, tools/testing/nvdimm/ makes heavy use of set_bit(), but that code
builds into a kernel module of sorts, i.e. pulls in all of the kernel's
header and so is getting the kernel's atomic set_bit().  The NVDIMM test
usage of atomics is likely unnecessary, e.g. ndtest_dimm_register() sets
bits in a local variable, but that's neither here nor there as far as
this change is concerned.

Signed-off-by: Sean Christopherson <seanjc@google.com>
---
 tools/include/asm-generic/bitops/atomic.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/tools/include/asm-generic/bitops/atomic.h b/tools/include/asm-generic/bitops/atomic.h
index 2f6ea28764a7..f64b049d236c 100644
--- a/tools/include/asm-generic/bitops/atomic.h
+++ b/tools/include/asm-generic/bitops/atomic.h
@@ -5,12 +5,12 @@
 #include <asm/types.h>
 #include <asm/bitsperlong.h>
 
-static inline void set_bit(int nr, unsigned long *addr)
+static inline void set_bit(unsigned long nr, unsigned long *addr)
 {
 	addr[nr / __BITS_PER_LONG] |= 1UL << (nr % __BITS_PER_LONG);
 }
 
-static inline void clear_bit(int nr, unsigned long *addr)
+static inline void clear_bit(unsigned long nr, unsigned long *addr)
 {
 	addr[nr / __BITS_PER_LONG] &= ~(1UL << (nr % __BITS_PER_LONG));
 }
-- 
2.38.1.584.g0f3c55d4c2-goog


  parent reply	other threads:[~2022-11-19  1:54 UTC|newest]

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-11-19  1:34 [PATCH 0/9] tools: Make {clear,set}_bit() atomic for reals Sean Christopherson
2022-11-19  1:34 ` [PATCH 1/9] KVM: selftests: Add rdmsr_from_l2() implementation in Hyper-V eVMCS test Sean Christopherson
2022-11-19  1:34 ` [PATCH 2/9] KVM: selftests: Remove unused "vcpu" param to fix build error Sean Christopherson
2022-11-19  1:34 ` [PATCH 3/9] KVM: arm64: selftests: Enable single-step without a "full" ucall() Sean Christopherson
2022-11-19  1:34 ` Sean Christopherson [this message]
2022-11-19  1:34 ` [PATCH 5/9] perf tools: Use dedicated non-atomic clear/set bit helpers Sean Christopherson
2022-12-02 18:16   ` Namhyung Kim
2022-12-02 18:27     ` Arnaldo Carvalho de Melo
2022-11-19  1:34 ` [PATCH 6/9] KVM: selftests: Use non-atomic clear/set bit helpers in KVM tests Sean Christopherson
2022-11-19  1:34 ` [PATCH 7/9] tools: Drop conflicting non-atomic test_and_{clear,set}_bit() helpers Sean Christopherson
2022-11-19  1:34 ` [PATCH 8/9] tools: Drop "atomic_" prefix from atomic test_and_set_bit() Sean Christopherson
2022-11-19  1:34 ` [PATCH 9/9] tools: KVM: selftests: Convert clear/set_bit() to actual atomics Sean Christopherson
2022-12-02 18:23 ` [PATCH 0/9] tools: Make {clear,set}_bit() atomic for reals Paolo Bonzini

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20221119013450.2643007-5-seanjc@google.com \
    --to=seanjc@google.com \
    --cc=acme@kernel.org \
    --cc=alexander.shishkin@linux.intel.com \
    --cc=alexandru.elisei@arm.com \
    --cc=andriy.shevchenko@linux.intel.com \
    --cc=james.morse@arm.com \
    --cc=jolsa@kernel.org \
    --cc=kvm@vger.kernel.org \
    --cc=kvmarm@lists.cs.columbia.edu \
    --cc=kvmarm@lists.linux.dev \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-perf-users@vger.kernel.org \
    --cc=linux@rasmusvillemoes.dk \
    --cc=mark.rutland@arm.com \
    --cc=maz@kernel.org \
    --cc=mingo@redhat.com \
    --cc=namhyung@kernel.org \
    --cc=oliver.upton@linux.dev \
    --cc=pbonzini@redhat.com \
    --cc=peterz@infradead.org \
    --cc=suzuki.poulose@arm.com \
    --cc=yury.norov@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).