LKML Archive on lore.kernel.org
 help / color / Atom feed
From: Peter Zijlstra <peterz@infradead.org>
To: mingo@kernel.org, tglx@linutronix.de, rostedt@goodmis.org
Cc: linux-kernel@vger.kernel.org, kan.liang@linux.intel.com,
	acme@kernel.org, mark.rutland@arm.com,
	alexander.shishkin@linux.intel.com, jolsa@redhat.com,
	namhyung@kernel.org, ak@linux.intel.com, eranian@google.com,
	peterz@infradead.org
Subject: [PATCH 5/6] perf/arch: Remove perf_sample_data::regs_user_copy
Date: Fri, 30 Oct 2020 16:13:50 +0100
Message-ID: <20201030151955.258178461@infradead.org> (raw)
In-Reply-To: <20201030151345.540479897@infradead.org>

struct perf_sample_data lives on-stack, we should be careful about it's
size. Furthermore, the pt_regs copy in there is only because x86_64 is a
trainwreck, solve it differently.

Reported-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Steven Rostedt <rostedt@goodmis.org>
---
 arch/arm/kernel/perf_regs.c   |    3 +--
 arch/arm64/kernel/perf_regs.c |    3 +--
 arch/csky/kernel/perf_regs.c  |    3 +--
 arch/powerpc/perf/perf_regs.c |    3 +--
 arch/riscv/kernel/perf_regs.c |    3 +--
 arch/s390/kernel/perf_regs.c  |    3 +--
 arch/x86/kernel/perf_regs.c   |   15 +++++++++++----
 include/linux/perf_event.h    |    6 ------
 include/linux/perf_regs.h     |    6 ++----
 kernel/events/core.c          |    8 +++-----
 10 files changed, 22 insertions(+), 31 deletions(-)

--- a/arch/arm/kernel/perf_regs.c
+++ b/arch/arm/kernel/perf_regs.c
@@ -32,8 +32,7 @@ u64 perf_reg_abi(struct task_struct *tas
 }
 
 void perf_get_regs_user(struct perf_regs *regs_user,
-			struct pt_regs *regs,
-			struct pt_regs *regs_user_copy)
+			struct pt_regs *regs)
 {
 	regs_user->regs = task_pt_regs(current);
 	regs_user->abi = perf_reg_abi(current);
--- a/arch/arm64/kernel/perf_regs.c
+++ b/arch/arm64/kernel/perf_regs.c
@@ -73,8 +73,7 @@ u64 perf_reg_abi(struct task_struct *tas
 }
 
 void perf_get_regs_user(struct perf_regs *regs_user,
-			struct pt_regs *regs,
-			struct pt_regs *regs_user_copy)
+			struct pt_regs *regs)
 {
 	regs_user->regs = task_pt_regs(current);
 	regs_user->abi = perf_reg_abi(current);
--- a/arch/csky/kernel/perf_regs.c
+++ b/arch/csky/kernel/perf_regs.c
@@ -32,8 +32,7 @@ u64 perf_reg_abi(struct task_struct *tas
 }
 
 void perf_get_regs_user(struct perf_regs *regs_user,
-			struct pt_regs *regs,
-			struct pt_regs *regs_user_copy)
+			struct pt_regs *regs)
 {
 	regs_user->regs = task_pt_regs(current);
 	regs_user->abi = perf_reg_abi(current);
--- a/arch/powerpc/perf/perf_regs.c
+++ b/arch/powerpc/perf/perf_regs.c
@@ -144,8 +144,7 @@ u64 perf_reg_abi(struct task_struct *tas
 }
 
 void perf_get_regs_user(struct perf_regs *regs_user,
-			struct pt_regs *regs,
-			struct pt_regs *regs_user_copy)
+			struct pt_regs *regs)
 {
 	regs_user->regs = task_pt_regs(current);
 	regs_user->abi = (regs_user->regs) ? perf_reg_abi(current) :
--- a/arch/riscv/kernel/perf_regs.c
+++ b/arch/riscv/kernel/perf_regs.c
@@ -36,8 +36,7 @@ u64 perf_reg_abi(struct task_struct *tas
 }
 
 void perf_get_regs_user(struct perf_regs *regs_user,
-			struct pt_regs *regs,
-			struct pt_regs *regs_user_copy)
+			struct pt_regs *regs)
 {
 	regs_user->regs = task_pt_regs(current);
 	regs_user->abi = perf_reg_abi(current);
--- a/arch/s390/kernel/perf_regs.c
+++ b/arch/s390/kernel/perf_regs.c
@@ -53,8 +53,7 @@ u64 perf_reg_abi(struct task_struct *tas
 }
 
 void perf_get_regs_user(struct perf_regs *regs_user,
-			struct pt_regs *regs,
-			struct pt_regs *regs_user_copy)
+			struct pt_regs *regs)
 {
 	/*
 	 * Use the regs from the first interruption and let
--- a/arch/x86/kernel/perf_regs.c
+++ b/arch/x86/kernel/perf_regs.c
@@ -101,8 +101,7 @@ u64 perf_reg_abi(struct task_struct *tas
 }
 
 void perf_get_regs_user(struct perf_regs *regs_user,
-			struct pt_regs *regs,
-			struct pt_regs *regs_user_copy)
+			struct pt_regs *regs)
 {
 	regs_user->regs = task_pt_regs(current);
 	regs_user->abi = perf_reg_abi(current);
@@ -129,12 +128,20 @@ u64 perf_reg_abi(struct task_struct *tas
 		return PERF_SAMPLE_REGS_ABI_64;
 }
 
+static DEFINE_PER_CPU(struct pt_regs, nmi_user_regs);
+
 void perf_get_regs_user(struct perf_regs *regs_user,
-			struct pt_regs *regs,
-			struct pt_regs *regs_user_copy)
+			struct pt_regs *regs)
 {
+	struct pt_regs *regs_user_copy = this_cpu_ptr(&nmi_user_regs);
 	struct pt_regs *user_regs = task_pt_regs(current);
 
+	if (!in_nmi()) {
+		regs_user->regs = user_regs;
+		regs_user->abi = perf_reg_abi(current);
+		return;
+	}
+
 	/*
 	 * If we're in an NMI that interrupted task_pt_regs setup, then
 	 * we can't sample user regs at all.  This check isn't really
--- a/include/linux/perf_event.h
+++ b/include/linux/perf_event.h
@@ -1022,13 +1022,7 @@ struct perf_sample_data {
 	struct perf_callchain_entry	*callchain;
 	u64				aux_size;
 
-	/*
-	 * regs_user may point to task_pt_regs or to regs_user_copy, depending
-	 * on arch details.
-	 */
 	struct perf_regs		regs_user;
-	struct pt_regs			regs_user_copy;
-
 	struct perf_regs		regs_intr;
 	u64				stack_user_size;
 
--- a/include/linux/perf_regs.h
+++ b/include/linux/perf_regs.h
@@ -20,8 +20,7 @@ u64 perf_reg_value(struct pt_regs *regs,
 int perf_reg_validate(u64 mask);
 u64 perf_reg_abi(struct task_struct *task);
 void perf_get_regs_user(struct perf_regs *regs_user,
-			struct pt_regs *regs,
-			struct pt_regs *regs_user_copy);
+			struct pt_regs *regs);
 #else
 
 #define PERF_REG_EXTENDED_MASK	0
@@ -42,8 +41,7 @@ static inline u64 perf_reg_abi(struct ta
 }
 
 static inline void perf_get_regs_user(struct perf_regs *regs_user,
-				      struct pt_regs *regs,
-				      struct pt_regs *regs_user_copy)
+				      struct pt_regs *regs);
 {
 	regs_user->regs = task_pt_regs(current);
 	regs_user->abi = perf_reg_abi(current);
--- a/kernel/events/core.c
+++ b/kernel/events/core.c
@@ -6374,14 +6374,13 @@ perf_output_sample_regs(struct perf_outp
 }
 
 static void perf_sample_regs_user(struct perf_regs *regs_user,
-				  struct pt_regs *regs,
-				  struct pt_regs *regs_user_copy)
+				  struct pt_regs *regs)
 {
 	if (user_mode(regs)) {
 		regs_user->abi = perf_reg_abi(current);
 		regs_user->regs = regs;
 	} else if (!(current->flags & PF_KTHREAD)) {
-		perf_get_regs_user(regs_user, regs, regs_user_copy);
+		perf_get_regs_user(regs_user, regs);
 	} else {
 		regs_user->abi = PERF_SAMPLE_REGS_ABI_NONE;
 		regs_user->regs = NULL;
@@ -7083,8 +7082,7 @@ void perf_prepare_sample(struct perf_eve
 	}
 
 	if (sample_type & (PERF_SAMPLE_REGS_USER | PERF_SAMPLE_STACK_USER))
-		perf_sample_regs_user(&data->regs_user, regs,
-				      &data->regs_user_copy);
+		perf_sample_regs_user(&data->regs_user, regs);
 
 	if (sample_type & PERF_SAMPLE_REGS_USER) {
 		/* regs dump ABI info */



  parent reply index

Thread overview: 23+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-10-30 15:13 [PATCH 0/6] perf: Reduce stack usage (and misc bits) Peter Zijlstra
2020-10-30 15:13 ` [PATCH 1/6] perf: Reduce stack usage of perf_output_begin() Peter Zijlstra
2020-11-10 12:45   ` [tip: perf/urgent] " tip-bot2 for Peter Zijlstra
2020-10-30 15:13 ` [PATCH 2/6] perf/x86: Reduce stack usage for x86_pmu::drain_pebs() Peter Zijlstra
2020-11-10 12:45   ` [tip: perf/urgent] " tip-bot2 for Peter Zijlstra
2020-10-30 15:13 ` [PATCH 3/6] perf: Fix get_recursion_context() Peter Zijlstra
2020-11-10 12:45   ` [tip: perf/urgent] " tip-bot2 for Peter Zijlstra
2020-10-30 15:13 ` [PATCH 4/6] perf: Optimize get_recursion_context() Peter Zijlstra
2020-10-30 17:11   ` Jesper Dangaard Brouer
2020-10-30 20:22     ` Steven Rostedt
2020-10-30 22:14       ` Thomas Gleixner
2020-10-30 23:31         ` Steven Rostedt
2020-10-31 11:23           ` Peter Zijlstra
2020-10-30 23:01       ` Peter Zijlstra
2020-10-31 12:11         ` David Laight
2020-10-31 13:18           ` David Laight
2020-11-09 12:12           ` Peter Zijlstra
2020-11-09 14:14             ` David Laight
2020-11-10 12:45   ` [tip: perf/urgent] " tip-bot2 for Peter Zijlstra
2020-10-30 15:13 ` Peter Zijlstra [this message]
2020-11-10 12:45   ` [tip: perf/urgent] perf/arch: Remove perf_sample_data::regs_user_copy tip-bot2 for Peter Zijlstra
2020-10-30 15:13 ` [PATCH 6/6] perf/x86: Make dummy_iregs static Peter Zijlstra
2020-11-10 12:45   ` [tip: perf/urgent] " tip-bot2 for Peter Zijlstra

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=20201030151955.258178461@infradead.org \
    --to=peterz@infradead.org \
    --cc=acme@kernel.org \
    --cc=ak@linux.intel.com \
    --cc=alexander.shishkin@linux.intel.com \
    --cc=eranian@google.com \
    --cc=jolsa@redhat.com \
    --cc=kan.liang@linux.intel.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mark.rutland@arm.com \
    --cc=mingo@kernel.org \
    --cc=namhyung@kernel.org \
    --cc=rostedt@goodmis.org \
    --cc=tglx@linutronix.de \
    /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

LKML Archive on lore.kernel.org

Archives are clonable:
	git clone --mirror https://lore.kernel.org/lkml/0 lkml/git/0.git
	git clone --mirror https://lore.kernel.org/lkml/1 lkml/git/1.git
	git clone --mirror https://lore.kernel.org/lkml/2 lkml/git/2.git
	git clone --mirror https://lore.kernel.org/lkml/3 lkml/git/3.git
	git clone --mirror https://lore.kernel.org/lkml/4 lkml/git/4.git
	git clone --mirror https://lore.kernel.org/lkml/5 lkml/git/5.git
	git clone --mirror https://lore.kernel.org/lkml/6 lkml/git/6.git
	git clone --mirror https://lore.kernel.org/lkml/7 lkml/git/7.git
	git clone --mirror https://lore.kernel.org/lkml/8 lkml/git/8.git
	git clone --mirror https://lore.kernel.org/lkml/9 lkml/git/9.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 lkml lkml/ https://lore.kernel.org/lkml \
		linux-kernel@vger.kernel.org
	public-inbox-index lkml

Example config snippet for mirrors

Newsgroup available over NNTP:
	nntp://nntp.lore.kernel.org/org.kernel.vger.linux-kernel


AGPL code for this site: git clone https://public-inbox.org/public-inbox.git