All of lore.kernel.org
 help / color / mirror / Atom feed
From: Peter Zijlstra <peterz@infradead.org>
To: David Woodhouse <dwmw2@infradead.org>,
	Thomas Gleixner <tglx@linutronix.de>,
	Josh Poimboeuf <jpoimboe@redhat.com>
Cc: linux-kernel@vger.kernel.org, Dave Hansen <dave.hansen@intel.com>,
	Ashok Raj <ashok.raj@intel.com>,
	Tim Chen <tim.c.chen@linux.intel.com>,
	Andy Lutomirski <luto@kernel.org>,
	Linus Torvalds <torvalds@linux-foundation.org>,
	Greg KH <gregkh@linuxfoundation.org>,
	Andrea Arcangeli <aarcange@redhat.com>,
	Andi Kleen <ak@linux.intel.com>,
	Arjan Van De Ven <arjan.van.de.ven@intel.com>,
	Dan Williams <dan.j.williams@intel.com>,
	Paolo Bonzini <pbonzini@redhat.com>,
	Jun Nakajima <jun.nakajima@intel.com>,
	Asit Mallick <asit.k.mallick@intel.com>,
	Jason Baron <jbaron@akamai.com>,
	Peter Zijlstra <peterz@infradead.org>
Subject: [PATCH 18/24] objtool: More complex static jump implementation
Date: Tue, 23 Jan 2018 16:25:57 +0100	[thread overview]
Message-ID: <20180123152639.049872416@infradead.org> (raw)
In-Reply-To: 20180123152539.374360046@infradead.org

[-- Attachment #1: peterz-objtool-fancy.patch --]
[-- Type: text/plain, Size: 3264 bytes --]

When using something like:

  -#define sched_feat(x) (static_branch_##x(&sched_feat_keys[__SCHED_FEAT_##x]))
  +#define sched_feat(x) (static_branch_##x(&sched_feat_keys[__SCHED_FEAT_##x]) && \
  +			(arch_static_assert(), true))

we get an objtool assertion fail like:

kernel/sched/fair.o: warning: objtool: hrtick_update()+0xd: static assert FAIL

where:

0000000000001140 <hrtick_update>:
    1140:       0f 1f 44 00 00          nopl   0x0(%rax,%rax,1)
    1145:       c3                      retq
    1146:       48 8b b7 30 09 00 00    mov    0x930(%rdi),%rsi
    114d:       8b 87 d8 09 00 00       mov    0x9d8(%rdi),%eax
    1153:       48 0f a3 05 00 00 00    bt     %rax,0x0(%rip)        # 115b <hrtick_update+0x1b>
    115a:       00
                        1157: R_X86_64_PC32     __cpu_active_mask-0x4

and:

RELOCATION RECORDS FOR [__jump_table]:
0000000000000150 R_X86_64_64       .text+0x0000000000001140
0000000000000158 R_X86_64_64       .text+0x0000000000001146

RELOCATION RECORDS FOR [.discard.jump_assert]:
0000000000000028 R_X86_64_64       .text+0x000000000000114d

IOW, GCC managed to place the assertion 1 instruction _after_ the
static jump target (it lifted a load over it).

So while the code generation is fine, the assertion gets placed wrong.
We can 'fix' this by not only considering the immediate static jump
locations but also all the unconditional code after it, terminating
the basic block on any unconditional instruction or branch entry
point.

Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
---
 tools/objtool/check.c |   41 +++++++++++++++++++++++++++++++++++++++++
 tools/objtool/check.h |    1 +
 2 files changed, 42 insertions(+)

--- a/tools/objtool/check.c
+++ b/tools/objtool/check.c
@@ -520,6 +520,8 @@ static int add_jump_destinations(struct
 				  dest_off);
 			return -1;
 		}
+
+		insn->jump_dest->branch_target++;
 	}
 
 	return 0;
@@ -1205,6 +1207,41 @@ static int assert_static_jumps(struct ob
 	return 0;
 }
 
+static int grow_static_blocks(struct objtool_file *file)
+{
+	struct instruction *insn;
+	bool static_block = false;
+
+	for_each_insn(file, insn) {
+		if (!static_block && !insn->static_jump_dest)
+			continue;
+
+		if (insn->static_jump_dest) {
+			static_block = true;
+			continue;
+		}
+
+		if (insn->branch_target) {
+			static_block = false;
+			continue;
+		} else switch (insn->type) {
+		case INSN_JUMP_CONDITIONAL:
+		case INSN_JUMP_UNCONDITIONAL:
+		case INSN_JUMP_DYNAMIC:
+		case INSN_CALL:
+		case INSN_CALL_DYNAMIC:
+		case INSN_RETURN:
+		case INSN_BUG:
+			static_block = false;
+			continue;
+		}
+
+		insn->static_jump_dest = static_block;
+	}
+
+	return 0;
+}
+
 static int decode_sections(struct objtool_file *file)
 {
 	int ret;
@@ -1247,6 +1284,10 @@ static int decode_sections(struct objtoo
 	if (ret)
 		return ret;
 
+	ret = grow_static_blocks(file);
+	if (ret)
+		return ret;
+
 	return 0;
 }
 
--- a/tools/objtool/check.h
+++ b/tools/objtool/check.h
@@ -46,6 +46,7 @@ struct instruction {
 	unsigned long immediate;
 	bool alt_group, visited, dead_end, ignore, hint, save, restore, ignore_alts;
 	bool retpoline_safe, static_jump_dest;
+	unsigned int branch_target;
 	struct symbol *call_dest;
 	struct instruction *jump_dest;
 	struct list_head alts;

  parent reply	other threads:[~2018-01-23 15:25 UTC|newest]

Thread overview: 71+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-01-23 15:25 [PATCH 00/24] objtool: retpoline and asm-goto validation Peter Zijlstra
2018-01-23 15:25 ` [PATCH 01/24] objtool: Use existing global variables for options Peter Zijlstra
2018-01-23 15:25 ` [PATCH 02/24] objtool: Add retpoline validation Peter Zijlstra
2018-01-23 18:21   ` Borislav Petkov
2018-01-26  9:54   ` David Woodhouse
2018-01-23 15:25 ` [PATCH 03/24] x86/paravirt: Annotate indirect calls Peter Zijlstra
2018-01-25 10:02   ` David Woodhouse
2018-01-25 10:22     ` Peter Zijlstra
2018-01-25 10:26       ` Juergen Gross
2018-01-25 10:52         ` David Woodhouse
2018-01-25 11:35           ` Peter Zijlstra
2018-01-26  9:57             ` David Woodhouse
2018-01-29 17:58   ` Josh Poimboeuf
2018-01-29 18:09     ` David Woodhouse
2018-01-29 18:17     ` Peter Zijlstra
2018-01-29 18:38   ` Josh Poimboeuf
2018-01-29 19:21     ` Peter Zijlstra
2018-01-30 16:02       ` Josh Poimboeuf
2018-01-31  4:13         ` [PATCH] x86/paravirt: Remove 'noreplace-paravirt' cmdline option Josh Poimboeuf
2018-01-31  5:59           ` Juergen Gross
2018-01-31  9:42           ` [tip:x86/pti] " tip-bot for Josh Poimboeuf
2018-01-23 15:25 ` [PATCH 04/24] x86,nospec: Annotate indirect calls/jumps Peter Zijlstra
2018-01-26 10:19   ` David Woodhouse
2018-01-29 17:44     ` Peter Zijlstra
2018-01-23 15:25 ` [PATCH 05/24] x86: Annotate indirect jump in head_64.S Peter Zijlstra
2018-01-26 10:24   ` David Woodhouse
2018-01-23 15:25 ` [PATCH 06/24] x86,kvm: Fix indirect calls in emulator Peter Zijlstra
2018-01-23 20:28   ` Borislav Petkov
2018-01-23 20:48     ` David Woodhouse
2018-01-24 10:35       ` Peter Zijlstra
2018-01-24 10:43         ` Paolo Bonzini
2018-01-25  9:34           ` Peter Zijlstra
2018-01-25  9:49             ` David Woodhouse
2018-01-26 10:57             ` Paolo Bonzini
2018-01-23 15:25 ` [PATCH 07/24] x86,vmx: Fix indirect call Peter Zijlstra
2018-01-25  9:36   ` Peter Zijlstra
2018-01-23 15:25 ` [PATCH 08/24] x86,sme: Annotate " Peter Zijlstra
2018-01-26 10:37   ` David Woodhouse
2018-01-29 17:49     ` Peter Zijlstra
2018-01-29 17:50       ` Peter Zijlstra
2018-01-31  9:29     ` Peter Zijlstra
2018-01-31 15:04       ` Josh Poimboeuf
2018-01-31 16:00         ` Peter Zijlstra
2018-01-23 15:25 ` [PATCH 09/24] jump_label: Add branch hints to static_branch_{un,}likely() Peter Zijlstra
2018-01-24 18:46   ` Borislav Petkov
2018-01-23 15:25 ` [PATCH 10/24] sched: Optimize ttwu_stat() Peter Zijlstra
2018-01-23 15:25 ` [PATCH 11/24] x86: Reindent _static_cpu_has Peter Zijlstra
2018-01-23 15:25 ` [PATCH 12/24] x86: Update _static_cpu_has to use all named variables Peter Zijlstra
2018-01-25 19:31   ` Borislav Petkov
2018-01-23 15:25 ` [PATCH 13/24] objtool: Implement base jump_assert support Peter Zijlstra
2018-01-26 10:45   ` David Woodhouse
2018-01-23 15:25 ` [PATCH 14/24] x86: Add a type field to alt_instr Peter Zijlstra
2018-01-23 15:25 ` [PATCH 15/24] x86: Annotate static_cpu_has alternative Peter Zijlstra
2018-01-23 15:25 ` [PATCH 16/24] objtool: Implement jump_assert for _static_cpu_has() Peter Zijlstra
2018-01-23 15:25 ` [PATCH 17/24] objtool: Introduce special_type Peter Zijlstra
2018-01-23 15:25 ` Peter Zijlstra [this message]
2018-01-23 15:25 ` [PATCH 19/24] objtool: Even more complex static block checks Peter Zijlstra
2018-01-23 15:25 ` [PATCH 20/24] objtool: Another static block fail Peter Zijlstra
2018-01-29 22:52   ` Josh Poimboeuf
2018-01-30  9:56     ` Peter Zijlstra
2018-01-31  3:12       ` Josh Poimboeuf
2018-01-31 10:01         ` Peter Zijlstra
2018-01-31 10:07           ` David Woodhouse
2018-01-31 10:27             ` Peter Zijlstra
2018-01-23 15:26 ` [PATCH 21/24] objtool: Skip static assert when KCOV/KASAN Peter Zijlstra
2018-01-23 15:26 ` [PATCH 22/24] x86/jump_label: Implement arch_static_assert() Peter Zijlstra
2018-01-23 15:26 ` [PATCH 23/24] x86: Force asm-goto Peter Zijlstra
2018-01-23 15:26 ` [PATCH 24/24] x86: Remove FAST_FEATURE_TESTS Peter Zijlstra
2018-01-23 15:42 ` [PATCH 00/24] objtool: retpoline and asm-goto validation Peter Zijlstra
2018-01-23 15:57   ` David Woodhouse
2018-01-23 16:03     ` 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=20180123152639.049872416@infradead.org \
    --to=peterz@infradead.org \
    --cc=aarcange@redhat.com \
    --cc=ak@linux.intel.com \
    --cc=arjan.van.de.ven@intel.com \
    --cc=ashok.raj@intel.com \
    --cc=asit.k.mallick@intel.com \
    --cc=dan.j.williams@intel.com \
    --cc=dave.hansen@intel.com \
    --cc=dwmw2@infradead.org \
    --cc=gregkh@linuxfoundation.org \
    --cc=jbaron@akamai.com \
    --cc=jpoimboe@redhat.com \
    --cc=jun.nakajima@intel.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=luto@kernel.org \
    --cc=pbonzini@redhat.com \
    --cc=tglx@linutronix.de \
    --cc=tim.c.chen@linux.intel.com \
    --cc=torvalds@linux-foundation.org \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.