From: Josh Poimboeuf <jpoimboe@redhat.com>
To: Borislav Petkov <bp@alien8.de>
Cc: Thomas Gleixner <tglx@linutronix.de>,
Ingo Molnar <mingo@redhat.com>, "H. Peter Anvin" <hpa@zytor.com>,
Michal Marek <mmarek@suse.cz>,
Peter Zijlstra <peterz@infradead.org>,
Andy Lutomirski <luto@kernel.org>,
Linus Torvalds <torvalds@linux-foundation.org>,
Andi Kleen <andi@firstfloor.org>, Pedro Alves <palves@redhat.com>,
x86@kernel.org, live-patching@vger.kernel.org,
linux-kernel@vger.kernel.org
Subject: Re: [RFC PATCH 04/21] x86/hweight: Add stack frame dependency for __arch_hweight*()
Date: Sat, 18 Jul 2015 08:44:15 -0500 [thread overview]
Message-ID: <20150718134415.GA24587@treble.redhat.com> (raw)
In-Reply-To: <20150718050536.GA23478@nazgul.tnic>
On Sat, Jul 18, 2015 at 07:05:36AM +0200, Borislav Petkov wrote:
> On Fri, Jul 17, 2015 at 12:32:20PM -0500, Josh Poimboeuf wrote:
> > Well, but this isn't some whitelist code to make stackvalidate happy.
> >
> > It's actually a real runtime frame pointer bug, and the rsp dependency
> > is real. If it does the call without first creating the stack frame
> > then it breaks frame pointer based stack traces.
>
> I think we can live with the stack trace being a little wrong in those
> __sw_* variants. And besides, we're talking about the very very small
> percentage of machines (which keeps getting smaller) which don't
> support POPCNT. And from those, only for the cases where the arg is not
> __builtin_constant_p() because there we do the __const_hweight* thing.
>
> I'd prefer to not clutter the code more in that case.
Ok, so would you rather adding a whitelist to tell stackvalidate to
ignore it? Something like this?
diff --git a/arch/x86/include/asm/arch_hweight.h b/arch/x86/include/asm/arch_hweight.h
index 9686c3d..d604691 100644
--- a/arch/x86/include/asm/arch_hweight.h
+++ b/arch/x86/include/asm/arch_hweight.h
@@ -1,6 +1,8 @@
#ifndef _ASM_X86_HWEIGHT_H
#define _ASM_X86_HWEIGHT_H
+#include <linux/stackvalidate.h>
+
#ifdef CONFIG_64BIT
/* popcnt %edi, %eax -- redundant REX prefix for alignment */
#define POPCNT32 ".byte 0xf3,0x40,0x0f,0xb8,0xc7"
@@ -25,7 +27,9 @@ static inline unsigned int __arch_hweight32(unsigned int w)
{
unsigned int res = 0;
- asm (ALTERNATIVE("call __sw_hweight32", POPCNT32, X86_FEATURE_POPCNT)
+ asm (ALTERNATIVE(STACKVALIDATE_IGNORE_INSN
+ "call __sw_hweight32",
+ POPCNT32, X86_FEATURE_POPCNT)
: "="REG_OUT (res)
: REG_IN (w));
@@ -50,7 +54,9 @@ static inline unsigned long __arch_hweight64(__u64 w)
return __arch_hweight32((u32)w) +
__arch_hweight32((u32)(w >> 32));
#else
- asm (ALTERNATIVE("call __sw_hweight64", POPCNT64, X86_FEATURE_POPCNT)
+ asm (ALTERNATIVE(STACKVALIDATE_IGNORE_INSN
+ "call __sw_hweight64",
+ POPCNT64, X86_FEATURE_POPCNT)
: "="REG_OUT (res)
: REG_IN (w));
#endif /* CONFIG_X86_32 */
--
Josh
next prev parent reply other threads:[~2015-07-18 13:44 UTC|newest]
Thread overview: 89+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-07-14 17:14 [PATCH v7 0/4] Compile-time stack validation Josh Poimboeuf
2015-07-14 17:14 ` [PATCH v7 1/4] x86/asm: Frame pointer macro cleanup Josh Poimboeuf
2015-07-14 17:14 ` [PATCH v7 2/4] x86/stackvalidate: Compile-time stack validation Josh Poimboeuf
2015-07-14 20:57 ` Peter Zijlstra
2015-07-14 21:11 ` Josh Poimboeuf
2015-07-14 21:08 ` Peter Zijlstra
2015-07-14 21:30 ` Josh Poimboeuf
2015-07-14 21:56 ` Peter Zijlstra
2015-07-14 22:32 ` Josh Poimboeuf
2015-07-20 16:53 ` Namhyung Kim
2015-07-20 17:50 ` Josh Poimboeuf
2015-07-21 8:02 ` Ingo Molnar
2015-07-21 12:04 ` Josh Poimboeuf
2015-07-21 8:42 ` Bernd Petrovitsch
2015-07-21 12:06 ` Josh Poimboeuf
2015-07-14 17:14 ` [PATCH v7 3/4] x86/stackvalidate: Add file and directory ignores Josh Poimboeuf
2015-07-14 17:14 ` [PATCH v7 4/4] x86/stackvalidate: Add ignore macros Josh Poimboeuf
2015-07-14 17:25 ` [PATCH v7 0/4] Compile-time stack validation Josh Poimboeuf
2015-07-15 10:16 ` Ingo Molnar
2015-07-15 16:05 ` Josh Poimboeuf
2015-07-17 16:47 ` [RFC PATCH 00/21] x86: Proposed fixes for stackvalidate warnings Josh Poimboeuf
2015-07-17 16:47 ` [RFC PATCH 01/21] stackvalidate: Process ignores earlier and add more ignore checks Josh Poimboeuf
2015-07-17 16:47 ` [RFC PATCH 02/21] stackvalidate: Add C version of STACKVALIDATE_IGNORE_INSN Josh Poimboeuf
2015-07-18 14:56 ` Borislav Petkov
2015-07-18 16:00 ` Josh Poimboeuf
[not found] ` <CA+55aFyoO75n-mQBrB_YBLx9yNpAjisFAqkO8+YsphD-xmgY+w@mail.gmail.com>
2015-07-18 16:40 ` Josh Poimboeuf
2015-07-17 16:47 ` [RFC PATCH 03/21] x86/asm: Add C versions of FRAME and ENDFRAME macros Josh Poimboeuf
2015-07-17 16:47 ` [RFC PATCH 04/21] x86/hweight: Add stack frame dependency for __arch_hweight*() Josh Poimboeuf
2015-07-17 17:17 ` Borislav Petkov
2015-07-17 17:32 ` Josh Poimboeuf
2015-07-18 5:05 ` Borislav Petkov
2015-07-18 13:44 ` Josh Poimboeuf [this message]
2015-07-18 14:56 ` Borislav Petkov
2015-07-18 15:57 ` Josh Poimboeuf
2015-07-19 4:12 ` Borislav Petkov
2015-07-22 0:13 ` Andy Lutomirski
2015-07-22 4:25 ` Borislav Petkov
2015-07-22 4:39 ` Andy Lutomirski
2015-07-22 4:45 ` Borislav Petkov
2015-07-17 16:47 ` [RFC PATCH 05/21] x86/xen: Add stack frame dependency to hypercall inline asm calls Josh Poimboeuf
2015-07-17 16:47 ` [RFC PATCH 06/21] x86/paravirt: Add stack frame dependency to PVOP " Josh Poimboeuf
2015-07-17 16:47 ` [RFC PATCH 07/21] x86/paravirt: Fix frame pointer usage in PV_CALLEE_SAVE_REGS_THUNK Josh Poimboeuf
2015-07-17 16:47 ` [RFC PATCH 08/21] x86/paravirt: Align paravirt thunk functions at 16-byte boundaries Josh Poimboeuf
2015-07-17 16:47 ` [RFC PATCH 09/21] x86/amd: Set ELF function type for vide() Josh Poimboeuf
2015-07-17 16:47 ` [RFC PATCH 10/21] x86/reboot: Add ljmp instructions to stackvalidate whitelist Josh Poimboeuf
2015-07-17 16:47 ` [RFC PATCH 11/21] x86/xen: Add xen_cpuid() and xen_setup_gdt() to stackvalidate whitelists Josh Poimboeuf
2015-07-17 16:47 ` [RFC PATCH 12/21] sched: Add __schedule() to stackvalidate whitelist Josh Poimboeuf
2015-07-17 19:46 ` Peter Zijlstra
2015-07-17 19:58 ` Andy Lutomirski
2015-07-17 21:03 ` Peter Zijlstra
2015-07-17 21:23 ` Josh Poimboeuf
2015-07-18 3:44 ` Ingo Molnar
2015-07-17 16:47 ` [RFC PATCH 13/21] x86/asm/crypto: Fix frame pointer usage in aesni-intel_asm.S Josh Poimboeuf
2015-07-17 19:43 ` Ingo Molnar
2015-07-17 19:44 ` Andy Lutomirski
2015-07-17 20:37 ` Josh Poimboeuf
2015-07-17 20:39 ` Andy Lutomirski
2015-07-17 20:44 ` Josh Poimboeuf
2015-07-17 20:46 ` Andy Lutomirski
2015-07-17 20:59 ` Josh Poimboeuf
2015-07-17 21:01 ` Andy Lutomirski
2015-07-17 21:10 ` Josh Poimboeuf
2015-07-18 8:42 ` Borislav Petkov
2015-07-18 13:46 ` Josh Poimboeuf
2015-07-18 14:25 ` Borislav Petkov
2015-07-18 15:40 ` Josh Poimboeuf
2015-07-18 2:51 ` Ingo Molnar
2015-07-18 3:56 ` Josh Poimboeuf
2015-07-20 7:56 ` Ingo Molnar
2015-07-20 13:59 ` Josh Poimboeuf
2015-07-20 17:21 ` Ingo Molnar
2015-07-20 18:00 ` Josh Poimboeuf
2015-07-22 11:52 ` Josh Poimboeuf
2015-07-20 15:30 ` Andy Lutomirski
2015-07-20 16:36 ` Josh Poimboeuf
2015-07-20 16:52 ` Peter Zijlstra
2015-07-20 17:19 ` Josh Poimboeuf
2015-07-21 8:00 ` Ingo Molnar
2015-07-21 12:06 ` Josh Poimboeuf
2015-07-17 16:47 ` [RFC PATCH 14/21] x86/asm/crypto: Move .Lbswap_mask data to .rodata section Josh Poimboeuf
2015-07-17 16:47 ` [RFC PATCH 15/21] x86/asm/crypto: Move jump_table " Josh Poimboeuf
2015-07-17 16:47 ` [RFC PATCH 16/21] x86/asm/crypto: Fix frame pointer usage in clmul_ghash_mul/update() Josh Poimboeuf
2015-07-17 16:47 ` [RFC PATCH 17/21] x86/asm/entry: Fix frame pointer usage in thunk functions Josh Poimboeuf
2015-07-17 16:47 ` [RFC PATCH 18/21] x86/asm/acpi: Fix frame pointer usage in do_suspend_lowlevel() Josh Poimboeuf
2015-07-17 16:47 ` [RFC PATCH 19/21] x86/asm: Fix frame pointer usage in rwsem functions Josh Poimboeuf
2015-07-17 16:47 ` [RFC PATCH 20/21] x86/asm/efi: Fix frame pointer usage in efi_call() Josh Poimboeuf
2015-07-17 16:47 ` [RFC PATCH 21/21] x86/asm/power: Fix frame pointer usage in hibernate_asm_64.S Josh Poimboeuf
2015-07-17 18:56 ` [RFC PATCH 00/21] x86: Proposed fixes for stackvalidate warnings Andy Lutomirski
2015-07-18 3:05 ` Ingo Molnar
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=20150718134415.GA24587@treble.redhat.com \
--to=jpoimboe@redhat.com \
--cc=andi@firstfloor.org \
--cc=bp@alien8.de \
--cc=hpa@zytor.com \
--cc=linux-kernel@vger.kernel.org \
--cc=live-patching@vger.kernel.org \
--cc=luto@kernel.org \
--cc=mingo@redhat.com \
--cc=mmarek@suse.cz \
--cc=palves@redhat.com \
--cc=peterz@infradead.org \
--cc=tglx@linutronix.de \
--cc=torvalds@linux-foundation.org \
--cc=x86@kernel.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 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).