mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Pawan Gupta <pawan.kumar.gupta@linux.intel.com>
To: Brian Gerst <brgerst@gmail.com>
Cc: Borislav Petkov <bp@alien8.de>,
	Dave Hansen <dave.hansen@linux.intel.com>,
	linux-kernel@vger.kernel.org, x86@kernel.org,
	Robert Gill <rtgill82@gmail.com>,
	Jari Ruusu <jariruusu@protonmail.com>,
	"Linux regression tracking (Thorsten Leemhuis)"
	<regressions@leemhuis.info>,
	antonio.gomez.iglesias@linux.intel.com,
	daniel.sneddon@linux.intel.com
Subject: Re: [PATCH v3] x86/entry_32: Move CLEAR_CPU_BUFFERS before restoring segments
Date: Mon, 1 Jul 2024 16:22:21 -0700	[thread overview]
Message-ID: <20240701232221.s4egkkyctgrasulr@desk> (raw)
In-Reply-To: <CAMzpN2iJHO01499VsarF1a3ZyjCLUn36FpNqEZeAaXuCoc9GwA@mail.gmail.com>

On Mon, Jul 01, 2024 at 03:22:44PM -0400, Brian Gerst wrote:
> Perhaps I should have been a bit clearer, but I meant adding the SS
> override to the VERW instructions in their present locations, instead
> of moving it into RESTORE_REGS.

I did consider that way, but moving it to RESTORE_REGS saves a callsite for
CLEAR_CPU_BUFFERS. I believe Dave prefered it within RESTORE_REGS, but that
was before the SS override. Unless anyone has a strong preference, I will
change the current callsites to include the SS override, and not move
CLEAR_CPU_BUFFERS to RESTORE_REGS.

> IIRC we don't want data reads (including stack pops) between the VERW
> instruction and IRET/SYSEXIT.

Stack pops are okay if the memory *and* the registers doesn't contain
sensitive data, which doesn't seem to be the case here. But, SS override is
a safer option.

BTW, Dave pointed out that CLEAR_CPU_BUFFERS needs to be done after
RESTORE_ALL_NMI in the NMI return path, that adds an extra callsite:

---
diff --git a/arch/x86/entry/entry_32.S b/arch/x86/entry/entry_32.S
index d3a814efbff6..9b58d2cbdfef 100644
--- a/arch/x86/entry/entry_32.S
+++ b/arch/x86/entry/entry_32.S
@@ -1144,7 +1144,6 @@ SYM_CODE_START(asm_exc_nmi)
 
 	/* Not on SYSENTER stack. */
 	call	exc_nmi
-	CLEAR_CPU_BUFFERS
 	jmp	.Lnmi_return
 
 .Lnmi_from_sysenter_stack:
@@ -1165,6 +1164,7 @@ SYM_CODE_START(asm_exc_nmi)
 
 	CHECK_AND_APPLY_ESPFIX
 	RESTORE_ALL_NMI cr3_reg=%edi pop=4
+	CLEAR_CPU_BUFFERS
 	jmp	.Lirq_return
 
 #ifdef CONFIG_X86_ESPFIX32
@@ -1206,6 +1206,7 @@ SYM_CODE_START(asm_exc_nmi)
 	 *  1 - orig_ax
 	 */
 	lss	(1+5+6)*4(%esp), %esp			# back to espfix stack
+	CLEAR_CPU_BUFFERS
 	jmp	.Lirq_return
 #endif
 SYM_CODE_END(asm_exc_nmi)

      reply	other threads:[~2024-07-01 23:22 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-07-01 18:34 Pawan Gupta
2024-07-01 19:22 ` Brian Gerst
2024-07-01 23:22   ` Pawan Gupta [this message]

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=20240701232221.s4egkkyctgrasulr@desk \
    --to=pawan.kumar.gupta@linux.intel.com \
    --cc=antonio.gomez.iglesias@linux.intel.com \
    --cc=bp@alien8.de \
    --cc=brgerst@gmail.com \
    --cc=daniel.sneddon@linux.intel.com \
    --cc=dave.hansen@linux.intel.com \
    --cc=jariruusu@protonmail.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=regressions@leemhuis.info \
    --cc=rtgill82@gmail.com \
    --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

all inboxes | Powered by JetHome®