Re: [PATCH] x86/bugs: Allow STIBP with IBRS
From: Pawan Gupta
Date: Tue Feb 21 2023 - 20:20:20 EST
On Mon, Feb 20, 2023 at 11:09:08AM -0800, Josh Poimboeuf wrote:
> On Mon, Feb 20, 2023 at 07:34:59PM +0100, Borislav Petkov wrote:
> > Drop stable@ again.
> >
> > On Mon, Feb 20, 2023 at 10:27:17AM -0800, Josh Poimboeuf wrote:
> > > IBRS is only enabled in kernel space. Since it's not enabled in user
> > > space, user space isn't protected from indirect branch prediction
> > > attacks from a sibling CPU thread.
> > >
> > > Allow STIBP to be enabled to protect against such attacks.
> > >
> > > Fixes: 7c693f54c873 ("x86/speculation: Add spectre_v2=ibrs option to support Kernel IBRS")
> >
> > Yah, look at that one:
> >
> > commit 7c693f54c873691a4b7da05c7e0f74e67745d144
> > Author: Pawan Gupta <pawan.kumar.gupta@xxxxxxxxxxxxxxx>
> > Date: Tue Jun 14 23:15:55 2022 +0200
> >
> > x86/speculation: Add spectre_v2=ibrs option to support Kernel IBRS
> >
> > Extend spectre_v2= boot option with Kernel IBRS.
> >
> > [jpoimboe: no STIBP with IBRS]
> > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> >
> > I'm assuming this was supposed to mean no STIBP in *kernel mode* when
> > IBRS is selected?
>
> No it was supposed to be "no STIBP with *eIBRS*".
Maybe not, "no STIBP with eIBRS" was the state before the said patch.
In an offlist discussion during Retbleed embargo(copied below), it
appears to mean "no STIBP *in kernel* with IBRS". But anyways, we missed
to consider userspace.
(BTW replying late because yesterday was a holiday in my geo).
---
> > Subject: [PATCH v5 26/30] x86/speculation: Add spectre_v2=ibrs option to support Kernel IBRS
> >
> > From: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
> >
> > From: Pawan Gupta <pawan.kumar.gupta@xxxxxxxxxxxxxxx>
> >
> > The "spectre_v2=" boot option is extended to enable Kernel IBRS.
> >
> > Signed-off-by: Pawan Gupta <pawan.kumar.gupta@xxxxxxxxxxxxxxx>
> > Signed-off-by: Peter Zijlstra (Intel) <peterz@xxxxxxxxxxxxx>
> > ---
> > Documentation/admin-guide/kernel-parameters.txt | 1
> > arch/x86/include/asm/nospec-branch.h | 1
> > arch/x86/kernel/cpu/bugs.c | 29 ++++++++++++++++++++++--
> > 3 files changed, 29 insertions(+), 2 deletions(-)
> >
> > @@ -1163,6 +1182,10 @@ static void __init spectre_v2_select_mit
> > case SPECTRE_V2_EIBRS:
> > break;
> >
> > + case SPECTRE_V2_IBRS:
> > + setup_force_cpu_cap(X86_FEATURE_KERNEL_IBRS);
> > + break;
>
> Don't we also need to set SPEC_CTRL_IBRS in x86_spec_ctrl_base?
Also, STIBP isn't needed with IBRS. Suggested changes:
diff --git a/arch/x86/kernel/cpu/bugs.c b/arch/x86/kernel/cpu/bugs.c
index 344ab7c9a4e2..498cb36587a3 100644
--- a/arch/x86/kernel/cpu/bugs.c
+++ b/arch/x86/kernel/cpu/bugs.c
@@ -897,11 +897,13 @@ spectre_v2_parse_user_cmdline(enum spectre_v2_mitigation_cmd v2_cmd)
return SPECTRE_V2_USER_CMD_AUTO;
}
-static inline bool spectre_v2_in_eibrs_mode(enum spectre_v2_mitigation mode)
+static inline bool spectre_v2_in_ibrs_mode(enum spectre_v2_mitigation mode)
{
- return (mode == SPECTRE_V2_EIBRS ||
- mode == SPECTRE_V2_EIBRS_RETPOLINE ||
- mode == SPECTRE_V2_EIBRS_LFENCE);
+
+ return spectre_v2_mode == SPECTRE_V2_IBRS
+ spectre_v2_mode == SPECTRE_V2_EIBRS ||
+ spectre_v2_mode == SPECTRE_V2_EIBRS_RETPOLINE ||
+ spectre_v2_mode == SPECTRE_V2_EIBRS_LFENCE;
}
static void __init
@@ -966,12 +968,12 @@ spectre_v2_user_select_mitigation(enum spectre_v2_mitigation_cmd v2_cmd)
}
/*
- * If no STIBP, enhanced IBRS is enabled or SMT impossible, STIBP is not
- * required.
+ * If no STIBP, IBRS or enhanced IBRS is enabled, or SMT impossible,
+ * STIBP is not required.
*/
if (!boot_cpu_has(X86_FEATURE_STIBP) ||
!smt_possible ||
- spectre_v2_in_eibrs_mode(spectre_v2_enabled))
+ spectre_v2_in_ibrs_mode(spectre_v2_enabled))
return;
/*
@@ -1171,7 +1173,7 @@ static void __init spectre_v2_select_mitigation(void)
if (mode == SPECTRE_V2_EIBRS && unprivileged_ebpf_enabled())
pr_err(SPECTRE_V2_EIBRS_EBPF_MSG);
- if (spectre_v2_in_eibrs_mode(mode)) {
+ if (spectre_v2_in_ibrs_mode(mode)) {
/* Force it so VMEXIT will restore correctly */
x86_spec_ctrl_base |= SPEC_CTRL_IBRS;
wr_spec_ctrl(x86_spec_ctrl_base, true);
@@ -1212,19 +1214,17 @@ static void __init spectre_v2_select_mitigation(void)
pr_info("Spectre v2 / SpectreRSB mitigation: Filling RSB on context switch\n");
/*
- * Retpoline means the kernel is safe because it has no indirect
- * branches. Enhanced IBRS protects firmware too, so, enable restricted
- * speculation around firmware calls only when Enhanced IBRS isn't
- * supported or kernel IBRS isn't enabled.
+ * Retpoline protects the kernel, but doesn't protect firmware. IBRS
+ * and Enhanced IBRS protect firmware too, so enable IBRS around
+ * firmware calls only when IBRS / Enhanced IBRS aren't otherwise
+ * enabled.
*
* Use "mode" to check Enhanced IBRS instead of boot_cpu_has(), because
* the user might select retpoline on the kernel command line and if
* the CPU supports Enhanced IBRS, kernel might un-intentionally not
* enable IBRS around firmware calls.
*/
- if (boot_cpu_has(X86_FEATURE_IBRS) &&
- !boot_cpu_has(X86_FEATURE_KERNEL_IBRS) &&
- !spectre_v2_in_eibrs_mode(mode)) {
+ if (boot_cpu_has(X86_FEATURE_IBRS) && !spectre_v2_in_ibrs_mode(mode)) {
setup_force_cpu_cap(X86_FEATURE_USE_IBRS_FW);
pr_info("Enabling Restricted Speculation for firmware calls\n");
}
@@ -1937,7 +1937,7 @@ static ssize_t tsx_async_abort_show_state(char *buf)
static char *stibp_state(void)
{
- if (spectre_v2_in_eibrs_mode(spectre_v2_enabled))
+ if (spectre_v2_in_ibrs_mode(spectre_v2_enabled))
return "";
switch (spectre_v2_user_stibp) {