[PATCH v3 0/2] Improve KASLR diagnostics

[PATCH v3 0/2] Improve KASLR diagnostics

[Mark Brown]

This series provides diagnostics on boot for KASLR to improve usability
at runtime.

v3: Still mask the seed from FDT when disabling from the command line.
v2: Defer the print to a core_initcall() so we don't try and print
    before printk() can cope, covering other less common error cases as
    well and also an explicit message when KASLR is enabled.

Mark Brown (2):
  arm64: kaslr: Announce KASLR status on boot
  arm64: kaslr: Check command line before looking for a seed

 arch/arm64/kernel/kaslr.c | 44 +++++++++++++++++++++++++++++++++++----
 1 file changed, 40 insertions(+), 4 deletions(-)

-- 
2.20.1


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH v3 1/2] arm64: kaslr: Announce KASLR status on boot

[Mark Brown]

Currently the KASLR code is silent at boot unless it forces on KPTI in
which case a message will be printed for that. This can lead to users
incorrectly believing their system has the feature enabled when it in
fact does not, and if they notice the problem the lack of any
diagnostics makes it harder to understand the problem. Add an initcall
which prints a message showing the status of KASLR during boot to make
the status clear.

This is particularly useful in cases where we don't have a seed. It
seems to be a relatively common error for system integrators and
administrators to enable KASLR in their configuration but not provide
the seed at runtime, often due to seed provisioning breaking at some
later point after it is initially enabled and verified.

Signed-off-by: Mark Brown <broonie@kernel.org>
---
 arch/arm64/kernel/kaslr.c | 41 ++++++++++++++++++++++++++++++++++++---
 1 file changed, 38 insertions(+), 3 deletions(-)

diff --git a/arch/arm64/kernel/kaslr.c b/arch/arm64/kernel/kaslr.c
index 416f537bf614..0039dc50e556 100644
--- a/arch/arm64/kernel/kaslr.c
+++ b/arch/arm64/kernel/kaslr.c
@@ -19,6 +19,14 @@
 #include <asm/pgtable.h>
 #include <asm/sections.h>
 
+enum kaslr_status {
+	KASLR_ENABLED,
+	KASLR_DISABLED_CMDLINE,
+	KASLR_DISABLED_NO_SEED,
+	KASLR_DISABLED_FDT_REMAP,
+};
+
+enum kaslr_status __ro_after_init kaslr_status;
 u64 __ro_after_init module_alloc_base;
 u16 __initdata memstart_offset_seed;
 
@@ -91,15 +99,19 @@ u64 __init kaslr_early_init(u64 dt_phys)
 	 */
 	early_fixmap_init();
 	fdt = fixmap_remap_fdt(dt_phys, &size, PAGE_KERNEL);
-	if (!fdt)
+	if (!fdt) {
+		kaslr_status = KASLR_DISABLED_FDT_REMAP;
 		return 0;
+	}
 
 	/*
 	 * Retrieve (and wipe) the seed from the FDT
 	 */
 	seed = get_kaslr_seed(fdt);
-	if (!seed)
+	if (!seed) {
+		kaslr_status = KASLR_DISABLED_NO_SEED;
 		return 0;
+	}
 
 	/*
 	 * Check if 'nokaslr' appears on the command line, and
@@ -107,8 +119,10 @@ u64 __init kaslr_early_init(u64 dt_phys)
 	 */
 	cmdline = kaslr_get_cmdline(fdt);
 	str = strstr(cmdline, "nokaslr");
-	if (str == cmdline || (str > cmdline && *(str - 1) == ' '))
+	if (str == cmdline || (str > cmdline && *(str - 1) == ' ')) {
+		kaslr_status = KASLR_DISABLED_CMDLINE;
 		return 0;
+	}
 
 	/*
 	 * OK, so we are proceeding with KASLR enabled. Calculate a suitable
@@ -170,3 +184,24 @@ u64 __init kaslr_early_init(u64 dt_phys)
 
 	return offset;
 }
+
+static int __init kaslr_init(void)
+{
+	switch (kaslr_status) {
+	case KASLR_ENABLED:
+		pr_info("KASLR enabled\n");
+		break;
+	case KASLR_DISABLED_CMDLINE:
+		pr_info("KASLR disabled on command line\n");
+		break;
+	case KASLR_DISABLED_NO_SEED:
+		pr_warn("KASLR disabled due to lack of seed\n");
+		break;
+	case KASLR_DISABLED_FDT_REMAP:
+		pr_warn("KASLR disabled due to FDT remapping failure\n");
+		break;
+	}
+
+	return 0;
+}
+core_initcall(kaslr_init)
-- 
2.20.1


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH v3 2/2] arm64: kaslr: Check command line before looking for a seed

[Mark Brown]

Now that we print diagnostics at boot the reason why we do not initialise
KASLR matters. Currently we check for a seed before we check if the user
has explicitly disabled KASLR on the command line which will result in
misleading diagnostics so reverse the order of those checks. We still
parse the seed from the DT early so that if the user has both provided a
seed and disabled KASLR on the command line we still mask the seed on
the command line.

Signed-off-by: Mark Brown <broonie@kernel.org>
---
 arch/arm64/kernel/kaslr.c | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/arch/arm64/kernel/kaslr.c b/arch/arm64/kernel/kaslr.c
index 0039dc50e556..c09f12ab6525 100644
--- a/arch/arm64/kernel/kaslr.c
+++ b/arch/arm64/kernel/kaslr.c
@@ -108,10 +108,6 @@ u64 __init kaslr_early_init(u64 dt_phys)
 	 * Retrieve (and wipe) the seed from the FDT
 	 */
 	seed = get_kaslr_seed(fdt);
-	if (!seed) {
-		kaslr_status = KASLR_DISABLED_NO_SEED;
-		return 0;
-	}
 
 	/*
 	 * Check if 'nokaslr' appears on the command line, and
@@ -124,6 +120,11 @@ u64 __init kaslr_early_init(u64 dt_phys)
 		return 0;
 	}
 
+	if (!seed) {
+		kaslr_status = KASLR_DISABLED_NO_SEED;
+		return 0;
+	}
+
 	/*
 	 * OK, so we are proceeding with KASLR enabled. Calculate a suitable
 	 * kernel image offset from the seed. Let's place the kernel in the
-- 
2.20.1


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH v3 1/2] arm64: kaslr: Announce KASLR status on boot

[Mark Rutland]

On Fri, Nov 08, 2019 at 04:20:00PM +0000, Mark Brown wrote:
> Currently the KASLR code is silent at boot unless it forces on KPTI in
> which case a message will be printed for that. This can lead to users
> incorrectly believing their system has the feature enabled when it in
> fact does not, and if they notice the problem the lack of any
> diagnostics makes it harder to understand the problem. Add an initcall
> which prints a message showing the status of KASLR during boot to make
> the status clear.
> 
> This is particularly useful in cases where we don't have a seed. It
> seems to be a relatively common error for system integrators and
> administrators to enable KASLR in their configuration but not provide
> the seed at runtime, often due to seed provisioning breaking at some
> later point after it is initially enabled and verified.
> 
> Signed-off-by: Mark Brown <broonie@kernel.org>
> ---
>  arch/arm64/kernel/kaslr.c | 41 ++++++++++++++++++++++++++++++++++++---
>  1 file changed, 38 insertions(+), 3 deletions(-)
> 
> diff --git a/arch/arm64/kernel/kaslr.c b/arch/arm64/kernel/kaslr.c
> index 416f537bf614..0039dc50e556 100644
> --- a/arch/arm64/kernel/kaslr.c
> +++ b/arch/arm64/kernel/kaslr.c
> @@ -19,6 +19,14 @@
>  #include <asm/pgtable.h>
>  #include <asm/sections.h>
>  
> +enum kaslr_status {
> +	KASLR_ENABLED,
> +	KASLR_DISABLED_CMDLINE,
> +	KASLR_DISABLED_NO_SEED,
> +	KASLR_DISABLED_FDT_REMAP,
> +};
> +
> +enum kaslr_status __ro_after_init kaslr_status;

This can probably be __initdata since it's only consumed by an __init
call.

Otherwise this looks sound to me; with that:

Acked-by: Mark Rutland <mark.rutland@arm.com>

Thanks,
Mark.

>  u64 __ro_after_init module_alloc_base;
>  u16 __initdata memstart_offset_seed;
>  
> @@ -91,15 +99,19 @@ u64 __init kaslr_early_init(u64 dt_phys)
>  	 */
>  	early_fixmap_init();
>  	fdt = fixmap_remap_fdt(dt_phys, &size, PAGE_KERNEL);
> -	if (!fdt)
> +	if (!fdt) {
> +		kaslr_status = KASLR_DISABLED_FDT_REMAP;
>  		return 0;
> +	}
>  
>  	/*
>  	 * Retrieve (and wipe) the seed from the FDT
>  	 */
>  	seed = get_kaslr_seed(fdt);
> -	if (!seed)
> +	if (!seed) {
> +		kaslr_status = KASLR_DISABLED_NO_SEED;
>  		return 0;
> +	}
>  
>  	/*
>  	 * Check if 'nokaslr' appears on the command line, and
> @@ -107,8 +119,10 @@ u64 __init kaslr_early_init(u64 dt_phys)
>  	 */
>  	cmdline = kaslr_get_cmdline(fdt);
>  	str = strstr(cmdline, "nokaslr");
> -	if (str == cmdline || (str > cmdline && *(str - 1) == ' '))
> +	if (str == cmdline || (str > cmdline && *(str - 1) == ' ')) {
> +		kaslr_status = KASLR_DISABLED_CMDLINE;
>  		return 0;
> +	}
>  
>  	/*
>  	 * OK, so we are proceeding with KASLR enabled. Calculate a suitable
> @@ -170,3 +184,24 @@ u64 __init kaslr_early_init(u64 dt_phys)
>  
>  	return offset;
>  }
> +
> +static int __init kaslr_init(void)
> +{
> +	switch (kaslr_status) {
> +	case KASLR_ENABLED:
> +		pr_info("KASLR enabled\n");
> +		break;
> +	case KASLR_DISABLED_CMDLINE:
> +		pr_info("KASLR disabled on command line\n");
> +		break;
> +	case KASLR_DISABLED_NO_SEED:
> +		pr_warn("KASLR disabled due to lack of seed\n");
> +		break;
> +	case KASLR_DISABLED_FDT_REMAP:
> +		pr_warn("KASLR disabled due to FDT remapping failure\n");
> +		break;
> +	}
> +
> +	return 0;
> +}
> +core_initcall(kaslr_init)
> -- 
> 2.20.1
> 

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH v3 2/2] arm64: kaslr: Check command line before looking for a seed

[Mark Rutland]

On Fri, Nov 08, 2019 at 04:20:01PM +0000, Mark Brown wrote:
> Now that we print diagnostics at boot the reason why we do not initialise
> KASLR matters. Currently we check for a seed before we check if the user
> has explicitly disabled KASLR on the command line which will result in
> misleading diagnostics so reverse the order of those checks. We still
> parse the seed from the DT early so that if the user has both provided a
> seed and disabled KASLR on the command line we still mask the seed on
> the command line.
> 
> Signed-off-by: Mark Brown <broonie@kernel.org>

Acked-by: Mark Rutland <mark.rutland@arm.com>

Mark.

> ---
>  arch/arm64/kernel/kaslr.c | 9 +++++----
>  1 file changed, 5 insertions(+), 4 deletions(-)
> 
> diff --git a/arch/arm64/kernel/kaslr.c b/arch/arm64/kernel/kaslr.c
> index 0039dc50e556..c09f12ab6525 100644
> --- a/arch/arm64/kernel/kaslr.c
> +++ b/arch/arm64/kernel/kaslr.c
> @@ -108,10 +108,6 @@ u64 __init kaslr_early_init(u64 dt_phys)
>  	 * Retrieve (and wipe) the seed from the FDT
>  	 */
>  	seed = get_kaslr_seed(fdt);
> -	if (!seed) {
> -		kaslr_status = KASLR_DISABLED_NO_SEED;
> -		return 0;
> -	}
>  
>  	/*
>  	 * Check if 'nokaslr' appears on the command line, and
> @@ -124,6 +120,11 @@ u64 __init kaslr_early_init(u64 dt_phys)
>  		return 0;
>  	}
>  
> +	if (!seed) {
> +		kaslr_status = KASLR_DISABLED_NO_SEED;
> +		return 0;
> +	}
> +
>  	/*
>  	 * OK, so we are proceeding with KASLR enabled. Calculate a suitable
>  	 * kernel image offset from the seed. Let's place the kernel in the
> -- 
> 2.20.1
> 

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH v3 1/2] arm64: kaslr: Announce KASLR status on boot

[Mark Brown]

[-- Attachment #1.1: Type: text/plain, Size: 273 bytes --]

On Fri, Nov 08, 2019 at 04:46:36PM +0000, Mark Rutland wrote:
> On Fri, Nov 08, 2019 at 04:20:00PM +0000, Mark Brown wrote:

> > +enum kaslr_status __ro_after_init kaslr_status;

> This can probably be __initdata since it's only consumed by an __init
> call.

Yes, it can.

[-- Attachment #1.2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

[-- Attachment #2: Type: text/plain, Size: 176 bytes --]

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel