Alexander Popov [Thu, 28 Nov 2019 21:11:07 +0000 (00:11 +0300)]
x86_32: INTEL_IOMMU is not enabled by default - fix the reason
Alexander Popov [Thu, 28 Nov 2019 21:07:48 +0000 (00:07 +0300)]
X86_INTEL_UMIP is now X86_UMIP
Alexander Popov [Thu, 28 Nov 2019 21:07:21 +0000 (00:07 +0300)]
x86_64: more hardening options are enabled by default - change the reason
Alexander Popov [Thu, 28 Nov 2019 17:24:55 +0000 (20:24 +0300)]
Improve the list of the kernel parameters in TODO
Alexander Popov [Thu, 28 Nov 2019 17:23:04 +0000 (20:23 +0300)]
Add CLIP OS links
Alexander Popov [Thu, 28 Nov 2019 16:56:13 +0000 (19:56 +0300)]
Update the column width
Alexander Popov [Thu, 28 Nov 2019 16:54:55 +0000 (19:54 +0300)]
Some of my recommendations are used by CLIP OS, change the `reason` field
Alexander Popov [Thu, 28 Nov 2019 16:30:01 +0000 (19:30 +0300)]
Don't recommend disabling IKCONFIG anymore
That info is needed for this script :)
Alexander Popov [Thu, 28 Nov 2019 16:28:52 +0000 (19:28 +0300)]
Save more hardening sysctls for TODO
Alexander Popov [Thu, 28 Nov 2019 16:27:53 +0000 (19:27 +0300)]
Update CLIP OS doc
Alexander Popov [Thu, 28 Nov 2019 09:09:36 +0000 (12:09 +0300)]
Group security policies together
Also update the name of the lockdown feature (merged into v5.4).
Alexander Popov [Thu, 28 Nov 2019 09:07:11 +0000 (12:07 +0300)]
Add INIT_ON_ALLOC_DEFAULT_ON and INIT_ON_FREE_DEFAULT_ON introduced in v5.3
Alexander Popov [Thu, 28 Nov 2019 09:06:27 +0000 (12:06 +0300)]
Add RODATA_FULL_DEFAULT_ENABLED for ARM64
Alexander Popov [Thu, 28 Nov 2019 07:32:49 +0000 (10:32 +0300)]
Add info about Debian and AOSP kernel configs to links.txt
Alexander Popov [Thu, 28 Nov 2019 07:17:57 +0000 (10:17 +0300)]
Add Debian Buster kernel config
Alexander Popov [Thu, 28 Nov 2019 07:17:07 +0000 (10:17 +0300)]
Add AOSP kernel config for Pixel 3a
Alexander Popov [Fri, 23 Aug 2019 16:09:36 +0000 (19:09 +0300)]
Introduce the versioning
At the Chaos Communication Camp 2019 @jelly told that it would be nice to add the kconfig-hardened-check to Arch Linux.
So I add versioning to make it happen.
Thanks @jelly, nice to meet you!
Alexander Popov [Fri, 23 Aug 2019 12:48:43 +0000 (15:48 +0300)]
Update the script output in the README
Alexander Popov [Fri, 23 Aug 2019 10:35:53 +0000 (13:35 +0300)]
Add HARDEN_BRANCH_PREDICTOR and HARDEN_EL2_VECTORS
Alexander Popov [Fri, 23 Aug 2019 11:40:41 +0000 (14:40 +0300)]
Bring more order to the offsets (style fix)
Alexander Popov [Thu, 22 Aug 2019 10:43:46 +0000 (13:43 +0300)]
Add INIT_STACK_ALL as an alternative to GCC_PLUGIN_STRUCTLEAK_BYREF_ALL
Alexander Popov [Thu, 22 Aug 2019 10:35:32 +0000 (13:35 +0300)]
Add SHUFFLE_PAGE_ALLOCATOR from v5.2
Alexander Popov [Thu, 22 Aug 2019 10:34:49 +0000 (13:34 +0300)]
Add some new sysctls (to remember them)
Alexander Popov [Mon, 8 Jul 2019 14:07:18 +0000 (17:07 +0300)]
Merge pull request #22 from adrianopol/master
#20 fix: use right quotes in json output
Thanks @adrianopol
Andrew Petelin [Sun, 7 Jul 2019 19:24:41 +0000 (22:24 +0300)]
#20 fix: use right quotes in json output
Alexander Popov [Mon, 24 Jun 2019 12:05:51 +0000 (15:05 +0300)]
Do code refactoring without changing the functionality
Changes:
- get rid of checklist global variable,
- improve print_checks().
Alexander Popov [Mon, 24 Jun 2019 11:04:11 +0000 (14:04 +0300)]
Merge branch 'json-support'
Thanks to @adrianopol
Alexander Popov [Mon, 24 Jun 2019 10:51:35 +0000 (13:51 +0300)]
json: Fix minor things and update the README
Andrew Petelin [Fri, 21 Jun 2019 19:56:23 +0000 (22:56 +0300)]
add --json option
Alexander Popov [Tue, 4 Jun 2019 22:04:07 +0000 (01:04 +0300)]
Drop CONFIG_X86_MSR from the recommendations
It exposes MSRs to the userspace, IMO it is not needed for mitigating
X86 CPU bugs.
Refers to the issue #19 (comment by @Bernhard40)
Alexander Popov [Mon, 3 Jun 2019 23:43:58 +0000 (02:43 +0300)]
Add the LDISC_AUTOLOAD check
In fact we have a false positive here because the absence
of the disabled CONFIG_LDISC_AUTOLOAD means FAIL (line
disciplines are automatically loaded).
TODO: Introduce a special check for this type of cases.
Alexander Popov [Mon, 3 Jun 2019 20:00:59 +0000 (23:00 +0300)]
Attribute some of my recommendations to CLIP OS - part II
They have a bigger authority :)
Refers to the issue #19 by @HacKurx
Alexander Popov [Mon, 3 Jun 2019 19:33:35 +0000 (22:33 +0300)]
Add the link to the CLIP OS kernel configuration
Refers to the issue #19 by @HacKurx
Alexander Popov [Mon, 3 Jun 2019 18:04:03 +0000 (21:04 +0300)]
Update the README
Refers to the issue #19 by @HacKurx
Alexander Popov [Mon, 3 Jun 2019 17:55:44 +0000 (20:55 +0300)]
Update the README (printing format)
Alexander Popov [Mon, 3 Jun 2019 17:41:13 +0000 (20:41 +0300)]
Add a snapshot of the CLIP OS config documentation
Refers to the issue #19 by @HacKurx
Alexander Popov [Mon, 3 Jun 2019 17:38:17 +0000 (20:38 +0300)]
Attribute some of my recommendations to CLIP OS
They have a bigger authority :)
Refers to the issue #19 by @HacKurx
Alexander Popov [Mon, 3 Jun 2019 17:27:51 +0000 (20:27 +0300)]
Add my recommendations for AMD (similar to CLIP OS recommendations for Intel)
Refers to the issue #19 by @HacKurx
Alexander Popov [Mon, 3 Jun 2019 17:24:21 +0000 (20:24 +0300)]
Add X86-specific CLIP OS recommendations for kernel self-protection
Refers to the issue #19 by @HacKurx
Alexander Popov [Mon, 3 Jun 2019 17:19:02 +0000 (20:19 +0300)]
Add arch-independent CLIP OS recommendations for kernel self-protection
Refers to the issue #19 by @HacKurx
Alexander Popov [Mon, 3 Jun 2019 17:13:32 +0000 (20:13 +0300)]
Add more details about STACKLEAK
Refers to the issue #19 by @HacKurx
Alexander Popov [Mon, 3 Jun 2019 17:03:58 +0000 (20:03 +0300)]
Don't recommend any particular LSM to avoid the holy war
Alexander Popov [Mon, 3 Jun 2019 17:02:42 +0000 (20:02 +0300)]
Add CLIP OS recommendations for cutting attack surface
Refers to the issue #19 by @HacKurx
Alexander Popov [Mon, 3 Jun 2019 16:59:25 +0000 (19:59 +0300)]
Improve printing of the results
Alexander Popov [Mon, 3 Jun 2019 10:12:35 +0000 (13:12 +0300)]
Update the link to Alpine config
Alexander Popov [Mon, 3 Jun 2019 10:10:28 +0000 (13:10 +0300)]
Merge pull request #18 from HacKurx/patch-1
Update pentoo config link
Loïc [Sat, 1 Jun 2019 12:02:36 +0000 (14:02 +0200)]
Update pentoo config link
Alexander Popov [Mon, 27 May 2019 14:42:53 +0000 (17:42 +0300)]
Add more kernel command line parameters to comments
Going to use them in future
Alexander Popov [Fri, 17 May 2019 15:13:40 +0000 (18:13 +0300)]
Merge remote-tracking branch 'hackurx/master'
Thanks to @HacKurx for updating the distro configs.
Loïc [Sun, 12 May 2019 15:04:25 +0000 (17:04 +0200)]
Create rhel-8.0.config
config check is finished: 'OK' - 41 / 'FAIL' - 62
Loïc [Sun, 12 May 2019 09:59:15 +0000 (11:59 +0200)]
Update and rename pentoo-4.17.11.config to pentoo-livecd.config
config check is finished: 'OK' - 71 / 'FAIL' - 32
Loïc [Sun, 12 May 2019 09:54:43 +0000 (11:54 +0200)]
Update Archlinux-hardened.config
config check is finished: 'OK' - 75 / 'FAIL' - 28
Loïc [Sun, 12 May 2019 09:51:39 +0000 (11:51 +0200)]
Update Alpinelinux-edge.config
config check is finished: 'OK' - 49 / 'FAIL' - 54
Loïc [Sun, 12 May 2019 09:46:53 +0000 (11:46 +0200)]
Update debian-stretch.config
config check is finished: 'OK' - 42 / 'FAIL' - 61
Loïc [Sun, 12 May 2019 09:38:12 +0000 (11:38 +0200)]
Create AmazonLinux2.config
config check is finished: 'OK' - 42 / 'FAIL' - 61
Alexander Popov [Wed, 20 Mar 2019 07:25:22 +0000 (10:25 +0300)]
Add Q&A to the README
Refers to the issue #14 by @jcberthon.
Alexander Popov [Wed, 13 Mar 2019 17:40:23 +0000 (20:40 +0300)]
Add the comment about kptr_restrict
Alexander Popov [Wed, 13 Mar 2019 13:45:34 +0000 (16:45 +0300)]
Add ARM64_PTR_AUTH check
Alexander Popov [Wed, 13 Mar 2019 09:02:19 +0000 (12:02 +0300)]
Add STACKPROTECTOR_PER_TASK check for ARM
Alexander Popov [Wed, 13 Mar 2019 08:37:13 +0000 (11:37 +0300)]
Add defconfigs for 5.0
Alexander Popov [Tue, 12 Mar 2019 21:46:32 +0000 (00:46 +0300)]
Don't hide AND check results if the requirements are not met
Report them as FAIL.
Thanks to @Bernhard40 for this nice idea.
Alexander Popov [Tue, 12 Mar 2019 15:11:56 +0000 (18:11 +0300)]
Update the README
Alexander Popov [Tue, 12 Mar 2019 14:29:20 +0000 (17:29 +0300)]
Improve the final result output
Refers to issue #13.
Alexander Popov [Tue, 12 Mar 2019 14:12:14 +0000 (17:12 +0300)]
Use the AND check for HARDENED_USERCOPY_FALLBACK
If HARDENED_USERCOPY is not set, HARDENED_USERCOPY_FALLBACK is not checked.
Refers to issue #13.
Alexander Popov [Tue, 12 Mar 2019 14:10:57 +0000 (17:10 +0300)]
Use the AND check for PAGE_POISONING_NO_SANITY and PAGE_POISONING_ZERO
If PAGE_POISONING is not set, PAGE_POISONING_NO_SANITY and
PAGE_POISONING_ZERO are not checked.
Refers to issue #13.
Alexander Popov [Tue, 12 Mar 2019 13:45:35 +0000 (16:45 +0300)]
Implement AND ComplexOptCheck
Use case: AND(<suboption>, <main_option>).
Suboption is not checked if checking of the main_option is failed.
It's needed to solve issue #13.
Alexander Popov [Tue, 12 Mar 2019 13:42:23 +0000 (16:42 +0300)]
Add a sanity check and do minor refactoring
Alexander Popov [Tue, 12 Mar 2019 12:02:49 +0000 (15:02 +0300)]
Introduce the ComplexOptCheck superclass
Alexander Popov [Mon, 11 Mar 2019 15:59:10 +0000 (18:59 +0300)]
Update the README
Alexander Popov [Mon, 11 Mar 2019 15:21:18 +0000 (18:21 +0300)]
Add explicit checks for CONFIG_MODULES and CONFIG_DEVMEM
I like this hack. Now the script recommends to disable modules and
devmem OR harden them at least.
Alexander Popov [Mon, 11 Mar 2019 15:08:59 +0000 (18:08 +0300)]
Add missing OR use case
Alexander Popov [Mon, 11 Mar 2019 15:33:11 +0000 (18:33 +0300)]
Improve the output of OR checks
Alexander Popov [Mon, 4 Mar 2019 18:24:45 +0000 (21:24 +0300)]
Add the RESET_ATTACK_MITIGATION check according to the feature request #11
Let's check the RESET_ATTACK_MITIGATION option.
The description of this security feature:
https://lwn.net/Articles/730006/
It needs support from the userspace side:
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=
a5c03c31af2291f13689d11760c0b59fb70c9a5a
Improve the comments about the userspace support by the way.
Alexander Popov [Mon, 4 Mar 2019 13:38:14 +0000 (16:38 +0300)]
Fix false positive about CONFIG_MODULE_SIG_FORCE.
CONFIG_MODULE_SIG_FORCE shouldn't be checked if CONFIG_MODULES is not set.
Fixes issue #12.
Thanks to @hannob.
Alexander Popov [Thu, 24 Jan 2019 07:43:58 +0000 (10:43 +0300)]
Update the README and comments after adding ARM support
Alexander Popov [Thu, 24 Jan 2019 07:34:00 +0000 (10:34 +0300)]
Fix typo in KSPP recommendations for ARM
Alexander Popov [Thu, 24 Jan 2019 07:33:25 +0000 (10:33 +0300)]
Add ARM support
Alexander Popov [Wed, 23 Jan 2019 17:02:53 +0000 (20:02 +0300)]
Update the README after adding ARM64 support
Alexander Popov [Wed, 23 Jan 2019 16:31:26 +0000 (19:31 +0300)]
Go through all the checks in debug mode
Alexander Popov [Wed, 23 Jan 2019 16:11:55 +0000 (19:11 +0300)]
Add ARM64 support
Alexander Popov [Tue, 22 Jan 2019 12:22:04 +0000 (15:22 +0300)]
Update the README after adding X86_32 support
And improve the style by the way.
Alexander Popov [Tue, 22 Jan 2019 11:55:47 +0000 (14:55 +0300)]
Add X86_32 support
Alexander Popov [Tue, 22 Jan 2019 11:55:07 +0000 (14:55 +0300)]
Merge branch 'arch-configs'
Alexander Popov [Tue, 22 Jan 2019 11:22:35 +0000 (14:22 +0300)]
Create a separate directory for distro configs
Alexander Popov [Tue, 22 Jan 2019 11:10:23 +0000 (14:10 +0300)]
Create a separate directory for defconfigs
Alexander Popov [Tue, 22 Jan 2019 11:09:26 +0000 (14:09 +0300)]
Add arm64 defconfig for v4.20
Alexander Popov [Tue, 22 Jan 2019 11:09:09 +0000 (14:09 +0300)]
Add arm defconfig for v4.20
Alexander Popov [Tue, 22 Jan 2019 10:47:31 +0000 (13:47 +0300)]
Add x86_32 defconfig for v4.20
Alexander Popov [Tue, 22 Jan 2019 11:04:51 +0000 (14:04 +0300)]
Create a separate directory for KSPP recommendations
Alexander Popov [Tue, 22 Jan 2019 11:03:11 +0000 (14:03 +0300)]
Specify the architecture in KSPP recommendations
Alexander Popov [Mon, 21 Jan 2019 22:18:36 +0000 (01:18 +0300)]
Update the README (arch support)
Alexander Popov [Mon, 21 Jan 2019 22:06:45 +0000 (01:06 +0300)]
Make the script aware of target architecture
Add the ability to parse the processor architecture from the config file.
Change '-p' command-line argument behaviour. Now it comes with the
name of architecture you want to print recommendations for.
Currently only X86_64 is supported. More architectures to come soon.
This is based heavily on work by @tyhicks.
Alexander Popov [Mon, 21 Jan 2019 16:16:00 +0000 (19:16 +0300)]
Merge branch 'from-tyhicks-1'
Create arch-dependent KSPP recommendations.
Thanks to @tyhicks.
Tyler Hicks [Thu, 17 Jan 2019 17:57:23 +0000 (17:57 +0000)]
Add a KSPP recommendations config for arm64
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Tyler Hicks [Thu, 17 Jan 2019 17:56:09 +0000 (17:56 +0000)]
Add a KSPP recommendations config for arm
The arm section of the KSPP Recommended_Settings wiki page contains the
following lines:
# If building an old out-of-tree Qualcomm kernel, this is similar to
# CONFIG_STRICT_KERNEL_RWX.
CONFIG_STRICT_MEMORY_RWX=y
Since this option only applies to an old out-of-tree Qualcomm kernel,
it is not included in the config file.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Tyler Hicks [Thu, 17 Jan 2019 17:53:43 +0000 (17:53 +0000)]
Add a KSPP recommendations config for x86_32
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Tyler Hicks [Thu, 17 Jan 2019 17:47:44 +0000 (17:47 +0000)]
Make KSPP recommendations config x86_64 specific
Rename the file so that it is clear that the recommendations are x86-64
specific.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Tyler Hicks [Sat, 12 Jan 2019 00:04:46 +0000 (00:04 +0000)]
Update kspp-recommendations.config to look like an x86_64 config
Add a header that will make the checker script think that it is dealing
with a x86_64 config file. Additionally, update the stackprotector
related options to reflect the >= 4.18 names.
Signed-off-by: Tyler Hicks <tyhicks@canonical.com>
Alexander Popov [Mon, 14 Jan 2019 22:45:35 +0000 (01:45 +0300)]
Add x86_64_defconfig for v4.20
Alexander Popov [Mon, 14 Jan 2019 22:28:03 +0000 (01:28 +0300)]
Merge branch 'decision-cleanup'
Thanks to @Bernhard40 for a nice idea.