| Message ID | 20201013081804.17332-1-ardb@kernel.org (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | [v2] ima: defer arch_ima_get_secureboot() call to IMA init time | expand |
[Cc'ing linuxppc-dev@lists.ozlabs.org] On Tue, 2020-10-13 at 10:18 +0200, Ard Biesheuvel wrote: > Chester reports that it is necessary to introduce a new way to pass > the EFI secure boot status between the EFI stub and the core kernel > on ARM systems. The usual way of obtaining this information is by > checking the SecureBoot and SetupMode EFI variables, but this can > only be done after the EFI variable workqueue is created, which > occurs in a subsys_initcall(), whereas arch_ima_get_secureboot() > is called much earlier by the IMA framework. > > However, the IMA framework itself is started as a late_initcall, > and the only reason the call to arch_ima_get_secureboot() occurs > so early is because it happens in the context of a __setup() > callback that parses the ima_appraise= command line parameter. > > So let's refactor this code a little bit, by using a core_param() > callback to capture the command line argument, and deferring any > reasoning based on its contents to the IMA init routine. > > Cc: Chester Lin <clin@suse.com> > Cc: Mimi Zohar <zohar@linux.ibm.com> > Cc: Dmitry Kasatkin <dmitry.kasatkin@gmail.com> > Cc: James Morris <jmorris@namei.org> > Cc: "Serge E. Hallyn" <serge@hallyn.com> > Link: https://lore.kernel.org/linux-arm-kernel/20200904072905.25332-2-clin@suse.com/ > Signed-off-by: Ard Biesheuvel <ardb@kernel.org> > --- > v2: rebase onto series 'integrity: improve user feedback for invalid bootparams' Thanks, Ard. Based on my initial, limited testing on Power, it looks good, but I'm hesistant to include it in the integrity 5.10 pull request without it having been in linux-next and some additional testing. It's now queued in the next-integrity-testing branch awaiting some tags. thanks, Mimi
On Tue, 13 Oct 2020 at 18:46, Mimi Zohar <zohar@linux.ibm.com> wrote: > > [Cc'ing linuxppc-dev@lists.ozlabs.org] > > On Tue, 2020-10-13 at 10:18 +0200, Ard Biesheuvel wrote: > > Chester reports that it is necessary to introduce a new way to pass > > the EFI secure boot status between the EFI stub and the core kernel > > on ARM systems. The usual way of obtaining this information is by > > checking the SecureBoot and SetupMode EFI variables, but this can > > only be done after the EFI variable workqueue is created, which > > occurs in a subsys_initcall(), whereas arch_ima_get_secureboot() > > is called much earlier by the IMA framework. > > > > However, the IMA framework itself is started as a late_initcall, > > and the only reason the call to arch_ima_get_secureboot() occurs > > so early is because it happens in the context of a __setup() > > callback that parses the ima_appraise= command line parameter. > > > > So let's refactor this code a little bit, by using a core_param() > > callback to capture the command line argument, and deferring any > > reasoning based on its contents to the IMA init routine. > > > > Cc: Chester Lin <clin@suse.com> > > Cc: Mimi Zohar <zohar@linux.ibm.com> > > Cc: Dmitry Kasatkin <dmitry.kasatkin@gmail.com> > > Cc: James Morris <jmorris@namei.org> > > Cc: "Serge E. Hallyn" <serge@hallyn.com> > > Link: https://lore.kernel.org/linux-arm-kernel/20200904072905.25332-2-clin@suse.com/ > > Signed-off-by: Ard Biesheuvel <ardb@kernel.org> > > --- > > v2: rebase onto series 'integrity: improve user feedback for invalid bootparams' > > Thanks, Ard. Based on my initial, limited testing on Power, it looks > good, but I'm hesistant to include it in the integrity 5.10 pull > request without it having been in linux-next and some additional > testing. It's now queued in the next-integrity-testing branch awaiting > some tags. > Thanks. No rush as far as I am concerned, although I suppose Chester may want to rebase his arm64 IMA enablement series on this. Suggestion: can we take the get_sb_mode() code from ima_arch.c in arch/x86, and generalize it for all EFI architectures? That way, we can enable 32-bit ARM and RISC-V seamlessly once someone gets around to enabling IMA on those platforms. In fact, get_sb_mode() itself should probably be factored out into a generic helper for use outside of IMA as well (Xen/x86 has code that does roughly the same already)
On Tue, 2020-10-13 at 18:59 +0200, Ard Biesheuvel wrote: > Suggestion: can we take the get_sb_mode() code from ima_arch.c in > arch/x86, and generalize it for all EFI architectures? That way, we > can enable 32-bit ARM and RISC-V seamlessly once someone gets around > to enabling IMA on those platforms. In fact, get_sb_mode() itself > should probably be factored out into a generic helper for use outside > of IMA as well (Xen/x86 has code that does roughly the same already) On Power, there are three different policies - secure, trusted, and secure & trusted boot policy rules. Based on whether secure or trusted boot is enabled, the appropriate policy is enabled. On x86, if secure_boot is enabled (and CONFIG_IMA_ARCH_POLICY is enabled) both the secure and trusted boot rules are defined. Is this design fine enough granularity or should should there be a get_trustedboot_mode() function as well? Agreed, the code should not be duplicated across arch's. As for making get_sb_mode() generic, not dependent on IMA, where would it reside? Would this be in EFI? thanks, Mimi
Hi Ard & Mimi, On Tue, Oct 13, 2020 at 06:59:21PM +0200, Ard Biesheuvel wrote: > On Tue, 13 Oct 2020 at 18:46, Mimi Zohar <zohar@linux.ibm.com> wrote: > > > > [Cc'ing linuxppc-dev@lists.ozlabs.org] > > > > On Tue, 2020-10-13 at 10:18 +0200, Ard Biesheuvel wrote: > > > Chester reports that it is necessary to introduce a new way to pass > > > the EFI secure boot status between the EFI stub and the core kernel > > > on ARM systems. The usual way of obtaining this information is by > > > checking the SecureBoot and SetupMode EFI variables, but this can > > > only be done after the EFI variable workqueue is created, which > > > occurs in a subsys_initcall(), whereas arch_ima_get_secureboot() > > > is called much earlier by the IMA framework. > > > > > > However, the IMA framework itself is started as a late_initcall, > > > and the only reason the call to arch_ima_get_secureboot() occurs > > > so early is because it happens in the context of a __setup() > > > callback that parses the ima_appraise= command line parameter. > > > > > > So let's refactor this code a little bit, by using a core_param() > > > callback to capture the command line argument, and deferring any > > > reasoning based on its contents to the IMA init routine. > > > > > > Cc: Chester Lin <clin@suse.com> > > > Cc: Mimi Zohar <zohar@linux.ibm.com> > > > Cc: Dmitry Kasatkin <dmitry.kasatkin@gmail.com> > > > Cc: James Morris <jmorris@namei.org> > > > Cc: "Serge E. Hallyn" <serge@hallyn.com> > > > Link: https://lore.kernel.org/linux-arm-kernel/20200904072905.25332-2-clin@suse.com/ > > > Signed-off-by: Ard Biesheuvel <ardb@kernel.org> > > > --- > > > v2: rebase onto series 'integrity: improve user feedback for invalid bootparams' > > > > Thanks, Ard. Based on my initial, limited testing on Power, it looks > > good, but I'm hesistant to include it in the integrity 5.10 pull > > request without it having been in linux-next and some additional > > testing. It's now queued in the next-integrity-testing branch awaiting > > some tags. > > Tested-by: Chester Lin <clin@suse.com> I have tested this patch on x86 VM. * System configuration: - Platform: QEMU/KVM - Firmware: EDK2/OVMF + secure boot enabled. - OS: SLE15-SP2 + SUSE's kernel-vanilla (=linux v5.9) + the follow commits from linux-next and upstream: * [PATCH v2] ima: defer arch_ima_get_secureboot() call to IMA init time https://www.spinics.net/lists/linux-efi/msg20645.html * e4d7e2df3a09 "ima: limit secure boot feedback scope for appraise" * 7fe2bb7e7e5c "integrity: invalid kernel parameters feedback" * 4afb28ab03d5 "ima: add check for enforced appraise option" * Logs with UEFI secure boot enabled: [ 0.000000] Linux version 5.9.0-858-g865c50e1d279-1.g8764d18-vanilla (geeko@b uildhost) (gcc (SUSE Linux) 10.2.1 20200825 [revision c0746a1beb1ba073c7981eb09f 55b3d993b32e5c], GNU ld (GNU Binutils; openSUSE Tumbleweed) 2.34.0.20200325-1) # 1 SMP Wed Oct 14 04:00:11 UTC 2020 (8764d18) [ 0.000000] Command line: BOOT_IMAGE=/boot/vmlinuz-5.9.0-858-g865c50e1d279-1. g8764d18-vanilla root=UUID=5304c03e-4d8a-4d67-873a-32a32e57cdeb console=ttyS0,11 5200 resume=/dev/disk/by-path/pci-0000:04:00.0-part4 mitigations=auto ignore_log level crashkernel=192M,high crashkernel=72M,low ima_appraise=off [ 0.000000] x86/fpu: Supporting XSAVE feature 0x001: 'x87 floating point regi sters' [ 0.000000] x86/fpu: Supporting XSAVE feature 0x002: 'SSE registers' [ 0.000000] x86/fpu: Supporting XSAVE feature 0x004: 'AVX registers' .... .... [ 1.720309] ima: Secure boot enabled: ignoring ima_appraise=off option [ 1.720314] ima: No TPM chip found, activating TPM-bypass! [ 1.722129] ima: Allocated hash algorithm: sha256 .... > > Thanks. No rush as far as I am concerned, although I suppose Chester > may want to rebase his arm64 IMA enablement series on this. Yes, I've finished coding but still verifying it. As you have suggested, My v2 patch will separate the get_sb_mode() from arch/x86 so that other architectures can reuse it. Thanks, Chester > > Suggestion: can we take the get_sb_mode() code from ima_arch.c in > arch/x86, and generalize it for all EFI architectures? That way, we > can enable 32-bit ARM and RISC-V seamlessly once someone gets around > to enabling IMA on those platforms. In fact, get_sb_mode() itself > should probably be factored out into a generic helper for use outside > of IMA as well (Xen/x86 has code that does roughly the same already) >
On Wed, 2020-10-14 at 17:35 +0800, Chester Lin wrote: > Hi Ard & Mimi, > > On Tue, Oct 13, 2020 at 06:59:21PM +0200, Ard Biesheuvel wrote: > > On Tue, 13 Oct 2020 at 18:46, Mimi Zohar <zohar@linux.ibm.com> wrote: > > > > > > [Cc'ing linuxppc-dev@lists.ozlabs.org] > > > > > > On Tue, 2020-10-13 at 10:18 +0200, Ard Biesheuvel wrote: > > > > Chester reports that it is necessary to introduce a new way to pass > > > > the EFI secure boot status between the EFI stub and the core kernel > > > > on ARM systems. The usual way of obtaining this information is by > > > > checking the SecureBoot and SetupMode EFI variables, but this can > > > > only be done after the EFI variable workqueue is created, which > > > > occurs in a subsys_initcall(), whereas arch_ima_get_secureboot() > > > > is called much earlier by the IMA framework. > > > > > > > > However, the IMA framework itself is started as a late_initcall, > > > > and the only reason the call to arch_ima_get_secureboot() occurs > > > > so early is because it happens in the context of a __setup() > > > > callback that parses the ima_appraise= command line parameter. > > > > > > > > So let's refactor this code a little bit, by using a core_param() > > > > callback to capture the command line argument, and deferring any > > > > reasoning based on its contents to the IMA init routine. > > > > > > > > Cc: Chester Lin <clin@suse.com> > > > > Cc: Mimi Zohar <zohar@linux.ibm.com> > > > > Cc: Dmitry Kasatkin <dmitry.kasatkin@gmail.com> > > > > Cc: James Morris <jmorris@namei.org> > > > > Cc: "Serge E. Hallyn" <serge@hallyn.com> > > > > Link: https://lore.kernel.org/linux-arm-kernel/20200904072905.25332-2-clin@suse.com/ > > > > Signed-off-by: Ard Biesheuvel <ardb@kernel.org> > > > > --- > > > > v2: rebase onto series 'integrity: improve user feedback for invalid bootparams' > > > > > > Thanks, Ard. Based on my initial, limited testing on Power, it looks > > > good, but I'm hesistant to include it in the integrity 5.10 pull > > > request without it having been in linux-next and some additional > > > testing. It's now queued in the next-integrity-testing branch awaiting > > > some tags. > > > > > Tested-by: Chester Lin <clin@suse.com> > > I have tested this patch on x86 VM. > > * System configuration: > - Platform: QEMU/KVM > - Firmware: EDK2/OVMF + secure boot enabled. > - OS: SLE15-SP2 + SUSE's kernel-vanilla (=linux v5.9) + the follow commits > from linux-next and upstream: > * [PATCH v2] ima: defer arch_ima_get_secureboot() call to IMA init time > https://www.spinics.net/lists/linux-efi/msg20645.html > * e4d7e2df3a09 "ima: limit secure boot feedback scope for appraise" > * 7fe2bb7e7e5c "integrity: invalid kernel parameters feedback" > * 4afb28ab03d5 "ima: add check for enforced appraise option" > > * Logs with UEFI secure boot enabled: > > [ 0.000000] Linux version 5.9.0-858-g865c50e1d279-1.g8764d18-vanilla (geeko@b > uildhost) (gcc (SUSE Linux) 10.2.1 20200825 [revision c0746a1beb1ba073c7981eb09f > 55b3d993b32e5c], GNU ld (GNU Binutils; openSUSE Tumbleweed) 2.34.0.20200325-1) # > 1 SMP Wed Oct 14 04:00:11 UTC 2020 (8764d18) > [ 0.000000] Command line: BOOT_IMAGE=/boot/vmlinuz-5.9.0-858-g865c50e1d279-1. > g8764d18-vanilla root=UUID=5304c03e-4d8a-4d67-873a-32a32e57cdeb console=ttyS0,11 > 5200 resume=/dev/disk/by-path/pci-0000:04:00.0-part4 mitigations=auto ignore_log > level crashkernel=192M,high crashkernel=72M,low ima_appraise=off > [ 0.000000] x86/fpu: Supporting XSAVE feature 0x001: 'x87 floating point regi > sters' > [ 0.000000] x86/fpu: Supporting XSAVE feature 0x002: 'SSE registers' > [ 0.000000] x86/fpu: Supporting XSAVE feature 0x004: 'AVX registers' > .... > .... > [ 1.720309] ima: Secure boot enabled: ignoring ima_appraise=off option > [ 1.720314] ima: No TPM chip found, activating TPM-bypass! > [ 1.722129] ima: Allocated hash algorithm: sha256 Thank you for testing the options aren't being set in secure boot mode. My main concern, however, is that IMA doesn't go into TPM-bypass mode. Does this system have a TPM? Mimi
On Wed, Oct 14, 2020 at 07:38:50AM -0400, Mimi Zohar wrote: > On Wed, 2020-10-14 at 17:35 +0800, Chester Lin wrote: > > Hi Ard & Mimi, > > > > On Tue, Oct 13, 2020 at 06:59:21PM +0200, Ard Biesheuvel wrote: > > > On Tue, 13 Oct 2020 at 18:46, Mimi Zohar <zohar@linux.ibm.com> wrote: > > > > > > > > [Cc'ing linuxppc-dev@lists.ozlabs.org] > > > > > > > > On Tue, 2020-10-13 at 10:18 +0200, Ard Biesheuvel wrote: > > > > > Chester reports that it is necessary to introduce a new way to pass > > > > > the EFI secure boot status between the EFI stub and the core kernel > > > > > on ARM systems. The usual way of obtaining this information is by > > > > > checking the SecureBoot and SetupMode EFI variables, but this can > > > > > only be done after the EFI variable workqueue is created, which > > > > > occurs in a subsys_initcall(), whereas arch_ima_get_secureboot() > > > > > is called much earlier by the IMA framework. > > > > > > > > > > However, the IMA framework itself is started as a late_initcall, > > > > > and the only reason the call to arch_ima_get_secureboot() occurs > > > > > so early is because it happens in the context of a __setup() > > > > > callback that parses the ima_appraise= command line parameter. > > > > > > > > > > So let's refactor this code a little bit, by using a core_param() > > > > > callback to capture the command line argument, and deferring any > > > > > reasoning based on its contents to the IMA init routine. > > > > > > > > > > Cc: Chester Lin <clin@suse.com> > > > > > Cc: Mimi Zohar <zohar@linux.ibm.com> > > > > > Cc: Dmitry Kasatkin <dmitry.kasatkin@gmail.com> > > > > > Cc: James Morris <jmorris@namei.org> > > > > > Cc: "Serge E. Hallyn" <serge@hallyn.com> > > > > > Link: https://lore.kernel.org/linux-arm-kernel/20200904072905.25332-2-clin@suse.com/ > > > > > Signed-off-by: Ard Biesheuvel <ardb@kernel.org> > > > > > --- > > > > > v2: rebase onto series 'integrity: improve user feedback for invalid bootparams' > > > > > > > > Thanks, Ard. Based on my initial, limited testing on Power, it looks > > > > good, but I'm hesistant to include it in the integrity 5.10 pull > > > > request without it having been in linux-next and some additional > > > > testing. It's now queued in the next-integrity-testing branch awaiting > > > > some tags. > > > > > > > > Tested-by: Chester Lin <clin@suse.com> > > > > I have tested this patch on x86 VM. > > > > * System configuration: > > - Platform: QEMU/KVM > > - Firmware: EDK2/OVMF + secure boot enabled. > > - OS: SLE15-SP2 + SUSE's kernel-vanilla (=linux v5.9) + the follow commits > > from linux-next and upstream: > > * [PATCH v2] ima: defer arch_ima_get_secureboot() call to IMA init time > > https://www.spinics.net/lists/linux-efi/msg20645.html > > * e4d7e2df3a09 "ima: limit secure boot feedback scope for appraise" > > * 7fe2bb7e7e5c "integrity: invalid kernel parameters feedback" > > * 4afb28ab03d5 "ima: add check for enforced appraise option" > > > > * Logs with UEFI secure boot enabled: > > > > [ 0.000000] Linux version 5.9.0-858-g865c50e1d279-1.g8764d18-vanilla (geeko@b > > uildhost) (gcc (SUSE Linux) 10.2.1 20200825 [revision c0746a1beb1ba073c7981eb09f > > 55b3d993b32e5c], GNU ld (GNU Binutils; openSUSE Tumbleweed) 2.34.0.20200325-1) # > > 1 SMP Wed Oct 14 04:00:11 UTC 2020 (8764d18) > > [ 0.000000] Command line: BOOT_IMAGE=/boot/vmlinuz-5.9.0-858-g865c50e1d279-1. > > g8764d18-vanilla root=UUID=5304c03e-4d8a-4d67-873a-32a32e57cdeb console=ttyS0,11 > > 5200 resume=/dev/disk/by-path/pci-0000:04:00.0-part4 mitigations=auto ignore_log > > level crashkernel=192M,high crashkernel=72M,low ima_appraise=off > > [ 0.000000] x86/fpu: Supporting XSAVE feature 0x001: 'x87 floating point regi > > sters' > > [ 0.000000] x86/fpu: Supporting XSAVE feature 0x002: 'SSE registers' > > [ 0.000000] x86/fpu: Supporting XSAVE feature 0x004: 'AVX registers' > > .... > > .... > > [ 1.720309] ima: Secure boot enabled: ignoring ima_appraise=off option > > [ 1.720314] ima: No TPM chip found, activating TPM-bypass! > > [ 1.722129] ima: Allocated hash algorithm: sha256 > > > Thank you for testing the options aren't being set in secure boot mode. > My main concern, however, is that IMA doesn't go into TPM-bypass mode. > Does this system have a TPM? > > Mimi > Last time I didn't emulate a TPM device in KVM so that's why the kernel couldn't find a TPM chip. I have tested this patch again with a virtual TPM 1.2 by running swtpm and here are some logs: ----- [ 0.000000] Linux version 5.9.0-858-g865c50e1d279-1.g8764d18-vanilla (geeko@b uildhost) (gcc (SUSE Linux) 10.2.1 20200825 [revision c0746a1beb1ba073c7981eb09f 55b3d993b32e5c], GNU ld (GNU Binutils; openSUSE Tumbleweed) 2.34.0.20200325-1) # 1 SMP Wed Oct 14 04:00:11 UTC 2020 (8764d18) [ 0.000000] Command line: BOOT_IMAGE=/boot/vmlinuz-5.9.0-858-g865c50e1d279-1. g8764d18-vanilla root=UUID=5304c03e-4d8a-4d67-873a-32a32e57cdeb console=ttyS0,11 5200 resume=/dev/disk/by-path/pci-0000:04:00.0-part4 mitigations=auto ignore_log level crashkernel=192M,high crashkernel=72M,low ima_appraise=off [ 0.000000] x86/fpu: Supporting XSAVE feature 0x001: 'x87 floating point regi sters' ..... ..... [ 0.000000] efi: EFI v2.60 by EDK II [ 0.000000] efi: SMBIOS=0x7fe82000 ACPI=0x7feb0000 ACPI 2.0=0x7feb0014 MEMATTT R=0x7ef89698 RNG=0x7ea12e18 [ 0.000000] efi: seeding entropy pool [ 0.000000] random: fast init done [ 0.000000] SMBIOS 2.8 present. [ 0.000000] DMI: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 0.0.0 02/06/2015 [ 0.000000] Hypervisor detected: KVM ..... ..... [ 1.957625] Linux agpgart interface v0.103 [ 1.961903] tpm_tis 00:04: 1.2 TPM (device-id 0x1, rev-id 1) [ 1.966306] tpm tpm0: starting up the TPM manually [ 1.981753] ahci 0000:00:1f.2: version 3.0 [ 1.984503] PCI Interrupt Link [GSIA] enabled at IRQ 16 ..... ..... [ 2.086645] ima: Secure boot enabled: ignoring ima_appraise=off option [ 2.086650] ima: Allocated hash algorithm: sha256 [ 2.100662] evm: Initialising EVM extended attributes: [ 2.100689] audit: type=1807 audit(1602760724.564:2): action=measure func=KEX EC_KERNEL_CHECK res=1 [ 2.101849] evm: security.selinux [ 2.101849] evm: security.apparmor [ 2.101849] evm: security.ima [ 2.101850] evm: security.capability [ 2.101850] evm: HMAC attrs: 0x1 [ 2.102083] PM: Magic number: 8:588:327 [ 2.106723] audit: type=1807 audit(1602760724.564:3): action=measure func=MODD ULE_CHECK res=1 [ 2.118761] RAS: Correctable Errors collector initialized. ..... ----- In terms of TPM 2.0, swtpm seems to have issues while running the selftest so I verified this patch on my laptop as well because it has a real TPM 2.0 chip: ----- [ 0.000000] microcode: microcode updated early to revision 0xd6, date = 2020- 04-27 [ 0.000000] Linux version 5.9.0-858-g865c50e1d279-1.g8764d18-vanilla (geeko@b uildhost) (gcc (SUSE Linux) 10.2.1 20200825 [revision c0746a1beb1ba073c7981eb09f 55b3d993b32e5c], GNU ld (GNU Binutils; openSUSE Tumbleweed) 2.34.0.20200325-1) # 1 SMP Wed Oct 1404:00:11 UTC 2020 (8764d18) [ 0.000000] Command line: BOOT_IMAGE=/boot/vmlinuz-5.9.0-858-g865c50e1d279-1. g8764d18-vanilla root=UUID=2184bbb7-780e-48a2-98b8-3a8cd3366e5e splash=silent re sume=/dev/disk/by-id/ata-INTEL_SSDSCKKF512G8_SATA_512GB_BTLA82850G3E512K-part3 i gnore_loglevel crashkernel=222M,high crashkernel=72M,low ima_appraise=log [ 0.000000] x86/fpu: Supporting XSAVE feature 0x001: 'x87 floating point regi sters' ..... ..... [ 0.000000] efi: EFI v2.60 by American Megatrends [ 0.000000] efi: ACPI=0x7a34e000 ACPI 2.0=0x7a34e000 SMBIOS=0xf0000 SMBIOS 3. 0=0xf0020 TPMFinalLog=0x7ac8b000 ESRT=0x7b0e0018 MEMATTR=0x7762d018 RNG=0x7b133f 18 TPMEventLog=0x6ab96018 [ 0.000000] efi: seeding entropy pool [ 0.000000] random: fast init done [ 0.000000] SMBIOS 3.1.0 present. [ 0.000000] DMI: Dell Inc. Latitude 7490/0KP0FT, BIOS 1.14.0 01/22/2020 ..... ..... [ 7.182711] Non-volatile memory driver v1.3 [ 7.182748] Linux agpgart interface v0.103 [ 7.210826] tpm_tis MSFT0101:00: 2.0 TPM (device-id 0xFC, rev-id 1) [ 7.234324] ahci 0000:00:17.0: version 3.0 ..... ..... [ 7.288573] integrity: Loading X.509 certificate: UEFI:MokListRT [ 7.289299] integrity: Loaded X.509 cert 'openSUSE Secure Boot CA: 6842600de22c4c477e95be23dfea9513e5971762' [ 7.289913] ima: Secure boot enabled: ignoring ima_appraise=log option [ 7.289921] ima: Allocated hash algorithm: sha256 [ 7.314854] evm: Initialising EVM extended attributes: [ 7.314878] audit: type=1807 audit(1602778176.572:2): action=measure func=KEXEC_KERNEL_CHECK res=1 [ 7.316656] evm: security.selinux [ 7.316659] evm: security.apparmor [ 7.318493] audit: type=1807 audit(1602778176.572:3): action=measure func=MODULE_CHECK res=1 [ 7.320092] evm: security.ima [ 7.320093] evm: security.capability [ 7.322554] evm: HMAC attrs: 0x1
diff --git a/include/linux/ima.h b/include/linux/ima.h index d15100de6cdd..8aefee9c36ab 100644 --- a/include/linux/ima.h +++ b/include/linux/ima.h @@ -27,6 +27,12 @@ extern void ima_post_path_mknod(struct dentry *dentry); extern int ima_file_hash(struct file *file, char *buf, size_t buf_size); extern void ima_kexec_cmdline(int kernel_fd, const void *buf, int size); +#ifdef CONFIG_IMA_APPRAISE_BOOTPARAM +extern void ima_appraise_parse_cmdline(void); +#else +static inline void ima_appraise_parse_cmdline(void) {} +#endif + #ifdef CONFIG_IMA_KEXEC extern void ima_add_kexec_buffer(struct kimage *image); #endif diff --git a/security/integrity/ima/ima_appraise.c b/security/integrity/ima/ima_appraise.c index 3dd8c2e4314e..dfe08489faea 100644 --- a/security/integrity/ima/ima_appraise.c +++ b/security/integrity/ima/ima_appraise.c @@ -16,12 +16,19 @@ #include "ima.h" -static int __init default_appraise_setup(char *str) -{ #ifdef CONFIG_IMA_APPRAISE_BOOTPARAM +static char *ima_appraise_cmdline_default __initdata; +core_param(ima_appraise, ima_appraise_cmdline_default, charp, 0); + +void __init ima_appraise_parse_cmdline(void) +{ + const char *str = ima_appraise_cmdline_default; bool sb_state = arch_ima_get_secureboot(); int appraisal_state = ima_appraise; + if (!str) + return; + if (strncmp(str, "off", 3) == 0) appraisal_state = 0; else if (strncmp(str, "log", 3) == 0) @@ -42,11 +49,8 @@ static int __init default_appraise_setup(char *str) } else { ima_appraise = appraisal_state; } -#endif - return 1; } - -__setup("ima_appraise=", default_appraise_setup); +#endif /* * is_ima_appraise_enabled - return appraise status diff --git a/security/integrity/ima/ima_main.c b/security/integrity/ima/ima_main.c index a5a2ae36a36d..2497412b1ad2 100644 --- a/security/integrity/ima/ima_main.c +++ b/security/integrity/ima/ima_main.c @@ -865,6 +865,7 @@ static int __init init_ima(void) { int error; + ima_appraise_parse_cmdline(); ima_init_template_list(); hash_setup(CONFIG_IMA_DEFAULT_HASH); error = ima_init();
Chester reports that it is necessary to introduce a new way to pass the EFI secure boot status between the EFI stub and the core kernel on ARM systems. The usual way of obtaining this information is by checking the SecureBoot and SetupMode EFI variables, but this can only be done after the EFI variable workqueue is created, which occurs in a subsys_initcall(), whereas arch_ima_get_secureboot() is called much earlier by the IMA framework. However, the IMA framework itself is started as a late_initcall, and the only reason the call to arch_ima_get_secureboot() occurs so early is because it happens in the context of a __setup() callback that parses the ima_appraise= command line parameter. So let's refactor this code a little bit, by using a core_param() callback to capture the command line argument, and deferring any reasoning based on its contents to the IMA init routine. Cc: Chester Lin <clin@suse.com> Cc: Mimi Zohar <zohar@linux.ibm.com> Cc: Dmitry Kasatkin <dmitry.kasatkin@gmail.com> Cc: James Morris <jmorris@namei.org> Cc: "Serge E. Hallyn" <serge@hallyn.com> Link: https://lore.kernel.org/linux-arm-kernel/20200904072905.25332-2-clin@suse.com/ Signed-off-by: Ard Biesheuvel <ardb@kernel.org> --- v2: rebase onto series 'integrity: improve user feedback for invalid bootparams' include/linux/ima.h | 6 ++++++ security/integrity/ima/ima_appraise.c | 16 ++++++++++------ security/integrity/ima/ima_main.c | 1 + 3 files changed, 17 insertions(+), 6 deletions(-)