Add constant time memcmp_ct function #3

jbech-linaro · 2014-06-22T20:42:54Z

This patch adds a constant time memory compare function that mitigates
timing attacks. This patch also fixes three locations where this could
be a problem.

Signed-off-by: Joakim Bech joakim.bech@linaro.org

pascal-brand38 · 2014-06-23T06:56:17Z

The original memcmp() returns an integer less than, equal to, or greater than zero if s1 is found, respectively, to be less than, to match, or be greater than s2.
In this version of memcmp_ct(), it returns 0 if the memories equal, and non-zero otherwise.

Is not it an issue not to have the very-same behavior, even if I agree we only look if the result is 0 or not?

jbech-linaro · 2014-06-23T07:08:22Z

True, due to the nature how to check this, it would be tricky to have the same behavior as the libc memcmp. Adding a comment that clarifies this might be enough?

pascal-brand38 · 2014-06-23T07:57:55Z

Yes, adding a comment is sufficient I guess.

jenswi-linaro · 2014-06-23T11:29:50Z

The return value of memcmp() makes it suitable as a comparison function for qsort() and one may expect the same of any function which has "memcmp" in the name. Perhaps we should name it differently to make it clear that it doesn't behave as an ordinary memcmp? What do you think of something like:
bool bufs_are_equal_ct(const void *s1, const void *s2, size_t n);

jbech-linaro · 2014-06-23T11:44:59Z

Added the comment and also renamed the function to buf_compare_ct. Let me know if/when this is OK for you and I'll squash and push using force to make it a single patch.

pascal-brand38 · 2014-06-23T11:59:05Z

Hi,

this is ok for me... but I did not run any tests nor try to compile.
I think there is no problem, but the process would require that we test it.

Can you push the squashed version, and then I'll test on tomorrow morning?

Regards,
Pascal.

On 23 June 2014 13:44, Joakim Bech notifications@github.com wrote:

Added the comment and also renamed the function to buf_compare_ct. Let me
know if/when this is OK for you and I'll squash and push using force to
make it a single patch.

—
Reply to this email directly or view it on GitHub
#3 (comment).

jbech-linaro · 2014-06-23T12:03:14Z

Yes, I can push the squashed version and indeed it must be tested on a device, which I haven't done (Jens and me needs to be able to share the Orly2 board we have so I also can test, at least using "xtest light", I'll sort that out with Jens).

This patch adds a constant time memory compare function that mitigates timing attacks. This patch also fixes three locations where this could be a problem. Signed-off-by: Joakim Bech <joakim.bech@linaro.org>

jbech-linaro · 2014-06-24T07:37:26Z

In a few minutes I will push (--force) the squashed version.

Address Etienne's comments (round #3). Signed-off-by: Jerome Forissier <jerome.forissier@linaro.org>

Re-work drvcrypt_get_alg_hash_oid() and move it to hash_oid.c Signed-off-by: Clement Faure <clement.faure@nxp.com>

Fix the RSA private key format OP-TEE#3. Signed-off-by: Clement Faure <clement.faure@nxp.com>

…river Fix comment typo. Signed-off-by: Clement Faure <clement.faure@nxp.com>

Size of each JR Output ring entry is of 12 bytes for CAAM using address pointer size as 64 bit. The descriptor address pointer thus lies at 32 bit boundary in second output ring entry. 64 bit access of descriptor pointer at 32 bit boundary generates alignment fault. To fix this, descriptor address pointer should be accessed as two 32 bit operations. regression_1004 Test User Crypt TA E/TC:03 00 Core data-abort at address 0xfc09e74c (alignment fault) E/TC:03 00 esr 0x96000021 ttbr0 0x20000fc0d7060 ttbr1 0x00000000 cidr 0x0 E/TC:03 00 cpu OP-TEE#3 cpsr 0x200001c4 E/TC:03 00 x0 00000000fc09e74c x1 0000000000000000 E/TC:03 00 x2 0000000000000050 x3 0000008000010100 E/TC:03 00 x4 0000000000000003 x5 00000000fc0e46e5 E/TC:03 00 x6 00000000fc09e74c x7 00000000fc09df78 E/TC:03 00 x8 0000000000000078 x9 00000000fc09c110 E/TC:03 00 x10 0000000041001900 x11 00000000ab12a911 E/TC:03 00 x12 0000000032e4d24d x13 00000000fc0e46e5 E/TC:03 00 x14 0000000000000000 x15 0000000000000000 E/TC:03 00 x16 00000000fc0e4b88 x17 0000000000000000 E/TC:03 00 x18 0000000000000000 x19 0000000000000000 E/TC:03 00 x20 000000000000270f x21 00000000fc07c000 E/TC:03 00 x22 00000000fc07c000 x23 0000000000000000 E/TC:03 00 x24 00000000fc09e74c x25 00000000fc0716d0 E/TC:03 00 x26 00000000fc09df78 x27 0000000000000000 E/TC:03 00 x28 0000000000000000 x29 00000000fc0e4900 E/TC:03 00 x30 00000000fc01ae8c elr 00000000fc01c124 E/TC:03 00 sp_el0 00000000fc0e4900 E/TC:03 00 TEE load address @ 0xfc000000 E/TC:03 00 Core data-abort at address 0xfc09e74c .debug_info+649036 (alignment fault) E/TC:03 00 Call stack: E/TC:03 00 0x00000000fc01c124 caam_desc_pop at core/drivers/crypto/caam/caam_desc.c:88 E/TC:03 00 0x00000000fc01b2ac caam_jr_enqueue at core/drivers/crypto/caam/caam_jr.c:510 E/TC:03 00 0x00000000fc02247c caam_cipher_block at core/drivers/crypto/caam/cipher/caam_cipher.c:331 E/TC:03 00 0x00000000fc022970 do_update_cipher at core/drivers/crypto/caam/cipher/caam_cipher.c:976 E/TC:03 00 0x00000000fc01a290 cipher_update at core/drivers/crypto/crypto_api/cipher/cipher.c:144 E/TC:03 00 0x00000000fc03562c tee_fs_fek_crypt at core/tee/tee_fs_key_manager.c:118 E/TC:03 00 0x00000000fc033dbc verify_root at core/tee/fs_htree.c:549 E/TC:03 00 0x00000000fc031edc ree_fs_open_primitive at core/tee/tee_ree_fs.c:416 E/TC:03 00 0x00000000fc0345d0 tee_fs_dirfile_open at core/tee/fs_dirfile.c:122 E/TC:03 00 0x00000000fc0321cc open_dirh at core/tee/tee_ree_fs.c:530 E/TC:03 00 0x00000000fc032498 ree_fs_open at core/tee/tee_ree_fs.c:604 E/TC:03 00 0x00000000fc0363dc tadb_open at core/tee/tadb.c:214 E/TC:03 00 0x00000000fc036c44 tee_tadb_ta_open at core/tee/tadb.c:633 E/TC:03 00 0x00000000fc00578c secstor_ta_open at core/arch/arm/kernel/secstor_ta.c:19 E/TC:03 00 0x00000000fc026658 system_open_ta_binary at core/pta/system.c:259 E/TC:03 00 0x00000000fc005e24 pseudo_ta_enter_invoke_cmd at core/arch/arm/kernel/pseudo_ta.c:199 E/TC:03 00 0x00000000fc0250dc tee_ta_invoke_command at core/kernel/tee_ta_manager.c:761 E/TC:03 00 0x00000000fc02b398 syscall_invoke_ta_command at core/tee/tee_svc.c:885 E/TC:03 00 0x00000000fc0123bc tee_svc_do_call at core/arch/arm/tee/arch_svc_a64.S:141 E/TC:03 00 0x00000000fc00811c thread_svc_handler at core/arch/arm/kernel/thread.c:1378 E/TC:03 00 0x00000000fc0039e0 el0_svc at core/arch/arm/kernel/thread_a64.S:639 Signed-off-by: Ruchika Gupta <ruchika.gupta@nxp.com> Reviewed-by: Sahil Malhotra <sahil.malhotra@nxp.com> Reviewed-by: Clement Faure <clement.faure@nxp.com>

Size of each JR Output ring entry is of 12 bytes for CAAM using address pointer size as 64 bit. The descriptor address pointer thus lies at 32 bit boundary in second output ring entry. 64 bit access of descriptor pointer at 32 bit boundary generates alignment fault. To fix this, descriptor address pointer should be accessed as two 32 bit operations. regression_1004 Test User Crypt TA E/TC:03 00 Core data-abort at address 0xfc09e74c (alignment fault) E/TC:03 00 esr 0x96000021 ttbr0 0x20000fc0d7060 ttbr1 0x00000000 cidr 0x0 E/TC:03 00 cpu OP-TEE#3 cpsr 0x200001c4 E/TC:03 00 x0 00000000fc09e74c x1 0000000000000000 E/TC:03 00 x2 0000000000000050 x3 0000008000010100 E/TC:03 00 x4 0000000000000003 x5 00000000fc0e46e5 E/TC:03 00 x6 00000000fc09e74c x7 00000000fc09df78 E/TC:03 00 x8 0000000000000078 x9 00000000fc09c110 E/TC:03 00 x10 0000000041001900 x11 00000000ab12a911 E/TC:03 00 x12 0000000032e4d24d x13 00000000fc0e46e5 E/TC:03 00 x14 0000000000000000 x15 0000000000000000 E/TC:03 00 x16 00000000fc0e4b88 x17 0000000000000000 E/TC:03 00 x18 0000000000000000 x19 0000000000000000 E/TC:03 00 x20 000000000000270f x21 00000000fc07c000 E/TC:03 00 x22 00000000fc07c000 x23 0000000000000000 E/TC:03 00 x24 00000000fc09e74c x25 00000000fc0716d0 E/TC:03 00 x26 00000000fc09df78 x27 0000000000000000 E/TC:03 00 x28 0000000000000000 x29 00000000fc0e4900 E/TC:03 00 x30 00000000fc01ae8c elr 00000000fc01c124 E/TC:03 00 sp_el0 00000000fc0e4900 E/TC:03 00 TEE load address @ 0xfc000000 E/TC:03 00 Core data-abort at address 0xfc09e74c .debug_info+649036 (alignment fault) E/TC:03 00 Call stack: E/TC:03 00 0x00000000fc01c124 caam_desc_pop at core/drivers/crypto/caam/caam_desc.c:88 E/TC:03 00 0x00000000fc01b2ac caam_jr_enqueue at core/drivers/crypto/caam/caam_jr.c:510 E/TC:03 00 0x00000000fc02247c caam_cipher_block at core/drivers/crypto/caam/cipher/caam_cipher.c:331 E/TC:03 00 0x00000000fc022970 do_update_cipher at core/drivers/crypto/caam/cipher/caam_cipher.c:976 E/TC:03 00 0x00000000fc01a290 cipher_update at core/drivers/crypto/crypto_api/cipher/cipher.c:144 E/TC:03 00 0x00000000fc03562c tee_fs_fek_crypt at core/tee/tee_fs_key_manager.c:118 E/TC:03 00 0x00000000fc033dbc verify_root at core/tee/fs_htree.c:549 E/TC:03 00 0x00000000fc031edc ree_fs_open_primitive at core/tee/tee_ree_fs.c:416 E/TC:03 00 0x00000000fc0345d0 tee_fs_dirfile_open at core/tee/fs_dirfile.c:122 E/TC:03 00 0x00000000fc0321cc open_dirh at core/tee/tee_ree_fs.c:530 E/TC:03 00 0x00000000fc032498 ree_fs_open at core/tee/tee_ree_fs.c:604 E/TC:03 00 0x00000000fc0363dc tadb_open at core/tee/tadb.c:214 E/TC:03 00 0x00000000fc036c44 tee_tadb_ta_open at core/tee/tadb.c:633 E/TC:03 00 0x00000000fc00578c secstor_ta_open at core/arch/arm/kernel/secstor_ta.c:19 E/TC:03 00 0x00000000fc026658 system_open_ta_binary at core/pta/system.c:259 E/TC:03 00 0x00000000fc005e24 pseudo_ta_enter_invoke_cmd at core/arch/arm/kernel/pseudo_ta.c:199 E/TC:03 00 0x00000000fc0250dc tee_ta_invoke_command at core/kernel/tee_ta_manager.c:761 E/TC:03 00 0x00000000fc02b398 syscall_invoke_ta_command at core/tee/tee_svc.c:885 E/TC:03 00 0x00000000fc0123bc tee_svc_do_call at core/arch/arm/tee/arch_svc_a64.S:141 E/TC:03 00 0x00000000fc00811c thread_svc_handler at core/arch/arm/kernel/thread.c:1378 E/TC:03 00 0x00000000fc0039e0 el0_svc at core/arch/arm/kernel/thread_a64.S:639 Signed-off-by: Ruchika Gupta <ruchika.gupta@nxp.com> Reviewed-by: Sahil Malhotra <sahil.malhotra@nxp.com> Reviewed-by: Clement Faure <clement.faure@nxp.com> Acked-by: Etienne Carriere <etienne.carriere@linaro.org>

Size of each JR Output ring entry is of 12 bytes for CAAM using address pointer size as 64 bit. The descriptor address pointer thus lies at 32 bit boundary in second output ring entry. 64 bit access of descriptor pointer at 32 bit boundary generates alignment fault. To fix this, descriptor address pointer should be accessed as two 32 bit operations. regression_1004 Test User Crypt TA E/TC:03 00 Core data-abort at address 0xfc09e74c (alignment fault) E/TC:03 00 esr 0x96000021 ttbr0 0x20000fc0d7060 ttbr1 0x00000000 cidr 0x0 E/TC:03 00 cpu #3 cpsr 0x200001c4 E/TC:03 00 x0 00000000fc09e74c x1 0000000000000000 E/TC:03 00 x2 0000000000000050 x3 0000008000010100 E/TC:03 00 x4 0000000000000003 x5 00000000fc0e46e5 E/TC:03 00 x6 00000000fc09e74c x7 00000000fc09df78 E/TC:03 00 x8 0000000000000078 x9 00000000fc09c110 E/TC:03 00 x10 0000000041001900 x11 00000000ab12a911 E/TC:03 00 x12 0000000032e4d24d x13 00000000fc0e46e5 E/TC:03 00 x14 0000000000000000 x15 0000000000000000 E/TC:03 00 x16 00000000fc0e4b88 x17 0000000000000000 E/TC:03 00 x18 0000000000000000 x19 0000000000000000 E/TC:03 00 x20 000000000000270f x21 00000000fc07c000 E/TC:03 00 x22 00000000fc07c000 x23 0000000000000000 E/TC:03 00 x24 00000000fc09e74c x25 00000000fc0716d0 E/TC:03 00 x26 00000000fc09df78 x27 0000000000000000 E/TC:03 00 x28 0000000000000000 x29 00000000fc0e4900 E/TC:03 00 x30 00000000fc01ae8c elr 00000000fc01c124 E/TC:03 00 sp_el0 00000000fc0e4900 E/TC:03 00 TEE load address @ 0xfc000000 E/TC:03 00 Core data-abort at address 0xfc09e74c .debug_info+649036 (alignment fault) E/TC:03 00 Call stack: E/TC:03 00 0x00000000fc01c124 caam_desc_pop at core/drivers/crypto/caam/caam_desc.c:88 E/TC:03 00 0x00000000fc01b2ac caam_jr_enqueue at core/drivers/crypto/caam/caam_jr.c:510 E/TC:03 00 0x00000000fc02247c caam_cipher_block at core/drivers/crypto/caam/cipher/caam_cipher.c:331 E/TC:03 00 0x00000000fc022970 do_update_cipher at core/drivers/crypto/caam/cipher/caam_cipher.c:976 E/TC:03 00 0x00000000fc01a290 cipher_update at core/drivers/crypto/crypto_api/cipher/cipher.c:144 E/TC:03 00 0x00000000fc03562c tee_fs_fek_crypt at core/tee/tee_fs_key_manager.c:118 E/TC:03 00 0x00000000fc033dbc verify_root at core/tee/fs_htree.c:549 E/TC:03 00 0x00000000fc031edc ree_fs_open_primitive at core/tee/tee_ree_fs.c:416 E/TC:03 00 0x00000000fc0345d0 tee_fs_dirfile_open at core/tee/fs_dirfile.c:122 E/TC:03 00 0x00000000fc0321cc open_dirh at core/tee/tee_ree_fs.c:530 E/TC:03 00 0x00000000fc032498 ree_fs_open at core/tee/tee_ree_fs.c:604 E/TC:03 00 0x00000000fc0363dc tadb_open at core/tee/tadb.c:214 E/TC:03 00 0x00000000fc036c44 tee_tadb_ta_open at core/tee/tadb.c:633 E/TC:03 00 0x00000000fc00578c secstor_ta_open at core/arch/arm/kernel/secstor_ta.c:19 E/TC:03 00 0x00000000fc026658 system_open_ta_binary at core/pta/system.c:259 E/TC:03 00 0x00000000fc005e24 pseudo_ta_enter_invoke_cmd at core/arch/arm/kernel/pseudo_ta.c:199 E/TC:03 00 0x00000000fc0250dc tee_ta_invoke_command at core/kernel/tee_ta_manager.c:761 E/TC:03 00 0x00000000fc02b398 syscall_invoke_ta_command at core/tee/tee_svc.c:885 E/TC:03 00 0x00000000fc0123bc tee_svc_do_call at core/arch/arm/tee/arch_svc_a64.S:141 E/TC:03 00 0x00000000fc00811c thread_svc_handler at core/arch/arm/kernel/thread.c:1378 E/TC:03 00 0x00000000fc0039e0 el0_svc at core/arch/arm/kernel/thread_a64.S:639 Signed-off-by: Ruchika Gupta <ruchika.gupta@nxp.com> Reviewed-by: Sahil Malhotra <sahil.malhotra@nxp.com> Reviewed-by: Clement Faure <clement.faure@nxp.com> Acked-by: Etienne Carriere <etienne.carriere@linaro.org>

A squashed series of changes in stm32mp1 clock driver for supporting SCMI server as a clock consumer. These changes have not been upstream yet in OP-TEE OS. Below the details. Signed-off-by: Etienne Carriere <etienne.carriere@linaro.org> ### Commit message OP-TEE#1 plat-stm32mp1: clock: remove oscillators and PLLs from shared resources Oscillators, PLL1 and PLL2 are not resources allocated upon platform configuration, these are always under secure world control. This change removes them fro the list of the shared resources. Since these resource are always secure, there is no need to look up clock tree when a leaf clock is registered as secure to know which parent clock(s) is/are secure. This removes functions from shared_resources.c and stm32mp1_clk.c. stm32mp_register_clock_parents_secure() can be removed and all its private dependencies. Signed-off-by: Etienne Carriere <etienne.carriere@st.com> ### Commit message OP-TEE#2 plat-stm32mp1: clock: fix mcu/axi parent clock Correct MCU clock parent selector: MCU subsystem clock is derived from clock PLL3_P, not PLL3. Correct AXI clock parent selector: AXI subsystem clock is derived from clock PLL2_P, not PLL2. This change also renames MCU clock and AXI clock resources to prevent confusion. Signed-off-by: Etienne Carriere <etienne.carriere@st.com> ### Commit message OP-TEE#3 plat-stm32mp1: clock: allow tree lookup for several system clocks Oscillators, PLLs and some system clocks can be related straight to a parent clock. Prior this change were only oscillators and few clocks supported by this look up. This changes adds PLLs and other system clocks. This enables for flexible use of clock tree exploration when computing a clock frequency value. Signed-off-by: Etienne Carriere <etienne.carriere@st.com> ### Commit message OP-TEE#4 plat-stm32mp1: clock: handle always on clocks Oscillators and PLLs are not gated on stm32mp_clk_enable/disable() calls. This change allows function to blindly call clock gating function on always on clocks. Gating these clock is out of the scope of this change even if preferred for power consumption optimization. Signed-off-by: Etienne Carriere <etienne.carriere@st.com> ### Commit message OP-TEE#5 plat-stm32mp1: clock: add rtc as gateable clock Add clock RTC as a clock one can access through the stm32_util.h API function stm32_clock_*(). Signed-off-by: Etienne Carriere <etienne.carriere@st.com> ### Commit message OP-TEE#6 plat-stm32mp1: clock: enable some secure clocks at init With this change some system clock are enabled by core at boot time and have a reference counter synchronized with the clock hardware state. Signed-off-by: Etienne Carriere <etienne.carriere@st.com> ### Commit message OP-TEE#7 plat-stm32mp1: factorize rtc clock gating bit position For consistency, define macro RCC_BDCR_RTCCKEN_POS in stm32mp1_rcc.h to factorize definition of the RTC clock gating resources. Signed-off-by: Etienne Carriere <etienne.carriere@st.com> ### Commit message OP-TEE#8 plat-stm32mp1: add mdma secure clock Add support for MDMA secure clock. Signed-off-by: Etienne Carriere <etienne.carriere@st.com> ### Commit message OP-TEE#9 plat-stm32mp1: remove unused usb non-secure clock Remove unused clocks USBO_CLK and USBPHY_K resources. Signed-off-by: Etienne Carriere <etienne.carriere@st.com> ### Commit message OP-TEE#10 plat-stm32mp1: clock: secure and non-secure gateable clocks Array stm32mp1_clk_gate[] defines the clock resources. This change add a secure attribute to the clock: secure upon RCC[TZEN] (SEC), secure upon RCC[TZEN] and RCC[MCKPROT] (MKP) or always accessible from non-secure (N_S). At init, lookup clock tree to ensure that parents of a secure clock are registered a secure resources in the shared_resources.c driver. Non-secure clock that OP-TEE expect to enable are enabled without increase the clock refcount. For consistency, such clocks are not disabled by core. Such clocks may be accessed by OP-TEE Core when the non-secure world is not executing, for example at boot time or could be when system is suspending/resuming. Signed-off-by: Etienne Carriere <etienne.carriere@st.com> ### Commit message OP-TEE#11 plat-stm32mp1: clock: fixup parent clock ids Use _UNKNOWN_ID macro rather than 0xff for clocks parent IDs that do not relate to a gateable clock. Fix parent clock ID _HSE_KER_DIV2 that relates to clock CK_HSE_KER_DIV2, not CK_HSE. Signed-off-by: Etienne Carriere <etienne.carriere@st.com> ### Commit message OP-TEE#12 plat-stm32mp1: clock: don't embed unused non-secure uart clock Embed UART parent clock resource upon CFG_WITH_NSEC_UARTS=y. This configuration switch was already used to embed or not the non-secure UART clocks but not the resources used to looks there ascendant clocks. Signed-off-by: Etienne Carriere <etienne.carriere@st.com> ### Commit message OP-TEE#13 plat-stm32mp1: shared resources: get shared clock controller state stm32mp_nsec_can_access_clock() reports whether a clock is assigned to the secure world only, or when it can be manipulated by the non-secure world. Signed-off-by: Etienne Carriere <etienne.carriere@st.com> ### Commit message OP-TEE#14 plat-stm32mp1: shared resource: remove unused stm32mp_clock_is_*() Remove unused functions stm32mp_clock_is_shareable(), stm32mp_clock_is_shared() and stm32mp_clock_is_non_secure(). Signed-off-by: Etienne Carriere <etienne.carriere@st.com>

Adds checks in e32_relocate() that sym_tab is assigned a symbol table before using it. This fixes coverity scan: CID 1501826 (#1 of 3): Explicit null dereferenced (FORWARD_NULL) CID 1501826 (#2 of 3): Explicit null dereferenced (FORWARD_NULL) CID 1501826 (#3 of 3): Explicit null dereferenced (FORWARD_NULL) Signed-off-by: Jens Wiklander <jens.wiklander@linaro.org>

Check only for (src|dst, sz) in tests/invoke.c Signed-off-by: Anton Rybakov <a.rybakov@omp.ru>

Adds checks in e32_relocate() that sym_tab is assigned a symbol table before using it. This fixes coverity scan: CID 1501826 (#1 of 3): Explicit null dereferenced (FORWARD_NULL) CID 1501826 (#2 of 3): Explicit null dereferenced (FORWARD_NULL) CID 1501826 (#3 of 3): Explicit null dereferenced (FORWARD_NULL) Acked-by: Jerome Forissier <jerome@forissier.org> Signed-off-by: Jens Wiklander <jens.wiklander@linaro.org>

Integrate FreeRTOS as secure scheduler in OP-TEE OS.

Use the job ring OP-TEE#3 on i.mx8dxl to avoid resource conflict with other software stacks. Signed-off-by: Clement Faure <clement.faure@nxp.com>

Use the job ring OP-TEE#3 on i.mx8dxl to avoid resource conflict with other software stacks. Signed-off-by: Clement Faure <clement.faure@nxp.com> Acked-by: Jerome Forissier <jerome.forissier@linaro.org>

Use the job ring #3 on i.mx8dxl to avoid resource conflict with other software stacks. Signed-off-by: Clement Faure <clement.faure@nxp.com> Acked-by: Jerome Forissier <jerome.forissier@linaro.org>

Add constant time buf_compare_ct function

1665420

This patch adds a constant time memory compare function that mitigates timing attacks. This patch also fixes three locations where this could be a problem. Signed-off-by: Joakim Bech <joakim.bech@linaro.org>

ghost merged commit 1665420 into OP-TEE:master Jun 24, 2014

jbech-linaro deleted the const_time_memcmp branch June 24, 2014 21:34

neil1899 mentioned this pull request Oct 22, 2015

tee-supplicant panic & ta panic #506

Closed

acy54321 mentioned this pull request Oct 26, 2016

32bit OPTEE OS System call interface Question #1139

Closed

This was referenced Mar 15, 2017

communication to hikey via network problems #1392

Closed

hikey_debian: tee-supplicant fails: please clarify the build instruction around tee-supplicant, xtest #1432

Closed

dhclouisyang mentioned this pull request Aug 4, 2017

after long time playback, tee encounters core data-abort #1734

Closed

jforissier referenced this pull request in jforissier/optee_os Aug 8, 2017

[Review] Add support for several user TA stores

1343282

Address Etienne's comments (round #3). Signed-off-by: Jerome Forissier <jerome.forissier@linaro.org>

jforissier referenced this pull request in jforissier/optee_os Aug 8, 2017

[Review] Add support for early Trusted Applications

1349800

Address Etienne's comments (round #3). Signed-off-by: Jerome Forissier <jerome.forissier@linaro.org>

jforissier referenced this pull request in jforissier/optee_os Aug 8, 2017

[Review] Add support for compressed early TAs

09adabb

Address Etienne's comments (round #3). Signed-off-by: Jerome Forissier <jerome.forissier@linaro.org>

0xB0D mentioned this pull request Aug 30, 2018

Booting OPTEE on iMX7 #2513

Closed

daniel-thompson mentioned this pull request Oct 19, 2018

synquacer: Add RNG pseudo TA #2564

Merged

lorc mentioned this pull request Jan 14, 2019

xtest 1013 hangs in normal world #2737

Closed

PratikGajera mentioned this pull request May 20, 2019

Unhandled fault: external abort on non-linefetch #3023

Closed

sxc1143106070 mentioned this pull request Aug 23, 2019

xtes and optee_example_hello_world all crashed, below is logs,please help me! #3218

Closed

nageshshamnur mentioned this pull request Nov 14, 2019

normal world CA is hanging at getentropy( ) syscall #3396

Closed

AnishKumarAkGk mentioned this pull request Feb 25, 2020

Accessing GPIO from TA in Raspberry Pi 3 #1496

Closed

clementfaure added a commit to clementfaure/optee_os that referenced this pull request Feb 27, 2020

[Review OP-TEE#3] Check OID hash algorithm returned

6fe3ff3

Re-work drvcrypt_get_alg_hash_oid() and move it to hash_oid.c Signed-off-by: Clement Faure <clement.faure@nxp.com>

clementfaure added a commit to clementfaure/optee_os that referenced this pull request Mar 6, 2020

drivers: caam: fix RSA key format OP-TEE#3

525d117

Fix the RSA private key format OP-TEE#3. Signed-off-by: Clement Faure <clement.faure@nxp.com>

jforissier mentioned this pull request Mar 6, 2020

Fix CAAM RSA inconsistent fails #3661

Merged

clementfaure added a commit to clementfaure/optee_os that referenced this pull request Mar 9, 2020

[Review OP-TEE#3] core: driver: generic resources for crypto device d…

97b2a6b

…river Fix comment typo. Signed-off-by: Clement Faure <clement.faure@nxp.com>

ZhikuiRen mentioned this pull request Apr 13, 2020

Kernel boot stuck at idle after optee. Need help. #3745

Closed

a-rybakov added a commit to a-rybakov/optee_os-1 that referenced this pull request Jun 3, 2021

fix comments by Etienne OP-TEE#3

5c77d9e

Check only for (src|dst, sz) in tests/invoke.c Signed-off-by: Anton Rybakov <a.rybakov@omp.ru>

DevendraDevadiga mentioned this pull request Jul 28, 2021

Increase TA/TEE RAM size #4768

Closed

panda-balarka mentioned this pull request Oct 11, 2021

SUNXI_A64 - itr_add and itr_enable causes kernel panic #4910

Closed

klcib508 mentioned this pull request Feb 22, 2022

optee for kernel5.10 #5185

Closed

huangcm0101 mentioned this pull request Jul 13, 2022

how to debug PTA code? #5438

Closed

weizhaojiang mentioned this pull request Nov 3, 2022

core: implement a method to dump user ta's runtime status. #5590

Closed

gianlucascopelliti pushed a commit to AuthenticExecution/optee_os that referenced this pull request Dec 15, 2022

Merge pull request OP-TEE#3 from Distrinet-TACOS/free-rtos

7916a03

Integrate FreeRTOS as secure scheduler in OP-TEE OS.

zzbios mentioned this pull request Jan 31, 2023

optee secure_storage example CA fails to run #5803

Closed

zzbios mentioned this pull request Feb 26, 2023

Failed to run xtest 1034 #5850

Closed

zzbios mentioned this pull request Apr 6, 2023

Failed to run large size TA #5943

Closed

clementfaure mentioned this pull request Nov 2, 2023

drivers: crypto: caam: use job ring 3 on i.mx8dxlevk #6422

Merged

Amit-Radur mentioned this pull request Nov 8, 2023

Reading secure object from OPTEE-OS core #6441

Closed

Amit-Radur mentioned this pull request Dec 8, 2023

crypto_authenc_update_aad on decrypt path in hash tree #6532

Closed

Muhammadhassan1234 mentioned this pull request Aug 15, 2024

optee rpmb secure storage example in u-boot: TEE_ERROR_ITEM_NOT_FOUND #6986

Closed

This pull request was closed.

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add constant time memcmp_ct function #3

Add constant time memcmp_ct function #3

jbech-linaro commented Jun 22, 2014

pascal-brand38 commented Jun 23, 2014

jbech-linaro commented Jun 23, 2014

pascal-brand38 commented Jun 23, 2014

jenswi-linaro commented Jun 23, 2014

jbech-linaro commented Jun 23, 2014

pascal-brand38 commented Jun 23, 2014

jbech-linaro commented Jun 23, 2014

jbech-linaro commented Jun 24, 2014

Add constant time memcmp_ct function #3

Add constant time memcmp_ct function #3

Conversation

jbech-linaro commented Jun 22, 2014

pascal-brand38 commented Jun 23, 2014

jbech-linaro commented Jun 23, 2014

pascal-brand38 commented Jun 23, 2014

jenswi-linaro commented Jun 23, 2014

jbech-linaro commented Jun 23, 2014

pascal-brand38 commented Jun 23, 2014

jbech-linaro commented Jun 23, 2014

jbech-linaro commented Jun 24, 2014