mbed-tls July 2023

mbed-tls@lists.trustedfirmware.org

7 participants
15 discussions

by Ruchika Gupta

Hi, I am currently working on PSA driver for our SoC's security crypto-processor. This processor needs to do some handling when a key is destroyed. In the current implementation in mbedTLS, I don't see a driver wrapper function available in psa_destroy_key(). Is there a specific reason for not having wrapper for driver function available for psa_destroy_key() ? Another query pertaining to the tests in the testsuite in mbedTLS. I was exploring if I could reuse the tests for the crypto-processor implementation. Have these been written with this kind of reuse in mind ? Basically I would like the ability to add driver location to the cases. The few cases I have looked at for psa seem to be very specific. Any pointers/suggestions if anyone is reusing this test suite to test their specific PSA drivers? Regards, Ruchika

3 years, 1 month

Crash observed while using MBED-TLS

by Ajay Mishra

Hi All, Thank you for sharing the MBED-TLS open source platform. I need help with on of the crashes that I am observing right now. Below are my setup details: Below are my setup details: System information Client side setup Platform : Micro controller based platform stm32 OS : FreeRTOS Compiler: gcc-arm-none-eabi-8-2019-q3-update Below is the backtrace: (gdb) bt #0 panicCB (file=0xc94d7 "memory", line=734, message=0x2000ff50 <sli_wisun_event_loop_task_stack+4224> "Ajay Malloc Failed 88") at build/3pp/freertos/FreeRTOS/Source/include/../../Source/portable/GCC/ARM_CM3/portmacro.h:237 #1 0x0003b2ae in panicHalt_ (file=file@entry=0xc94d7 "memory", line=line@entry=734, message=<optimized out>) at src/platform/core/assert.c:34 #2 0x0003bf0a in reallocInternal (p=<optimized out>, size=68, zeroFill=<optimized out>, file=0xc94d7 "memory", line=734) at src/platform/core/memory.c:317 #3 0x0003c2d8 in reallocWrapper (p=p@entry=0x0, size=<optimized out>, zeroFill=zeroFill@entry=true, file=file@entry=0xc94d7 "memory", line=line@entry=734) at src/platform/core/memory.c:673 #4 0x0003c350 in platformCalloc (n=<optimized out>, size=<optimized out>) at src/platform/core/memory.c:734 #5 0x00043dac in mbedtls_mpi_grow (nblimbs=17, X=0x2001001c <sli_wisun_event_loop_task_stack+4428>) at build/3pp/mbedtls/library/bignum.c:125 #6 mbedtls_mpi_grow (X=0x2001001c <sli_wisun_event_loop_task_stack+4428>, nblimbs=17) at build/3pp/mbedtls/library/bignum.c:115 #7 0x0004414a in mbedtls_mpi_shift_l (X=X@entry=0x2001001c <sli_wisun_event_loop_task_stack+4428>, count=count@entry=256) at build/3pp/mbedtls/library/bignum.c:1016 #8 0x00044818 in mbedtls_mpi_div_mpi (Q=Q@entry=0x0, R=R@entry=0x200100bc <sli_wisun_event_loop_task_stack+4588>, A=A@entry=0x200100bc <sli_wisun_event_loop_task_stack+4588>, B=B@entry=0x200101d0 <sli_wisun_event_loop_task_stack+4864>) at build/3pp/mbedtls/library/bignum.c:1812 #9 0x00044a46 in mbedtls_mpi_mod_mpi (R=R@entry=0x200100bc <sli_wisun_event_loop_task_stack+4588>, A=A@entry=0x200100bc <sli_wisun_event_loop_task_stack+4588>, B=B@entry=0x200101d0 <sli_wisun_event_loop_task_stack+4864>) at build/3pp/mbedtls/library/bignum.c:1917 #10 0x00046486 in ecdsa_verify_restartable (grp=grp@entry=0x20010184 <sli_wisun_event_loop_task_stack+4788>, buf=buf@entry=0x200102a4 <sli_wisun_event_loop_task_stack+5076> "c;\371\274\231xF\004 \207\005\271\257\177\266\224\234\212\205Є\331\366D\304L<\243\303f\246j\250\370\001 ", blen=blen@entry=32, Q=Q@entry=0x2001020c <sli_wisun_event_loop_task_stack+4924>, r=r@entry=0x20010120 <sli_wisun_event_loop_task_stack+4688>, s=s@entry=0x2001012c <sli_wisun_event_loop_task_stack+4700>, rs_ctx=0x0) at build/3pp/mbedtls/library/ecdsa.c:649 #11 0x00046660 in mbedtls_ecdsa_read_signature_restartable (ctx=ctx@entry=0x20010184 <sli_wisun_event_loop_task_stack+4788>, hash=hash@entry=0x200102a4 <sli_wisun_event_loop_task_stack+5076> "c;\371\274\231xF\004 \207\005\271\257\177\266\224\234\212\205Є\331\366D\304L<\243\303f\246j\250\370\001 ", hlen=hlen@entry=32, sig=sig@entry=0x2001d2e1 "0D\002 r\203\237\331\v\346J*m\312\326\346\004+\a\377\373bX\016\233خ\346\030\356\311Nk\270\341L\002 h;\b\301\206\215\346\005\021c\"|\026\353𦕟\327\300ģ\221\\\324s\n^GW", <incomplete sequence \314>, slen=slen@entry=70, rs_ctx=rs_ctx@entry=0x0) at build/3pp/mbedtls/library/ecdsa.c:881 #12 0x0004668e in mbedtls_ecdsa_read_signature (ctx=ctx@entry=0x20010184 <sli_wisun_event_loop_task_stack+4788>, hash=hash@entry=0x200102a4 <sli_wisun_event_loop_task_stack+5076> "c;\371\274\231xF\004 \207\005\271\257\177\266\224\234\212\205Є\331\366D\304L<\243\303f\246j\250\370\001 ", hlen=hlen@entry=32, sig=sig@entry=0x2001d2e1 "0D\002 r\203\237\331\v\346J*m\312\326\346\004+\a\377\373bX\016\233خ\346\030\356\311Nk\270\341L\002 h;\b\301\206\215\346\005\021c\"|\026\353𦕟\327\300ģ\221\\\324s\n^GW", <incomplete sequence \314>, slen=slen@entry=70) at build/3pp/mbedtls/library/ecdsa.c:832 #13 0x0004a830 in ecdsa_verify_wrap (ctx=ctx@entry=0x20010184 <sli_wisun_event_loop_task_stack+4788>, md_alg=md_alg@entry=MBEDTLS_MD_SHA256, hash=hash@entry=0x200102a4 <sli_wisun_event_loop_task_stack+5076> "c;\371\274\231xF\004 \207\005\271\257\177\266\224\234\212\205Є\331\366D\304L<\243\303f\246j\250\370\001 ", hash_len=hash_len@entry=32, sig=sig@entry=0x2001d2e1 "0D\002 r\203\237\331\v\346J*m\312\326\346\004+\a\377\373bX\016\233خ\346\030\356\311Nk\270\341L\002 h;\b\301\206\215\346\005\021c\"|\026\353𦕟\327\300ģ\221\\\324s\n^GW", <incomplete sequence \314>, sig_len=sig_len@entry=70) at build/3pp/mbedtls/library/pk_wrap.c:633 #14 0x0004a872 in eckey_verify_wrap (ctx=<optimized out>, md_alg=<optimized out>, hash=0x200102a4 <sli_wisun_event_loop_task_stack+5076> "c;\371\274\231xF\004 \207\005\271\257\177\266\224\234\212\205Є\331\366D\304L<\243\303f\246j\250\370\001 ", hash_len=32, sig=0x2001d2e1 "0D\002 r\203\237\331\v\346J*m\312\326\346\004+\a\377\373bX\016\233خ\346\030\356\311Nk\270\341L\002 h;\b\301\206\215\346\005\021c\"|\026\353𦕟\327\300ģ\221\\\324s\n^GW", <incomplete sequence \314>, sig_len=70) at build/3pp/mbedtls/library/pk_wrap.c:252 #15 0x0004a3de in mbedtls_pk_verify (ctx=ctx@entry=0x2001ce48, md_alg=<optimized out>, hash=<optimized out>, hash_len=<optimized out>, sig=0x2001d2e1 "0D\002 r\203\237\331\v\346J*m\312\326\346\004+\a\377\373bX\016\233خ\346\030\356\311Nk\270\341L\002 h;\b\301\206\215\346\005\021c\"|\026\353𦕟\327\300ģ\221\\\324s\n^GW", <incomplete sequence \314>, sig_len=70) at build/3pp/mbedtls/library/pk.c:326 #16 0x0004a55c in mbedtls_pk_verify_ext (type=<optimized out>, options=<optimized out>, ctx=ctx@entry=0x2001ce48, md_alg=<optimized out>, hash=<optimized out>, hash@entry=0x200102a4 <sli_wisun_event_loop_task_stack+5076> "c;\371\274\231xF\004 \207\005\271\257\177\266\224\234\212\205Є\331\366D\304L<\243\303f\246j\250\370\001 ", hash_len=<optimized out>, hash_len@entry=32, sig=0x2001d2e1 "0D\002 r\203\237\331\v\346J*m\312\326\346\004+\a\377\373bX\016\233خ\346\030\356\311Nk\270\341L\002 h;\b\301\206\215\346\005\021c\"|\026\353𦕟\327\300ģ\221\\\324s\n^GW", <incomplete sequence \314>, sig_len=70) at build/3pp/mbedtls/library/pk.c:390 #17 0x0004feb4 in x509_crt_check_signature (rs_ctx=<optimized out>, parent=0x2001cd7c, child=0x2001f8a8) at build/3pp/mbedtls/library/x509_crt.c:2474 #18 x509_crt_find_parent_in (rs_ctx=<optimized out>, self_cnt=<optimized out>, path_cnt=<optimized out>, top=<optimized out>, r_signature_is_good=<optimized out>, r_parent=<optimized out>, candidates=<optimized out>, child=<optimized out>) at build/3pp/mbedtls/library/x509_crt.c:2612 #19 x509_crt_find_parent (rs_ctx=<optimized out>, self_cnt=<optimized out>, path_cnt=<optimized out>, signature_is_good=<optimized out>, parent_is_trusted=<optimized out>, parent=<optimized out>, trust_ca=<optimized out>, child=<optimized out>) at build/3pp/mbedtls/library/x509_crt.c:2709 #20 x509_crt_verify_chain (f_ca_cb=0x0, p_ca_cb=0x0, rs_ctx=0x0, ver_chain=0x200102c4 <sli_wisun_event_loop_task_stack+5108>, profile=0xcd08c <mbedtls_x509_crt_profile_default>, ca_crl=0x0, trust_ca=0x2001cd7c, crt=0x2001f8a8) at build/3pp/mbedtls/library/x509_crt.c:2919 #21 x509_crt_verify_restartable_ca_cb (crt=crt@entry=0x2001f8a8, trust_ca=trust_ca@entry=0x2001cd7c, ca_crl=0x0, profile=0xcd08c <mbedtls_x509_crt_profile_default>, cn=0x0, flags=0x2001dd30, f_vrfy=0x695f1 <tls_sec_prot_lib_x509_crt_verify>, p_vrfy=0x2001c8a0, rs_ctx=0x0, p_ca_cb=0x0, f_ca_cb=0x0) at build/3pp/mbedtls/library/x509_crt.c:3150 --Type <RET> for more, q to quit, c to continue without paging-- #22 0x00050090 in mbedtls_x509_crt_verify_restartable (crt=crt@entry=0x2001f8a8, trust_ca=trust_ca@entry=0x2001cd7c, ca_crl=<optimized out>, profile=<optimized out>, cn=0x0, flags=0x2001dd30, f_vrfy=0x695f1 <tls_sec_prot_lib_x509_crt_verify>, p_vrfy=0x2001c8a0, rs_ctx=rs_ctx@entry=0x0) at build/3pp/mbedtls/library/x509_crt.c:3258 #23 0x00058f7c in ssl_parse_certificate_verify (rs_ctx=0x0, chain=0x2001f8a8, authmode=<optimized out>, ssl=0x2001c918) at build/3pp/mbedtls/library/ssl_tls.c:2542 #24 mbedtls_ssl_parse_certificate (ssl=ssl@entry=0x2001c918) at build/3pp/mbedtls/library/ssl_tls.c:2795 #25 0x00051d1c in mbedtls_ssl_handshake_client_step (ssl=0x2001c918) at build/3pp/mbedtls/library/ssl_cli.c:4295 #26 0x00058210 in mbedtls_ssl_handshake_step (ssl=0x2001c918) at build/3pp/mbedtls/library/ssl_tls.c:5668 #27 0x00069922 in tls_sec_prot_lib_process () #28 0x0006b442 in client_tls_sec_prot_state_machine () #29 0x00069af8 in tls_sec_prot_receive () #30 0x00068072 in supp_eap_tls_sec_prot_state_machine () #31 0x000627b8 in supp_eap_tls_sec_prot_receive () #32 0x0006d28c in kmp_eapol_pdu_if_receive () #33 0x0007d602 in ws_eapol_pdu_mpx_data_indication () #34 0x00080f2e in ws_llc_mac_indication_cb.lto_priv () #35 0x000a8df6 in mac_mcps_sap_data_tasklet.lto_priv () #36 0x000ae3f8 in event_loop_thread.lto_priv () #37 0x000414c0 in xEventGroupSetBitsFromISR (xEventGroup=<optimized out>, uxBitsToSet=<optimized out>, pxHigherPriorityTaskWoken=<optimized out>) at build/3pp/freertos/FreeRTOS/Source/event_groups.c:724 Backtrace stopped: previous frame identical to this frame (corrupt stack?) (gdb) (gdb) I see a crash while my node is trying to authenticate using EAP-TLS to a gateway. I am able to consistently reproduce observe this issue. Is this something that has been already solved or is seen by anybody ? Any inputs on this will help. With regards, Ajay.

3 years, 1 month

mbedTLS hardware cryptography support

by Achraf BENLETAIFA

Hi, i wanted to ask if the hardware accelerators and TRNG of the Hardware Security Engine (HSE) of NXP S32K3XX microcontrollers are supported by the mbedTLS Library ? because i can't find it in the "MBEDTLS HARDWARE CRYPTOGRAPHY SUPPORT" list. Best regards, Ben

3 years, 1 month

Re: [EXTERNAL] Re: mbedtls_platform_setup/teardown in TF-M ?

by Shebu Varghese Kuriakose

+ Mbed TLS mailing list for visibility… Regards, Shebu From: Rehan Malak via TF-M <tf-m(a)lists.trustedfirmware.org> Sent: Monday, July 10, 2023 9:44 AM To: Antonio De Angelis <Antonio.DeAngelis(a)arm.com>; tf-m(a)lists.trustedfirmware.org Subject: [TF-M] Re: [EXTERNAL] Re: mbedtls_platform_setup/teardown in TF-M ? Hi Antonio, Thanks a lot for your answer ! Looking at the MbedTLS github link you sent (and the one therein) - https://github.com/Mbed-TLS/mbedtls/issues/6228 PSA: separate driver initialization with a nicer fixed ordering - https://github.com/Mbed-TLS/mbedtls/issues/6007 Define the PSA subsystem initialization interface in Mbed TLS it seems that for the best would be indeed to provide feedback to the MbedTLS/PSA team such that TF-M could just run the adequate psa_* functions. Thanks ! Rehan Intrinsic ID ________________________________ From: Antonio De Angelis <Antonio.DeAngelis(a)arm.com<mailto:Antonio.DeAngelis@arm.com>> Sent: Saturday, July 8, 2023 12:13 AM To: tf-m(a)lists.trustedfirmware.org<mailto:tf-m@lists.trustedfirmware.org> <tf-m(a)lists.trustedfirmware.org<mailto:tf-m@lists.trustedfirmware.org>>; Rehan Malak <Rehan.Malak(a)intrinsic-id.com<mailto:Rehan.Malak@intrinsic-id.com>> Subject: [EXTERNAL] Re: mbedtls_platform_setup/teardown in TF-M ? Hi Rehan, This looks like a non-PSA standardized feature specific to mbed TLS. We don't have any platform in TF-M that requires such setup/teardown so I can't comment from experience, but to me it looks like this would be an mbed TLS specific feature that needs to be hooked underneath the service and into the library. The most natural choice would be indeed to have them as part of the psa_crypto_init() and lower functions, but at this stage I think is not possible to implement this without patching the source code (i.e. there are no options to allow this at build time in TF-M, at least). Note also that mbed TLS has a on open ongoing issue to better define the initialisation sequence for the various operations in psa_crypto_init(): PSA: separate driver initialization with a nicer fixed ordering · Issue #6228 · Mbed-TLS/mbedtls (github.com)<https://github.com/Mbed-TLS/mbedtls/issues/6228> It might be a good place to start to provide feedback regarding this particular aspect of custom platform initialisation if is not being considered there. To conclude, if you want to propose a patch for TF-M to allow such functions to be plugged in (in the meantime that mbed TLS agrees on the long term course of action), I will be happy to review it. Thanks, Antonio ________________________________ From: Rehan Malak via TF-M <tf-m(a)lists.trustedfirmware.org<mailto:tf-m@lists.trustedfirmware.org>> Sent: Friday, July 7, 2023 13:15 To: tf-m(a)lists.trustedfirmware.org<mailto:tf-m@lists.trustedfirmware.org> <tf-m(a)lists.trustedfirmware.org<mailto:tf-m@lists.trustedfirmware.org>> Subject: [TF-M] mbedtls_platform_setup/teardown in TF-M ? Dear TF-M developers, I am currently adapting a basic MbedTLS / PSA Crypto example such that it would run on the NS side with TF-M doing the crypto. At the end, this is very similar to this psa_sign_verify_message_test from the NS crypto test suite : https://git.trustedfirmware.org/TF-M/tf-m-tests.git/tree/test/secure_fw/sui… But my build config of MbedTLS has MBEDTLS_PLATFORM_SETUP_TEARDOWN_ALT enabled because I have a custom mbedtls_platform_setup / mbedtls_platform_teardown. And I can't see any place in TF-M where mbedtls_platform_setup/mbedtls_platform_teardown are called : ? -> mbedtls_platform_setup ? -> mbedtls_platform_teardown At first, I tried to put this code into the psa_driver_wrapper_init/psa_driver_wrapper_free but I have a similar problem : tfm_crypto_engine_init -> psa_crypto_init -> psa_driver_wrapper_init ? -> mbedtls_psa_crypto_free -> psa_driver_wrapper_free Is there any cmake/Kconfig option or any C macros to hook TF-M initialization/shutdown with mbedtls_platform_setup/mbedtls_platform_teardown without patching TF-M ? If not, could mbedtls_platform_setup be called here ? https://git.trustedfirmware.org/TF-M/trusted-firmware-m.git/tree/secure_fw/… or is there a nicer way of doing this ? (btw, I am currently experimenting on qemu mps2-an521) Thanks for any advice ! 🙂 Best regards, Rehan MALAK Intrinsic ID

3 years, 1 month

Python version requirements

by Gilles Peskine

Hello, This is a question for integrators and packagers of Mbed TLS, especially if you're integrating the library in some embedded OS or BSP. To build, configure and test Mbed TLS, we use Python for several purposes: * To configure the library with scripts/config.py, unless you use the default configuration or write your own configuration from scratch. * To generate some configuration-independent library source files, but only if you use the development branch, not if you use a release or an LTS branch. * To generate some glue code for PSA drivers, if you use PSA drivers and don't write the glue code by hand. * To generate the unit test source files. * (We have many more maintenance and test scripts but they're out of scope here.) (Python is not necessary, and will remain unnecessary, to configure and build the library with a given set of hardware drivers, so that a typical BSP will not have to depend on Python.) For each of these purposes, how problematic is it if we require a recent version of Python? We're currently planning to drop support for older versions of Python as soon as they become unsupported upstream (so officially dropping 3.8 now — although right now all scripts still work on 3.5). This includes versions of Python that are still shipped in e.g. Linux distributions that are themselves officially supported. Is there demand for supporting older Python versions for some scripts? Also, how problematic is it if some of these purposes require third-party Python packages? Best regards, -- Gilles Peskine Mbed TLS developer

3 years, 1 month

2026

2025

2024

2023

2022

2021

2020

mbed-tls July 2023