Release candidate for mbed-os-5.11.5 #9805

cmonr · 2019-02-21T20:25:49Z

No description provided.

Like all HAL APIs, the calls in trng_api.h are not expected to be thread-safe. All current accesses to the TRNG HAL are currently via `mbedtls_hardware_poll`. Mbed TLS does not currently serialise these calls itself, as `MBEDTLS_THREADING_C` is not enabled. But even if Mbed TLS's own accesses were serialised, there are other direct users of `mbedtls_hardware_poll` such as randLIB, that need to use direct calls due to lack of API to extract entropy from Mbed TLS. As such it makes sense to treat `mbedtls_hardware_poll` as a de facto public Mbed OS API, akin to the C++ veneers on top of the HAL, and add a PlatformMutex there so that it is safe for multithreaded use.

Use PSA_NON_SECURE_ROM_START configuration value instead of hardcoded CY_CORTEX_M4_APPL_ADDR.

1. Removed random i/o glitches occurring during device reconfiguration 2. Fixed hazardous reads occurring at the end of transfer resulting in incorrect values being received 3. Added spi_free() function 4. Replaced default M0 image with a one ignoring i/o reservation. This is a workaround for missing proper destructors in Mbed drivers and BlockDevice tests failing on repeated initialization Fixes issue #9620.

When building secure targets, we would like compile only files that are in the the mbe-os root folder. This Fixes an issue when sending multiple --source parameters that are all inside the mbed-os tree

Previously, the CryptoToolbox was allocated once as part of the security manager. This was inneficient memory wise as it is only use to prepare key at initialization and when we need to compute shared keys. This was also inneficient power consumption wise as the Crypto cell was kept enabled even when it wasn't used. This fix creates a CryptoToolbox whenever it is needed and release it once it has fulfilled its purpose. Note that CryptoToolbox allocation happens on the heap as mbed tls data structure are huge and there's an high risk of crushing the stack.

In case mbedtls fails to execute handshake advertising MBEDTLS_ERR_SSL_WANT_READ or MBEDTLS_ERR_SSL_WANT_WRITE, TLSSocketWrapper::continue_handshake returns NSAPI_ERROR_WOULD_BLOCK.

Removed by mistake in 763cb4c

…nt have to change each run, just be unique. Because address seeds repeat themselves each run, rand() will produce repeating results that should not collide between different threads.

Port of NUC472/M487 crypto AC mgmt work to M2351: 1. Choose mutex to synchronize access to crypto non-SHA AC 2. Choose atomic flag to synchronize access to crypto SHA AC

Fix the following warning seen when built with GCC_ARM with mbed-cli v1.8.3 $ mbed compile -c -t GCC_ARM -m K64F <snip> [Warning] QUECTEL_M26_CellularStack.cpp@364,9: variable 'sent_acked' set but not used [-Wunused-but-set-variable] <snip>

Fixes a bug where after dropping bad packets receiver doesn't restore the socket back to non-blocking state as it's supposed to.

Test case is allowed to take not more than a half what has been given to the whole UDP suite. UDP test cases reorganized so that the longest running ones are executed last.

cmonr · 2019-02-21T20:26:57Z

CI started

This reverts commit 2a7f5e2.

cmonr · 2019-02-21T21:30:23Z

@JanneKiiskila Please review and OK when able.

mbed-ci · 2019-02-21T21:54:58Z

Test run: FAILED

Summary: 3 of 8 test jobs failed
Build number : 1
Build artifacts

Failed test jobs:

jenkins-ci/mbed-os-ci_build-IAR
jenkins-ci/mbed-os-ci_build-ARM
jenkins-ci/mbed-os-ci_build-GCC_ARM

cmonr · 2019-02-22T01:46:51Z

Reverted PR: #9616 (comment)

cmonr · 2019-02-22T01:47:32Z

CI restarted

mbed-ci · 2019-02-22T03:39:42Z

Test run: FAILED

Summary: 3 of 8 test jobs failed
Build number : 2
Build artifacts

Failed test jobs:

jenkins-ci/mbed-os-ci_build-IAR
jenkins-ci/mbed-os-ci_build-GCC_ARM
jenkins-ci/mbed-os-ci_build-ARM

This reverts commit 7824c79.

This reverts commit b705934.

This reverts commit 035ecc1.

cmonr · 2019-02-22T04:25:48Z

CI restarted.
Had to revert a couple more PRs to get the F411 to compile locally

cmonr · 2019-02-22T04:29:06Z

@ARMmbed/mbed-os-psa Please review.

Had to revert hte following PRs to get a locall F411 build to pass:

mbed-ci · 2019-02-22T07:23:08Z

Test run: SUCCESS

Summary: 12 of 12 test jobs passed
Build number : 3
Build artifacts

0xc0170 · 2019-02-22T10:56:55Z

@ARMmbed/mbed-os-psa Please review.

I've left comments in the PRs that they moved to 5.12. Not certain why but might be some dependancy. All of these should be fine for 5.12. But we should rather wait for psa team to confirm.

orenc17 · 2019-02-22T11:09:14Z

@Patater could you confirm that the target configuration is fine?

Patater · 2019-02-22T13:31:39Z

I had a look at targets.json between 5.11.0 and 5.11.5.

The label PSA was added to F411RE in a patch release, but no other PSA APIs like Crypto. The only benefit I see this having is enabling PSA ITS for this target. The FLASHIAP component was also added to this target in a patch release.
The label PSA was added to K66F, F429ZI, UBLOX_EVK_ODIN_W2 in a patch release, but no other PSA APIs like Crypto. If the point in adding PSA to these other targets is to make PSA available by default on platforms that Client wants, we should additionally add MBEDTLS_PSA_CRYPTO_C. Unlike the F411RE, we can enable Crypto on these targets by default because these targets do have a guaranteed source of entropy that Mbed OS can always assume exists: a TRNG.
The SD component was added to a few targets. Will this impact existing applications for these targets by making their file access code default to using SD instead of the previous storage option? Is there any possibility of a functional change here (for example, if, say, an SD card is not inserted)? CC @ARMmbed/mbed-os-storage

0xc0170 · 2019-02-22T16:21:24Z

@Patater To the problem we had on master, by not having the PRs referenced above in 5.11.5 - are we fine? will client work as it was for previous patch release? My assumption would be yes.

Patater · 2019-02-22T16:49:42Z

@0xc0170 Yes, the problem we had on master will not come up in 5.11.5 because there are no targets without guaranteed entropy having MBEDTLS_PSA_CRYPTO_C enabled.

0xc0170 · 2019-02-25T08:17:04Z

Confirmed, client tests are OK. This is ready for merge

kjbracey and others added 30 commits February 21, 2019 14:25

FUTURE_SEQUANA_M0_PSA - fix CM4 starting address

d7b58e9

Use PSA_NON_SECURE_ROM_START configuration value instead of hardcoded CY_CORTEX_M4_APPL_ADDR.

tests-netsocket-tls: TLSSOCKET_ECHOTEST deletes socket on failure case

275572c

tests-network-interface: drops unreferenced variable

70d8214

platform: includes string.h needed by memcpy and strlen

f886fc6

K64/ARM: Update to 2-region model for HEAP and Stack Memory

bd056c6

Revert "MIMXRT1050_EVK: Add IAR support in the exporter"

8482fb7

Force adding mbed-os root to include dirs

a4ed281

Check source relative to ROOT on secure targets

2636b7e

When building secure targets, we would like compile only files that are in the the mbe-os root folder. This Fixes an issue when sending multiple --source parameters that are all inside the mbed-os tree

Update test.py

1b8fe3e

Greentea UDPSOCKET_RECV_TIMEOUT increase time tolerance

42d9fdd

Greentea Socket bind tests made ipv6-aware

b88d69d

Greentea DNS test: added debug print

4a92a13

TLSSocket returns WOULD_BLOCK error instead of ALREADY

3f11bf5

In case mbedtls fails to execute handshake advertising MBEDTLS_ERR_SSL_WANT_READ or MBEDTLS_ERR_SSL_WANT_WRITE, TLSSocketWrapper::continue_handshake returns NSAPI_ERROR_WOULD_BLOCK.

Restore MBEDTLS_PSA_CRYPTO_C for PSA targets

035ecc1

Removed by mistake in 763cb4c

Fix general block device test to ensure unique block address per test

b2b7afe

Ensure unique block address for multi threaded test. The addresses do…

19b50ef

…nt have to change each run, just be unique. Because address seeds repeat themselves each run, rand() will produce repeating results that should not collide between different threads.

Fix crypto AC mgmt

693245d

Port of NUC472/M487 crypto AC mgmt work to M2351: 1. Choose mutex to synchronize access to crypto non-SHA AC 2. Choose atomic flag to synchronize access to crypto SHA AC

Hook crypto interrupt handler without NVIC_SetVector

e5e9e4f

Remove unnecessary header files in crypto-misc.cpp

3926223

Cellular: Fix socket bind tests for BG96

306943a

Cellular: Fix socket's send and receive boundaries for BG96

0ed5e80

Cellular: Fix UDPSOCKET_SENDTO_INVALID test for BG96

b8712b3

REALTEK_RTL8195AM daplink timestamp from os.environ

182f229

targets: QUECTEL: Fix build warning

3930d2e

Fix the following warning seen when built with GCC_ARM with mbed-cli v1.8.3 $ mbed compile -c -t GCC_ARM -m K64F <snip> [Warning] QUECTEL_M26_CellularStack.cpp@364,9: variable 'sent_acked' set but not used [-Wunused-but-set-variable] <snip>

STM32 GPIO: use maximum speed

e8ce42a

tests-netsocket-udp: UDPSOCKET_ECHOTEST_NONBLOCK restores non-blocking

46478ab

Fixes a bug where after dropping bad packets receiver doesn't restore the socket back to non-blocking state as it's supposed to.

tests-netsocket-udp: UDPSOCKET_ECHOTEST_NONBLOCK execution time limit

1b5fdaf

Test case is allowed to take not more than a half what has been given to the whole UDP suite. UDP test cases reorganized so that the longest running ones are executed last.

Revert "nRF52 serial: Tighten/simplify atomics"

f5eb87c

This reverts commit 2a7f5e2.

cmonr added needs: work needs: CI and removed needs: CI needs: work labels Feb 21, 2019

cmonr force-pushed the release-candidate branch from ab03bcd to f5eb87c Compare February 22, 2019 03:46

Cruz Monrreal II added 3 commits February 21, 2019 22:02

Revert "support in more types then just RSA_KEYPAIR"

42b1afa

This reverts commit 7824c79.

Revert "Add MBEDTLS_ENTROPY_NV_SEED to NUCLEO_F411RE to fix the build"

71ebf82

This reverts commit b705934.

Revert "Restore MBEDTLS_PSA_CRYPTO_C for PSA targets"

b83fb2b

This reverts commit 035ecc1.

0xc0170 approved these changes Feb 22, 2019

View reviewed changes

NirSonnenschein added ready for merge and removed needs: CI labels Feb 24, 2019

0xc0170 merged commit 51d5550 into mbed-os-5.11 Feb 25, 2019

0xc0170 removed the ready for merge label Feb 25, 2019

martinichka mentioned this pull request Jun 4, 2019

Assertion failed: pcache->block == 0xffffffff #10754

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Release candidate for mbed-os-5.11.5 #9805

Release candidate for mbed-os-5.11.5 #9805

Uh oh!

cmonr commented Feb 21, 2019

Uh oh!

cmonr commented Feb 21, 2019

Uh oh!

cmonr commented Feb 21, 2019

Uh oh!

mbed-ci commented Feb 21, 2019

Uh oh!

cmonr commented Feb 22, 2019

Uh oh!

cmonr commented Feb 22, 2019

Uh oh!

mbed-ci commented Feb 22, 2019

Uh oh!

cmonr commented Feb 22, 2019

Uh oh!

cmonr commented Feb 22, 2019

Uh oh!

mbed-ci commented Feb 22, 2019

Uh oh!

0xc0170 commented Feb 22, 2019

Uh oh!

orenc17 commented Feb 22, 2019

Uh oh!

Patater commented Feb 22, 2019

Uh oh!

0xc0170 commented Feb 22, 2019

Uh oh!

Patater commented Feb 22, 2019

Uh oh!

0xc0170 commented Feb 25, 2019

Uh oh!

Uh oh!

Release candidate for mbed-os-5.11.5 #9805

Release candidate for mbed-os-5.11.5 #9805

Uh oh!

Conversation

cmonr commented Feb 21, 2019

Uh oh!

cmonr commented Feb 21, 2019

Uh oh!

cmonr commented Feb 21, 2019

Uh oh!

mbed-ci commented Feb 21, 2019

Test run: FAILED

Uh oh!

cmonr commented Feb 22, 2019

Uh oh!

cmonr commented Feb 22, 2019

Uh oh!

mbed-ci commented Feb 22, 2019

Test run: FAILED

Uh oh!

cmonr commented Feb 22, 2019

Uh oh!

cmonr commented Feb 22, 2019

Uh oh!

mbed-ci commented Feb 22, 2019

Test run: SUCCESS

Uh oh!

0xc0170 commented Feb 22, 2019

Uh oh!

orenc17 commented Feb 22, 2019

Uh oh!

Patater commented Feb 22, 2019

Uh oh!

0xc0170 commented Feb 22, 2019

Uh oh!

Patater commented Feb 22, 2019

Uh oh!

0xc0170 commented Feb 25, 2019

Uh oh!

Uh oh!