Move TI hardware acceleration into Matter repo

[srickardti]

Problem

Fixes #9176

Change overview

Moved the SHA, AES, and ECJPAKE alt headers and source files into the Matter
repository. Fixed an issue with the semantics of SHA256 context cloning.
Changed the CHIPCryptoPALmbedTLS.cpp implementation to clone the temporary
context back over the original context when done with the non-final
finalization. Increased the SHA context opaque size in CHIPCryptoPAL.h to fit
the TI driver accelerated SHA context.

In future commits it may be possible to share a single SHA driver instance
between all the contexts in mbedtls. This has the disadvantage of requiring a
mutex for driver usage and reference counting. This would also increase
processing time due to the thrashing of the internal digest buffer. But has the
potential advantage that the SHA context can be copied directly with a memcpy.

Testing

Manually tested.

• lock-app was built
• example was commissioned onto a Thread network
• CASE session was established

[jmartinez-silabs]

Approved, but I second Andrei comment. The worst case should not be applied to kMAX_Hash_SHA256_Context_Size for all platforms. We should use a defined default value that a platform specific define can overrite.

[srickardti]

I was also very confused why the Crypto PAL component seemed to compile in the size of context objects and treated them as copy-able objects. But I was not quite confident in my understanding of the side-effects to OpenSSL to start messing with those files.

[woody-apple]

@Damian-Nordic @saurabhst @msandstedt ?

[tcarmelveilleux]

Cost of shared context everyone must use is too large.

Suggest:

• Add a CHIP_CONFIG variable for SHA256 Hardware Context size, used as a default define for kMAX_Hash_SHA256_Context_Size and set to previous value before this change.
• Update TI examples to use CHIP_CONFIG_xxxx new define and override it to (sizeof(unsigned int) * 76) which would mean ONLY TI platform users would pay that cost.

[srickardti]

The failures seem to be CIPD setup in the github pipeline. I don't think I can manually rerun those tests. Should I push a small change to kick them off again?

[github-actions]

Size increase report for "gn_qpg-example-build" from 8788f6f

File	Section	File	VM
chip-qpg6100-lighting-example.out	.text	16	16

Full report output

BLOAT REPORT

Files found only in the build output:
    report.csv

Comparing ./master_artifact/chip-qpg6100-lighting-example.out.map and ./pull_artifact/chip-qpg6100-lighting-example.out.map:

BLOAT EXECUTION FAILED WITH CODE 1:
bloaty: unknown file type for file './pull_artifact/chip-qpg6100-lighting-example.out.map'

Comparing ./master_artifact/chip-qpg6100-lighting-example.out and ./pull_artifact/chip-qpg6100-lighting-example.out:

sections,vmsize,filesize
.debug_info,0,184
.debug_loc,0,95
.debug_ranges,0,64
.text,16,16
.debug_line,0,13
.debug_abbrev,0,10
.debug_frame,0,8
.debug_str,0,2
[Unmapped],0,-16

[github-actions]

Size increase report for "esp32-example-build" from 8788f6f

File	Section	File	VM
chip-all-clusters-app.elf	.flash.text	16	16

Full report output

BLOAT REPORT

Files found only in the build output:
    report.csv

Comparing ./master_artifact/chip-all-clusters-app.elf and ./pull_artifact/chip-all-clusters-app.elf:

sections,vmsize,filesize
.debug_info,0,113
.debug_loc,0,96
.debug_line,0,70
.debug_ranges,0,32
.debug_abbrev,0,23
.debug_frame,0,16
.flash.text,16,16
.riscv.attributes,0,2
[Unmapped],0,-16

[github-actions]

Size increase report for "nrfconnect-example-build" from 8788f6f

File	Section	File	VM
chip-lock.elf	text	16	16

Full report output

BLOAT REPORT

Files found only in the build output:
    report.csv

Comparing ./master_artifact/chip-lock.elf and ./pull_artifact/chip-lock.elf:

sections,vmsize,filesize
.debug_info,0,174
.debug_loc,0,124
.debug_ranges,0,64
.debug_line,0,20
text,16,16
.debug_abbrev,0,10
.debug_frame,0,8

Comparing ./master_artifact/chip-shell.elf and ./pull_artifact/chip-shell.elf:

sections,vmsize,filesize
.debug_info,0,188
.debug_loc,0,120
.debug_ranges,0,64
.debug_line,0,26
.debug_abbrev,0,10
.debug_frame,0,8

[ti-tpan]

Hi all,

Thanks for the reviews here!

Any further actions needed before merging this PR (by an authorized user)?

Thanks,
Toby