image: update Cosmo A to Turin 1.0.0.a

[luqmana]

No description provided.

[rmustacc]

I performed initial turbo-boost validation on a 9355P which has a 4.4 GHz max clock boost. To do this I used cpustat -c PAPI_tot_cyc 1 and then in another shell did a pbind -b 23 $$ followed by a while :; do :; done. I watched the cycle count on CPU 23. I confirmed it quickly ramped up to the approximate range. An example value was 4482172528. And I also saw that as that workload was suspended and killed that the value ramped down appropriately.

[rmustacc]

Next I proceeded to verify hotplug through the following ways:

• Used cfgadm to do graceful power control and verified the slot was powered off, drive detached, and the LED turned off.
• Next I did the same in reverse, powering on the slot and verified that the drive was once again in nvmeadm.
• I used fmtopo to turn on a drive attention LED via /usr/lib/fm/fmd/fmtopo -P facility.mode=uint32:1 hc://:product-id=913-0000023:product-sn=BRM22250002:server-id=BRM22250002:chassis-id=BRM22250002/chassis=0/bay=1?indicator=attention on an empty slot and confirmed that worked.
• I inserted a drive into slot N7, which was empty since boot, and confirmed that it powered on and enumerated.
• I surprise removed a drive. Verified it no longer enumerated. Waited, reinstalled it, and verified it enumerated just fine.

[luqmana]

Also got this installed on a racklette last week atop the R18 branch to have product assurance run through it. From chat:

angela: The tests on london didn't uncover any issue. I did mostly sysbench runs and general CP orchestration things.