Recognize AArch64 and RISC-V in DDR TraceFileHeaderWriter

[knn-k]

Fixes: #18647

[knn-k]

This commit depends on eclipse-omr/omr#7208 and #18643.

[knn-k]

Jenkins compile alinux64 jdk17

[knn-k]

How can I test this? Are the steps below correct?

1. Generate a core file
2. Run !snaptrace command in jdmpview to generate a .trc file
3. Run traceformat to generate a .fmt file from the .trc file

[pshipton]

Perhaps @keithc-ca will comment in the new year.

[keithc-ca]

Are the steps below correct?

Yes, that is the way you would test this.

[knn-k]

Jenkins compile alinux64 jdk17

[knn-k]

Jenkins compile alinux64 jdk17

The previous build failed with the following error:

[2024-01-10T23:27:33.999Z] curl: (56) OpenSSL SSL_read: Connection reset by peer, errno 104

[knn-k]

Build failed again with the same curl error:

[2024-01-10T23:38:01.457Z] + curl -LJkO 'https://api.adoptopenjdk.net/v3/binary/latest/17/ga/linux/aarch64/jdk/openj9/normal/adoptopenjdk?project=jdk'
...
 60  197M   60  120M    0     0   140k      0  0:23:54  0:14:34  0:09:20     0
[2024-01-10T23:52:36.874Z] curl: (56) OpenSSL SSL_read: Connection reset by peer, errno 104

[knn-k]

I can see the following information in the .fmt file generated from a .trc file on AArch64 Linux, using my local build. Arch family is set correctly. Processor Sub-type is wrong, but it seems other architectures do not set it correctly, as I showed an example on x86-64 Linux in #18647 (comment).

Processor information:
        Arch family:         AArch64
        Processor Sub-type:  i486
        Num Processors:      8
        Big Endian:          false
        Word size:           64
        Using Trace Counter: J9 timer

[knn-k]

Jenkins compile alinux64 jdk17

[knn-k]

I haven't tested this on RISC-V.