Fix up core count detection logic on android

[eduardo-vp]

Parse /sys/devices/system/cpu/present as a comma-separated cpulist (single indices and ranges) rather than assuming a single 0-(n-1) range.

Closes #126417.

[azure-pipelines]

Azure Pipelines:
Successfully started running 3 pipeline(s).
13 pipeline(s) were filtered out due to trigger conditions.
There may be pipelines that require an authorized user to comment /azp run to run.

[dotnet-policy-service]

Tagging subscribers to this area: @JulieLeeMSFT, @VSadov
See info in area-owners.md if you want to be subscribed.

[Copilot]

Pull request overview

This PR updates Mono’s Android CPU core-count detection to correctly parse the kernel’s /sys/devices/system/cpu/present “cpulist” format (including non-contiguous ranges), improving the accuracy of mono_cpu_count() on Android.

Changes:

• Parse /sys/devices/system/cpu/present as a comma-separated cpulist (single indices and ranges) rather than assuming a single 0-(n-1) range.
• Increase the read buffer size and add parsing success gating so malformed/partial reads fall back to existing platform logic.
• Clarify the comment describing the /sys entry format with concrete examples.

[jkotas]

Do we need the same logic for CoreCLR/NAOT on Android?

[akoeplinger]

judging from my long comment a few lines below the changed lines it sounds like we should be fine with _SC_NPROCESSORS_CONF which is already used by coreclr from what I can see.

though I see some inconsistency there, gcenv.unix.cpp and sysinfo.cpp use _SC_NPROCESSORS_CONF on arm/arm64/loongarch64/riscv64 and _SC_NPROCESSORS_ONLN on other architectures