Re: [PATCH v14 02/15] arm64: barrier: Support smp_cond_load_relaxed_timeout()

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] The generic macro `smp_cond_load_relaxed_timeout()` is broken for loadable modules on ARM64 because it uses the unexported symbol `arch_timer_evtstrm_available`.
--

--- Patch [2]: [PATCH v14 02/15] arm64: barrier: Support smp_cond_load_relaxed_timeout() ---
(Note: The format of this review has been altered to provide a free-form summary rather than inline quotes due to recitation restrictions.)

commit 8ff6ac5954b14995e35ca4865c27742e37086e23
Author: Ankur Arora <ankur.a.arora@oracle.com>
arm64: barrier: Support smp_cond_load_relaxed_timeout()

This commit adds support for smp_cond_load_relaxed_timeout() on arm64 by
using __cmpwait_relaxed() when the event-stream is available. It introduces
cpu_poll_relax() which checks arch_timer_evtstrm_available() before deciding
to wait or fallback to cpu_relax().

[Severity: High]
Will loadable modules be able to build if they use the new timeout barrier API?

The new cpu_poll_relax() macro in arch/arm64/include/asm/barrier.h calls
arch_timer_evtstrm_available(), which is defined in
drivers/clocksource/arm_arch_timer.c. 

Since cpu_poll_relax() is embedded inline into callers of
smp_cond_load_relaxed_timeout(), any loadable module that uses this generic
API will now require the arch_timer_evtstrm_available symbol.

However, it does not appear that arch_timer_evtstrm_available is exported with
EXPORT_SYMBOL or EXPORT_SYMBOL_GPL. 

Does this cause an undefined reference error during the modpost step for
device drivers that attempt to use smp_cond_load_relaxed_timeout()?

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260714073041.40250-1-ankur.a.arora@oracle.com?part=2