[kas:sdei-nmi/diag-cpuoff-direct 5/5] arch/arm64/kernel/smp.c:893:14: warning: unused variable 'crash'

[kernel test robot <lkp@intel.com>]

tree:   https://git.kernel.org/pub/scm/linux/kernel/git/kas/linux.git sdei-nmi/diag-cpuoff-direct
head:   256121e44ded1ba35c2d5a495426fdc4f18440be
commit: 256121e44ded1ba35c2d5a495426fdc4f18440be [5/5] DIAG: call CPU_OFF directly from the SDEI handler (do not submit)
config: arm64-allnoconfig (https://download.01.org/0day-ci/archive/20260630/202606301044.xcGHJhez-lkp@intel.com/config)
compiler: aarch64-linux-gcc (GCC) 16.1.0
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20260630/202606301044.xcGHJhez-lkp@intel.com/reproduce)

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202606301044.xcGHJhez-lkp@intel.com/

All warnings (new ones prefixed by >>):

   arch/arm64/kernel/smp.c: In function 'arm64_nmi_cpu_stop':
>> arch/arm64/kernel/smp.c:893:14: warning: unused variable 'crash' [-Wunused-variable]
     893 |         bool crash = IS_ENABLED(CONFIG_KEXEC_CORE) && crash_stop;
         |              ^~~~~


vim +/crash +893 arch/arm64/kernel/smp.c

45ed695ac10a23 Nicolas Pitre          2014-07-25  865  
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02  866) /**
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02  867)  * arm64_nmi_cpu_stop() - stop the local CPU after it is told to stop.
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02  868)  * @regs: register state to record in the vmcore on a crash stop, or NULL for
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02  869)  *        panic_smp_self_stop(), which has no interrupted context to save.
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02  870)  * @die_on_crash: on the kdump crash path, power the CPU off via PSCI CPU_OFF
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02  871)  *                (so a capture kernel can reclaim it) rather than parking it.
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02  872)  *
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02  873)  * The single point every arm64 stop path funnels through, keeping the
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02  874)  * bookkeeping (mask interrupts, save the crash context, mark offline, mask
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02  875)  * SDEI, optionally power off) in one place:
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02  876)  *
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02  877)  *   - the regular IPI_CPU_STOP and pseudo-NMI IPI_CPU_STOP_NMI handlers;
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02  878)  *   - panic_smp_self_stop(), a CPU parking itself on a parallel panic();
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02  879)  *   - the SDEI cross-CPU NMI handler (drivers/firmware/arm_sdei_nmi.c),
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02  880)  *     which reaches CPUs the stop IPIs could not.
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02  881)  *
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02  882)  * The IPI stop handlers pass @die_on_crash true. The SDEI handler and
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02  883)  * panic_smp_self_stop() pass false and only park. For SDEI that is required,
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02  884)  * not just conservative: it runs inside an SDEI event that is deliberately
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02  885)  * never completed (completing it has firmware resume the wedged context), and
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02  886)  * a CPU_OFF from that not-yet-completed context wedges EL3 on some firmware --
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02  887)  * a documented follow-up. Parking also matches this path's own fallback when
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02  888)  * CPU_OFF is unavailable.
d914d4d4974529 Aaro Koskinen          2019-06-17  889   */
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02  890) void __noreturn arm64_nmi_cpu_stop(struct pt_regs *regs, bool die_on_crash)
d914d4d4974529 Aaro Koskinen          2019-06-17  891  {
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02  892) 	unsigned int cpu = smp_processor_id();
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02 @893) 	bool crash = IS_ENABLED(CONFIG_KEXEC_CORE) && crash_stop;
08e875c16a16c9 Catalin Marinas        2012-03-05  894  
fdfa588124b635 Douglas Anderson       2024-08-21  895  	/*
fdfa588124b635 Douglas Anderson       2024-08-21  896  	 * Use local_daif_mask() instead of local_irq_disable() to make sure
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02  897) 	 * that pseudo-NMIs are disabled. The "stop" code starts with an IRQ
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02  898) 	 * and falls back to NMI (which might be pseudo). If the IRQ finally
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02  899) 	 * goes through right as we're timing out then the NMI could interrupt
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02  900) 	 * us. It's better to prevent the NMI and let the IRQ finish since the
3032112bb2ab2d Kiryl Shutsemau (Meta  2026-06-02  901) 	 * pt_regs will be better.
fdfa588124b635 Douglas Anderson       2024-08-21  902  	 */
fdfa588124b635 Douglas Anderson       2024-08-21  903  	local_daif_mask();
fdfa588124b635 Douglas Anderson       2024-08-21  904  

:::::: The code at line 893 was first introduced by commit
:::::: 3032112bb2ab2d5149f2d38b97f85caf3a8bb757 arm64: escalate smp_send_stop() to an SDEI NMI as a last resort

:::::: TO: Kiryl Shutsemau (Meta) <kas@kernel.org>
:::::: CC: Kiryl Shutsemau (Meta) <kas@kernel.org>

--
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki