NCLnclNCL wrote:
> I am seeing a new build failure when building the Linux kernel as a result of
> this change.
>
> ```
> drivers/tty/sysrq.c:486:2: error: initializer element is not a compile-time
> constant
> 486 | &sysrq_showlocks_op, /* d */
> | ^~~~~~~~~~~~~~~~~~~
> ```
>
> as a result of
>
> ```c
> #ifdef CONFIG_LOCKDEP
> static void sysrq_handle_showlocks(u8 key)
> {
> debug_show_all_locks();
> }
>
> static const struct sysrq_key_op sysrq_showlocks_op = {
> .handler = sysrq_handle_showlocks,
> .help_msg = "show-all-locks(d)",
> .action_msg = "Show Locks Held",
> };
> #else
> #define sysrq_showlocks_op (*(const struct sysrq_key_op *)NULL)
> #endif
> ```
>
> This build does not have LOCKDEP enabled so `&sysrq_showlocks_op` becomes
> `&(*(const struct sysrq_key_op *)((void *)0))`. Is the error expected in this
> case? I am not sure if this construct is used elsewhere in the kernel but it
> was in the first build that was attempted during a compiler upgrade.
How to fix ?
https://github.com/llvm/llvm-project/pull/143667
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
