On 2/25/24 06:58, Kirill A. Shutemov wrote:
> On Fri, Feb 23, 2024 at 11:39:07AM -0800, Dave Hansen wrote:
...
>> I'd really prefer we find a way to do this with actual locks, especially
>> 'conversion_allowed'.
>>
>> This is _awfully_ close to being able to be handled by a rwsem where the
>> readers are the converters and tdx_kexec_stop_conversion() takes a write.
>
> Okay, here's what I come up with. It needs more testing.
>
> Any comments?
Looks a heck of a lot more straightforward to me.
> +static void tdx_kexec_stop_conversion(bool crash)
> +{
> + /* Stop new private<->shared conversions */
> + if (!stop_memory_enc_conversion(!crash))
> + pr_warn("Failed to finish shared<->private conversions\n");
> +}
FWIW, this is one of those places that could use a temporary variable to
help explain what is going on:
bool wait_for_lock = !crash;
/*
* ... explain why it doesn't or shouldn't take the lock here
*/
if (!stop_memory_enc_conversion(wait_for_lock))
...
This makes it understandable without looking at the called function.
> +bool stop_memory_enc_conversion(bool wait)
> +{
> + if (!wait)
> + return down_write_trylock(&mem_enc_lock);
> +
> + down_write(&mem_enc_lock);
> +
> + return true;
> +}
This also needs a comment about why the lock isn't being released.
_______________________________________________
kexec mailing list
kexec@xxxxxxxxxxxxxxxxxxx
http://lists.infradead.org/mailman/listinfo/kexec
