[Crash-utility] [PATCH 2/2] Fix "irq [-a|-s]" options on Linux 6.5-rc1 and later
HAGIO KAZUHITO(萩尾 一仁)
k-hagio-ab at nec.com
Wed Jul 12 08:30:01 UTC 2023
On 2023/07/12 17:17, lijiang wrote:
> On Wed, Jul 12, 2023 at 3:49 PM HAGIO KAZUHITO(萩尾 一仁) <k-hagio-ab at nec.com>
> wrote:
>
>> On 2023/07/12 16:30, lijiang wrote:
>>> On Fri, Jul 7, 2023 at 2:17 PM HAGIO KAZUHITO(萩尾 一仁) <k-hagio-ab at nec.com
>>>
>>> wrote:
>>>
>>>> From: Kazuhito Hagio <k-hagio-ab at nec.com>
>>>>
>>>> Kernel commit 721255b982 ("genirq: Use a maple tree for interrupt
>>>> descriptor management"), which is contained in Linux 6.5-rc1 and later
>>>> kernels, replaced irq_desc_tree with a maple tree sparse_irqs.
>>>>
>>>>
>>> Thank you for the fix, Kazu. Only two comments:
>>>
>>> Without the patch, "irq [-a|-s]" options fails with an error, e.g. the
>>>> following on x86_64, on kernels configured with CONFIG_SPARSE_IRQS=y.
>>>>
>>>
>>> ^^^^^
>>>
>>> It should be "CONFIG_SPARSE_IRQ=y", the following code comment is right.
>>> Maybe it's a typo?
>>> + /* 6.5 and later: CONFIG_SPARSE_IRQ */
>>
>> Yes, you are right. Will fix the commit log.
>>
>>>> @@ -5419,8 +5421,8 @@ x86_64_get_irq_affinity(int irq)
>>>> static void
>>>> x86_64_show_interrupts(int irq, ulong *cpus)
>>>> {
>>>> - if (symbol_exists("irq_desc") ||
>>>> - kernel_symbol_exists("irq_desc_ptrs") ||
>>>>
>>>
>>> The check for kernel_symbol_exists("irq_desc_ptrs") is removed in this
>>> function, is that expected behavior? But the above two functions still
>>> retain the kernel_symbol_exists("irq_desc_ptrs") condition check.
>>
>> ugh, thank you for pointing this out, I didn't intend to remove it.
>> Will fix this and post a v2.
>>
>>
> In addition, I also saw that there is a similar case on IA64 arch, but I
> can not test it. Maybe leave it there?
>
> static void
> ia64_dump_irq(int irq)
> {
> if (symbol_exists("irq_desc") || symbol_exists("_irq_desc") ||
> kernel_symbol_exists("irq_desc_ptrs")) {
> machdep->dump_irq = generic_dump_irq;
> return(generic_dump_irq(irq));
> }
> ...
> }
I also cannot test ia64, but it will be needed, let's add it here too.
Thanks,
Kazu
More information about the Crash-utility
mailing list