Re: [PATCH v1 03/19] xen/sysctl: wrap around XEN_SYSCTL_readconsole

[Jan Beulich <jbeulich@xxxxxxxx>]

On 20.03.2025 09:02, Penny, Zheng wrote:
>> -----Original Message-----
>> From: Alejandro Vallejo <alejandro.vallejo@xxxxxxxxx>
>> Sent: Thursday, March 13, 2025 8:03 PM
>>
>> Ok, so readconsole is done here. I see how if you're also removing the 
>> console
>> handler for the sysctl that's a bit unwiledly to do in one go.
>>
>> I think my earlier remarks still hold in terms of removal of else branches 
>> of ifdefs.
>>
>> On Wed Mar 12, 2025 at 4:06 AM GMT, Penny Zheng wrote:
>>> The following functions is to deal with XEN_SYSCTL_readconsole sub-op,
>>> and shall be wrapped:
>>> - xsm_readconsole
>>> - read_console_ring
>>>
>>> Signed-off-by: Penny Zheng <Penny.Zheng@xxxxxxx>
>>> ---
>>>  xen/drivers/char/console.c |  2 ++
>>>  xen/include/xen/console.h  |  8 ++++++++
>>>  xen/include/xsm/dummy.h    | 11 ++++++++---
>>>  xen/include/xsm/xsm.h      | 11 ++++++++---
>>>  xen/xsm/dummy.c            |  2 +-
>>>  xen/xsm/flask/hooks.c      |  4 ++--
>>>  6 files changed, 29 insertions(+), 9 deletions(-)
>>>
>>> diff --git a/xen/drivers/char/console.c b/xen/drivers/char/console.c
>>> index 2f028c5d44..6e4f3c4659 100644
>>> --- a/xen/drivers/char/console.c
>>> +++ b/xen/drivers/char/console.c
>>> @@ -336,6 +336,7 @@ static void conring_puts(const char *str, size_t len)
>>>          conringc = conringp - conring_size;  }
>>>
>>> +#ifdef CONFIG_SYSCTL
>>>  long read_console_ring(struct xen_sysctl_readconsole *op)  {
>>>      XEN_GUEST_HANDLE_PARAM(char) str; @@ -378,6 +379,7 @@ long
>>> read_console_ring(struct xen_sysctl_readconsole *op)
>>>
>>>      return 0;
>>>  }
>>> +#endif /* CONFIG_SYSCTL */
>>>
>>>
>>>  /*
>>> diff --git a/xen/include/xen/console.h b/xen/include/xen/console.h
>>> index 83cbc9fbda..e7d5063d82 100644
>>> --- a/xen/include/xen/console.h
>>> +++ b/xen/include/xen/console.h
>>> @@ -7,12 +7,20 @@
>>>  #ifndef __CONSOLE_H__
>>>  #define __CONSOLE_H__
>>>
>>> +#include <xen/errno.h>
>>>  #include <xen/inttypes.h>
>>>  #include <xen/ctype.h>
>>>  #include <public/xen.h>
>>>
>>>  struct xen_sysctl_readconsole;
>>
>> That forward declaration should probably be inside the ifdef
>>
>>> +#ifdef CONFIG_SYSCTL
>>>  long read_console_ring(struct xen_sysctl_readconsole *op);
>>> +#else
>>> +static inline long read_console_ring(struct xen_sysctl_readconsole
>>> +*op) {
>>> +    return -EOPNOTSUPP;
>>> +}
>>> +#endif
>>
>> This is only called from sysctl.c, which will be compiled out. Why is the 
>> else
>> needed?
>>
> 
> Because I wrapped the sysctl.c in the last commit.
> If removing the else condition here, the compilation will fail on this commit.
> So either I add #ifdef into read_console_ring function body, or in the last 
> commit,
> I draw back all these unnecessary else conditions, or combine all commits 
> into one
> Any preference? Or any other suggestion?

Munging everything in a single commit may yield unwieldily big a result. 
Transiently
adding #ifdef in sysctl.c would seem like the cleanest approach to me. In the 
final
commit it'll (hopefully) be obvious enough then why all of the #ifdef-s are 
dropped
again.

Jan