[bug report] mm,hwpoison: refactor get_any_page

[bug report] mm,hwpoison: refactor get_any_page

[Dan Carpenter]

Hello Oscar Salvador,

The patch 8295d535e2aa: "mm,hwpoison: refactor get_any_page" from Dec
14, 2020, leads to the following static checker warning:

	mm/memory-failure.c:1943 soft_offline_page()
	error: '%pG' must be followed by one of [gpv]

mm/memory-failure.c
  1933          put_online_mems();
  1934  
  1935          if (ret > 0) {
  1936                  ret = soft_offline_in_use_page(page);
  1937          } else if (ret == 0) {
  1938                  if (soft_offline_free_page(page) && try_again) {
  1939                          try_again = false;
  1940                          goto retry;
  1941                  }
  1942          } else if (ret == -EIO) {
  1943                  pr_info("%s: %#lx: unknown page type: %lx (%pGP)\n",
                                                                   ^^^^
Presumably lower case 'p' was intended.  "%pGp".  See the flags_string()
function for more details.

  1944                           __func__, pfn, page->flags, &page->flags);
  1945          }
  1946  
  1947          return ret;
  1948  }

regards,
dan carpenter

Re: [bug report] mm,hwpoison: refactor get_any_page

[Anshuman Khandual]

On 1/4/21 3:25 PM, Dan Carpenter wrote:
> Hello Oscar Salvador,
> 
> The patch 8295d535e2aa: "mm,hwpoison: refactor get_any_page" from Dec
> 14, 2020, leads to the following static checker warning:
> 
> 	mm/memory-failure.c:1943 soft_offline_page()
> 	error: '%pG' must be followed by one of [gpv]
> 
> mm/memory-failure.c
>   1933          put_online_mems();
>   1934  
>   1935          if (ret > 0) {
>   1936                  ret = soft_offline_in_use_page(page);
>   1937          } else if (ret == 0) {
>   1938                  if (soft_offline_free_page(page) && try_again) {
>   1939                          try_again = false;
>   1940                          goto retry;
>   1941                  }
>   1942          } else if (ret == -EIO) {
>   1943                  pr_info("%s: %#lx: unknown page type: %lx (%pGP)\n",
>                                                                    ^^^^
> Presumably lower case 'p' was intended.  "%pGp".  See the flags_string()
> function for more details.

Yeah right, s/P/p is required. Otherwise apart from this static checker
warning, this might not even print the page flags correctly.

Re: [bug report] mm,hwpoison: refactor get_any_page

[Oscar Salvador]

On 2021-01-04 10:55, Dan Carpenter wrote:
> Hello Oscar Salvador,
> 
> The patch 8295d535e2aa: "mm,hwpoison: refactor get_any_page" from Dec
> 14, 2020, leads to the following static checker warning:
> 
> 	mm/memory-failure.c:1943 soft_offline_page()
> 	error: '%pG' must be followed by one of [gpv]
> 
> mm/memory-failure.c
>   1933          put_online_mems();
>   1934
>   1935          if (ret > 0) {
>   1936                  ret = soft_offline_in_use_page(page);
>   1937          } else if (ret == 0) {
>   1938                  if (soft_offline_free_page(page) && try_again) 
> {
>   1939                          try_again = false;
>   1940                          goto retry;
>   1941                  }
>   1942          } else if (ret == -EIO) {
>   1943                  pr_info("%s: %#lx: unknown page type: %lx 
> (%pGP)\n",
>                                                                    ^^^^
> Presumably lower case 'p' was intended.  "%pGp".  See the 
> flags_string()
> function for more details.
> 
>   1944                           __func__, pfn, page->flags, 
> &page->flags);
>   1945          }
>   1946
>   1947          return ret;
>   1948  }
> 
> regards,
> dan carpenter

Thanks for the report, I'll have a look!

-- 
Oscar Salvador
SUSE L3