[PATCH] linux-user: limit check to HOST_LONG_BITS == 32

[PATCH] linux-user: limit check to HOST_LONG_BITS == 32

[Alex Bennée]

Newer clangs rightly spot that you can never exceed the full address
space of 64 bit hosts with:

  linux-user/elfload.c:2076:41: error: result of comparison 'unsigned
  long' > 18446744073709551615 is always false
  [-Werror,-Wtautological-type-limit-compare]
  4685         if ((guest_hiaddr - guest_base) > ~(uintptr_t)0) {
  4686             ~~~~~~~~~~~~~~~~~~~~~~~~~~~ ^ ~~~~~~~~~~~~~
  4687 1 error generated.

So lets limit the check to 32 bit hosts only.

Fixes: ee94743034bf
Reported-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
---
 linux-user/elfload.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/linux-user/elfload.c b/linux-user/elfload.c
index 01a9323a637..797fbf2337a 100644
--- a/linux-user/elfload.c
+++ b/linux-user/elfload.c
@@ -2072,6 +2072,7 @@ static void pgb_have_guest_base(const char *image_name, abi_ulong guest_loaddr,
                          image_name, (uint64_t)guest_hiaddr, reserved_va);
             exit(EXIT_FAILURE);
         }
+#if HOST_LONG_BITS == 32
     } else {
         if ((guest_hiaddr - guest_base) > ~(uintptr_t)0) {
             error_report("%s: requires more virtual address space "
@@ -2079,6 +2080,7 @@ static void pgb_have_guest_base(const char *image_name, abi_ulong guest_loaddr,
                          image_name, (uint64_t)guest_hiaddr - guest_base);
             exit(EXIT_FAILURE);
         }
+#endif
     }
 
     /*
-- 
2.20.1

Re: [PATCH] linux-user: limit check to HOST_LONG_BITS == 32

[Peter Maydell]

On Thu, 21 May 2020 at 11:22, Alex Bennée <alex.bennee@linaro.org> wrote:
>
> Newer clangs rightly spot that you can never exceed the full address
> space of 64 bit hosts with:
>
>   linux-user/elfload.c:2076:41: error: result of comparison 'unsigned
>   long' > 18446744073709551615 is always false
>   [-Werror,-Wtautological-type-limit-compare]
>   4685         if ((guest_hiaddr - guest_base) > ~(uintptr_t)0) {
>   4686             ~~~~~~~~~~~~~~~~~~~~~~~~~~~ ^ ~~~~~~~~~~~~~
>   4687 1 error generated.
>
> So lets limit the check to 32 bit hosts only.
>
> Fixes: ee94743034bf
> Reported-by: Thomas Huth <thuth@redhat.com>
> Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
> ---
>  linux-user/elfload.c | 2 ++
>  1 file changed, 2 insertions(+)
>
> diff --git a/linux-user/elfload.c b/linux-user/elfload.c
> index 01a9323a637..797fbf2337a 100644
> --- a/linux-user/elfload.c
> +++ b/linux-user/elfload.c
> @@ -2072,6 +2072,7 @@ static void pgb_have_guest_base(const char *image_name, abi_ulong guest_loaddr,
>                           image_name, (uint64_t)guest_hiaddr, reserved_va);
>              exit(EXIT_FAILURE);
>          }
> +#if HOST_LONG_BITS == 32
>      } else {
>          if ((guest_hiaddr - guest_base) > ~(uintptr_t)0) {
>              error_report("%s: requires more virtual address space "
> @@ -2079,6 +2080,7 @@ static void pgb_have_guest_base(const char *image_name, abi_ulong guest_loaddr,
>                           image_name, (uint64_t)guest_hiaddr - guest_base);
>              exit(EXIT_FAILURE);
>          }
> +#endif

Could we write this so that we have the #if...#endif
conditional nested inside the if { ... } else { ... },
rather than having the two conditional constructs
oddly intermeshed?

thanks
-- PMM

Re: [PATCH] linux-user: limit check to HOST_LONG_BITS == 32

[Alex Bennée]

Peter Maydell <peter.maydell@linaro.org> writes:

> On Thu, 21 May 2020 at 11:22, Alex Bennée <alex.bennee@linaro.org> wrote:
>>
>> Newer clangs rightly spot that you can never exceed the full address
>> space of 64 bit hosts with:
>>
>>   linux-user/elfload.c:2076:41: error: result of comparison 'unsigned
>>   long' > 18446744073709551615 is always false
>>   [-Werror,-Wtautological-type-limit-compare]
>>   4685         if ((guest_hiaddr - guest_base) > ~(uintptr_t)0) {
>>   4686             ~~~~~~~~~~~~~~~~~~~~~~~~~~~ ^ ~~~~~~~~~~~~~
>>   4687 1 error generated.
>>
>> So lets limit the check to 32 bit hosts only.
>>
>> Fixes: ee94743034bf
>> Reported-by: Thomas Huth <thuth@redhat.com>
>> Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
>> ---
>>  linux-user/elfload.c | 2 ++
>>  1 file changed, 2 insertions(+)
>>
>> diff --git a/linux-user/elfload.c b/linux-user/elfload.c
>> index 01a9323a637..797fbf2337a 100644
>> --- a/linux-user/elfload.c
>> +++ b/linux-user/elfload.c
>> @@ -2072,6 +2072,7 @@ static void pgb_have_guest_base(const char *image_name, abi_ulong guest_loaddr,
>>                           image_name, (uint64_t)guest_hiaddr, reserved_va);
>>              exit(EXIT_FAILURE);
>>          }
>> +#if HOST_LONG_BITS == 32
>>      } else {
>>          if ((guest_hiaddr - guest_base) > ~(uintptr_t)0) {
>>              error_report("%s: requires more virtual address space "
>> @@ -2079,6 +2080,7 @@ static void pgb_have_guest_base(const char *image_name, abi_ulong guest_loaddr,
>>                           image_name, (uint64_t)guest_hiaddr - guest_base);
>>              exit(EXIT_FAILURE);
>>          }
>> +#endif
>
> Could we write this so that we have the #if...#endif
> conditional nested inside the if { ... } else { ... },
> rather than having the two conditional constructs
> oddly intermeshed?

Sure - I thought the compiler would complain about having an empty else
leg but I've just checked and it doesn't seem to care.

>
> thanks
> -- PMM


-- 
Alex Bennée