Build failure on illumos after 8238988

David Holmes david.holmes at oracle.com
Fri Feb 28 22:44:26 UTC 2020


Ignore this - my email was slow to update this morning and I didn't see 
the rest of the thread.

Thanks Dan for taking care of this.

David

On 29/02/2020 8:40 am, David Holmes wrote:
> Hi Peter,
> 
> On 28/02/2020 11:16 pm, Peter Tribble wrote:
>> Hi,
>>
>> I'm building from source on illumos. After
>>
>> 8238988: Rename thread "in stack" methods and add in_stack_range
>>
>> I get a build failure:
>>
>> "/export/home/ptribble/ud/jdk-jdk-15+12/src/hotspot/os_cpu/solaris_x86/thread_solaris_x86.cpp", 
>>
>> line 92: Error: Formal argument limit of type unsigned char* in call to
>> Thread::is_in_stack_range_incl(unsigned char*, unsigned char*) const is
>> being passed long*.
>> 1 Error(s) detected.
>>
>> Looking at line 92 of 
>> src/hotspot/os_cpu/solaris_x86/thread_solaris_x86.cpp:
>>
>>      if (!jt->is_in_stack_range_incl((address)ret_fp, ret_sp)) {
>>
>> shouldn't the 2nd argument also be cast, as the signature has both being
>> the same type? Like so:
>>
>>      if (!jt->is_in_stack_range_incl((address)ret_fp, (address)ret_sp)) {
>>
>> with that change, I get a clean build again, and a working jdk.
> 
> Yep. I'll file a bug and get this fixed ASAP but that will be Monday.
> 
> Thanks,
> David
> 
>> Thanks,
>>


More information about the hotspot-runtime-dev mailing list