[ 
https://issues.apache.org/jira/browse/THRIFT-1398?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Nevo Hed updated THRIFT-1398:
-----------------------------

    Description: 
I use an older version of g++ (4.1.2) but some of my fellow team members who 
are using a more recent version (4.6.1)
they are reporting that (when used with '-Wall') g++, justifiably complains 
about possible lack of return


    int32_t calculate(const int32_t logid, const Work& w) {
      size_t sz = ifaces_.size();
      for (size_t i = 0; i < sz; ++i) {
        if (i == sz - 1) {
          return ifaces_[i]->calculate(logid, w);
        } else {
          ifaces_[i]->calculate(logid, w);
        }
      }
    }


Though a human reading this code can see that the last loop execution results 
in a return, the compiler can't
     "error: control reaches end of non-void function [-Werror=return-type]"
IMHO we should have a single return at the end of the function


I would add a generation-time option to suppress multiface if that seems 
reasonable

At this time we find that we need to either disable the warning or up the 
optimization (or both?)


  was:
I use an older version of g++ (4.1.2) but some of my fellow team members who 
are using a more recent version (4.6.1)
they are reporting that (when used with '-Wall') g++, justifiably complains 
about the second 


  int32_t calculate(const int32_t logid, const Work& w) {
    size_t sz = ifaces_.size();
    for (size_t i = 0; i < sz; ++i) {
      if (i == sz - 1) {
        return ifaces_[i]->calculate(logid, w);
      } else {
        ifaces_[i]->calculate(logid, w);
      }
    }
  }


Though a human reading this code can see that the last loop execution results 
in a return, the compiler can't
     "error: control reaches end of non-void function [-Werror=return-type]"

I would add a generation-time option to suppress multiface if that seems 
reasonable

At this time we find that we need to eother disable the warning or up the 
optimization (or both?)


    
> Compiler warning on generated code (Multiface kills -Wall -Werror on gcc 
> 4.6.1)
> -------------------------------------------------------------------------------
>
>                 Key: THRIFT-1398
>                 URL: https://issues.apache.org/jira/browse/THRIFT-1398
>             Project: Thrift
>          Issue Type: Bug
>          Components: C++ - Compiler
>    Affects Versions: 0.6, 0.7
>         Environment: g++ v4.6.1, with all warnings enabled -Wall, using the 
> purist "treat warnings as errors" -Werror flag.  No optimization flags
>            Reporter: Nevo Hed
>
> I use an older version of g++ (4.1.2) but some of my fellow team members who 
> are using a more recent version (4.6.1)
> they are reporting that (when used with '-Wall') g++, justifiably complains 
> about possible lack of return
>     int32_t calculate(const int32_t logid, const Work& w) {
>       size_t sz = ifaces_.size();
>       for (size_t i = 0; i < sz; ++i) {
>         if (i == sz - 1) {
>           return ifaces_[i]->calculate(logid, w);
>         } else {
>           ifaces_[i]->calculate(logid, w);
>         }
>       }
>     }
> Though a human reading this code can see that the last loop execution results 
> in a return, the compiler can't
>      "error: control reaches end of non-void function [-Werror=return-type]"
> IMHO we should have a single return at the end of the function
> I would add a generation-time option to suppress multiface if that seems 
> reasonable
> At this time we find that we need to either disable the warning or up the 
> optimization (or both?)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: 
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

Reply via email to