I have compiled VM successfully with this commit (it's required to update
classlib).
But with this commit I have SIGSEGV on Linux in NLS code added with
commit...
--

Ilya

2007/2/16, Salikh Zakirov <[EMAIL PROTECTED]>:


Is it just me, or is someone else having this error?
(Linux/i686, SUSE9)


----8<----
build.native.init:
    [echo] ## Building native of 'vm.port'

build.native.c:
      [cc] 11 total files to be compiled.

build.native.cpp:
      [cc] 3 total files to be compiled.
      [cc]
/files/sszakiro/harmony/drlvm/trunk/vm/port/src/logger/logger.cpp:
In member
      [cc]    function `const char* LogParams::release()':
      [cc]
/files/sszakiro/harmony/drlvm/trunk/vm/port/src/logger/logger.cpp:85:
error: `
      [cc]    buf_write_text' undeclared (first use this function)
      [cc]
/files/sszakiro/harmony/drlvm/trunk/vm/port/src/logger/logger.cpp:85:
error: (Each
      [cc]    undeclared identifier is reported only once for each
function it
appears
      [cc]    in.)
----8<----

The text was introduced by the commit

   Author: Gregory Shimansky <[EMAIL PROTECTED]>
   Date:   Tue Feb 13 13:15:34 2007 +0600

[r506992] Applied HARMONY-3026 [drlvm] Add Internationalization support to
the drlvm. Tests passed on Ubuntu6 x86, Windows 2003 server x86 and SuSE9
x86_64


The commenting code out seems to work, but I would like to hear opinion of
someone more knowledgeable in this area.

-----8<----
--- vm/port/src/logger/logger.cpp
+++ vm/port/src/logger/logger.cpp
@@ -81,9 +81,14 @@ const char* LogParams::release() {
    } else {
        messageId = def_messageId;
    }
+    /*
+       XXX: there is no buf_write_text function in HyPortLibrary,
+       and copying messageId into messageId makes no sense.
+
    if (portlib_for_logger) {
        messageId = ((HyPortLibrary
*)portlib_for_logger)->buf_write_text((struct HyPortLibrary
*)portlib_for_logger, (const char *)messageId, (IDATA) strlen(messageId));
    }
+    */
    int i = 0;
    while(messageId[i] != '\0') {
        if (messageId[i] == '{' && messageId[i + 1] >= '0' &&


--
Salikh Zakirov


Reply via email to