Обсуждение: pgadmin 1.10.2 pgscript Build issue on USLC Compiler
The version of pgAdmin you are running. If it is a development snapshot, please include the build date from the Help ->
Aboutscreen, and SVN repository version if known.
1.10.2 Source tarball
------------------------------------------------------------------------------
The configure options used (if compiled from source)
HOST=i586-sco-sysv5 CC="cc -Kpthread -Kalloca" CXX="CC -Kpthread -Kalloca"
./configure --prefix=/usr/local-test \ --with-wx-version=2.8 \ --with-wx=/usr/local-test \
--host=$HOST --target=$HOST --build=$HOST 2>&1 | tee mout-configure
-------------------------------------------------------------------------------
The operating system and version details - e.g. Windows XP Pro, SP2, or Slackware Linux 10, kernel 2.4.21
SCO Unix OpenServer 6.0 MP 4 __USLC__ compiler (UnixWare Base)
-------------------------------------------------------------------------------
The wxWidgets version and configure options used (if compiled from source)
wxGTK 2.8
HOST=i586-sco-sysv5
CFLAGS="-O2" CC="cc -Kpthread -Kalloca" CXX="CC -Kpthread -Kalloca"
./configure --prefix=/usr/local-test \ --enable-unicode --enable-shared --with-gtk \
--enable-gtk2\ --with-opengl \ --host=$HOST --target=$HOST --build=$HOST 2>&1 | tee
mout-configure
--------------------------------------------------------------------------------
The PostgreSQL version and configure options used (if compiled from source)
PostgreSQL 8.2.9
--------------------------------------------------------------------------------
Steps required to reproduce the problem. Please include a simple scheme with optional data if required. This may be
sentprivately upon request from a developer if you are concerned about posting sample code or data to a public mailing
list.
Failing compile command:
CC -Kpthread -Kalloca -DHAVE_CONFIG_H -I. -I.. -I/usr/local-test/include/glib-2.0 -I/usr/local-test/include
-I/usr/gnu/include-I/usr/include -DSSL -I/usr/local-test/lib/wx/include/gtk2-unicode-release-2.8
-I/usr/local-test/include/wx-2.8
-D_FILE_OFFSET_BITS=64 -D_LARGE_FILES -D__WXGTK__ -O2 -DEMBED_XRC -I/usr/include/libxml2 -I/usr/include/libxml2
-DDATA_DIR=\"/usr/local-test/share/pgadmin3/\" -I../pgadmin/include -c -o frmQuery.o ./frm/frmQuery.cpp
Build fails with the following:
"../pgadmin/include/frm/frmMain.h", line 76: warning: function "wxFrameBase::SetStatusText(const wxString &, int)" is
hiddenby "frmMain::SetStatusText" -- virtual function override intended? void SetStatusText(const wxString &msg);
^
"../pgadmin/include/frm/frmMain.h", line 81: warning: function "wxWindowBase::Refresh(bool, const wxRect *)" is hidden
by"frmMain::Refresh" -- virtual function override intended? void Refresh(pgObject *data); ^
"../pgadmin/include/gqb/gqbObject.h", line 36: warning: type qualifier on return type is meaningless const
type_gqbObjectgetType() { return Type; } ^
"../pgadmin/include/ctl/explainCanvas.h", line 96: warning: function "wxTopLevelWindowBase::SetShape(const wxRegion &)"
ishidden by "ExplainPopup::SetShape" -- virtual function override intended? void SetShape(ExplainShape *s);
^
"../pgadmin/include/schema/pgServer.h", line 99: warning: function "pgObject::GetFullIdentifier() const" is hidden by
"pgServer::GetFullIdentifier"-- virtual function override intended? wxString GetFullIdentifier(); ^
"../pgadmin/include/pgscript/utilities/pgsCopiedPtr.h", line 38: error: pointer to incomplete class type is not allowed
p(that.p == 0 ? 0 : that.p->clone()) ^ detected during instantiation of
"pgsCopiedPtr<T>::pgsCopiedPtr(constpgsCopiedPtr<T> &) [with T=pgsVariable]" at line 20 of
"../pgadmin/include/pgscript/expressions/pgsExpression.h"
"../pgadmin/include/pgscript/utilities/pgsCopiedPtr.h", line 45: warning: delete of pointer to incomplete class
pdelete(p); ^ detected during instantiation of "pgsCopiedPtr<T>::~pgsCopiedPtr() [with
T=pgsVariable]"at line 20 of "../pgadmin/include/pgscript/expressions/pgsExpression.h"
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
This is only problem of the build.
If I kludge the above
p(that.p == 0 ? 0 : that.p->clone())
and force a build past this with:
p(0)
all of the normal pgadmin3 functionality seems to build and operate correctly.
I am assuming that pgscript will fail because of the above kludge.
I am not a C++ guru but I have had an SCO engineer that is, look at this and
he says that p is not defined at the time the above failing statement is
encountered. It it failing during the first pass of the USLC compiler.
Any help would be greatly appreciated.
Thank you
Paul McNary
pmcnary@cameron.net
I have not seen a bug ticket on the following that I reported last month.
How do I get that accomplished?
Thank you
Paul McNary
pmcnary@cameron.net
The version of pgAdmin you are running. If it is a development snapshot, please include the build date from the Help ->
Aboutscreen, and SVN repository version if known.
1.10.2 Source tarball
------------------------------------------------------------------------------
The configure options used (if compiled from source)
HOST=i586-sco-sysv5 CC="cc -Kpthread -Kalloca" CXX="CC -Kpthread -Kalloca"
./configure --prefix=/usr/local-test \ --with-wx-version=2.8 \ --with-wx=/usr/local-test \
--host=$HOST --target=$HOST --build=$HOST 2>&1 | tee mout-configure
-------------------------------------------------------------------------------
The operating system and version details - e.g. Windows XP Pro, SP2, or Slackware Linux 10, kernel 2.4.21
SCO Unix OpenServer 6.0 MP 4 __USLC__ compiler (UnixWare Base)
-------------------------------------------------------------------------------
The wxWidgets version and configure options used (if compiled from source)
wxGTK 2.8
HOST=i586-sco-sysv5
CFLAGS="-O2" CC="cc -Kpthread -Kalloca" CXX="CC -Kpthread -Kalloca"
./configure --prefix=/usr/local-test \ --enable-unicode --enable-shared --with-gtk \
--enable-gtk2\ --with-opengl \ --host=$HOST --target=$HOST --build=$HOST 2>&1 | tee
mout-configure
--------------------------------------------------------------------------------
The PostgreSQL version and configure options used (if compiled from source)
PostgreSQL 8.2.9
--------------------------------------------------------------------------------
Steps required to reproduce the problem. Please include a simple scheme with optional data if required. This may be
sentprivately upon request from a developer if you are concerned about posting sample code or data to a public mailing
list.
Failing compile command:
CC -Kpthread -Kalloca -DHAVE_CONFIG_H -I. -I.. -I/usr/local-test/include/glib-2.0 -I/usr/local-test/include
-I/usr/gnu/include-I/usr/include -DSSL -I/usr/local-test/lib/wx/include/gtk2-unicode-release-2.8
-I/usr/local-test/include/wx-2.8
-D_FILE_OFFSET_BITS=64 -D_LARGE_FILES -D__WXGTK__ -O2 -DEMBED_XRC -I/usr/include/libxml2 -I/usr/include/libxml2
-DDATA_DIR=\"/usr/local-test/share/pgadmin3/\" -I../pgadmin/include -c -o frmQuery.o ./frm/frmQuery.cpp
Build fails with the following:
"../pgadmin/include/frm/frmMain.h", line 76: warning: function "wxFrameBase::SetStatusText(const wxString &, int)" is
hiddenby "frmMain::SetStatusText" -- virtual function override intended? void SetStatusText(const wxString &msg);
^
"../pgadmin/include/frm/frmMain.h", line 81: warning: function "wxWindowBase::Refresh(bool, const wxRect *)" is hidden
by"frmMain::Refresh" -- virtual function override intended? void Refresh(pgObject *data); ^
"../pgadmin/include/gqb/gqbObject.h", line 36: warning: type qualifier on return type is meaningless const
type_gqbObjectgetType() { return Type; } ^
"../pgadmin/include/ctl/explainCanvas.h", line 96: warning: function "wxTopLevelWindowBase::SetShape(const wxRegion &)"
ishidden by "ExplainPopup::SetShape" -- virtual function override intended? void SetShape(ExplainShape *s);
^
"../pgadmin/include/schema/pgServer.h", line 99: warning: function "pgObject::GetFullIdentifier() const" is hidden by
"pgServer::GetFullIdentifier"-- virtual function override intended? wxString GetFullIdentifier(); ^
"../pgadmin/include/pgscript/utilities/pgsCopiedPtr.h", line 38: error: pointer to incomplete class type is not allowed
p(that.p == 0 ? 0 : that.p->clone()) ^ detected during instantiation of
"pgsCopiedPtr<T>::pgsCopiedPtr(constpgsCopiedPtr<T> &) [with T=pgsVariable]" at line 20 of
"../pgadmin/include/pgscript/expressions/pgsExpression.h"
"../pgadmin/include/pgscript/utilities/pgsCopiedPtr.h", line 45: warning: delete of pointer to incomplete class
pdelete(p); ^ detected during instantiation of "pgsCopiedPtr<T>::~pgsCopiedPtr() [with
T=pgsVariable]"at line 20 of "../pgadmin/include/pgscript/expressions/pgsExpression.h"
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
This is only problem of the build.
If I kludge the above
p(that.p == 0 ? 0 : that.p->clone())
and force a build past this with:
p(0)
all of the normal pgadmin3 functionality seems to build and operate correctly.
I am assuming that pgscript will fail because of the above kludge.
I am not a C++ guru but I have had an SCO engineer that is, look at this and
he says that p is not defined at the time the above failing statement is
encountered. It it failing during the first pass of the USLC compiler.
Any help would be greatly appreciated.
Thank you
Paul McNary
pmcnary@cameron.net
Le 05/06/2010 03:29, Paul McNary a écrit : > I have not seen a bug ticket on the following that I reported last month. > How do I get that accomplished? > Probably because I didn't have the time to reply to your question. > This is only problem of the build. > If I kludge the above > p(that.p == 0 ? 0 : that.p->clone()) > > and force a build past this with: > p(0) > > all of the normal pgadmin3 functionality seems to build and operate > correctly. > I am assuming that pgscript will fail because of the above kludge. > > I am not a C++ guru but I have had an SCO engineer that is, look at this > and > he says that p is not defined at the time the above failing statement is > encountered. It it failing during the first pass of the USLC compiler. > I wouldn't be surprise that this is a bug with the SCO compiler. You should try with gcc. We never had any issues on this part of the code with the gcc compiler and the Visual C++ compiler. -- Guillaumehttp://www.postgresql.frhttp://dalibo.com
On Sat, Jun 5, 2010 at 10:41 PM, Guillaume Lelarge <guillaume@lelarge.info> wrote: > > I wouldn't be surprise that this is a bug with the SCO compiler. You > should try with gcc. We never had any issues on this part of the code > with the gcc compiler and the Visual C++ compiler. Which are the only two we support I'm afraid. Note that SCO isn't a platform we've tested or claim to support in any way either. -- Dave Page EnterpriseDB UK: http://www.enterprisedb.com The Enterprise Postgres Company