The problem was in the source of random related functions,
which are used in this test by several thread in parallel,
while this code is known and advertised as non-thread safe.
The access by multiple threads led to multiple increments of the
mt_index static variable, which was then never reset to zero and
led to all successive calls to random function to return zero.
This fix simply insures that the mt_state array is never
accessed past its end, and that mt_index is reset correctly even
if multiple increments do happen.
It does not make the random function thread-safe.
* Use local variable l_index to avoid accessing mt_state array past last element.
* Change mt_index=MTWIST_N+1 into l_index>=MTWIST_N+1
to insure that mt_init will be called again if index is past end.
git-svn-id: trunk@34320 -
rtl/inc/system.inc:
+ ResHeader variable that provides the public _FPC_ResLocation symbol used by intres.inc; this variable will always be available and without indirect entry information it will simply hook up to FPC_RESLOCATION itself
* SetupEntryInformation: setup FPCResLocation
rtl/inc/intres.inc:
* no need to cater for indirect entry or not anymore
git-svn-id: trunk@33964 -
Switch Mac OS X to indirect entry information.
compiler/systems/t_bsd.pas, TLinkerBSD:
+ implement InitSysInitUnitName
compiler/system.pas:
* systems_internal_sysinit: add Darwin systems
compiler/hlcgobj.pas, thlcgobj:
* gen_proc_symbol_end: for Darwin systems don't directly call PascalMain for libraries, but FPC_LIBMAIN instead
rtl/darwin:
+ add sysinit.pas unit which contains the executable and library entry points for Darwin
rtl/darwin/Makefile.fpc:
* add sysinit unit
rtl/bsd/sysosh.inc:
+ add a Darwin specific TPlatformEntryInformation (could probably be used for all Unix systems...)
rtl/bsd/system.pp:
* define FPC_HAS_INDIRECT_MAIN_INFORMATION for Darwin systems once we're no longer bootstrapping with 2.6.x
+ add EntryInformation variable (this could maybe moved to system.inc...)
+ add new procedure SysEntry that is called from the entrypoint and which sets up necessary information required by other parts of the RTL
* the old FPC_SYSTEMMAIN is still in place for bootstrapping
rtl/inc/systemh.inc, TEntryInformation:
* PascalMain is cdecl on non-Windows systems (ToDo: really?)
rtl/inc/system.inc:
* initialstklen is provided by the indirect entry information, so no external here
........
Fix bootstrapping.
rtl/darwin/sysinit.pas:
* enable the code only when no longer bootstrapping with 2.6.x
........
Fix bootstrapping.
rtl/objpas/objpas.pp:
* define FPC_HAS_INDIRECT_MAIN_INFORMATION for Darwin only while not bootstrapping with 2.6.x
........
rtl/darwin/sysinit.pas:
* reference PASCALMAIN correctly (I wonder how this even worked previously :/ )
git-svn-id: trunk@33949 -
rtl/inc/system.inc:
* SetupEntryInformation: call new, optional function OSSetupEntryInformation to handle platform specific entry information initialization
rtl/win/sysosh.inc, TEntryInformationOS:
+ new field TlsKeyAddr which will hold the address to the main binary's TlsKey variable
win32/sysinit.inc:
+ provide the variable holding the TlsKey and pass that on to the entry information record
win32/system.pp:
+ new OS specific entry information initialization (currently only the TlsKey)
win/systhrd.inc:
* declare TlsKey as a pointer to a DWord value instead of a DWord; on non-indirect entry platforms this is initialized with the address of new variable TlsKeyVar, on indirect entry platforms it will be initialized by the entry information initialization
* adjust usages of TlsKey from DWord to PDWord
win/systlsdir.inc:
* TlsKey is now a PDWord and (in sysinit) points to TlsKeyVar
win/syswin.inc:
* adjust TlsKey usage
inc/heaptrc.pp:
* TlsKey is now a PDWord, thus adjust the import and the usage
git-svn-id: trunk@33091 -
rtl/inc/system.inc:
* provide a public alias for SetupEntryInformation so that it can be used in SysInit as well without globally exporting it
rtl/win/systlsdir.inc:
* Exec_tls_callback: call SetupEntryInformation before initalizing the heap and multithreading so that the threadvar tables can be referenced correctly
git-svn-id: trunk@33035 -
rtl/inc/system.inc:
+ new variables that hold pointers to the tables
+ new procedure SetupEntryInformation() that should be used in the entry points of indirect entry targets to setup cross target fields
+ for direct entry targets the two resource string tables are imported here and supplied as initialization to the table pointers
objpas/objpas.pp:
* adjust table declarations so that the pointers provided from the System unit are used
* adjust usages of the tables as they are now pointers
win32/system.pp:
* Exe_Entry: use SetupEntryInformation()
win/syswin.inc:
* Dll_Entry: use SetupEntryInformation()
git-svn-id: trunk@33028 -
+ add (uninitialized) global variable EntryInformation for systems that use indirect main information
win32/system.pp:
- remove initialized EntryInformation and instead rely on the one passed by the sysinit* units
git-svn-id: trunk@33026 -
so that we can still find the type definitions in it once we don't store
the entire implementation symtable to the ppu anymore
git-svn-id: trunk@32151 -
type-safe way (for LLVM, and also internal consistency checking between
the VMT as generated in nobj.pas and ncgvmt.pas)
o also converted the VMT validity checking to the node level
git-svn-id: trunk@30950 -
* Unit finalization executes arbitrarily large amount of code. Doing it when error occurred can cause
other errors, overwriting the original error information.
* Code that prints error message depends on everything and a kitchen sink (most notably, Unicode manager
and lineinfo unit). Running it after finalizing units can be successful only by coincidence.
* Last but not least, this sequence (ExitProc -> print RTE -> finalize units) is same as one used in Delphi.
git-svn-id: trunk@29733 -
Each port that allows the usage of dynamic libraries should call InitSystemDynLibs during initialization of the System unit.
The DynLibs unit has been adjusted to be a forwarder for the functions that now reside in the System unit (for backwards compatiblity just in case). Additionally the unit will register the DynLibsManager if it doesn't reside in the System unit anyway.
Currently only the Windows targets (Win32, Win64 and WinCE) implement the DynLibsManager inside the System unit. For other systems using the LoadLibrary, etc. functions will lead to a run error/exception.
If a port wants to implement its own DynLibsManager then it needs to define DISABLE_NO_DYNLIBS_MANAGER. TLibHandle, NilHandle and optionally TOrdinalEntry (it's set to SizeUInt otherwise) should be defined in sysdlh.inc which needs to be enabled using the define FPC_SYSTEM_HAS_SYSDLH (because there are targets which have FEATURE_DYNLIBS enabled, but don't support them... e.g. powerpc-wii -.-).
The DynLibsManager contains methods for loading a library based on a RawByteString and based on a UnicodeString. BOTH should be implemented, but internally one can forward to the other.
The loading by ordinal *can* be be implemented. If it is set to Nil then the implementation of GetProcAddress(lib,ordinal) will return Nil.
+ new functions SetDynLibsManager and GetDynLibsManager to set and retrieve the current DynLibsManager
* provide interface of DynLibs unit in unit System
git-svn-id: trunk@29613 -
be included at the start of the implementation of the system unit (before
the rest of except.inc)
* catch declarations in/loading from the system unit of the TExceptAddr type
* use this type instead of hardcoded size constants in the compiler
* in generic code that is active for all targets, puts its use in a virtual
method since it's only valid for targets using setjmp/longjmp-style
exception handling (and the record is not defined at all in the JVM RTL)
git-svn-id: branches/hlcgllvm@28376 -
- rm ifdefs for i8086 far code memory model when doing stack traces. The
compiler will now choose the correct overload of HexStr according to the
CodePointer type.
git-svn-id: trunk@27163 -
* As a first part of cleanup, replaced StackTop with function returning StackBottom+StackLength.
* On Win32 and Win64, StackTop returns the stack base from TIB instead.
git-svn-id: trunk@27119 -
+ codepage support for textrec/filerec and the above routines
* textrec/filerec now store the filename by default using widechar. It is
possible to switch back to ansichars using the FPC_ANSI_TEXTFILEREC define.
In that case, from now on the filename will always be stored in
DefaultFileSystemEncoding
* fixed potential buffer overflows and non-null-terminated file names in
textrec/filerec
+ dodirseparators(pwidechar), changed the dodirseparators(pchar/pwidechar)
parameters into var-parameters and gave those routines an extra parameter
that indicates whether the p(wide)char can be changed in place if
necessary or whether a copy must be made first (avoids us having to make
all strings always unique everywhere, because they might be changed on
some platforms via a pchar)
* do_open/do_erase/do_rename got extra boolean parameters indicating whether
the passed pchars point to data that can be freely changed (to pass on to
dodirseparators() if applicable)
* objects.pp: force assign(pchar) to be called, because
assign(array[0..255]) cannot choose between pchar and rawbytestring
versions (and removing the pchar version means that assign(pchar) will
be mapped to assign(shortstring) in {$h-})
* fixed up some routines in other units that depend on the format of
the textrec/filerec.name field
git-svn-id: branches/cpstrrtl@25137 -
* renamed platform-specific pchar versions of those rouines to do_*() and
changed them to either rawbytestring or unicodestring depending on the
FPCRTL_FILESYSTEM_SINGLE_BYTE_API/FPCRTL_FILESYSTEM_TWO_BYTE_API setting
* implemented generic shortstring versions of those routines on top of either
rawbytestring or unicodestring depending on the API-kind (in case of the
embedded target, if ansistring are not supported they will map directly
to shortstring routines instead)
* all platform-specific *dir() routines with rawbytestring parameters now
receive their parameters in DefaultFileSystemCodePage
- removed no longer required ansistring variants from the objpas unit
- removed no longer required FPC_SYS_MKDIR etc aliases
* factored out empty string and inoutres<>0 checks from platform-specific
*dir() routines to generic ones
o platform-specific notes:
o amiga/morphos: check new pathconv(rawbytestring) function
o macos TODO: convert PathArgToFSSpec (and the routines it calls) to
rawbytestring
o nativent: added SysUnicodeStringToNtStr() function
o wii: convert dirio callbacks to use rawbytestring to avoid conversion
+ test for unicode mk/ch/rm/getdir()
git-svn-id: branches/cpstrrtl@25048 -
getdir(rawbytestring):rawbytestring so it can accept strings in any
encoding and cleanly return results in DefaultRTLFileSystemCodePage
+ getdir(unicodestring):unicodestring
* renamed the getdir implementation of all platforms except for embedded-
without-ansistring-support to do_getdir(), and depending on the
FPCRTL_FILESYSTEM_SINGLE_BYTE_API/FPCRTL_FILESYSTEM_TWO_BYTE_API define
changed its shortstring parameter to ansistring or unicodestring. The
do_getdir(rawbytestring) routine should just set the code page of the
return value to DefaultFileSystemCodePage without conversion (not
DefaultRTLFileSystemCodePage with conversion, that conversion is performed
in getdir if necessary; this avoids double conversions in case
getdir(unicodestring) is called)
+ generic getdir(shortstring) for platforms supporting either ansistrings or widestrings
o platform maintainers:
o OS/2: adjust code to supports paths > 255 characters if those are supported
o Wii: adjust used callback to use rawbytestring to support paths > 255 characters and
avoid shortstring->rawbytestring conversion overhead
o Windows: GetCurrentDirectoryW is now always used (to prevent data loss)
git-svn-id: branches/cpstrrtl@24993 -
that the type of the parameters can be determined automatically
o added compilerproc declarations for all helpers called in the compiler
via their assembler name, so we can look up the corresponding procdef
git-svn-id: trunk@23325 -
+ CPUARM_HAS_BX is defined if the CPU supports the BX* instruction
+ CPUARM_HAS_REV is defined if the CPU supports the REV instruction. Note that you still have to check for compiler versions > 2.6.0 since the assembler reader of 2.6.0 does not understand that instruction.
+ CPUARM_HAS_IDIV is defined if the CPU supports the sdiv, udiv instructions. Use of this fixes a bug where previously these instruction were only used for armv7-m, while cortex3m cpus also support it.
+ CPUARM_HAS_LDREX is defined if the CPU supports the ldrex/strex instructions. Use of this fixes a bug with armv7(-a) cpus where this path has not been used.
+ SYSTEM_HAS_KUSER_CMPXCHG is defined if the system (mainly OS) support the kuser_cmpxchg functions. Use of this fixes a bug where ARMHF systems did not use it for synchronization (although ARMHF is armv7+ only, i.e. the LDREX path is used anyway)
git-svn-id: trunk@22081 -
get_caller_frame, get_caller_addr and dump_stack
with default NIL value to systemh.inc.
+ Added new get_addr function.
system.inc: Use get_addr and get_frame to call
HandleErrorAddrFrame instead of HandleErrorFrame
in several error functions.
Modify dump_stack to use frame and addr parameters.
Provide a dummy get_addr function returning nil.
i386/i386.inc, x86_64./x86_64.inc: Provide real
implementation of get_addr function.
git-svn-id: trunk@21697 -
and also use this code to trash local variables in inlined routines
o fixes mantis #22088
o makes it possible to also implement it for the jvm target in the future
git-svn-id: trunk@21393 -
+ implement trashing of local variables if subroutine is inlined
* fix some errors releated to interproc. gotos and inlining
+ node_count function
* inline cannot be used with iochecking and safecall calling conventions
* track inherited usage
* don't inline if inherited is used
git-svn-id: trunk@21335 -
to low addresses in case the target cpu/OS does not support this (enabled
for reads on AIX/PPC because on AIX the first 16KB are readable)
git-svn-id: trunk@20832 -
in code and not in constants. In the case of primitive types constant nodes
are used while complex types like arrays, records and objects use a local
variable which is initialized to zero once at the entry of the method (the
variable is reused if Default() is used for the same type multiple times in
the same method). For this a new compilerproc was added which uses FillChar
to initialize the given memory area to zero.
This fixes Mantis #9420.
+ psystem.pas: Added Default symbol to system unit
+ htypechk.pas: Added function "is_valid_for_default" which checks recursively
whether the given type can be used with Default at all.
Forbidden types are files, helpers, ObjC and C++ types. This
check is used for records, arrays and objects only if the mode
is a non-Delphi one, as Delphi ignores these types on lower
levels.
+ msg/errore.msg: Added error message for unsupported types for Default()
+ symconst.pas: Added a new enum value vo_is_default_var which is used for the
local variables utilized by Default() so their initalization
and finalization can be avoided.
+ pexpr.pas: Add handling of Default() intrinsic to "statement_syssym"
+ ninl.pas: Extended tinlinenode by a method which returns the correct node for
a Default() and used that method in handle_typecheck.
* ncgutil.pas: Check for new flag "vo_is_default_var" when initializing and
finalizing local variables.
* ppu.pas: increase PPU version
+ psub.pas:
* Added a new routine which zeros defaultvars of a symtable.
* Use this routine inside "initializevars".
* Also use this routine to initialize the staticsymtable of the unit/program.
* Adjusted ppudump, because of the new enum value.
+ Added implementation of fpc_zeromem to system unit.
+ Added tests for Default()
git-svn-id: trunk@20629 -
* moved console io on embedded systems into a seperate unit, this unit might setup input/output e.g. to be redirected to a serial port
* cleanup of the embedded system unit
git-svn-id: trunk@19168 -
* InternalExit: now calls new procedure SysFlushStdIO
* DoneThread: Also calls SysFlushStdIO to avoid loosing output
if standard files are redirected.
git-svn-id: trunk@17798 -
