Logo Search packages:      
Sourcecode: virtualbox-ose version File versions  Download package

Todo List

page VirtualBox
Make links to the components.

Class __DRIversionRec
There are several data structures that explicitly store a major version, minor version, and patch level. These structures should be modified to have a __DRIversionRec instead.

Member _ADAPT::fProcessingPacketFilter
join all boolean states to one field treated as flags bitmap

Member _OP_PARAMETER::param
switch param and parval and move disp64 and flags up here with the other 64-bit vars to get more natural alignment and save space.

Member _SHFLDIRINFO::name
malc, a description, please.

Class _VBoxSFInformation
different file info structures

Class _VBOXSTATSCONTEXT
port me.

Member _XF86DRISAREA::drawable_lock
Use readers/writer lock for drawable_lock

Class Backupable< D >
(dmik) remove after we switch to VirtualBoxBaseNEXT completely

Member BUSLOGIC::IBase
use PDMDEVINS::IBase

Class CIShared< D >
put the "original" state definition here...

Member CombinedProgress::Cancel ()
Teleportation: Shouldn't this be propagated to mProgresses? If powerUp creates passes a combined progress object to the client, I won't get called back since I'm only getting the powerupProgress ... Or what?

Member VirtualBoxBase::translate (const char *context, const char *sourceText, const char *comment=0)
(dmik) incorporate Qt translation file parsing and lookup

Member Console::configMedium (PCFGMNODE pLunL0, bool fPassthrough, DeviceType_T enmType, bool fUseHostIOCache, bool fSetupMerge, unsigned uMergeSource, unsigned uMergeTarget, IMedium *pMedium, MachineState_T aMachineState, HRESULT *phrc)
Live Migration: I don't quite like this, we risk screwing up when we're resuming the VM if some 3rd dude have any of the VDIs open with write sharing denied. However, if the two VMs are sharing a image it really is necessary....

Member Console::configNetwork (const char *pszDevice, unsigned uInstance, unsigned uLun, INetworkAdapter *aNetworkAdapter, PCFGMNODE pCfg, PCFGMNODE pLunL0, PCFGMNODE pInst, bool fAttachDetach)
pritesh: get the dhcp server name from the previous network configuration and then stop the server else it may conflict with the dhcp server running with the current attachment type

Member Console::CreateSharedFolder (IN_BSTR aName, IN_BSTR aHostPath, BOOL aWritable)
see

Member Console::CreateSharedFolder (IN_BSTR aName, IN_BSTR aHostPath, BOOL aWritable)
in AttachUSBDevice() about the Paused state

Member Console::doCPUAdd (ULONG aCpu)
r=bird: This should be allowed on paused VMs as well. Later.

Member Console::doCPUAdd (ULONG aCpu)
warning if the guest doesn't support it

Member Console::fetchSharedFolders (BOOL aGlobal)
grab & process global folders when they are done

Member Console::fetchSharedFolders (BOOL aGlobal)
report errors as runtime warnings through VMSetError

Member Console::onAdditionsOutdated ()
Use the On-Screen Display feature to report the fact. The user should be told to install additions that are provided with the current VBox build: VBOX_VERSION_MAJOR.VBOX_VERSION_MINOR.VBOX_VERSION_BUILD

Member Console::onUSBControllerChange ()
implement one day.

Member Console::powerDown (Progress *aProgress=NULL)
r=bird: Doesn't make sense. Please remove after 3.1 has been branched off.

Member Console::powerUp (IProgress **aProgress, bool aPaused)
move down to powerDown();

Member Console::powerUp (IProgress **aProgress, bool aPaused)
we should save it afterwards, but that isn't necessarily a good idea. Find a better place for this (VBoxSVC).

Member Console::powerUp (IProgress **aProgress, bool aPaused)
later applies to floppies as well

Member Console::processRemoteUSBDevices (uint32_t u32ClientId, VRDPUSBDEVICEDESC *pDevList, uint32_t cbDevList)
(sunlover) REMOTE_USB Strict validation of the pDevList.

Member Console::processRemoteUSBDevices (uint32_t u32ClientId, VRDPUSBDEVICEDESC *pDevList, uint32_t cbDevList)
(r=dmik) warning reporting subsystem

Member Console::RemoveSharedFolder (IN_BSTR aName)
see

Member Console::RemoveSharedFolder (IN_BSTR aName)
in AttachUSBDevice() about the Paused state

Member Console::Reset ()
r=bird: This should be allowed on paused VMs as well. Later.

Member Console::setMachineState (MachineState_T aMachineState, bool aUpdateServer=true)
(dmik)

Member Console::SleepButton ()
Live Migration: ???

Member CPUMCTX::fpu
This doesn't have to be in X86FXSTATE on CPUs without fxsr - we need a type for the actual format or convert it (waste of time).

Member CPUMCTX_VER1_6::dr4
remove dr4 and dr5.

Member CPUMCTX_VER1_6::fpu
This doesn't have to be in X86FXSTATE on CPUs without fxsr - we need a type for the actual format or convert it (waste of time).

Member CPUMCTX_VER1_6::idtr
fix this hack

Member CPUMCTX_VER1_6::ldtr
fix this hack

Class CRVBOXHGCMDATA
Try use the Vbgl interface instead of talking directly to the driver?

Member DRVNETSNIFFER::INetworkConfig
this is a main interface and shouldn't be here...

Member DRVTAP::XmitLock
The transmit thread.

Member DRVVDE::XmitLock
The transmit thread.

Class E1kIpHeader
use+extend RTNETIPV4

Class E1kTcpHeader
use+extend RTNETTCP

Member EEPROM93C46::save (PSSMHANDLE pSSM)
save and load methods

Class EPTPML4EBITS
uint64_t isn't safe for bitfields (gcc pedantic warnings, and IIRC, this did cause trouble with one compiler/version).

Member com::EventQueue::interruptEventQueueProcessing ()
: rethink me!

Member xml::File::File (Mode aMode, const char *aFileName, bool aFlushIt=false)
change to RTFILE_O_DENY_WRITE where appropriate.

Class GCFGMVALUEREQ
Move got GCFGM.h when it's implemented.

Class gl_tex_env_combine_state
If GL_NV_texture_env_combine4 is ever supported, the arrays in this structure will need to be expanded for 4 elements.

Member Global::vboxStatusCodeFromCOM (HRESULT aComStatus)
Check for the win32 facility and use the RTErrConvertFromWin32 function on windows.

Member HGCMMsgCore::m_pPrev
seems not necessary. Previous element in a message queue.

Member HGCMMsgCore::m_pPrev
seems not necessary. Previous element in a message queue.

Member HostUSBDevice::compare (PCUSBDEVICE pDev1, PCUSBDEVICE pDev2)
Sander, will this work on windows as well? Linux won't reuse an address for quite a while.

Member HostUSBDevice::reset ()
this operation might fail and cause the device to the reattached with a different address and all that.

Class INITBLK16
All structs: big endian?

Class INTNETMACTAB
Having this in a separate structure didn't work out as well as it should. Consider merging it into INTNETNETWORK.

Member INTNETTRUNKSWPORT::DECLR0CALLBACKMEMBER (bool, pfnRecv,(PINTNETTRUNKSWPORT pSwitchPort, void *pvIf, PINTNETSG pSG, uint32_t fSrc))
Do any of the host require notification before frame modifications? If so, we'll add a callback to INTNETTRUNKIFPORT for this (pfnSGModifying) and a SG flag.

Member KDBGSYMBOL::cbSelf
type info.

Class KLDRMODMACHO
interpret the other MH_* formats.

Member Machine::HWData::HWData ()
get the default from the schema if that is possible.

Member Medium::compareLocationTo (const char *aLocation, int &aResult)
NEWMEDIA delegate the comparison to the backend?

Member Medium::GetProperties (IN_BSTR aNames, ComSafeArrayOut(BSTR, aReturnNames), ComSafeArrayOut(BSTR, aReturnValues))
make use of aNames according to the documentation

Member Medium::init (VirtualBox *aVirtualBox, DeviceType_T aDeviceType, CBSTR aLocation, CBSTR aDescription=NULL)
optimize all callers to avoid reconstructing objects with the same information over and over again - in the typical case each VM referring to a particular host drive has its own instance.

Member Medium::init (VirtualBox *aVirtualBox, DeviceType_T aDeviceType, CBSTR aLocation, CBSTR aDescription=NULL)
generate uuid (similarly to host network interface uuid) from location and device type

Member Medium::init (VirtualBox *aVirtualBox, Medium *aParent, DeviceType_T aDeviceType, const settings::Medium &data)
handle host drive settings here as well?

Member Medium::queryInfo ()
This kind of opening of images is assuming that diff images can be opened as base images. Should be documented if it must work for all medium format backends.

Member Medium::queryInfo ()
NEWMEDIA what to do if the parent is not accessible while the diff is? Probably nothing. The real code will detect the mismatch anyway.

Member Medium::startThread (Medium::Task *pTask)
use a more descriptive task name

Member Medium::taskCloneHandler (Medium::CloneTask &task)
r=klaus target isn't locked, race getting the state

Member Medium::taskCreateDiffHandler (Medium::CreateDiffTask &task)
r=klaus neither target nor base() are locked, potential race!

Member Medium::taskResetHandler (Medium::ResetTask &task)
Below we use a pair of delete/create operations to reset the diff contents but the most efficient way will of course be to add a VDResetDiff() API call

Member Medium::taskResetHandler (Medium::ResetTask &task)
use the same image variant as before

Member MediumFormat::init (const VDBACKENDINFO *aVDInfo)
add extendedFlags to Property when we reach the 32 bit limit (or make the argument ULONG64 after checking that COM is capable of defining enums (used to represent bit flags) that contain 64-bit values)

Member iprt::MiniString::contains (const iprt::MiniString &that, CaseSensitivity cs=CaseSensitive) const
r-bird: Not checking for NULL strings like startsWith does (and endsWith only does half way).

Member iprt::MiniString::endsWith (const iprt::MiniString &that, CaseSensitivity cs=CaseSensitive) const
r=bird: If l2 is 0, then m_psz can be NULL and we will crash. See also handling of l2 == in startsWith.

Member iprt::MiniString::startsWith (const iprt::MiniString &that, CaseSensitivity cs=CaseSensitive) const
r=bird: this differs from endsWith, and I think other IPRT code. If l2 == 0, it matches anything.

Member MMHYPERSTAT::Core
The type is wrong! Get your lazy a$$ over and create that offsetted uint32_t version we need here!

Class my_tm
Replace struct my_tm with RTTIME.

Member matching::ParsedIntervalFilter::parseValue (const char *aFilter, size_t aStart, size_t aEnd, Mode aMode)
(dmik): future optimizations:

Member PDMDEVREG::pfnSoftReset

Member PDMIDISPLAYCONNECTOR::DECLR3CALLBACKMEMBER (int, pfnVBVAResize,(PPDMIDISPLAYCONNECTOR pInterface, const PVBVAINFOVIEW pView, const PVBVAINFOSCREEN pScreen, void *pvVRAM))
Merge with pfnResize.

Member PDMIDISPLAYVBVACALLBACKS::DECLR3CALLBACKMEMBER (int, pfnVHWACommandCompleteAsynch,(PPDMIDISPLAYVBVACALLBACKS pInterface, PVBOXVHWACMD pCmd))
r=bird: if asynch means asyncronous; then s/pfnVHWACommandCompleteAsynch/pfnVHWACommandCompleteAsync/; fi

Class PDMSCATTERGATHER
Promote this to VBox/types.h, VBox/pdmcommon.h or some such place.

Member PDMUSBINS::R3PTRTYPE (void *) pvVUsbDev2
Moved this to PDMUSBINSINT.

Member PDMUSERPERVM::R3PTRTYPE (struct PDMQUEUE *) pQueuesTimer
move more stuff over here.

Class PGMCHUNKR3MAPTLB
Generalize this TLB + AVL stuff, shouldn't be all that difficult when we switch to the new inlined AVL trees (from kStuff).

Class PGMLIVESAVEMMIO2PAGE
Implement dirty MMIO2 page reporting that can be enabled during live save but normally is disabled. Since we can write monitore guest accesses on our own, we only need this for host accesses. Shouldn't be too difficult for DevVGA, VMMDev might be doable, the planned networking fun will be fun since it involves ring-0.

Member PGMPAGE::u16MiscY
Change this to a union with both bitfields, u8 and u accessors. That'll help deal with some of the hardcoded accesses.

Member PGMPAGE::u16MiscY
Include uStateY and uTypeY as well so it becomes 32-bit. This will make it possible to turn some of the 16-bit accesses into 32-bit ones, which may be efficient (stalls).

Member PGMPAGEMAPLOCK::uPageAndType
see PGMPhysIsPageMappingLockValid for possibly incorrect assumptions

Class PGMPOOLPHYSEXT
Calling this an 'extent' is not quite right, find a better name.

Class PGMPOOLPHYSEXT
find out the optimal size of the aidx array

Class PointerShapeChangeData
someone please explain why this is necessary. This breaks darwin solid.

Class RTIDTR
find a more proper place for this structure?

Member RTLDRMODPE::pvBits
the reader should take care of this.

Member RTLOCKVALCLASSREFCHUNK::aRefs [2]
for testing alloction of new chunks.

Member RTMAC::au8 [6]
add a bitfield view of this stuff.

Member RTMEMCACHEINT::pFreeTop
This doesn't scale well when several threads are beating on the cache. Also, it totally doesn't work when we've got a constructor/destructor around or the objects are too small.

Class RTNETARPHDR
add ICMPv6 when needed.

Class RTNETICMPV4
add more ICMPv4 as needed.

Member RTNETIPV6::ip6_vfc
this is probably mislabeled - ip6_flow vs. ip6_vfc, fix later.

Member RTRANDINT::DECLCALLBACKMEMBER (void, pfnGetBytes)(PRTRANDINT pThis
later.

Class RTREQQUEUE
hide this

Class RTS3BUCKETENTRY
Consider making the entire list const unless there are plans for more APIs using this structure which requires the caller to create or modify it.

Class RTS3KEYENTRY
Consider making the entire list const unless there are plans for more APIs using this structure which requires the caller to create or modify it.

Member RTTIMER::pvUser
We have to take the signals on a dedicated timer thread as we (might) have code assuming that signals doesn't screw around on existing threads. (It would be sufficient to have one thread per signal of course since the signal will be masked while it's running, however, it may just cause more compilcations than its worth - sigwait/sigwaitinfo work atomically anyway...) Also, must block the signal in the thread main procedure too.

Class RTZIPLZFHDR
FastLZ? QuickLZ? Others?

Member com::SafeArray::cloneTo (ComSafeArrayOut(T, aArg)) const
Implement me!

Member SCMSTREAMLINE::cch
This could be derived from the offset of the next line if that wasn't so tedious.

Member SELM::Tss
I/O bitmap & interrupt redirection table?

Member guestControl::Service::paramBufferAssign (PVBOXGUESTCTRPARAMBUFFER pBuf, uint32_t cParms, VBOXHGCMSVCPARM paParms[])
Add check to verify if the HGCM request is the same *type* as the buffered one!

Member guestControl::Service::retrieveNextHostCmd (uint32_t u32ClientID, VBOXHGCMCALLHANDLE callHandle, uint32_t cParms, VBOXHGCMSVCPARM paParms[])
r=bird: check if already in the list.

Member guestProp::Service::mpfnHostCallback
we should have classes for thread and request handler thread

Member SessionMachine::AdoptSavedState (IN_BSTR aSavedStateFile)
setError.

Member SessionMachine::BeginTakingSnapshot (IConsole *aInitiator, IN_BSTR aName, IN_BSTR aDescription, IProgress *aConsoleProgress, BOOL fTakingSnapshotOnline, BSTR *aStateFilePath)
Confusing! Saving is used for both online and offline snapshots.

Member SessionMachine::BeginTakingSnapshot (IConsole *aInitiator, IN_BSTR aName, IN_BSTR aDescription, IProgress *aConsoleProgress, BOOL fTakingSnapshotOnline, BSTR *aStateFilePath)
r=bird: Delete stateTo when appropriate.

Member SessionMachine::deleteSnapshotHandler (DeleteSnapshotTask &aTask)
NEWMEDIA turn the following errors into warnings because the snapshot itself has been already deleted (and interpret these warnings properly on the GUI side)

Member SessionMachine::deleteSnapshotHandler (DeleteSnapshotTask &aTask)
VBoxHDD cannot handle backward merges where source==active disk yet

Member SessionMachine::deleteSnapshotHandler (DeleteSnapshotTask &aTask)
r=klaus the SaveS_Force is right now a workaround,

Member SessionMachine::setMachineState (MachineState_T aMachineState)
NEWMEDIA don't let mDVDDrive and other children change anything when in the Starting/Restoring state

Member SessionMachine::setMachineState (MachineState_T aMachineState)
Live Migration: clean up (lazy bird)

Member SessionMachine::setMachineState (MachineState_T aMachineState)
(dmik)

Class Shareable< D >
(dmik) remove after we switch to VirtualBoxBaseNEXT completely

Class SMLSFIXTURE
sort this). We expect that the set of visible windows will be the same whether we start the code before the event and handle it or start the code after the event.

Class SSM
Move this to UVM.

Class symtab_command
routines_command (LC_ROUTINES)

Class symtab_command
routines_command_64 (LC_ROUTINES_64)

Class thread_command
fvmlib

Class thread_command
fvmlib_command (LC_IDFVMLIB or LC_LOADFVMLIB)

Class thread_command
dylib

Class thread_command
dylib_command (LC_ID_DYLIB, LC_LOAD_DYLIB, LC_LOAD_WEAK_DYLIB)

Class thread_command
sub_framework_command (LC_SUB_FRAMEWORK)

Class thread_command
sub_client_command (LC_SUB_CLIENT)

Class thread_command
sub_umbrella_command (LC_SUB_UMBRELLA)

Class thread_command
sub_library_command (LC_SUB_LIBRARY)

Class thread_command
prebound_dylib_command (LC_PREBOUND_DYLIB)

Class thread_command
dylinker_command (LC_ID_DYLINKER or LC_LOAD_DYLINKER)

Member TM::offUTC
Implement warpdrive on UTC.

Member TRPM::fSafeToDropGuestIDTMonitoring
r=bird: bool and 7 byte achPadding1.

Member TRPMCPU::uActiveErrorCode
don't use RTGCUINT

Member TRPMCPU::uSavedVector
don't use RTGCUINT

Member USBProxyService::USBProxyService (HostUSB *aHost)
add the required locking.

Member USBProxyService::processChanges (void)
detect status changes!

Member USBProxyService::processChanges (void)
add details .

Member com::Utf8Str::copyFrom (CBSTR s)
r=bird: This isn't throwing std::bad_alloc / handling return codes. Also, this technically requires using RTStrFree, ministring::cleanup() uses RTMemFree.

Member com::Utf8Str::copyFrom (CBSTR s)
optimize by using a different RTUtf* function

Class uuid_command
dysymtab_command (LC_DYSYMTAB)

Class uuid_command
dylib_table_of_contents

Class uuid_command
dylib_module_32

Class uuid_command
dylib_module_64

Class uuid_command
dylib_reference

Class uuid_command
twolevel_hints_command (LC_TWOLEVEL_HINTS)

Class uuid_command
twolevel_hint

Class uuid_command
prebind_cksum_command (LC_PREBIND_CKSUM)

Member UVM::pNext
move this into vm.s!

Member VBGLR3GUESTPROPENUM::pchBuf
add a magic and validate the handle.

Member VBoxGlobal::centerWidget (QWidget *aWidget, QWidget *aRelative, bool aCanResize=true)
(r=dmik) not sure if we really need this

Member VBoxGlobal::details (const CMedium &aHD, bool aPredictDiff)
in VBoxMedium::details()).

Member VBoxGlobal::init ()
Live Migration: New icon? (not really important)

Member VBoxGlobal::init ()
(dmik) separate icon?

Member VBoxGlobal::init ()
live snapshot deletion: new icon?

Member VBoxGlobal::init ()
add an else { msgbox(syntax error); exit(1); } here, pretty please...

Member VBoxGlobal::locationForHTML (const QString &aFileName)
(dmik) remove?

Member VBoxGuestSeamlessHostThread::stop (void)
is this reasonable? If the thread is in the event loop then the cancelEvent() will cause it to exit. If it enters or exits the event loop it will also notice that we wish it to exit. And if it is somewhere in-between, the yield() should give it time to get to one of places mentioned above.

Class VBOXIDTE_TASKGATE
VBOXIDT -> VBOXDESCIDT, skip the complex variations. We'll never use them.

Member VBoxNetDhcp::VBoxNetDhcp ()
tune to 64 KB with help from SrvIntR0

Member VBoxNetDhcp::findOption (uint8_t uOption, PCRTNETBOOTP pDhcpMsg, size_t cb, size_t *pcbMaxOpt)
search extended dhcp option field(s) when present

Member VBoxNetDhcp::handleDhcpReqDecline (PCRTNETBOOTP pDhcpMsg, size_t cb)
Probably need to match the server IP here to work correctly with other servers.

Member VBoxNetDhcp::handleDhcpReqDecline (PCRTNETBOOTP pDhcpMsg, size_t cb)
this is not required in the initial implementation, do it later.

Member VBoxNetDhcp::handleDhcpReqRelease (PCRTNETBOOTP pDhcpMsg, size_t cb)
Probably need to match the server IP here to work correctly with other servers.

Member VBoxNetDhcp::handleDhcpReqRelease (PCRTNETBOOTP pDhcpMsg, size_t cb)
this is not required in the initial implementation, do it later.

Member VBoxNetDhcp::handleDhcpReqRequest (PCRTNETBOOTP pDhcpMsg, size_t cb)
Probably need to match the server IP here to work correctly with other servers.

Member VBoxNetDhcp::handleDhcpReqRequest (PCRTNETBOOTP pDhcpMsg, size_t cb)
This code isn't entirely correct and quite a bit of a hack, but it will have to do for now as the right thing (tm) is very complex. Part of the fun is verifying that the request is something we can and should handle.

Member VBoxNetDhcp::handleDhcpReqRequest (PCRTNETBOOTP pDhcpMsg, size_t cb)
client id stuff - it doesn't make sense here imho, we need IP + MAC. What would make sense though is to compare the client id with what we've got in the lease and use it to root out bad requests.

Member VBoxNetDhcp::handleDhcpReqRequest (PCRTNETBOOTP pDhcpMsg, size_t cb)
this ain't safe.

Member VBoxNetDhcp::makeDhcpReply (uint8_t uMsgType, VBoxNetDhcpLease *pLease, PCRTNETBOOTP pDhcpMsg, size_t cb)
respect the RTNET_DHCP_OPT_MAX_DHCP_MSG_SIZE option

Member VBoxNetDhcp::newLease (PCRTNETBOOTP pDhcpMsg, size_t cb)
extract the client id.

Member VBoxNetDhcp::run (void)
fix interruptability in SrvIntNet!

Member VBoxNetDhcp::run (void)
handle pure bootp clients too?

Member VBoxNetDhcp::tryGoOnline (void)
check this

Member VBoxNetDhcpLease::activate (uint32_t xid)
check if this is really necessary.

Member VBoxNetDhcpWriteCursor::overloadMore (void)
not sure if this stuff is at all correct...

Member VBOXNETFLTINS::u
move MacAddr out of this structure!

Member VBoxProblemReporter::cannotCloseMedium (QWidget *aParent, const VBoxMedium &aMedium, const COMResult &aResult)
(translation-related): the gender of "the" in translations will depend on the gender of aMedium.type().

Member VBoxProblemReporter::cannotOpenMedium (QWidget *aParent, const CVirtualBox &aVBox, VBoxDefs::MediumType aType, const QString &aLocation)
(translation-related): the gender of "the" in translations will depend on the gender of aMedium.type().

Member VBoxProblemReporter::cannotRemountMedium (QWidget *aParent, const CMachine &aMachine, const VBoxMedium &aMedium, bool aMount, bool aRetry)
(translation-related): the gender of "the" in translations will depend on the gender of aMedium.type().

Member VBoxProblemReporter::confirmReleaseMedium (QWidget *aParent, const VBoxMedium &aMedium, const QString &aUsage)
(translation-related): the gender of "the" in translations will depend on the gender of aMedium.type().

Member VBoxProblemReporter::confirmRemoveMedium (QWidget *aParent, const VBoxMedium &aMedium)
(translation-related): the gender of "the" in translations will depend on the gender of aMedium.type().

Member VBoxProblemReporter::showRuntimeError (const CConsole &console, bool fatal, const QString &errorID, const QString &errorMsg) const
(r=dmik) it's just a preliminary box. We need to:

Member VBoxSelectorWnd::vmListViewCurrentChanged (bool aRefreshDetails=true, bool aRefreshSnapshots=true, bool aRefreshDescription=true)
Live Migration: does this make sense?

Class VBOXTSS
Move this to SELM! Use X86TSS32 instead.

Member VBoxVMDescriptionPage::updateState ()
disabling the edit button for a saved VM will not be necessary when we implement the selective VM Settings dialog, where only fields that can be changed in the saved state, can be changed.

Class VirtualBoxBaseWithChildrenNEXT
This is a VirtualBoxBaseWithChildren equivalent that uses the VirtualBoxBaseNEXT implementation. Will completely supersede VirtualBoxBaseWithChildren after the old VirtualBoxBase implementation has gone.

Class com::VirtualBoxErrorInfo
Rename IVirtualBoxErrorInfo/VirtualBoxErrorInfo to something like IExtendedErrorInfo since it's not actually VirtualBox-dependent any more. This will also require to create IExtendedErrorInfo.idl/h etc to let adding this class to custom type libraries.

Class VirtualBoxSupportErrorInfoImpl< C, I >
switch to com::SupportErrorInfo* and remove

Class VirtualBoxSupportErrorInfoImplBase
switch to com::SupportErrorInfo* and remove

Member VMCPU::dbgf
Combine this with other tiny structures.

Class VMINT
Move this and all related things to VMM. The VM component was, to some extent at least, a bad ad hoc design which should all have been put in VMM.

Member VMINTUSERPERVM::Halt
Do NOT add new members here or resue the current, we need to store the config for each halt method seperately because we're racing on SMP guest rigs.

Class VMTask
Move the TAP mess back into the driver!

Class VUSBHUB
eliminate this (PDM / roothubs only).

Member VUSBROOTHUB::Hub
remove this?

Member DECL_FORCE_INLINE
quite possibly incorrect

Member DECL_FORCE_INLINE
possibly incorrect

Member DBGCVAR_ISPOINTER
Rename to DBGCVAR_IS_xyz.

Member rtFsModeIsValid
more checks!

Member rtFsModeIsValidPermissions
more checks!

Member rtldrOpenWithReader
find the aout magics in emx or binutils.

Member DECLHIDDEN
(dmik) well, it's better to consider there's no volume name

Member rtPathFromNative
We must compose the codepoints in the string here. We get file names in normalization form D so we'll end up with normalization form C whatever approach we take.

Member rtPathFromNativeCopy
We must compose the codepoints in the string here. We get file names in normalization form D so we'll end up with normalization form C whatever approach we take.

Member rtPathToNative
We should decompose the string here, but the file system will do that for us if we don't, so why bother.

Member rtThreadNativeSetPriority
Just keep quiet about failures now - we'll fail here because we're not allowed to raise our own priority. This is a problem when starting the threads with higher priority from EMT (i.e. most threads it starts). This is apparently inherited from the parent in some cases and not in other cases. I guess this would come down to which kind of pthread implementation is actually in use, and how many sensible patches which are installed. I need to find a system where this problem shows up in order to come up with a proper fix. There's an pthread_create attribute for not inherting scheduler stuff I think...

Member PDMNETWORKGSOTYPE_IPV4_IPV6_TCP
Needs checking against facts, this is just an outline of the idea.

Member PDMNETWORKGSOTYPE_IPV4_IPV6_UDP
Needs checking against facts, this is just an outline of the idea.

Member com::GetVBoxUserHomeDirectory (char *aDir, size_t aDirLen)
r=bird: RTPathUserHome doesn't necessarily return a full (abs) path like the comment above seems to indicate.

Member com::Initialize ()
the below rough method of changing the aparment type doesn't work on some systems for unknown reason (CoUninitialize() simply does nothing there, or at least all 10 000 of subsequent CoInitializeEx() continue to return RPC_E_CHANGED_MODE there). The problem on those systems is related to the "Extend support for advanced text services to all programs" checkbox in the advanced language settings dialog, i.e. the problem appears when this checkbox is checked and disappears if you clear it. For this reason, we disable the code below and instead initialize COM in MTA as early as possible, before 3rd party libraries we use have done so (i.e. Qt).

Member DISDECL
Define output callback.

Member DISDECL
Using signed integers as sizes is a bit odd. There are still some GCC warnings about mixing signed and unsigend integers.

Member DISDECL
Need to extend this interface to include a code address so we can dissassemble GC code. Perhaps a new function is better...

Member DISDECL
Define output callback.

Member DISDECL
Define output callback.

Member vboxClipboardUtf16GetWinSize
use const where appropriate; delinuxifiy the code (*Lin* -> *Host*); use AssertLogRel*.

Member ClipCompleteDataRequestFromX11
change this to deal with the buffer issues rather than offloading them onto the caller

Member VBOX_HGCM_SVC_VERSION_MAJOR
proper comments.

File pdmsrv.h
This has not been implemented, consider dropping the concept.

Member SHFL_CF_OPEN_TARGET_DIRECTORY
possibly not needed???

Member SHFL_LIST_NONE
might be necessary for future.

Member SHFL_FN_SET_UTF8
macl, a description, please.

Member MSR_K6_EFER_SCE
document EFER

Member X86_PDPE_PG_MASK_FULL
Get rid of the above hack; makes code unreadable.

Member X86_PTE_PAE_PG_MASK_FULL
Get rid of the above hack; makes code unreadable.

Member ADPCTLERR_BAD_NAME
Error codes must be moved to some header file

Member VBOXNETADP_CTL_DEV_NAME
These are duplicates from src/VBox/HostDrivers/VBoxNetAdp/VBoxNetAdpInternal.h

Member ScmStreamGetLine
this doesn't work when pStream->off != pStream->paLines[pStream->iLine-1].pff.

Member ScmStreamSeekAbsolute
Should do a binary search here, but I'm too darn lazy tonight.

Member krdrRTFileGenericProtect
drop cast

Member krdrRTFileGenericRefresh
only zero the areas not covered by raw file bits.

Member KCPU_DECL
check it using cpu.

Member kDbgModPeRVAToSegOff
should probably be less strict about address in the alignment gaps.

Member kdbgModDHFindDbgHelp
find the actual path specs, I'm probably not doing this 100% correctly here.

Member kdbgModDHQuerySymbol
SYMFLAG_CONSTANT must be tested - documentation is too brief to say what is really meant here.

Member kdbgModDHRVAToSegOff
should probably be less strict about address in the alignment gaps.

Member kdbgModDHTryDbgHelp
somehow reject 64-bit .dlls when in 32-bit mode... dwFileOS is completely useless.

Member KHLP_DECL
need to figure out where the stuff is or how it's inherited on darwin ...

Member KHLP_DECL
convert errno to kErrors

Member kldrDyldDoLoadExeStackSwitch
implement this properly!

Member kldrDyldFindDoDllSearch
Deal with DT_RUNPATH and DT_RPATH.

Member kldrDyldFindDoDllSearch
ld.so.cache?

Member kldrDyldFindDoDllSearch
get rid of the strlen() on the various paths here!

Member kldrDyldFindGetDefaults
*penmSearch = KLDRDYLD_SEARCH_DARWIN;

Member kldrDyldFindLazyInitAppDir
implement this!

Member kldrDyldFindTryOpen
deal with return codes properly.

Member kldrModLXDoNameLookup
hash name table for speed.

Member kldrModLXDoNameLookup
non-resident name table.

Member kldrModLXDoProtect
the gap page should be marked NOACCESS!

Member kldrModLXDoReloc
fix error, add more checks!

Member kldrModLXEnumSymbols
implement enumeration of forwarders properly.

Member kldrModLXEnumSymbols
lazy load the non-resident name table.

Member kldrModLXEnumSymbols
figure out TYPEINFO.

Member kldrModLXQuerySymbol
figure out TYPEINFO.

Member kldrModMachODoEnumSymbols32Bit
implement indirect and prebound symbols.

Member kldrModMachODoEnumSymbols64Bit
implement indirect and prebound symbols.

Member kldrModMachODoQuerySymbol32Bit
implement indirect and prebound symbols.

Member kldrModMachODoQuerySymbol64Bit
implement indirect and prebound symbols.

Member kldrModMachOFixupSectionAMD64
Deal with unaligned accesses on non x86 platforms.

Member kldrModMachOFixupSectionGeneric32Bit
Deal with unaligned accesses on non x86 platforms.

Member kldrModMachOObjDoFixups
this aint right.

Member kldrModMachOObjDoImports
Implement N_REF_TO_WEAK.

Member kldrModMachOObjDoImports
implement weak symbols.

Member kldrModMachOParseLoadCommands

Member kldrModMachOParseLoadCommands
fixme!

Member kldrModMachOParseLoadCommands
move up!

Member kldrModMachOParseLoadCommands
update the protection...

Member kldrModMachOPreParseLoadCommands
This doesn't work because of BSS.

Member kldrModMachOPreParseLoadCommands
this requires a query API or flag...

Member kldrModMachOPreParseLoadCommands
deal with this!

Member kldrModMachOPreParseLoadCommands

Member kldrModMachOPreParseLoadCommands
Check anything here need converting?

Member kldrModNativeEnumDbgInfo
implement me on OS/2.

Member kldrModNativeEnumSymbols
implement export enumeration on OS/2.

Member kldrModNativeGetImport
implement import enumeration on OS/2.

Member kldrModNativeGetStackInfo
implement stack info on OS/2.

Member kldrModNativeHasDbgInfo
implement me on OS/2.

Member kldrModNativeNumberOfImports
implement import counting on OS/2.

Member kldrModNativeQueryMainEntrypoint
implement me on OS/2.

Member kldrModPEAllocTLS
implement TLS.

Member kldrModPEDoCall
try/except

Member kldrModPEDoCallTLS
implement TLS support.

Member kLdrModPEDoOptionalHeaderValidation
validate more

Member kLdrModPEDoSectionHeadersValidation
validate shdrs

Member kldrModPEFreeTLS
implement TLS.

Member main

Member main
Create a tstFileLock-2 which doesn't require interaction and counts errors.

Member UnwindInt
need to deal with alternative stacks!

Member UnwindOne
adjust overhead

Member KPRF_TYPE
recursion

Member KPRF_TYPE
bad names and descriptions!

Member IsValid
more validation here..

Member krdrFileCreate
normalize the filename.

Member krdrFileGenericProtect
drop cast

Member krdrFileGenericRefresh
only zero the areas not covered by raw file bits.

Member krdrFileMap

Member krdrFilePageSize
find some header somewhere...

Member krdrFileProtect
implement me.

Member krdrFileRefresh
implement me.

Member krdrFileUnmap
implement me.

Member rtThreadAlloc
get a better error code!

Member rtThreadRemoveLocked
this asserts for threads created by NSPR

Member cpu_apic_rdmsr
: exception ?

Member cpu_apic_wrmsr
: exception if error ?

Member cpu_get_pic_interrupt
r=bird: In the long run we should just do the interrupt handling in EM/CPUM/TRPM/somewhere and if we cannot execute the interrupt handler in raw-mode just reschedule to REM. Once that is done we remove this kludge.

Member DECLCALLBACK
Move stats to REM::s some rainy day we have nothing do to.

Member disas
remove when DISInstr starts supporing 64-bit code.

Member REMR3DECL
this isn't ensuring correct replay order.

Member REMR3DECL
fix the solaris and freebsd math stuff.

Member REMR3DECL
: is it correct? No!

Member REMR3DECL
we could probably benefit from using a CPUM_CHANGED_DRx flag too!

Member REMR3DECL
need to find a way to communicate potential GDT/LDT changes and thread switches. The selector might be the same but not the base/limit.

Member REMR3DECL
this won't work :/

Member REMR3DECL
FOP

Member REMR3DECL
FPUIP

Member REMR3DECL
CS

Member REMR3DECL
FPUDP

Member REMR3DECL
DS

Member REMR3DECL
check if FPU/XMM was actually used in the recompiler

Member REMR3DECL
These values could still be out of sync!

Member remR3NotifyHandlerPhysicalDeregister
this isn't right, MMIO can (in theory) be restored as RAM.

Member remR3StateUpdate
FOP

Member remR3StateUpdate
FPUIP

Member remR3StateUpdate
CS

Member remR3StateUpdate
FPUDP

Member remR3StateUpdate
DS

Member remR3StateUpdate
Fix MXCSR support in QEMU so we don't overwrite MXCSR with 0 when we shouldn't!

Member remR3StateUpdate
check if FPU/XMM was actually used in the recompiler

Member remR3StateUpdate
These values could still be out of sync!

Member need_GL_ARB_occlusion_query
some of those are or'ed with GL_VERSIONS and ain't needed here

Member VBOX_GL_FUNC
change stub's createcontext to reuse driver private part of mesa's ctx to store stub ctx info.

Member vbox_extensions
add more which are supported by chromium like GL_NV_vertex_program etc.

Member vboxdriTexOffsetExtension
not sure we need it from start

Member vboxdriCreateContext
we should also pass some information from the visual back to the host.

Member vboxdriCreateContext
save ctx, or be more confident that we can don't need to.

Member vboxdriInitExtensions
have to check extensions supported by host here first

Member vboxdriInitFuncs
deal with texnames and gl_texture_object pointers which are passed here

Member vboxdriInitFuncs

Member vboxdriInitScreen
check size of DRIRec (passed from X.Org driver), allocate private structure if necessary, parse options if necessary, map VRAM if necessary.

Member vboxdriInitScreen
This should depend on what the host can do, not the guest. However, we will probably want to discover that in the X.Org driver, not here.

Member driCalculateSwapUsage
Instead of caching the glXGetMscRateOML function pointer, would it be possible to cache the sync rate?

Member pam_vbox_do_check
Add handling domains as well.

Member pam_vbox_do_check
implement RTR3Term, or create some hack to force anything containing RTR3Init to stay loaded.

Member VBGLR3DECL
A scatter-gather version of vbglR3GRPerform would be nice, so that we don't have to copy our rectangle and header data into a single structure and perform an additional allocation.

Member main
r=bird: This ALWAYS goes first, the only exception is when you have to parse args to figure out which to call!

Member DEVICE_NAME
r=bird: This must merge with SUPDrv-freebsd.c before long. The two source files should only differ on prefixes and the extra bits wrt to the pci device. I.e. it should be diffable so that fixes to one can easily be applied to the other.

Member vboxguestLinuxInitPci
why aren't we requesting ownership of the I/O ports as well?

Member vboxGuestOS2DetectVersion
dig up the version stuff from GIS later and verify that the numbers are actually decimal.

Member VBoxGuestSolarisAttach
implement resume for guest driver.

Member VBoxGuestSolarisDetach
implement suspend for guest driver.

Member VBoxGuestCloseSession
RTr0proc

Member VBoxGuestCreateKernelSession
RTr0proc

Member VBoxGuestCreateUserSession
RTr0proc

Member DECLVBGL
copy buffers to physical or mapped memory.

Member DECLVBGL
Use some header that we have in common with VBoxGuestLib.h...

Member vbglHGCMHandleAlloc
bitmap for faster search and other obvious optimizations.

Member dprintf
dprintf() -> Log()

Member vbglR3DoIOCtl
explain why this is here.

Member vbglR3DoIOCtl
test status code passing!

Member VBGLR3DECL
Use RTProcDaemonize instead of this.

Member vmmdevSetIRQ_Legacy_EMT
make IRQ pin configurable, at least a symbolic constant

Member vboxServiceWinAddAceToObjectsSecurityDescriptor
Format code style.

Member vboxServiceWinAddAceToObjectsSecurityDescriptor
Add full unicode support.

Member vboxServiceWinAddAceToObjectsSecurityDescriptor
Add event log capabilities / check return values.

Member vboxServiceWinAddAceToObjectsSecurityDescriptor
get rid of that spaghetti jump ...

Member vboxServiceWinSetDesc
On Vista+ SERVICE_DESCRIPTION also supports localized strings!

Member vboxServiceWinStart
r=bird: rc = RTErrConvertFromWin32(GetLastError()); ?

Member vboxServiceWinStart
r=bird: else vboxServiceWinSetStatus(SERVICE_STOPPED, 0); ?

Member g_pszProgName
LOG_GROUP

Member VBoxServiceControlExecProcHandleOutputEvent
Put the last 4 parameters into a struct!

Member VBoxServiceControlExecProcHandleStdInWritableEvent
do we need to do something about this error condition?

Member VBoxServiceControlExecWritePipeBuffer
Use RTMemCache or RTMemObj here?

Member VBoxServiceControlExecWritePipeBuffer
Add offset clamping!

Member VBoxServicePropCacheCreate
Prevent init the cache twice! r=bird: Use a magic.

Member vboxServicePropCacheFindInternal
Docs

Member vboxServicePropCacheFindInternal
This is a O(n) lookup, maybe improve this later to O(1) using a map. r=bird: Use a string space (RTstrSpace*). That is O(log n) in its current implementation (AVL tree). However, this is not important at the moment.

Member vboxServicePropCacheInsertEntryInternal
Docs

Member VBoxServicePropCacheUpdateEntry
Docs

Member VBoxServiceVMInfoWinIsLoggedIn
On Vista and W2K, always "old" user name are still there. Filter out the old one!

Member VBoxServiceVMInfoWinProcessesGetTokenInfo
Implement more token classes here.

Member VBoxServiceVMInfoWinProcessesGetTokenInfo
Add more information of TOKEN_STATISTICS as needed.

Member vboxvideo_PCI_IDS
Take PCI IDs from VBox/param.h; VBOX_VESA_VENDORID, VBOX_VESA_DEVICEID.

Member process_mount_opts
convert string to id.

Member sf_reg_read_aux
bird: yes, kmap() and kmalloc() input only. Since the buffer is contiguous in physical memory (kmalloc or single page), we should use a physical address here to speed things up.

Member sf_reg_write_aux
bird: yes, kmap() and kmalloc() input only. Since the buffer is contiguous in physical memory (kmalloc or single page), we should use a physical address here to speed things up.

Member initDisplay
this should probably be replaced by something IPRT

Member main
Use RTGetOpt() if the arguments become more complex.

Member main
explain why we aren't using RTPathUserHome here so it doesn't get changed accidentally during some cleanup effort.

Member showHelp
Show this as a dialog on windows.

Member VBoxMouseInit
later Linux should also use R3 lib for this

Member vbox_vmmcall
this should work fine on other platforms too, but it needs to be checked for each one.

Member reset_bm_regs
do we need to do that?

Member update_sr
is IRQ deasserted when only one of status bits is cleared?

File DevSB16.cpp
hiccups on NT4 and Win98.

Member pci_data_read
do lookup in R0/RC too!

Member pci_data_write
do lookup in R0/RC too!

Member DebugAssert
fix this.

Member DebugCodeEnabled
ditto

Member DebugPrintEnabled
some PCD for this so we can disable it in release builds.

Member fsw_hfs_dir_lookup
: read additional info

Member VBoxFsCN2_GetControllerName
try query the protocol from the controller and forward the query.

Member VBoxFsCN_GetControllerName
try query the protocol from the controller and forward the query.

Member VBoxVgaMiniPortCN2_GetControllerName
try query the protocol from the controller and forward the query.

Member VBoxVgaMiniPortCN_GetControllerName
try query the protocol from the controller and forward the query.

Member VBoxVgaMiniPortDB_Supported
this doesn't quite work with our DevVGA since it doesn't flag I/O access.

Member kbd_write_mouse
send error if we NACKed the previous command?

File DevE1000.cpp
IPv6 checksum offloading support

File DevE1000.cpp
VLAN checksum offloading support

File DevE1000.cpp
Flexible Filter / Wakeup (optional?)

Member PhyLog
Remove me! For now I want asserts to work in release code.

Member pcnetRdtePoll
NNRD

Member pcnetTdtePoll
consistency with previous receive descriptor

Member DECLINLINE
if (!ASMAtomicXchgBool(&pThis->fXmitSignalled, true)) - needs careful optimizing.

Member VBOX_NAT_DELAY_HACK
: This is a bad hack to prevent freezing the guest during high network activity. Windows host only. This needs to be fixed properly.

File asn1_dec.c
not optimised (yet), favor correctness over speed, favor speed over size

File asn1_enc.c
not optimised (yet), favor correctness over speed, favor speed over size

Member VBOXNET_IPV4ADDR_DEFAULT
r=bird: The inlined code below that drags in asm.h here. I doubt speed is very important here, so move it into a .cpp file, please.

File snmp_structs.h
namespace prefixes

Member LWIPMUTEXTYPE
during my tests on Debian Lenny 64 bit I ran into trouble using mutex semaphores (crash deep down in the pthreads lib). Using the write case of rw semaphores also gives mutual exclusion, and didn't show those crashes. Should be investigated, because this "fix" might be just covering the symptoms of a bug elsewhere.

Member m_inc
better error reporting.

Member activate_port_forwarding
finish this for service case

Member find_guest_ip
see rt_lookup_in_cache

Member tcp_input
(vvl) clarify why it might happens

Member tcp_do_rfc1323
check if it really works, it was turned off before.

Member tstIntNetTestFrame
ICMP?

Member PDMBOTHCBDECL
r=bird: Move this down to where it's used.

Member getBiosBootCode
r=bird: This is a rather silly function since the conversion is 1:1.

Member pic_update_imr
Is this correct?

Member pit_get_count
check this value.

Member pit_get_count
This is WRONG! But I'm too tired to fix it properly and just want to shut up a DIV/0 trap now.

Member pit_get_count
check that we don't return 0, in most modes (all?) the counter shouldn't be zero.

File ATAController.cpp
Actually share this code?

Member vboxscsiWriteRegister
Better return value to indicate ready command?

Member vhdGetVersion
use correct version

Member vscsiReqSenseErrorSet
: Provide more info.

Member vusbRhNewUrb
try find a best fit, MSD which submits rather big URBs intermixed with small control messages ends up with a 2+ of these big URBs when a single one is sufficient.

Member vusbRhNewUrb
The allocations should be done by the device, at least as an option, since the devices frequently wish to associate their own stuff with the in-flight URB or need special buffering (isochronous on Darwin for instance).

Member FindCachedString
binary lookups!

Member ReadCachedConfigDesc
See {2693}

Member vusbDevCancelAllUrbs
Windows doesn't do cancelling, thus this kludge to prevent really bad things from happening if we leave a pending URB behinds.

Member vusbUrbCompletionRh
r=bird: OHCI say "If the Transfer Descriptor is being retired because of an error, the Host Controller must update the Halt bit of the Endpoint Descriptor."

Member DISDECL
Define output callback.

Member DISDECL
Using signed integers as sizes is a bit odd. There are still some GCC warnings about mixing signed and unsigend integers.

Member DISDECL
Need to extend this interface to include a code address so we can dissassemble GC code. Perhaps a new function is better...

Member ParseThreeByteEsc4
Should we take the first or last prefix byte in case of multiple prefix bytes???

Member ParseThreeByteEsc5
Should we take the first or last prefix byte in case of multiple prefix bytes???

Member ParseTwoByteEsc
Should we take the first or last prefix byte in case of multiple prefix bytes???

Member g_apThreeByteMapX86_660F3A
remainder missing (too lazy now)

Member g_apThreeByteMapX86_F20F38
remainder missing (too lazy now)

Member g_aThreeByteMapX86_660F38_8
remainder missing (too lazy now)

Member DECLCALLBACK
This should check that the disassembler doesn't do unnecessary reads, however the current doesn't do this and is just complicated...

Member MyDisasYasmFormatterIsOddEncoding
correct?

Member MyDisasYasmFormatterIsOddEncoding
branch hinting 0x2e/0x3e...

Member LONG
64-bit: ULONG is 32-bit.

Member USHORT
64-bit: ULONG is 32-bit.

Member DECLCALLBACK
consider use of RTReq

Member hgcmThreadInit
error processing.

Member DECLCALLBACK
consider use of RTReq

Member hgcmThreadInit
error processing.

Member FNHGCMNEWMSGALLOC
comment

Member FNHGCMNEWMSGALLOC
comment

Member Port
implement

Member Singlestep

Member Singlestep

Member CSAMEnabled
handle error case

Member Singlestep

Member Singlestep

Member VirtualTimeRate
handle error code.

Member MACAddress
the runtime lacks an ascii lower/upper conv

Member szError
currently this is only set but never read.

Member for
filename

Member while
that somehow doesn't seem to work!

Member RTEnvUnset
this should really be taken care of in Main.

Member NewUniqueKey
this is no longer necessary, we can enumerate extra data

Member handleAddiSCSIDisk
move the location stuff to Main, which can use pfnComposeName from the disk backends to construct the location properly. Also do not use slashes to separate the parts, as otherwise only the last element comtaining information will be shown.

Member handleControlVM
RTGETOPT_FLAG_MANDATORY

Member handleConvertFromRaw
cleaner solution than assuming 1:1 mapping?

Member handleModifyVM
generalize this, allow arbitrary number of DVD drives and as a consequence multiple attachments and different storage controllers.

Member handleOpenMedium
remove this hack! First try opening the image as is (using the regular API semantics for images with relative path or without path), and if that fails with a file related error then try it again with what the client thinks the relative path would mean. Requires doing the command twice in certain cases. This is an ugly hack and needs to be removed whevever we have a chance to clean up the API semantics.

Member handleRegisterVM
Ugly hack to get both the API interpretation of relative paths and the client's interpretation of relative paths. Remove after the API has been redesigned.

Member handleShowHardDiskInfo
NEWMEDIA check accessibility of all parents

Member showVMInfo
the quoting is not yet implemented!

Member g_fExecCanceled

Member handleExecProgram
Add fancy piping stuff here.

Member handleExecProgram
Check for uint64_t vs. uint32_t here!

Member HandleHostKey
process gui events.

Member SaveState
process gui events.

Member SendMouseEvent
PutMouseEventAbsolute() expects x and y starting from 1,1. should we do the increment internally in PutMouseEventAbsolute() or state it in PutMouseEventAbsolute() docs?

Member SendMouseEvent
PutMouseEventAbsolute() expects x and y starting from 1,1. should we do the increment internally in PutMouseEventAbsolute() or state it in PutMouseEventAbsolute() docs?

Member SDL_USER_EVENT_UPDATERECT
check why we need to do this on windows.

Member ESC_NORM
someone please explain why we don't follow the book!

File DarwinKeyboard.cpp
Move this up somewhere so that the two SDL GUIs can use parts of this code too (-HID crap).

Member DarwinKeyCodeToDarwinModifierMask
find symbols for these keycodes...

Member DarwinModifierMaskToDarwinKeycode
find symbols for these keycodes...

File DarwinKeyboard.h
Move this up somewhere so that the two SDL GUIs can use this code too.

File tstDarwinKeyboard.cpp
Move this up somewhere so that the two SDL GUIs can use parts of this code too (-HID crap).

Member darwinToCGImageRef
this code assumes 32-bit image input, the lazy bird convert image to 32-bit method is anything but optimal...

Member Q_LONG
bird: Use (U)INT_PTR, (U)LONG_PTR, DWORD_PTR, or (u)intptr_t.

Member crVBoxHGCMCall
use vbglR3DoIOCtl here instead

Member vboxClipboardDisconnect
handle this slightly more reasonably, or be really sure it won't go wrong.

Member VBoxDrvDarwinStop
I've got a nagging feeling that we'll have to keep track of users and refuse unloading if we're busy. Investigate and implement this!

Member SUPR0DECL
switch this bugger to RTHCPHYS

Member SUPR0DECL
document me

Member VBoxDrvFreeBSDOpen
get (r)uid and (r)gid. pSession->Uid = stuff; pSession->Gid = stuff;

Member VBOX_WITH_SUSPEND_NOTIFICATION
figure out the exact version number

Member supdrvOSLdrOpen
This is something that shouldn't be impossible to implement here and would make a few people happy.

Member DECLCALLBACK
needs wrapping?

Member supdrvCleanupSession
figure out how to handle this.

Member supdrvGipCreate
Disabled this as we didn't used to do it before and causes unnecessary stress on laptops. It only applies to Windows and should probably revisited later, if possible made part of the timer code (return min granularity in RTTimerGetSystemGranularity and set it in RTTimerStart).

Member supdrvGipDoUpdateCpu
validate the NanoTS delta, don't trust the OS to call us when it should...

Member supdrvGipUpdate
fix 64-bit div here to work on x86 linux.

Member supdrvInitDevExt
make this random?

Member supdrvIOCtl
Somehow validate the client and negotiate a secure cookie...

Member supdrvIOCtl_LdrFree
BRANCH-2.1: remove this after branching.

Member supdrvIOCtl_LdrOpen
check cbImageBits and cbImageWithTabs here, if they differs that indicates that the images are different.

Member supdrvMemRelease
figure out how to handle this.

Member SUPDRV_IOC_VERSION
Pending work on next major version change:
  • Nothing.

Member DECLCALLBACK
trim this down.

Member DECLEXPORT
This will probably deserve it's own session or some other good solution...

Member suplibOsInit
someone must test what is actually returned.

Member g_aSupInstallFiles
This needs reviewing against the linux packages.

Member g_aSupInstallFiles
The excessive use of kSupID_SharedLib needs to be reviewed at some point. For the time being we're building the linux packages with SharedLib pointing to AppPrivArch (lazy bird).

Member supR3HardenedMakePath
fix this AppBin crap (uncertain wtf some binaries actually are installed).

Member supSvcDisplayErrorV
format it!

Member supSvcLogError
This should later be replaced by the release logger and callback destination(s).

Member supSvcLogErrorV
This should later be replaced by the release logger and callback destination(s).

Member supSvcOsLogErrorStr
mc

Member mymemcmp
use the nocrt stuff?

Member VBoxDrvNtInternalDeviceControl
IDC on NT: figure when to create the session and that stuff...

Member supSvcWinCreate
Set the service description or it'll look weird in the vista service manager. Anything else that should be configured? Start access or something?

Member supSvcWinRunIt
release logging of the system-wide service.

Member VBoxNetAdpLinuxIOCtl
remove the adapter again?

Member VBoxNetAdpLinuxUnload
We have to prevent this if we're busy!

Member vboxNetFltFreeBsdIsPromiscuous
This isn't taking into account that we put the interface in promiscuous mode.

Member vboxNetFltOsInitInstance
keep these reports up to date, either by polling for changes or intercept some control flow if possible.

Member VBOXNETFLT_WITH_GSO_RECV
This is currently disable because it causes performance loss of 5-10%.

Member vboxNetFltLinuxCanForwardAsGso
Figure out how to handle SKB_GSO_TCP_ECN!

Member vboxNetFltLinuxCanForwardAsGso
IPv6: 4to6 tunneling

Member vboxNetFltLinuxCanForwardAsGso
IPv6: Dig our way out of the other headers.

Member vboxNetFltLinuxSkBufFromSG
We should use fragments mapping the SG buffers with large packets. 256 bytes seems to be the a threshold used a lot for this. It requires some nasty work on the intnet side though...

Member vboxNetFltLinuxSkBufFromSG
VLAN: +4 if VLAN?

Member VBoxNetFltLinuxUnload
We have to prevent this if we're busy!

Member vboxNetFltOsConnectIt
duplicate work here now? Attach

Member vboxNetFltOsConnectIt
GSO: Test UDP offloading (UFO) on linux.

Member vboxNetFltOsDeleteInstance
This code may race vboxNetFltLinuxUnregisterDevice (very very unlikely, but none the less). Since it doesn't actually update the state (just reads it), it is likely to panic in some interesting ways.

Member vboxNetFltSolarisAttachToInterface
There is no easy way of obtaining the global host side promiscuous counter. Currently we just return false.

Member vboxNetFltSolarisQueueLoopback
consider if this is worth switching to a double linked list...

Member vboxNetFltSolarisReportInfo
Promisc

Member vboxNetFltMaybeRediscovered
this isn't 100% serialized.

Member DECLHIDDEN
r=bird: This is too slow and is probably returning the wrong information. What we're interested in is whether someone besides us has put the interface into promiscuous mode.

Member InstallNetFlt
add size check for (sizeof(PtInf)/sizeof(PtInf[0])) == dwSize (string length in sizeof(PtInf[0]))

Member InstallNetFlt
add size check for (sizeof(MpInf)/sizeof(MpInf[0])) == dwSize (string length in sizeof(MpInf[0]))

Member DECLHIDDEN
Hrmpf, how can we fix this assumption? I fear this'll cause data corruption and maybe even BSODs ...

Member DECLHIDDEN
Need to report changes to the switch via: pThis->pSwitchPort->pfnReportPromiscuousMode(pThis->pSwitchPort, fPromisc);

Member vboxNetFltWinReportStuff
Keep these up to date, esp. the promiscuous mode bit.

Member vboxNetFltWinReportStuff
We should be able to do pfnXmit at DISPATCH_LEVEL...

Member guestControl::gctrlAddToExecEnvv (const char *pszEnv, void **ppvList, uint32_t *pcbList, uint32_t *pcEnv)

Member guestControl::gctrlAddToExecEnvv (const char *pszEnv, void **ppvList, uint32_t *pcbList, uint32_t *pcEnv)
write docs!

Member guestControl::gctrlPrepareExecArgv (char *pszArgs, void **ppvList, uint32_t *pcbList, uint32_t *pcArgs)

Member guestControl::gctrlPrepareExecArgv (char *pszArgs, void **ppvList, uint32_t *pcbList, uint32_t *pcArgs)
Respect spaces when quoting for arguments, e.g. "c:\\program files\\".

Member guestControl::gctrlPrepareExecArgv (char *pszArgs, void **ppvList, uint32_t *pcbList, uint32_t *pcArgs)
Handle empty ("") argguments.

Member guestControl::gctrlPrepareExecArgv (char *pszArgs, void **ppvList, uint32_t *pcbList, uint32_t *pcArgs)
write docs!

Member guestExecProcess
consider supporting: gcc stuff.c >file 2>&1.

Member guestExecProcHandleOutputEvent
Put the last 4 parameters into a struct!

Member guestExecProcHandleStdInWritableEvent
do we need to do something about this error condition?

Member hostExecCmd
Put some assert macros here!

Member hostExecGetStatus
Put some assert macros here!

Member testCase
should depend on the guest OS

Member vbsfCorrectCasing
this is quite inefficient, especially for directories with many files

Member vbsfBuildFullPath
don't check when creating files or directories; waste of time

Member vbsfConvertFileOpenFlags
treat this as read access, but theoretically this could be a no access request.

Member vbsfCorrectCasing
this is quite inefficient, especially for directories with many files

Member vbsfCreate

Member vbsfCreate
I left this in in order not to change the behaviour of the function too much. Is it really needed, and should it really be here?

Member vbsfCreate
This should be in the Windows Guest Additions, as no-one else needs it.

Member vbsfOpenDir
Can anyone think of a sensible, race-less way to do this? Although I suspect that the race is inherent, due to the API available...

Member vbsfOpenDir
render supplied attributes. bird: The guest should specify this. For windows guests RTFS_DOS_DIRECTORY should suffice.

Member vbsfOpenFile
Possible race left here.

Member vbsfStripLastComponent
find a better solution for supporting the execute bit for non-windows guests on windows host. Search for "0111" to find all the relevant places.

Member vbsfWrite
correct?

Member renderspu_SystemGetMaxWindowSize
port to 64-bit darwin.

Member vboxfuseNodeAlloc
make this safe!

Member vboxfuseNodeFillStat
file times

Member vboxfuseOp_opendir
access checks.

Member vboxfuseOp_opendir
update the accessed TS?

Member getSmcDeviceKey
implement + test RTSYSDMISTR_MANUFACTURER on all hosts. Currently falling back on the product name.

Member DarwinGetEthernetControllers
IOPacketFilters / IONetworkFilterGroup?

Member getDriveInfoFromEnv
r=bird: missing a call to devValidateDevice() here and szUdi wasn't initialized because of that. Need proper fixing.

Member com::processPendingEvents (nsIEventQueue *pQueue)
: rethink interruption events, current NULL event approach is bad

Member OperatingSystem
error reporting?

Member OSVersion
error reporting?

Member Accelerate2DVideoEnabled
check validity!

Member Accelerate3DEnabled
check validity!

Member MemoryBalloonSize
this method should not be public

Member Capabilities
add COMGETTER(ExtendedCapabilities) when we reach the 32 bit limit (or make the argument ULONG64 after checking that COM is capable of defining enums (used to represent bit flags) that contain 64-bit values)

Member Description
update m->description and save the global registry (and local registries of portable VMs referring to this medium), this will also require to add the mRegistered flag to data

Member Location
NEWMEDIA for file names, add the default extension if no extension is present (using the information from the VD backend which also implies that one more parameter should be passed to setLocation() requesting that functionality since it is only allwed when called from this method

Member Location
NEWMEDIA rename the file and set m->location on success, then save the global registry (and local registries of portable VMs referring to this medium), this will also require to add the mRegistered flag to data

Member Type
implement this case later

Member MetricName
(r=dmik) why do all these getters not do AutoCaller and AutoReadLock? Is the underlying metric a constant object?

Member Console
VBox 3.3: replace E_FAIL with rc here.

Member Machine
VBox 3.3: replace E_FAIL with rc here.

Member PortCount
allow 2 floppies later

Member USBStandard
This is no longer correct

Member main
There is almost no need to make a special case of OS/2 here. Just the execv call needs to be told to create a background process...

Member HOSTSUFF_EXE
move this to RT headers (and use them in MachineImpl.cpp as well)

Member VBoxNetArpHandleIt
RTNET_ARPOP_INVREQUEST

Member VBoxNetUDPMatch
IPv6 UDP support, goes for this entire function really. Not really important yet since this is currently only used by the DHCP server.

Member vusbProxyLinuxUrbDoTimeouts
Make the HCI do proper timeout handling! Current timeout is 3 min and 20 seconds as not to break bloomberg which queues IN packages with 3 min timeouts.

Member rtCrc32ProcessTooBig
Check if we can't just use the zlib code directly here.

Member ValidateSectionHeader
find specs and check up which sh_info fields indicates section table entries

Member rtldrPEValidateDirectories
Rainy Day: Implement further verfication using openssl.

Member rtldrPEValidateSectionHeaders
only if SizeOfRawData > 0 ?

Member rtldrPEValidateSectionHeaders
pSH->PointerToRawData >= cbRawImage ?

Member rtldrPEValidateSectionHeaders
r=bird: more sanity checks!

Member DECLCALLBACK
consider fixing the config instead.

Member rtgetoptConvertIPv4Addr
This should be move to some generic part of the runtime.

Member rtgetoptConvertMacAddr
This should be move to some generic part of the runtime.

Member rtGetOptProcessValue
CIDR

Member rtLockValidatorLazyInit
register some cleanup callback if we care.

Member rtLockValidatorRecSharedMakeRoom
enable this after making sure growing works flawlessly.

Member rtTarCopyFileFrom
r=bird: Use a bigger buffer here, see comment in rtTarCopyFileTo.

Member rtTarCopyFileTo
r=bird: using a 64KB buffer here instead of 0.5KB would probably be a good thing.

Member rtStrFormatNumber
implement RTSTR_F_ZEROPAD + RTSTR_F_THOUSAND_SEP.

Member CalcFlags
read the specs on this other vs standard stuff, and check out the finer points

Member GenerateExcludedData
generate the decomposition: http://unicode.org/reports/tr15/#Hangul

Member GenerateExcludedData
CJK Ideographs Extension A (U+3400 - U+4DB5) CJK Ideographs (U+4E00 - U+9FA5) CJK Ideograph Extension B (U+20000 - U+2A6D6) CJK Ideograph Extension C (U+2A700 - U+2B734)

Member rtStrGetCpExFailure
code this!

Member rtStrGetCpNExFailure
code this!

Member DECLINLINE
Perhaps some of these KAVL_SET_POINTER_NULL() cases could be optimized away..

Member RT_ASSERT_PREEMPT_CPUID_VAR
enable on windows after testing

Member rtR0MemAlloc
drop the r0drv/alloc-r0drv.cpp stuff on OS/2.

Member DECLINLINE
put this where it belongs! (i.e. in a separate file and prototype in iprt/err.h)

Member RTR0DECL
implement RTR0AssertPanicSystem.

Member rtR0MemObjNativeLockInMap
fix mach -> vbox error conversion for freebsd.

Member rtR0MemObjNativeReserveInMap
fix translation (borrow from darwin)

Member rtSemEventWait
this isn't necessarily correct, we've could've woken up just before destruction...

Member rtSemEventMultiWait
this isn't necessarily correct, we've could've woken up just before destruction...

Member rtSemSpinMutexEnter
Later #1: When entering in interrupt context and we're not able to wake up threads from it, we could try switch the lock into pure spinlock mode. This would require that there are no other threads currently waiting on it and that the RTSEMSPINMUTEX_FLAGS_IRQ_SAFE flag is set.

Member RTR0DECL
implement thread2-r0drv-linux.c

Member rtR0Term
implement thread2-r0drv-linux.c

Member RTR0DECL
find the exact kernel where change_page_attr was introduced.

Member RTR0DECL
find the exact kernel where change_page_attr was introduced.

Member rtR0MemObjLinuxDoMmap
implement uAlignment properly... We'll probably need to make some dummy mappings to fill up alignment gaps. This is of course complicated by fragmentation (which we might have cause ourselves) and further by there begin two mmap strategies (top / bottom).

Member rtR0ProcessToLinuxTask
fix rtR0ProcessToLinuxTask!!

Member RTR0DECL
implement RTR0AssertPanicSystem.

Member rtR0InitNative
array w/ data + script for extracting a row. (save space + readability; table will be short.)

Member rtR0InitNative
this handler doesn't seem to work... Because of Irql?

Member rtR0InitNative
fix above for all Nt versions.

Member rtR0MemObjNtMap
uAlignment

Member rtR0MemObjNtMap
How to set the protection on the pages?

Member rtR0MemObjNtMap
add cache type to fProt.

Member rtMpCall
Consider changing this to an active wait using some atomic inc/dec stuff (and check for the current cpu above in the specific case).

Member rtR0MemObjNativeMapUser
implement this...

Member rtR0PowerNotificationInit
OS specific init here

Member rtR0PowerNotificationTerm
OS specific term here

Member rtMpNotificationSolarisCallback
are these relevant?

Member LOG_GROUP
r=bird: What's this about "must be opened with O_DIRECT"? An explanation would be nice, esp. seeing what Linus is quoted saying about it in the open man page...

Member rtLinuxConstructPathV
r=bird: This whole API should be rewritten to use IPRT status codes. using errno was a mistake and only results in horrible code.

Member rtLinuxConstructPathV
While RTStrPrintfV prevents overflows, it doesn't make it easy to check for truncations. RTPath should provide some formatters and joiners which can take over this rather common task that is performed here.

Member AssertCompileSize
move this to r3/posix/something.h.

Member RTLOCALIPC_WIN_SDDL
Double check this!

Member RTLOCALIPC_WIN_SDDL
Drop the EA rights too? Since they doesn't mean anything to PIPS according to the docs.

Member RTLOCALIPC_WIN_SDDL
EVERYONE -> AUTHENTICATED USERS or something more appropriate?

Member RTLOCALIPC_WIN_SDDL
Have trouble allowing the owner FILE_CREATE_PIPE_INSTANCE access, so for now I'm hacking it just to get progress - the service runs as local system. The CREATOR OWNER and PERSONAL SELF works (the former is only involved in inheriting it seems, which is why it won't work. The latter I've no idea about. Perhaps the solution is to go the annoying route of OpenProcessToken, QueryTokenInformation, ConvertSidToStringSid and then use the result... Suggestions are very welcome

Member rtLocalIpcServerWinCreatePipeInstance
dynamically resolve this as it is the only thing that prevents loading IPRT on NT4.

File poll-win.cpp
merge poll-win.cpp and poll-posix.cpp, there is lots of common code.

Member rtFileNativeSetAttributes
rename to rtFileWinSetAttributes

Member rtFileNativeSetAttributes
resolve dynamically to avoid dragging in NtDll?

Member rtSystemWinQueryOSVersion
fixme

Member main
sticy bits++

Member main
RTRand smoke testing.

Member tstBase64
negative testing.

Member TSTFILEAIO_MAX_REQS_IN_FLIGHT
make configurable through cmd line.

Member main
Add some detail info output with a signature identical to RTPrintf.

Member main
Add some detail info output with a signature identical to RTPrintf.

Member main
This is a bit weird/wrong, but not half as weird as the way Windows+OS/2 deals with a trailing dots.

Member main
Cover files, ++

Member tstRTCreateProcEx5Child
Lookup UID/effective UID, maybe GID?

Member main
make a bit string/dumper similar to Rhxs/d.

Member main
add a testcase where some stuff times out.

Member main
{ "-1", 0, VWRN_NEGATIVE_UNSIGNED, ~0 }, - no longer true. bad idea?

Member main
This isn't really x86 or AMD64 specific...

Member main
wrapping detection.

Member main
This isn't really x86 or AMD64 specific...

Member Benchmarks
add RTTest* methods for reporting benchmark results.

Member Benchmarks
figure this stuff into the test framework.

Member GetRandUtf16
use our random.

Member test2
RTCpsToUtf8 or something.

Member testLatin1
Either fix the documentation or fix the code - cchActual is set to the number of bytes actually encoded.

Member TstRTStrXCmp
test the non-ascii bits.

Member RTR3DECL
better translation

Member RTR3DECL
fix ERROR_RESOURCE_DATA_NOT_FOUND translation

Member RTR3DECL
fix ERROR_INVALID_ADDRESS translation - dbghelp returns it on some line number queries.

Member cpumR3DetectVendor
detect the other buggers...

Member dbgfR3DisasInstrFirst
Deal more explicitly with RC in DBGFR3Disas*.

Member PDBGFMOD
Only use DBGHELP for reading modules since it doesn't do all we want (relocations), or is way to slow in some cases (add symbol)!

Member PDBGFMOD
use iprt/stream.h!

Member dbgfR3ModuleLocateAndOpen
RTPathCompose, RTPathComposeN(). This code isn't right for 'E:' on DOS systems. It may also create unwanted double slashes.

Member dbgfR3ModuleLocateAndOpen
do a depth search using the specified path.

Member mmR3HeapDestroy
free all nodes in a AVL tree.

Member mmR3HyperInit
{1865}, {3202}: Change the cbHyperHeap default depending on whether VT-x/AMD-V is enabled or not! Don't waste precious kernel space on heap for the PATM.

Member mmPagePoolPhys2Ptr
have to fix the debugger, but until then this is going on my nerves.

Member mmR3PagePoolInit
{1865},{3202}: mapping the page pool page into ring-0. Need to change the wasy we allocate it...

Member fInCSAMCodePageInvalidate
Temporary for debugging.

Member patmReadBytes
could change in the future!

Member PATMInstallGuestSpecificPatch
might have to check if the patch crosses a page boundary. Currently not necessary, but that might change in the future!!

Member patmPatchGenCall
wasting memory as the complex search is overkill and we need only one lookup slot...

Member patmPatchGenIntEntry
must check if virtual IF is already cleared on entry!!!!!!!!!!!!!!!!!!!!!!!

Member patmPatchGenMovControl
: make this an array in the context structure

Member patmPatchGenMovDebug
: if we ever want to support hardware debug registers natively, then this will need to be changed!

Member patmPatchGenSldtStr
segment prefix (untested)

Member PPDMASYNCCOMPLETIONENDPOINTFILE
: Revise the caching of tasks. We have currently four caches: Per endpoint task cache Per class cache Per endpoint task segment cache Per class task segment cache

Member pdmR3CritSectInitOne
plug the "leak"...

Member pdmR3QueueCreate
we should add a priority to the queues so we don't have to rely on the initialization order to deal with problems like #1605 (pgm/pcnet deadlock caused by the critsect queue to be last in the chain).
  • Update, the critical sections are no longer using queues, so this isn't a real problem any longer. The priority might be a nice feature for later though.

Member DECLCALLBACK
define LOG_GROUP LOG_GROUP_PDM_THREAD

Member PGM_SAVED_STATE_VERSION
remove the guest mappings from the saved state at next version change!

Member pgmR3GetRomPage
change this to take a hint.

Member TRPMHANDLER
need to specialize this as we go along.

Member selmGCSyncGDTEntry
Fix this by loading ds and cs before turning off WP.

Member selmGCSyncGDTEntry
what about interrupt gates and rawr0?

Member selmGCSyncGDTEntry
this isn't correct in two(+) ways! 1. It shouldn't be done until the LDTR is reloaded. 2. It caused the next instruction to be emulated.

Member DTRACE_EXPERIMENT
fix/remove/permanent-enable this when DIS/PATM handles invalid lock sequences.

Member vmmR0InitVM
rename to CPUMR0InitVM

Member vmmR3SwitcherGenericRelocate
case FIX_CR4_MASK:

Member vmmR3SwitcherGenericRelocate
case FIX_CR4_OSFSXR:

Member FIX_HC_2_GC_NEAR_REL
document what arguments these take and what they do.

Namespace com
this is for IVirtualBoxErrorInfo, see the

Namespace com
below.

Namespace guestControl
Remove unused header files below!

Member RTDECL
darwin R0 MP

Member RTDECL
(not used on non-Windows platforms yet).

Member RTDECL
Darwin: Implement RTThreadIsInInterrupt. Required for guest additions!

Member RTDECL
is there a native freebsd function/macro for this?

Member RTDECL
FreeBSD: Implement RTThreadIsInInterrupt. Required for guest additions!

Member RTDECL
(not used on non-Windows platforms yet).

Member RTDECL
online/present cpu stuff must be corrected for proper W2K8 support (KeQueryMaximumProcessorCount).

Member RTDECL
not implemented

Member RTDECL
Is there a more correct way?

Member RTDECL
Check if DHGETDOSV_YIELDFLAG includes TCYIELDFLAG.

Member RTDECL
OS/2: verify the usage of DHGETDOSV_INTERRUPTLEV.

Member RTDECL
check this up with hotplugging!

Member DECLCALLBACK
r=bird: We should just drop all these items, restoring doesn't make sense.

Member DECLCALLBACK
Add IGNNE support.

Member DECLCALLBACK
figure out what the difference between FPU ports 0xf0 and 0xf1 are...

Member DECLCALLBACK
Check if the A20 enable/disable method implemented here in any way should cooperate with the one implemented in the PS/2 keyboard device. This probably belongs together in the PS/2 keyboard device (since that is where the "port B" mentioned by Ralph Brown is implemented).

Member DECLCALLBACK
Check this cast!

Member DECLCALLBACK
make interruptible?

Member DECLCALLBACK
this dumps everything attached to this device instance, which is more than necessary. Dumping the changed LUN would be enough.

Member DECLCALLBACK
(r=dmik) if we failed with drive mount, then the VMR3Resume

Member DECLCALLBACK
register another VMSetError callback?

Member DECLCALLBACK
this dumps everything attached to this device instance, which is more than necessary. Dumping the changed LUN would be enough.

Member DECLCALLBACK
this could probably be made more generic and reused elsewhere.

Member DECLCALLBACK
Check this cast!

Member DECLCALLBACK
Have to think more about this ....

Member RTDECL
Implement timeouts in RTSemMutexRequest.

Member RTDECL
Implement timeouts and interrupt checks in RTSemMutexRequestNoResume.

Member RTDECL
enable this when the logger has been fixed.

Member RTDECL
darwin R0 MP (rainy day)

Member RT_LOCK_NO_STRICT
remove this:

Group grp_rt_dbgmod
Missing some bits here.

Member RTDECL
mscount will roll over after ~48 days.

Member RTDECL
darwin R0 MP

Member RTDECL
use KeQueryMaximumProcessorCount on vista+

Member RTDECL
online/present cpu stuff must be corrected for proper W2K8 support (KeQueryMaximumProcessorCount).

Member DECLCALLBACK
: If we got the callback then VM if definitely running. But a better method should be implemented.

Member DECLCALLBACK
machineState is set to VMSTATE_RUNNING before all devices have received power on events this can prematurely allow the main thread to enter the event loop

Member DECLCALLBACK
reset the transaction number and whatnot when iTick == 1.

Member DECLCALLBACK
fix vboxNetFltPortOsDisconnectInterface.

Member DECLCALLBACK
why is Additions/WINNT/Graphics/Miniport/VBoxVideo.cpp pushing this to us?

Member RTDECL
figure out why

Member RTDECL
huh? we're using posix_memalign in the next function...

Group grp_rt_mem
remove when spitting.

Member RTDECL
darwin R0 MP

Member PRTNETDHCP
AssertCompileSize(RTNETDHCP, );

Group grp_rt_param
Much of the PAGE_* stuff here is obsolete and highly risky to have around. As for component configs (MM_*), either we gather all in here or we move those bits away!

Member RTDECL
Handled bad encodings correctly+quietly, remove assumption, optimize.

Member RTDECL
Handled bad encodings correctly+quietly, remove assumption, optimize.

Member RTDECL
move this to assembly where a simple "jmp printf" will to the trick.

Member DECLINLINE
Maybe we could figure out what to poll for depending on the kind of pipe we're dealing with.

Member R3PTRTYPE
the following three definitions may move the iprt/types.h later.

Member RTDECL
import alien threads?

Member RTDECL
convert from local to UCT.

Member NIL_RTGID
check this for portability!

Member NIL_RTUID
check this for portability!

Member PRTGCINT
this isn't right, natural int is 32-bit, see RTHCINT.

Member PRTGCUINT
this isn't right, natural int is 32-bit, see RTHCUINT.

Member DECLEXPORT

Member DBGFRETURNTYPE_IRET64
64-bit iret return.

Member GMMR0DECL
IPRT function for query ram size and such.

Member GVMMR0DECL
should be adjusted to relative to the cpu count or something...

Member IOM_NO_PDMINS_CHECKS
Check this again, now.

Member IOM_SUCCESS
Will have to add VINF_EM_DBG_HYPER_BREAKPOINT if the I/O port and MMIO breakpoints should trigger before the I/O is done. Currently, we don't implement these kind of breakpoints.

Member MM_HYPER_AREA_MAX_SIZE
Readjust when floating RAMRANGEs have been implemented. Used to be 20 * _1MB

Member VBOX_DEVICEID
ditto.

Member VBOX_VENDORID
wonderful choice of name! Please squeeze a _VGA_ or something in there, please.

Member PCIADDRESSSPACE
we should introduce 32 & 64 bits physical address types

Member VMMR3DECL
selector relocations should be a seperate operation?

Member VMMR3DECL
How to test for the 40 bits support? Long mode seems to be the test criterium.

Member PDMDRVREG::pfnSoftReset

Member PPDMIAUDIOSNIFFERPORT
r=bird: the two following interfaces are hacks to work around the missing audio driver interface. This should be addressed rather than making more temporary hacks.

Member PVBOXVHWACMD
r=bird: _VBOXVHWACMD -> VBOXVHWACMD; avoid using 1 or 2 leading underscores. Also, a line what it is to make doxygen happy.

Group grp_types_idt
This all belongs in x86.h!

Group grp_USBLib
merge the usblib-win.h interface into the darwin and linux ports where suitable.

Member VBOXGUEST_IOCTL_CODE
r=bird: Please do it like SUPDRVIOC to keep it as similar as possible.

Member VBOXGUEST_DEVICE_NAME
it would be nice if we could have two define without paths.

Member DECLCALLBACK
missing prefix.

Member PGCFGMVALUEREQ
Move got GCFGM.h when it's implemented.

Member VMMR0DECL
check this with with VC7!

Member VMMDevReqHypervisorInfo::hypervisorStart
Make this 64-bit compatible?

Member VMMDevVideoSetVisibleRegion::Rect
array is spelled aRects[1].

Member VMMDevSeamlessMode
DARN! DARN! DARN! Who forgot to do the 32-bit hack here??? FIXME! XXX!

Member KHLP_DECL
convert errno -> kErrors

Member KHLP_DECL
convert errno -> kErrors

Member kLdrModMostlyDone
define exactly what it possible to do after this call.

Member kLdrModOpenNativeByHandle
DosQueryHeaderInfo should be able to get us what we want on OS/2.

Member kLdrDyldLoadExe
fix this mess...

Member KLDRDYLD_SEARCH_KLDR
invent me.

Group grp_macho_o_syms
symseg_command (LC_SYMSEG)

Group grp_macho_o_syms
ident_command (LC_IDENT)

Group grp_macho_o_syms
fvmfile_command (LC_FVMFILE)

Member kldrDyldFailure
print it.

Member kldrDyldModCreate
deal with fFlags (exec/dll)

Member kldrDyldModFixup
fixme.

Member kldrDyldModMap
TLS

Member kldrInit
atomic test and set if we care.

Member EfiInfoIndex
Put this in DEVEFIINFO, that's much easier to access.

Member VBOX_EFI_TOP_OF_STACK
Move this to 1MB + 128KB and drop the stack relocation the firmware does. It expects the stack to be within the temporary memory that SEC hands to PEI and the VBoxAutoScan PEIM reports.

Member rtR0MemObjNativeAllocPage
fix translation (borrow from darwin)

Member rtR0MemObjNativeAllocPhys
alignment

Member rtR0MemObjNativeAllocPhysNC
rtR0MemObjNativeAllocPhys / darwin. This might be a bit problematic and may very well require having to create our own object which we populate with pages but without mapping it into any address space. Estimate is 2-3 days.

Member rtR0MemObjNativeLockUser
The Linux fork() protection will require more work if this API is to be used for anything but locking VM pages.

Member rtR0MemObjNativeMapKernel
First, the memory should've been mapped by now, and second, it shouild have the wired attribute in the PTE (bit 9). Neither is seems to be the case. The disabled locking code doesn't make any difference, which is extremely odd, and breaks rtR0MemObjNativeGetPagePhysAddr (getPhysicalSegment64 -> 64 for the lock descriptor.

Member dbgfR3BpInit

Member iomGetRegImmData
Get rid of this code. Use DISQueryParamVal instead

Member PGM_PTFLAGS_CSAM_VALIDATED
Move PGM_PTFLAGS_* and PGM_PDFLAGS_* to VBox/pgm.h.

Member PGM_SYNC_MONITOR_CR3
This is obsolete now. Remove after 2.2.0 is branched off.

Member PGMPAGETYPE
So, convert to #defines!

Member DECLINLINE
Split out all the inline stuff into a separate file. Then we can include it later when VM and VMCPU are defined and so avoid all that &pVM->pgm.s and &pVCpu->pgm.s stuff. It also chops ~1600 lines off this file and will make it somewhat easier to navigate...

Member PGM_GCPHYS_2_PTR_BY_PGMCPU
this isn't asserting, use PGMRamGCPhys2HCPtr!

Member PGM_INVL_PG
this isn't asserting, use PGMRamGCPhys2HCPtr!

Member RAWEx_ProfileStart
r=bird: clean up the RAWEx stats.

Page CMOS Assignments (BIOS)
Mark which bits are compatible with which BIOSes and which are our own definitions.

Page VUSB - Virtual USB
read thru this and correct typos. Merge with old docs.

Page VBoxNetFlt - Network Interface Filter
Need to figure which host allow unloading and block/fix it.

Page Guest Control HGCM Service
Write up some nice text here.

Page Runtime - Logging
Update this section...

Page PDM - The Pluggable Device & Driver Manager
more details.

Page PDM - The Pluggable Device & Driver Manager
implement and describe

Page Raw-mode Code Execution
Write raw-mode chapter!


Generated by  Doxygen 1.6.0   Back to index