TCP Stop Working

TheMSsForum.com: The Microsoft Software Forum

The MSS Forum ‹ Drivers
- Archive
  - Biz
  - MCSE
  - CRM
  - Drivers
  - Framework
  - ADO
  - ASP
  - Compact
  - Forms
  - Dotnet
  - C#
  - VB
  - FontpageGen
  - Excel
  - WorkSheet
  - Exchange
  - Setup
  - Fox
  - Fontpage
  - ASP
  - IIS
  - Entourage
  - Money
  - Messanger
  - PocketPC
  - Powerpoint
  - Project
  - Publisher
  - Excel
  - VB
  - Security
  - Portal
  - Services
  - SQLServerDev
  - SVCS
  - SQLServer
  - VB
  - VC
  - MFC
  - ExcelGen

- Previous
  - 1
    - Driver to work around WriteFile access denied issue in Vista I wrote a program, several years old, which uses WriteFile to write to sector 0 on compact flash disks. As has been noted in other threads this fails under Vista with an access denied message. (I haven't seen this change noted in MSDN documentation.) The work around given in the threads is to write a driver to do the write. Since this is an apparent common problem and I have never before written a driver myself, I was hoping there might be some example source code or hopefully an installable driver already avaialble. Does anybody know of any? Thanks in advance. Tag: TCP Stop Working Tag: 102739
  - 2
    - PtSendComplete Adapter=NULL My intermediate driver is crashing on random occations at PtSendComplete() it seem like the Adapter become NULL when the driver call NdisMSendComplete( TransportHandle, OriginalPacket, Status ); . However the Adapter is always valid until NdisMSendComplete get call. below is my seudocode: VOID PtSendComplete(IN NDIS_HANDLE ProtocolBindingContext, IN PNDIS_PACKET Packet, IN NDIS_STATUS Status ) { -passed lower MP adapter context -get fully iniialized context -complete the original packet back to the higher layer -get the original packet, passed down from the transport -deallocate the chained buffers -Check if this is the duplicate packet -recyle the resources if done witht the copy of the packet -complete the original packet if it came from an upper transport -need to use the original send adapter to pass the completed packet back to the one that sent it. TransportHandle = OriginalAdapter->IMNdisHandle; NdisInterlockedIncrement(&Adapter->PtSendCompleteCount); ===>invalid handle happend here.<=== NdisMSendComplete( TransportHandle, OriginalPacket, Status ); } If I have a condition to check for NULL adapter, what will it happend if it NULL? is this become memory leak? Why is the adapter become NULL in the middle of the process? how do I avoid the Adapter become NULL so it not going to BSOD? Thanks Tag: TCP Stop Working Tag: 102738
  - 3
    - How do you make a WDF driver create .Raw & .Translated resources? I'm trying to convert a legacy device driver, which was created with BlueWater's DriverWizard, to WDF. The legacy driver creates MyDriverName.Raw & MyDriverName.Translated resources under MACHINE\HARDWARE\RESOURCEMAP\PnP Manager\PnpManager, but the WDF driver does not. How do I make the WDF driver create these resources? /Russ Tag: TCP Stop Working Tag: 102735
  - 4
    - Dispatch and Completion routine synchronisation Dear all, I am a little confused because I keep finding contradicting information about this issue. In a dispatch routine, I set a completion routine and then forward the Irp to the lower driver. Something like this: IoSetCompletionRoutine(Irp, CompletionRoutine, &Event, NULL, NULL, NULL); Status = IoCallDriver(...); if (Status == STATUS_PENDING) KeWaitForSingleObject(...); ... Do some processing... IoCompleteRequest(...); return (Status); So, at the moment my completion routine is tiny and looks a little like this: KeSetEvent(...); return (STATUS_MORE_PROCESSING_REQUIRED); Requirements for the completion routine is to just signal an event so my dispatch routine has an opportunity to do some processing. I've looked at lots of sample code, DDK docs and newsgroup postings and I still am not exactly sure what I should do in terms of what I do in the completion routine and what NTSTATUS value I return from the dispatch routine. 1. Should I check for Irp->PendingReturned and call IoMarkIrpPending(Irp) in the completion routine? (The WDK docs imply that I do not because my completion routine is just signalling an event, but other resources state that the synchronicity of the Irp must be propogated). 2. I assume that because I have halted completion processing with STATUS_MORE_PROCESSING_REQUIRED) , I need to call IoCompleteRequest(...) to allow completion processing to continue. 3. What do I return from my dispatch routine? The status of IoCallDriver or Irp->IoStatus.Status? It feels like I have changed the synchronicity of the Irp so it should be Irp->IoStatus.Status. Many thanks in advance! Carl Tag: TCP Stop Working Tag: 102730
  - 5
    - IRP_MJ_READ/IRP_MJ_WRITE Buffer Confusion. Dear all, I am confused about which buffer is of interest in read/write dispatch routines (and their completion routines). I understand the differences between direct, buffered and neither buffers. So far, I've concluded that: 1. If Irp->MdlAdress exists, use that. 2. If the lower device object Flags specify DO_BUFFERED_IO then use Irp->SystemAddress.Buffered. 3. If none of the above hold true, then use Irp->UserBuffer. Quite frankly, the above rules makes me nervous because I made them up as I went along! What are the exact rules I should follow? Many thanks in advance! Carl Tag: TCP Stop Working Tag: 102729
  - 6
    - AVStream driver show upside-down video for RGB24 format Dear All, I wrote a avstream video capture driver based on the avshws driver sample code. But I found that the FormatRGB24Bpp_Capture structure's bimHeader.biHeight is -D_Y. It's negative value. It will fail to connect the Video renderer in graphedit. If I change it to positive D_Y, then it can render OK. But the video is upside-down. I searched the answer in the web, it looks the negative Height is indicate the bottom-up bitmap. The imageSynth will create a bottom-up frame data to fill into the mapping buffer of the stream queue. But my PCI device can generate only top-down video. And I don't want to use CPU to flip the video data. Is there any way to fix this problem? I ever tried to return my pin's descriptor VideoInfoHeader in the IntersectHandler routine. If can pass the pin connection with negative Height value and the video is correct showing. But the negative Height setting will cause black display in YUY2 format when use AMCAP to preview. Any response is appreciated! Best regards Jerry Pan Tag: TCP Stop Working Tag: 102724
  - 7
    - Regarding DTM set up for LAN. Hi, I am doing whql test for LAN device driver (Vista driver). For the DTM set up of LAN, is it Ok if I use one client machine with XP OS and another client machine with Vista OS ? More precisely the NDIS test server will be mchine with XP loaded and the NDIS client machine will be with vista loaded. Is it ok? This is the same setup for DTM setuo of LAN provided in whql help docs . Need immediate help. Rgds, Ram Tag: TCP Stop Working Tag: 102717
  - 8
    - Newbee: Debugging Blue Screen? Hello, I'm getting a "blue screen" on an IBM x3950 running Windows 2003 Server when reading a word from a custom PCI board: *** Hardware Malfunction. Call your hardware vendor for support. *** The system has halted. *** What is the best way to get started debugging this problem? Thanks, Kent Tag: TCP Stop Working Tag: 102710
  - 9
    - How to get the device instance id of a disk or volume? I am developing a software where I store the Instance ID of every disk in the system. How can I get the Instance ID from a disk number? e.g I want Instance ID of DIsk0 DIsk1 etc.? here is what I am doing 1] SetupDiClassGuidsFromNameA("DiskDrive",***); to get the GUID for disk drives 2]SetupDiEnumDeviceInf(***) to enumrate all the disks.This gives me Instance ID's of all the disks but how can I get the device instance ID of spesific disk(like Disk0)? Tag: TCP Stop Working Tag: 102699
  - 10
    - How to get the device instance id of a disk or volume? I am developing a software where I store the Instance ID of every disk in the system. How can I get the Instance ID from a disk number? e.g I want Instance ID of DIsk0 DIsk1 etc.? here is what I am doing 1] SetupDiClassGuidsFromNameA("DiskDrive",***); to get the GUID for disk drives 2]SetupDiEnumDeviceInf(***) to enumrate all the disks.This gives me Instance ID's of all the disks but how can I get the device instance ID of spesific disk(like Disk0)? Tag: TCP Stop Working Tag: 102698
  - 11
    - Creating a Silent Install for VPN I'm trying to create a silent install for my works VPN program and am having trouble creating or recording the answer file. I've read a few walktroughs on silent installations and have been trying to use setup.exe /r to record the installation process with all the information required to do a custom install. When trying to use /r a window pops up with all available commands and combinations. I'm stuck at the moment and would greatly apperciate some direction from the pros. Any help would be appreciated. Thanks, Lorenzo Tag: TCP Stop Working Tag: 102697
  - 12
    - ndis------Leaking the resource When i use "prefast build -c", i got this message, which i could not understand. NdisAcquireSpinLock(&pAdapter->Lock); protocol.c(679) : warning 8103: Leaking the resource stored in 'NdisSpinLock:pAdapter->Lock'. problem occurs in function 'ProtocolSendComplete' Path includes 17 statements on the following lines: 640 640 641 642 652 654 656 667 668 669 674 675 679 680 681 682 683 That to say there is leaking resource, but i just write the code like this: NdisAcquireSpinLock(&pAdapter->Lock); pAdapter->OutstandingSends--; NdisFreeSpinLock(&pAdapter->Lock); so, how the leak happened? Tag: TCP Stop Working Tag: 102693
  - 13
    - Creating a virtual network card (VLAN) from a script I'm looking for a way to create a virtual network card from a script instead of Windows UI.. I succeeded to install the Protocol driver from a script but didn't find a way to create new VLANs without using the UI. I used the notify dll that comes with mux sample in WDDK. My questions are: - Can I expose a new method in my notify dll for adding a new VLAN, and call it from a script? - Is there a generic tool that makes it? - Do you have a sample code? Thanks Uri Tag: TCP Stop Working Tag: 102692
  - 14
    - Driver installation warning, driver signing and WHQL Hi group, I have a device driver from FTDI which I am using in my software product (for Windows XP). The driver is already WHQL cedrtified. I changed the driver .inf file such that it contains my company's vendor- and product ID which results in loosing the WHQL certification. Software and driver work as expected but: The driver is not installed automatically during the software installation procedure. Instead, Windows complains about the driver and says it is not signed/certified and may harm my computer. 1) How do I get rid of those warning dialogs during driver installation? My guess is that the driver has to be (re-) certified/ signed. Is that true? 2) How do re-certify an already WHQL certified device driver? I figured out that there is the WHQL reseller program. To participate, I (my company) has to sign a lot of legal stuff. Before I continue, I would like to know if I am on the right track... 3) If the WHQL reseller program is the right thing to do to get rid of the warning message during installation, what is the result of this? How is my driver resigned/recertified? Will I get some file from Microsoft with magic information telling the driver is fine? If so, what do I do with this file? 4) I searched the Microsoft WHQL internet pages but it drives me crazy. I cannot find the information I need. Is there a source of useful (!) information about driver signing/certifying and WHQL you can point me to? Thanks for your answers! Regards, Paul Tag: TCP Stop Working Tag: 102689
  - 15
    - firewire hard drive & isochronous transfer How do the firewire hard drives work with isochronous transfer? The isochronous transfers can guarantee constant speed but there is possibility of data loss. Do the firewire hard drives use the isochronous transfer? If yes, will be data get lost? thanks in advance min Tag: TCP Stop Working Tag: 102681
  - 16
    - COM ports disappear with install / remove cycle my driver installs one or more COM ports. After I have installed the devices and removed them again, the next install sometimes uses higher COM port numbers. There must be a map somewhere of free COM numbers, which I have to reset, but so far I have not found it yet. Anyone have an idea? Thanks, Jan Tag: TCP Stop Working Tag: 102666
  - 17
    - keyboard/mice problem after using Hauppauge USB TV Card I use a Hauppauge USB analog TV Card. After a lot channel switching or just use I can't use any Keyboard or Mice(just Mouse Key 1 and 2 are in function) anymore. I can use all type ov viewing software(hauppauge software or DScaler), everywhere the same problem. If I log off an relogin in Windows XP all is ok. How can I check if it is an Driver Problem? It seems so. regards Tobias Philipp Tag: TCP Stop Working Tag: 102664
  - 18
    - how can I get a vista device handle? maybe it is easy when you run the function in a wdf device driver. but I have a wdm driver, want to talk to wdf driver. how can I do that? IoGetDeviceObjectPointer( &unicodeName, FILE_READ_ATTRIBUTES, &pFile, &pDevice ); wdfFile = *((WDFFILEOBJECT *)pFile); wdfDevice = WdfFileObjectGetDevice(wdfFile); it can get the fileobject and device object when I know the wdf device name-"unicodeName", and later I force the pointer to WDFFILEOBJCT, it tell me access memory error. Tag: TCP Stop Working Tag: 102661
  - 19
    - pIsochDetachData->AttachStatus This is a multi-part message in MIME format. ------=_NextPart_000_0008_01C86377.FDF355A0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable In the 1394 kmdf sample driver in WDK6000, the function 't1394_IsochAttachBuffers' uses a variable = pIsochDetachData. -------------------------------------------------------------------------= -------------- // need to save hResource, numDescriptors and Request to use when = detaching. // this needs to be done before we submit the irp, since the isoch = callback // can be called before the submitirpsynch call completes. PISOCH_DETACH_DATA pIsochDetachData; pIsochDetachData =3D ExAllocatePoolWithTag(NonPagedPool, sizeof(ISOCH_DETACH_DATA), POOLTAG_1394); -------------------------------------------------------------------------= -------------------------------------------------------- PISOCH_DETACH_DATA is defined as -------------------------------------------------------------------------= -------------------------------------------------------- typedef struct _ISOCH_DETACH_DATA { LIST_ENTRY IsochDetachList; PDEVICE_EXTENSION DeviceExtension; PISOCH_DESCRIPTOR IsochDescriptor; WDFREQUEST Request; PIRP newIrp; PIRB DetachIrb; PIRB AttachIrb; NTSTATUS AttachStatus; KTIMER Timer; KDPC TimerDpc; HANDLE hResource; ULONG numIsochDescriptors; ULONG outputBufferLength; ULONG bDetach; } ISOCH_DETACH_DATA, *PISOCH_DETACH_DATA; = -------------------------------------------------------------------------= -------------------------------------------------------- What exactly says the NTSTATUS value 'AttachStatus' from = pIsochDetachData?=20 It seems, it gets already a value direct after calling = 'ExAllocatePoolWithTag'... thanks in advance min ------=_NextPart_000_0008_01C86377.FDF355A0 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN"> <HTML><HEAD> <META http-equiv=3DContent-Type content=3D"text/html; = charset=3Diso-8859-1"> <META content=3D"MSHTML 6.00.6000.16587" name=3DGENERATOR> <STYLE></STYLE> </HEAD> <BODY> <DIV><FONT face=3DArial size=3D2>In the 1394 kmdf sample driver in=20 WDK6000,</FONT></DIV> <DIV><FONT face=3DArial size=3D2></FONT> </DIV> <DIV><FONT face=3DArial size=3D2>the function=20 't1394_IsochAttachBuffers' uses a variable=20 pIsochDetachData.</FONT></DIV> <DIV><FONT face=3DArial size=3D2></FONT> </DIV> <DIV><FONT face=3DArial=20 size=3D2>----------------------------------------------------------------= -----------------------</FONT></DIV> <P><FONT face=3DArial size=3D2>// need to save hResource, numDescriptors = and Request=20 to use when detaching.</FONT></P> <P><FONT face=3DArial size=3D2>// this needs to be done before we submit = the irp,=20 since the isoch callback</FONT></P> <P><FONT face=3DArial size=3D2>// can be called before the = submitirpsynch call=20 completes.</FONT></P> <P><FONT face=3DArial size=3D2>PISOCH_DETACH_DATA = pIsochDetachData;</FONT></P> <P><FONT face=3DArial size=3D2>pIsochDetachData =3D=20 ExAllocatePoolWithTag(NonPagedPool,</FONT></P> <P><FONT face=3DArial size=3D2>sizeof(ISOCH_DETACH_DATA),</FONT></P> <P><FONT face=3DArial size=3D2>POOLTAG_1394);</FONT></P> <P><FONT face=3DArial=20 size=3D2>----------------------------------------------------------------= -----------------------------------------------------------------</FONT><= /P> <P><FONT face=3DArial size=3D2></FONT> </P> <P><FONT face=3DArial size=3D2>PISOCH_DETACH_DATA is defined = as</FONT></P> <P><FONT face=3DArial=20 size=3D2>----------------------------------------------------------------= -----------------------------------------------------------------</FONT><= /P> <P><FONT face=3DArial size=3D2>typedef struct _ISOCH_DETACH_DATA = {</FONT></P> <P><FONT face=3DArial size=3D2>LIST_ENTRY IsochDetachList;</FONT></P> <P><FONT face=3DArial size=3D2>PDEVICE_EXTENSION = DeviceExtension;</FONT></P> <P><FONT face=3DArial size=3D2>PISOCH_DESCRIPTOR = IsochDescriptor;</FONT></P> <P><FONT face=3DArial size=3D2>WDFREQUEST Request;</FONT></P> <P><FONT face=3DArial size=3D2>PIRP newIrp;</FONT></P> <P><FONT face=3DArial size=3D2>PIRB DetachIrb;</FONT></P> <P><FONT face=3DArial size=3D2>PIRB AttachIrb;</FONT></P> <P><FONT face=3DArial size=3D2>NTSTATUS AttachStatus;</FONT></P> <P><FONT face=3DArial size=3D2>KTIMER Timer;</FONT></P> <P><FONT face=3DArial size=3D2>KDPC TimerDpc;</FONT></P> <P><FONT face=3DArial size=3D2>HANDLE hResource;</FONT></P> <P><FONT face=3DArial size=3D2>ULONG numIsochDescriptors;</FONT></P> <P><FONT face=3DArial size=3D2>ULONG outputBufferLength;</FONT></P> <P><FONT face=3DArial size=3D2>ULONG bDetach;</FONT></P> <P><FONT face=3DArial size=3D2>} ISOCH_DETACH_DATA, = *PISOCH_DETACH_DATA;</FONT></P> <DIV><FONT face=3DArial=20 size=3D2> ----------------------------------------------------------= -----------------------------------------------------------------------</= FONT></DIV> <DIV><FONT face=3DArial size=3D2></FONT> </DIV> <DIV><FONT face=3DArial size=3D2></FONT> </DIV> <DIV><FONT face=3DArial size=3D2></FONT> </DIV> <DIV><FONT face=3DArial size=3D2>What exactly says the NTSTATUS value = 'AttachStatus'=20 from pIsochDetachData? </FONT></DIV> <DIV><FONT face=3DArial size=3D2></FONT> </DIV> <DIV><FONT face=3DArial size=3D2>It seems,  it gets already a = value direct=20 after calling 'ExAllocatePoolWithTag'...</FONT></DIV> <DIV><FONT face=3DArial size=3D2></FONT> </DIV> <DIV><FONT face=3DArial size=3D2></FONT> </DIV> <DIV><FONT face=3DArial size=3D2>thanks in advance</FONT></DIV> <DIV><FONT face=3DArial size=3D2>min</FONT></DIV> <DIV><FONT face=3DArial size=3D2></FONT> </DIV> <DIV><FONT face=3DArial size=3D2></FONT> </DIV></BODY></HTML> ------=_NextPart_000_0008_01C86377.FDF355A0-- Tag: TCP Stop Working Tag: 102660
  - 20
    - kernel memory Hi, i have a very basic question regarding kernel memory allocation. is allocating kernel memory always performed on some kind of pool (paged or nopaged) ? is there any memory allocation methods that acts on a "regular" memory, meaning type of memory other than paged or nopaged ? since kernel memory space is about 1G, i'm a bit confused since both the paged pool and the non paged pool are smaller than 1G. so what with the rest of the memory ? can i allocate more than the pools have to offer ? if so, how ? since all the routines deal with a pool type (paged or non paged). thanks a lot, kobi. Tag: TCP Stop Working Tag: 102655
  - 21
    - MSI IRQ is negative in device manager Running on Vista32, I have a PCIe device that supports MSI. The CM_PARTIAL_RESOURCE_DESCRIPTOR has a value of 0x8 for the MessageInterrupt.Translated.Level member, but device manager shows that the IRQ is 0xFFFFFFFC ( -4?). There is a KB that describes this situation but does not indicate whether this is expected behavior or anything of that nature. It merely says it is aware of it. http://support.microsoft.com/kb/940394 Can anyone shed some light on this? Tag: TCP Stop Working Tag: 102654
  - 22
    - staticdv -> This is not a valid driver folder. Hi, IÂ´m trying to use staticdv for a audio codec and i get the follow error (tested in checked and free): --------------------------------------------------------------------- Microsoft (R) Windows (R) Static Driver Verifier Version 1.4.251.0 Copyright (C) Microsoft Corporation. All rights reserved. --------------------------------------------------------------------- Warning: A driver (or a library) will be processed in the CHK build environment. Build 'ACM.AudioCodec' ...Done This is not a valid driver folder. which is the correct structure for this check ? the build and prefast works fine. Thanx, Daniel Tag: TCP Stop Working Tag: 102649
  - 23
    - NDIS6 - Renew IP Address Hello, I'm writing a miniport driver for NDIS6 (Vista) which emulates 802.3 network device, but the actual physical device is different. Within a certain scenario I need the overlying protocol(s) to renegotiate the IP Address. I'm trying to follow the MSDN documentaion thus sending the status NDIS_STATUS_NETWORK_CHANGE with the specific NDIS_NETWORK_CHANGE_TYPE (NdisDefinitelyNetworkChange or NdisPossibleNetworkChange, i've tried them both). Unfortunatly, I couldn't obtain a new IP Address. when I run "ipconfig /renew" manualy I get a new IP easily. I've also seen in a sniffer that no DHCP Request is sent when my miniport indicates the status above (NDIS_STATUS_NETWORK_CHANGE ). In the past, for my previous miniport driver for NDIS5 (XP) the "trick" to renew the IP from the miniport was to indicate "link down" and "link up" in a row. in that case it worked fine. But now, running with NDIS6 I dont know the correct method for doing it. Can anybody help? Thanks Yogev Tag: TCP Stop Working Tag: 102648
  - 24
    - File filter driver or MUP network redirector? Hi I have a design session coming up where the ultimate aim is to protect parts of the file system from prying eyes using either a MUP network redirector ( assuming file system mounted through remoted shares) or a file system filter driver using filter manager. I need to clear up some assumptions : Is mup a client or server side ( or both ? ) technology ? I understand I need to write a network redirector driver that owns a VNC path and will manage access to that path. If the unc path is on a file server - what needs to be done at the file server ? The file mini filter will need to intercept all calls to the filesystem. I have big fears about all the different MJ call types I will have to handle - fast IO, asynch io etc. What is the best plan to handle this - my opinion is to route them all through a common procedure and allow/disallow access as appropriate ut this can be a lot of different calls which expect different handling. This may lead to deadlocks and performance problems if not very carefully handled on a busy fileserver. The major benefit to me of the latter is that it can be installed everywhere - so a common build. Any commentsw or guidance as when to use one and when to use the other. The WDK docs on MUP haven't been terribly clear to me yet. regards Mike Tag: TCP Stop Working Tag: 102645
  - 25
    - Bulkusb.sys performance hi, I have adapted bulkusb.sys (WDK 6000) for my usb device driver. With the driver communicating on interrupt endpoint i am not able to get performance more than 15Mbits/sec. However with the same device and application but driver from "thesycon" i.e. USBIO, performance is about 80Mbits/sec. Can some one tell why ? or if someone has the performance figures of bulkusb.sys or if some can guide how to improve the performance. Regards, Tag: TCP Stop Working Tag: 102644
- Next
  - 1
    - Winusb.sys and winusb.lib source Hi , Is the source code of winusb.sys and winusb.lib available. It could be needed if some optimization is to be done. Regards, anshul Tag: TCP Stop Working Tag: 102642
  - 2
    - NDIS setup and DTM seup for LAN is different Hi, 1. Are the NDIS setup and DTM setup for LAN is different ?Please let me know as I observed the difference between them. 2. In the DTM setup for LAN, local message device is not mentioned but in running the LAN test this is mentioned. This is confusing so please help to understand it. Rgds, Ram Tag: TCP Stop Working Tag: 102639
  - 3
    - Problem reconfiguring PCI-e device in Vista Hi, I'm writing a driver for a proprietary PCI-e adapter. The board contains multiple FPGA-based PCI-e devices behind a PCI-e bridge. From time to time, we need to reconfigure the FPGA's while the system is alive. We do not however want the system to reboot (think of it as hot- swapping without physically replacing the device). The problem is that the FPGA's incorporate the PCI-e interface so that they cannot be reconfigured while they are active. In XP we finally managed to reconfigure the FPGA's by first calling CM_Query_And_Remove_SubTree() on the bridge (being the parent devices of our FPGA's). This causes the FPGA-devices to be switched to the D3- state, after which they can safely be reprogrammed. We then call CM_Reenumerate_DevNode() on the parent of the bridge device so that the new FPGA-devices are recognised again. In Vista however, the system seems to automatically re-enumerate the PCI bus immediately after the CM_Query_And_Remove_SubTree() has completed. This causes the FPGA's to be accessed while they are being configured resulting in a BSOD (Machine-check-excpetion). - Any idea how to disable (or delay) the re-enumeration of the PCI- bus? - Any link to more extensive documentation of the internals of the PnP system? I've found many ways to manage PnP devices, but little or no information about what is happening behind the scenes. - Any suggestions for alternative methods to achieve the reconfiguration? Note: redesigning our hardware is no option. Thanks in advance for your help, Koen Tag: TCP Stop Working Tag: 102638
  - 4
    - Vista 32 bit display driver for small lcd's on USB Hi I need to make a solution which permits having secondary monitors on USB (small ones - vga/qvga) As far as I seen the display driver model has changed completly from the XP. Does anybody know a good startpoint for a vista display driver? The R200Sample from the WDK is a good starting point? A second problem is that Vista does not permits having two WDDM display drivers from different vendors. Does anybody knows solutions for this? XPDM is not really an option (only last one) since it switches off aero on primary monitor too which is something would need to avoid ... regards, -- -Barni Tag: TCP Stop Working Tag: 102637
  - 5
    - INF file for SMBus driver Folks, I am writing a client driver for an SMBus device on a PCIe card. The SMBus device gets detected by the host controller. I can see the device listed by utilities like SpeedFan. The SMBus device uses a persistent slave address (actually an I2C device posing as an SMBus device). My question is how does one write an INF file for such an SMBus device - more specifically, how does one describe the device. I found this entry in the acpi.inf file - %SMBALI.DeviceDesc% = SMBALI_Inst,ACPI\SMBUS&VEN_10B9&DEV_7101 I guess this will work for devices that use dynamic address and get their address using ARP because the device id and vendor id will be available to the SMBus host controller from the device's UDID. But for SMBus devices that do not have things like UDID how does one specify the device description. Thanks, Ashwat Tag: TCP Stop Working Tag: 102634
  - 6
    - Bluetooth profile driver in vista Hi all, I have some confussed questions. Yes, I read some msdn about bluetooth profile driver development in MSDN. But unfortunately, I haven't find source code in WDK sample. There are some questions : 1) If I write a headset or handsfree profile driver, I should communiate with SPP profile or RFCOMM stack. But I can not find the programming interface. 2) What interface should profile driver provide? For handsfree or headset profile, it should cowork with windows audio stack in bluetooth vista stack. But how? And the last question, do we need develope alll profile driver for user? or Microsoft will provide all profile driver in vista? Tag: TCP Stop Working Tag: 102631
  - 7
    - Network bridge interface configuration I am trying to configure a network bridge programmatically. I can create and remove one fine using devcon, but how can I add my interface adapters to the existing bridge programmatically? Can I use devcon, or must I do some DLL calls through hnetcfg.dll? If I have to go the latter route, how do I get the function signatures (and any associated info I might need, if they take complex structures as arguments)? I know netsh does a lot, but there isn't much/any bridge configuration capability with it. Thanks in advance, -- Kevin Tag: TCP Stop Working Tag: 102619
  - 8
    - DDK Halftone Functions Hi, I notice that print drivers, through the DDK, have access to some pretty neat halftone functions, with various halftone patters like these... HT_PATSIZE_2x2 HT_PATSIZE_2x2_M HT_PATSIZE_4x4 HT_PATSIZE_4x4_M HT_PATSIZE_6x6 HT_PATSIZE_6x6_M HT_PATSIZE_8x8 HT_PATSIZE_8x8_M HT_PATSIZE_10x10 HT_PATSIZE_10x10_M HT_PATSIZE_12x12 HT_PATSIZE_12x12_M HT_PATSIZE_14x14 HT_PATSIZE_14x14_M HT_PATSIZE_16x16 HT_PATSIZE_16x16_M HT_PATSIZE_SUPERCELL HT_PATSIZE_SUPERCELL_M HT_PATSIZE_AUTO Is it possible to access these same halftone patterns throguh the regular Win32 API or SDK? I am not writing a driver, but I am trying to add halftone capabilities to an old legacy C Win32 application, and I thought there might be a way to leverage these pre existing microsoft halftone functions in the app. Is it possible to get these same patterns through BitBlt, or StretchBlt, etc? Thanks in advance, Skeet Tag: TCP Stop Working Tag: 102618
  - 9
    - question about 64-bit compliant DDK samples I am playing around with the samples in the DDK. In the HTML document, it says "this is 64-bit compliant". Does this mean that it won't run on my 32-bit Windows XP system? I tried to install 64-bit compliant Msplotter sample on my machine. I could not see it after a successful installation. On the other hand, a 32-bit complaint software can run on 64-bit system. Am I right? I am wondering which option I should change so that I will build a 32-bit compliant samples so that they can run on my XP. Thanks in advance! Tag: TCP Stop Working Tag: 102615
  - 10
    - Obtainning USB_DEVICE_DESCRIPTOR How can I go about obtainning the USB_DEVICE_DESCRIPTOR for my custom device without enumerating through all hubs, ports and connected devices (ddk usbview example)? I also would like to avoid creating a custom IOCTL (i.e. IOCTL_MYDEVICE_GET_DEVICE_DESCRIPTOR). Tag: TCP Stop Working Tag: 102614
  - 11
    - SET_LINE_CODING and usbser.sys I'm trying to write a USB to serial device (and who isn't<g>?) Right now I'm using usbser.sys. I've read all the warnings but a custom driver is currently out of the question. I don't understand why I'm not getting SET_LINE_CODING messages. When I go into Device Manager and change the baud rate, I would think that would be the time to call into usbser.sys and pass on a SET_LINE_CODING. My ultimate goal is to have the converter get the baud rate, etc on startup from Windows. Tag: TCP Stop Working Tag: 102613
  - 12
    - Validating parameters for IoDeviceControl in KMDF -> KMDF verifier error When testing my driver to be crash safe for various user input, I came across an issue in my driver that I don't know how to handle. To eliminate my own driver, I tested the same thing with the original WDK (WDK 6000) sample osrusbfx2 driver. By modifying the testapp to send erronous input to the driver, I could get the error/break even there. The sample uses the parameters\wdf\DbgBreakOnError = 1. Now I wonder how it should be handled by the driver, as the KMDF says it's an error. How to reproduce: To have some common code to look at, I use the sample KMDF osrusbfx2. Modify the testapp to call the driver with IOCTL_OSRUSBFX2_GET_7_SEGMENT_DISPLAY and put erronous parameters in the call: if (!DeviceIoControl(deviceHandle, IOCTL_OSRUSBFX2_GET_7_SEGMENT_DISPLAY, &sevenSegment, // Ptr to InBuffer, NULL in example sizeof(UCHAR), // Length of InBuffer, 0 in example &sevenSegment, // Ptr to OutBuffer 0, // Length of OutBuffer &index, // BytesReturned 0)) { // Ptr to Overlapped structure I've only seen it when input buffer pointer and length is valid and the output buffer pointer is correct but oputput buffer length is zero. By turning on Parameters\Wdf\DbgBreakOnError, I get: >KMDF detected potentially invalid operation by canusb.sys. Turn on >framework verifier to break into the debugger. Turning on the verifier gives me >410: FxRequest::CompleteInternal - Caller of this WDFREQUEST 7A104170, Irp >0x862BB398 is passing an OuputBuffer address even though the >OutputBufferLength is zero <yes, there is a spell error in the log: OuputBuffer> So, here's my question: if this is a driver error, how can it be corrected? If this is just invalid parameters, why does the KMDF verifier break to the debugger? Regards, Carl-Magnus Segerlund Tag: TCP Stop Working Tag: 102601
  - 13
    - Got few questions/doubts about MDL (after browsing archives/web) on this subject Hi, Working on a problem at hand with miniport network driver. I have few questions related to MDL & NET_BUFFER. My questions are: 1. What is the maximum size of virtual address-range that a single MDL can map? (I am under the impression that only one page of contiguous virtual address range can be mapped by a **Single** MDL. Let me know if this is wrong). 2. If a single buffer whose virtual address range is more than 8 K (size of receive buffer is more than 2 page assuming 4K page size), how many MDLs would be required in the *LEAST* to map the entire virtual address range? Answer to Q#1 will probably answer Q#2 as well. If not pls include answer for this as well. 3. Can ByteOffset be more than size of one physical page? I see that ByteCount value of Mdl returned by NdisAllocateMdl (length = 9k) set to 9K. 4. I have got a buffer whose size is more than 8K and I use NdisAllocateMdl to allocate Mdl to map this buffer passing start address of this buffer and complete length of this buffer. But I receive a single MDL in response to this operation? Does that not mean its mapping only part of buffer I passed? When I tried changing virtual address from (start address) to (start address + 5k), Mdl->StartVa & Mdl->ByteOffset changes to reflect next page which seems to answer my aobve question that it maps only part of the buffer. But still want to get it verified from experts on this list. How to get MDL chain mapping for the complete virtual address range and not just the first page of the receive buffer? Should I explicitly call NdisAllocateMdl() in loop for entire range starting from the start-address stepping through at size of 4k? 5. If I call NdisAllocateMdl on a buffer with address say X. I receive an Mdl and after executing some code, I realize that I need to indicate data to NDIS starting from X+5K. Can I pass the above allocated Mdl & NET_BUFFER to do this by manipulating DataOffset/ByteOffset? Thanks, -Praveen I see the following information in OSRonline doc on MDL: // Notice that while in the context of the subject // thread, the base virtual address of a buffer mapped // by an MDL may be referenced using the following: // // Mdl->StartVa | Mdl->ByteOffset typedef struct _MDL { struct _MDL *Next; CSHORT Size; CSHORT MdlFlags; struct _EPROCESS *Process; PVOID MappedSystemVa; PVOID StartVa; ULONG ByteCount; ULONG ByteOffset; } MDL, *PMDL; Also you know that StartVa is the page aligned virtual address of the start of the range and ByteOffset is the starting offset into the first page. TRIVIA POINT: MmBuildMdlForNonPagedPool sets the MappedSystemVa of the MDL to (Mdl->StartVa | Mdl->ByteOffset) and then sets the MDL_SOURCE_IS_ NONPAGED_POOL bit in the MdlFlags. Tag: TCP Stop Working Tag: 102593
  - 14
    - Log Mejor Functions I want to log (in memory in circular list type space- so I get it on BSOD) all mejor function calls with some call parameters. I am thinking about setting all mejor function to a single function, log it there and forward to real function and log again on return. Would that be a good idea? Any idea or comment? -- Sincerely, Maruf Maniruzzaman, Software Engineer, KAZ Software Limited, Dhaka, Bangladesh. http://kaz.com.bd http://kuashaonline.com This posting is provided "AS IS", and confers no rights. Tag: TCP Stop Working Tag: 102587
  - 15
    - How to delete the Tab "Layout" or "Paper/Quality" in Postscript Dr I'm writing a Customised Postscript Driver which is based on "PS5UI.DLL" and "PSCRIPT5.DLL". When Printing Perference window displays, I want to display the tab that I created by myself and I do not want to display the tab "Layout" or "Paper/Quality". The question is, how can I delete the tabs "Layout" or "Paper/Quality" from Printing Perference window? the following method I used is failed,because I can't get the handle of tab "Layout" or "Paper/Quality" . HRESULT __stdcall IPrintOemUI::DocumentPropertySheets( PPROPSHEETUI_INFO pPSUIInfo, LPARAM lParam) { if (pPSUIInfo) { HANDLE hPage[10]; LONG iPageCount; memset(hPage,0,sizeof(hPage)); iPageCount = (LONG)pPSUIInfo->pfnComPropSheet( pPSUIInfo->hComPropSheet, CPSFUNC_GET_HPSUIPAGES,(LPARAM)hPage,(LPARAM)10); if (iPageCount) { iPageCount = (LONG)pPSUIInfo->pfnComPropSheet( pPSUIInfo->hComPropSheet, CPSFUNC_DELETE_HCOMPROPSHEET,(LPARAM)hPage[1],(LPARAM)0); } } } Tag: TCP Stop Working Tag: 102584
  - 16
    - PrinterDriver UI and MFC Hi NG, I'm writing a MsPlot based monotholitic printer driver. The question is wether it's possible, and if so what are the required steps, or not to use Mfc from withing Configuration Dll of a printer driver. I've already done some investigations, but weren't able to solve all Linker errors. - LNK2001 unresolved external symbol __afxForceEXLUDE - LNK2001 unresolved external symbol _afxPcNil I've included "afx.h". Which preprocessor defines must be set? Which header do I have to include? Which settings do I've to change in my sources file? Best regards, Lela Tag: TCP Stop Working Tag: 102582
  - 17
    - suggestions needed for developing a printer driver I have no expereince on wrting a printer driver and need your advices or suggestions on developing a unidrive based image driver. I have downloaded WDK/WDF, built the samples by using "build -cZ", and tested the sample image driver in Print section. Now I need to modify the src code, integrate more stuff together, and do debugging soon. Is Visual Studio 6.0 a good IDE for this project? Do I need SDK for this? Is WDK the only one I need? Do you have any other advices or suggestions for me? Thanks in advance! Tag: TCP Stop Working Tag: 102571
  - 18
    - Synchronising read and write ops All, I am writing a very basic on the fly encryption file system filter driver as a learning exercise. I was planning on: 1. Storing the real length of the file and other information in a header 2. Reading and writing the header as part of non-cached and paging IRP_MJ_READ and IRP_MJ_WRITE operations. 3. Returning the length from the header for IRP_MJ_QUERY_INFORMATION Irps. I was planning to use a driver managed queue of Irps so that I could ensure the I could update the header in response to a write before a another thread read the header; classic synchronization problem. However, I discovered that most (if not all) of the operations use Irp->UserBuffer, and in my driver dedicated thread I would have to change process context all the time to access the buffer, which strikes me as both dangerous and extremely slow. So, my question really is how would one generally synchronize this kind of thing? Could I use an ERESOURCE and have any read operations acquire shared and write operations acquire exclusively? Is it safe to enter a critical region for all read and write Irps? Many thanks in advance! Carl Tag: TCP Stop Working Tag: 102570
  - 19
    - DiskId32 has trouble reading some SATA drives My DiskId32 tool, http://www.winsim.com/diskid32/diskid32.html , reads hard drive information directly from the hard drive in Windows Vista, XP, Win2K, Win9X, etc. DiskId32 works in three modes: direct using administrator rights, indirect using the scsi bypass and direct using user mode rights. The scsi bypass was dropped in Windows server 2003 and Vista. Please note that the source code for DiskId32 is there also. For some reason, the direct reading of the hard drive using user mode rights is failing for some SATA hard drives. However, the direct reading of the hard drive works fine in administrator mode. Specifically, what is failing is the reading of the serial number of the hard drive. The identify bank being returned by the hard drive is only 121 bytes long and the serial number pointer is at position 120 which has a zero in it. I would appreciate any insights or thoughts people might have for this issue. Thanks, Lynn Tag: TCP Stop Working Tag: 102562
  - 20
    - KMDF filter driver not loading in safe boot mode We have several filter drivers using KMDF, and we noticed that under Windows XP SP2 they fail to load when you boot in safe mode. Seems that wdf01000.sys is not loaded in safe mode, causing the filter driver to not load. Issue is not present in Vista. Was able to fix this issue by adding the following registry key: HKLM\SYSTEM\CurrentControlSet\Control\SafeBoot\Minimal\WdfLoadGroup Is this a known issue? Was not able to find any information on Microsoft, but seems that some other people had the same issue. Thanks for any help, Olivier Tag: TCP Stop Working Tag: 102560
  - 21
    - Storport driver dropping HBA API events? Hi, I have taken the TempConsumer SDK example and converted it to consume the MSFC_AdapterEvent class. I have a Fibre Channel driver that fires multiple events (using ScsiPortFireWmiAdapterEvent) in rapid succession. However, the consumer only sees the first event. If the events are fired a few seconds later, again I only see the first event. Is there a limitation for firing events? If so, how is an event like this one supposed to work when there is unique data (the event type) each time the event is fired? Thanks, Scott Tag: TCP Stop Working Tag: 102545
  - 22
    - Mysterious change in DbgPrint functionality I just solved a problem that was baffling me. I was trying to debug a driver on Vista Home Premium (32-bit) and, following advice I found online, I disabled User Account Control (UAC). All of a sudden, DebugView stopped showing messages from my driver. Turning UAC back on didn't fix it. I eventually solved my problem by creating a REG_DWORD value named DEFAULT, value -1, in HKLM\System\CurrentControlSet\Control\Session Manager\Debug Print Filter. My assumption is that disabling UAC deletes that key and re-enabling UAC doesn't put it back. -- Walter Oney, Consulting and Training http://www.oneysoft.com Tag: TCP Stop Working Tag: 102541
  - 23
    - CreateFile not working in vista-32 Application is calling CreateFile("\\\\.\\xxxx", GENERIC_READ | GENERIC_WRITE, etc.). Works great on Vista Ultimate (64-bit) and all pre-Vista 32-bit systems. Fails on Vista Home Premium (32-bit) with ERROR_ACCESS_DENIED. IRP, if ever generated, is not getting to my driver. App is running from command prompt running as administrator. Device object is created with call to IoCreateDeviceSecure specifying a valid setup GUID and SDDL_DEVOBJ_SYS_ALL_ADM_ALL. Objdir shows that there is no DACL on the device object, which presumably explains why CreateFile is failing. (There is a DACL on the symbolic link, and it allows R/W access by administrators). -- Walter Oney, Consulting and Training http://www.oneysoft.com Tag: TCP Stop Working Tag: 102540
  - 24
    - Synchronizing MiniportResetEx with other Miniport entry points Hi, When MiniportResetEx is pending in the Miniport Driver, will NDIS call any other MiniportXXX entry points till Miniport says NdisMResetComplete (mainly oid, tx, adapter state transition and tx/ oid cancel routines)? If not then this will greatly reduce the synchronization effort in the Miniport. Thanks, Aniketa Tag: TCP Stop Working Tag: 102538
  - 25
    - Error on Smart Card Service Hi, this may be offtopic but this problem involves, I think, low level Windows knowledge: I've developed a smart card application; it works ok, but when the session of Windows XP is locked and the application is running and tries to call a winscard API, it hangs in that call; never returns. I have found this error in the System log of the Event Viewer: Smart Card Reader 'OMNIKEY CardMan 3x21 1' rejected IOCTL GET_STATE: Access is denied. Source: SCardSvr Event_ID: 610 I would really appreciate any hint about this problem or where to find information / a forum to ask these questions. Thanks, -- Federico Tag: TCP Stop Working Tag: 102537

I have an ndis miniport driver that I've developed. ARP, TCP, ICMP, etc, all
seem to work fine. But when I run an extened test that sends data over a
socket for about 10 minutes straight, TCP stops working on the entire system.
Other protcols are still working, because ping works just fine. It's not
just my miniport interface that doesn't work. I can't get TCP to work on any
network interface. When I try to run ipconfig I get the following error "The
application failed to initialize properly (0xc0000142). Click on OK to
terminate the application." I also tried running Ethereal after the error
occurred, and it can't even find any interfaces to bind to. I have driver
verifier turned on my driver, and no errors were detected. Has anyone ran
into this condition before? How do you debug a system in this state?

Thanks,
Michael

Top

Re: TCP Stop Working by Alireza

Alireza
Tue Feb 05 02:05:15 CST 2008

Are you sure you are completing all the send requests?

Or maybe you are dropping ACKs?

what does your IM driver do?

-ali

--
This posting is provided "AS IS" with no warranties, and confers no rights.

"Michael" <Michael@discussions.microsoft.com> wrote in message
news:C6F10B8B-5B47-4455-85F8-D54035B5A961@microsoft.com...
>I have an ndis miniport driver that I've developed. ARP, TCP, ICMP, etc,
>all
> seem to work fine. But when I run an extened test that sends data over a
> socket for about 10 minutes straight, TCP stops working on the entire
> system.
> Other protcols are still working, because ping works just fine. It's not
> just my miniport interface that doesn't work. I can't get TCP to work on
> any
> network interface. When I try to run ipconfig I get the following error
> "The
> application failed to initialize properly (0xc0000142). Click on OK to
> terminate the application." I also tried running Ethereal after the error
> occurred, and it can't even find any interfaces to bind to. I have driver
> verifier turned on my driver, and no errors were detected. Has anyone ran
> into this condition before? How do you debug a system in this state?
>
> Thanks,
> Michael
>

Top