|
Greetings,
I have "gitted" the libwdi and compiled it using MinGW. I've tested it on a virtual box running windows xp sp3 and it seems to work quite well. When I insert a second device of the same type (different serial number, same VID, PID, interface number) the device did not attach to libusb/winusb. When I run zadic.exe on the second device it installed the drivers properly. For most of my users installing the driver for each hardware instance will not be a problem, but some of my users have dozens of devices that must be inserted into the same machine and reinstall for each instance would be a considerable inconvenience. Is there any way to install the libusb/winusb driver in a way that allows all devices to be recognized? By the way, in installer.c, main() I commented out "update_driver(device_id)", see below, to see if the device_id had any affect. This change did not seem to effect anything. ... if (b == true) { // Success plog("driver update completed"); // TODO: remove this? plog("joey does not update drivers"); //update_driver(device_id); ret = WDI_SUCCESS; goto out; } ... Thanks in advance for any insight you can provide. ------------------------------------------------------------------------------ _______________________________________________ Libusb-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/libusb-devel |
I've found the answer to my question in Jan Axelson's USB Complete 3rd Edition on page 276. Since our device is a HID which is a supported class, windows will always match the compatible ID (which contains the class of the device) with the class of windows generic HID drivers rather than libusb/winusb. This is because windows HID driver is digitally signed where as the .inf zadic.exe uses is digitally unsigned. Even though zadic's inf contains information to match itself with our device, windows will search for any match that exists in its digitally signed .inf files before attempting to associate unsigned .inf files to a device.Greetings, I have "gitted" the libwdi and compiled it using MinGW. I've tested it on a virtual box running windows xp sp3 and it seems to work quite well. When I insert a second device of the same type (different serial number, same VID, PID, interface number) the device did not attach to libusb/winusb. When I run zadic.exe on the second device it installed the drivers properly. For most of my users installing the driver for each hardware instance will not be a problem, but some of my users have dozens of devices that must be inserted into the same machine and reinstall for each instance would be a considerable inconvenience. Is there any way to install the libusb/winusb driver in a way that allows all devices to be recognized? To get a Microsoft digital signature it seems that you must pass a WHQL test. What does it take to get a WHQL certification?? I will be thankful for any insight that can be provided by someone who knows anything about obtaining digital signatures from Microsoft. Joseph, ------------------------------------------------------------------------------ _______________________________________________ Libusb-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/libusb-devel ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ Libusb-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/libusb-devel |
|
Hi Joseph,
Apologies for missing your first post (and thanks for testing libwdi). I must admit that I hadn't considered that installation of driver for all devices of the same type would be an issue. On 2010.06.04 15:28, joey wrote: >> When I insert a second device of the same type (different serial number, >> same VID, PID, interface number) the device did not attach to >> libusb/winusb. Interesting result. I checked what happened with a couple of Logitech USB mice, and found that plugging one or the other in the same USB port would keep the same driver (but these mice don't report a serial number in their USB strings, and I believe if that wasn't the case, I would have observed different results), while plugging any of these mice onto a different port, after the WinUSB driver has been installed, would still install the HidUSB driver. Other devices, like an iPod where the default USBSTOR driver was replaced with WinUSB will on the other hand use WinUSB regardless of the slot. >> For most of my users installing the driver for each hardware >> instance will not be a problem, but some of my users have dozens of >> devices that must be inserted into the same machine and reinstall for >> each instance would be a considerable inconvenience. Yes, I agree with that. I had not really considered that scenario. >> Is there any way to install the libusb/winusb driver in a way that >> allows all devices to be recognized? >> > I've found the answer to my question in Jan Axelson's _USB Complete 3rd > Edition_ on page 276. Since our device is a HID which is a supported > class, windows will always match the compatible ID (which contains the > class of the device) with the class of windows generic HID drivers > rather than libusb/winusb. This is because windows HID driver is > digitally signed where as the .inf zadic.exe uses is digitally unsigned. > Even though zadic's inf contains information to match itself with our > device, windows will search for any match that exists in its digitally > signed .inf files before attempting to associate unsigned .inf files to > a device. That seems to be matching the results I am observing with my mice, and I'm quite glad you found that information. This only seems to apply to HID devices however, as the USBSTOR driver, which is digitally signed by Microsoft, does not overwrite other drivers. > To get a Microsoft digital signature it seems that you must pass a WHQL > test. Yes, to sign the inf, you have no choice but to go through the WHQL process. Just a reminder that this is different from digitally signing the driver itself, which can be done without Microsoft's intervention, with a code signing certificate. > What does it take to get a WHQL certification?? I think you probably want to read the WHQL related information from the Windows Logo Program System and Device Requirements, Version 2.2.1a from http://www.microsoft.com/whdc/winlogo/hwrequirements/v22.mspx The document seems to list http://www.microsoft.com/whdc/winlogo/getstart/testing.mspx as the starting point. Otherwise, I believe Orin went through the WHQL process, so he might be able to give some pointers. Regards, /Pete ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ Libusb-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/libusb-devel |
|
Pete Batard wrote:
> Other devices, like an iPod where the default USBSTOR driver was > replaced with WinUSB will on the other hand use WinUSB regardless > of the slot. iPod has a serial number set right? Then I would expect Windows to remember which driver to use for that device. But what if another iPod was plugged in? //Peter ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ Libusb-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/libusb-devel |
|
On 2010.06.04 16:51, Peter Stuge wrote:
> iPod has a serial number set right? Then I would expect Windows to > remember which driver to use for that device. But what if another > iPod was plugged in? That's something I'd like to figure out as well. I'm trying to get my hands on two USB Mass Storage devices with the same VID/PID (and a valid serial) to test that. Or, if someone has two such devices, they can just use zadig to install the WinUSB driver against one, and see what happens when they plug in the other. Regards, /Pete ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ Libusb-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/libusb-devel |
|
In reply to this post by Pete Batard
On Fri, Jun 4, 2010 at 8:21 AM, Pete Batard <[hidden email]> wrote:
I'd say start by reading the docs and then ask again... last year you needed a 2003 Server to host the tests. What the latest set of tests require, I don't know, but I'd guess they still require a server OS - expensive if you don't have one already. The device must comply with the USB specs or it won't pass. I couldn't get our 1.1 device through the tests as it claimed it supported remote wakeup. The tests failed as I couldn't get the device to send the wakeup. Some form of certificate is required to sign WHQL requests. Whether the special certificate that they were selling 'cheaply' is all you need, or whether you need a code-signing certificate now, I don't know. You certainly need a code-signing certificate for drivers that load at boot time. Orin. ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ Libusb-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/libusb-devel |
|
In reply to this post by Pete Batard
Pete Batard wrote:
> On 2010.06.04 16:51, Peter Stuge wrote: > >> iPod has a serial number set right? Then I would expect Windows to >> remember which driver to use for that device. But what if another >> iPod was plugged in? >> It is a new install -- a new "devnode", in Windows terms. Windows will go back to the INF file and use the driver specified therein. If you had done registry magic on some other device, you'd have to redo it for the new device This is one common use for a "class filter" driver. Class filters get loaded with all devices of a particular class, regardless of hardware ID. The class filter can then install a device filter if needed. -- Tim Roberts, [hidden email] Providenza & Boekelheide, Inc. ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ Libusb-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/libusb-devel |
|
In reply to this post by joey-44
joey wrote:
>> > I've found the answer to my question in Jan Axelson's _USB Complete > 3rd Edition_ on page 276. Since our device is a HID which is a > supported class, windows will always match the compatible ID (which > contains the class of the device) with the class of windows generic > HID drivers rather than libusb/winusb. This is because windows HID > driver is digitally signed where as the .inf zadic.exe uses is > digitally unsigned. Even though zadic's inf contains information to > match itself with our device, windows will search for any match that > exists in its digitally signed .inf files before attempting to > associate unsigned .inf files to a device. > > To get a Microsoft digital signature it seems that you must pass a > WHQL test. What does it take to get a WHQL certification?? Time and money. You need to fetch the Windows Logo Kit, then install and run the Driver Test Manager. This is not a trivial exercise. It requires two machines, one of which (the controller) much be running Windows Server 2003 or 2008. You put your driver and your device on the other machine, and DTM runs an extensive set of automated tests for whatever class your device is in. Note that you must have both 32-bit and 64-bit drivers, and that means you'll have to sign the 64-bit driver with an "approved" class 3 code-signing certificate. Assuming you pass all the tests, DTM will help you create a submission package, including all of the passing test logs. Now, you create a Winqual account. To do that, you must have a Verisign certificate (no other brand will do), although not necessarily a code-signing certificate. You then submit your submission package to Winqual. A few days later, you'll get your driver package back with the WHQL signature. You'll also have the right to use the "Designed for Windows 7" logo on your marketing material. It costs $250 per operating system "family". 32 and 64-bit variants are all part of one "family". XP comes for free if you submit for Win 7 and ask for it. -- Tim Roberts, [hidden email] Providenza & Boekelheide, Inc. ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ Libusb-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/libusb-devel |
|
In reply to this post by joey-44
On Wed, Jun 2, 2010 at 11:30 PM, joey <[hidden email]> wrote:
> I have "gitted" the libwdi and compiled it using MinGW. I've tested it > on a virtual box running windows xp sp3 and it seems to work quite well. > When I insert a second device of the same type (different serial number, > same VID, PID, interface number) the device did not attach to > libusb/winusb. I can confirm this under a real Windows Vista 32bit box. I added serial number 1234 and switched the Jan Axelson's Generic HID Firmware to use WinUSB instead of HID driver. libwdi:debug [wdi_create_list] got hardware ID: USB\VID_0925&PID_7001&REV_0001 libwdi:debug [wdi_create_list] WinUSB USB device (6): USB\VID_0925&PID_7001\1234 libwdi:debug [wdi_create_list] Device description: 'USB Human Interface Device' I changed the serial number to 5678, downloaded the firmware and then reset the device, it comes back to use HID driver instead of WinUSB. libwdi:debug [wdi_create_list] got hardware ID: USB\VID_0925&PID_7001&REV_0001 libwdi:debug [wdi_create_list] HidUsb USB device (6): USB\VID_0925&PID_7001\5678 libwdi:debug [wdi_create_list] Device description: 'USB Human Interface Device' Next test will be trying out the Microchip USB Mass Storage Firmware. -- Xiaofan http://mcuee.blogspot.com ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ Libusb-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/libusb-devel |
|
On Sat, Jun 5, 2010 at 10:17 AM, Xiaofan Chen <[hidden email]> wrote:
> Next test will be trying out the Microchip USB Mass Storage Firmware. > Original it was using USBStor driver and the serial number is 1234. libwdi:debug [wdi_create_list] USBSTOR USB device (3): USB\VID_04D8&PID_0009\1234 libwdi:debug [wdi_create_list] Device description: 'USB Mass Storage Device' libwdi:debug [wdi_create_list] got hardware ID: USB\VID_04D8&PID_000B&REV_0000 I changed to use WinUSB. libwdi:debug [wdi_create_list] got hardware ID: USB\VID_04D8&PID_0009&REV_0001 libwdi:debug [wdi_create_list] WinUSB USB device (3): USB\VID_04D8&PID_0009\1234 libwdi:debug [wdi_create_list] Device description: 'USB Mass Storage Device' Then I changed the serial number to 5678, downloaded the firmware and then reset the device. It goes back to use the default USB Mass Storage Device. libwdi:debug [wdi_create_list] USBSTOR USB device (3): USB\VID_04D8&PID_0009\5678 libwdi:debug [wdi_create_list] Device description: 'USB Mass Storage Device' libwdi:debug [wdi_create_list] got hardware ID: USB\VID_04D8&PID_000B&REV_0000 So this is just what Windows match the best device driver to use. If Windows detects multiple drivers for the same device, Windows assigns each driver a rank that is based on several factors. Windows first compares the signing scores of the competing drivers. If one driver has the highest signing score, Windows installs that driver. That is happening here. http://www.microsoft.com/whdc/driver/install/drvsign/pnp-driver.mspx -- Xiaofan http://mcuee.blogspot.com ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ Libusb-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/libusb-devel |
|
On 2010.06.05 04:57, Xiaofan Chen wrote:
> Then I changed the serial number to 5678, downloaded the firmware and then > reset the device. It goes back to use the default USB Mass Storage Device. Thanks for these tests. I was trying to avoid having to recreate the Mass Storage firmware for my PIC18F2550 board, as the one from LakeView Research is for a 18F4550. > Windows first compares the signing > scores of the competing drivers. If one driver has the highest signing score, > Windows installs that driver. That is happening here. > > http://www.microsoft.com/whdc/driver/install/drvsign/pnp-driver.mspx Yup. And Microsoft treats two devices that only differ on the serial number as two completely different devices. Kind of defeats the whole point of asking manufacturers to use the same PIDs for devices that are identical in USB functionality. If end users installed a driver for a specific VID:PID, there's no reason to believe they don't want to use the same driver for a similar device with a different serial. I'm therefore not sure where Microsoft is coming from on this one... Looks like going through a class filter or WHQL (provided WHQL results in the same score as Microsoft's own drivers, and that the most recent driver wins) are going to be the options to avoid reinstalling the driver over and over again. Either that, or you would need to have a service running that detects USB devices as soon as they are plugged in and installs the driver, if needed, whenever there's a VID:PID match, which is similar to what the Apple iTunes service seems to be doing. Oh, and it doesn't look like Microsoft provides any official means of temporarily disabling restore points, at least on Vista and later. Moreover, disabling them manually on these platforms, while greatly speeding up the driver installation process, does result in all restore points to be deleted => not a viable option. It looks unlikely that we'll be able to speed up the installation process from libwdi then, as long as we use the official driver installation API. Regards, /Pete ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ Libusb-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/libusb-devel |
|
Pete Batard wrote:
>> If end users installed a driver for a specific VID:PID, there's no >> reason to believe they don't want to use the same driver for a similar >> device with a different serial. I'm therefore not sure where Microsoft >> is coming from on this one... I use different drivers for the same VID:PID sometimes (and perhaps I'm wrong not to change the PID, but here goes). I have two FT2232D's(*), and I simultaneously use one with the virtual COM port driver, while I use the other one with libftdi (libusb-win32-0.1 driver). Both chips have the same VID:PID. Note that for the former, I can't even easily obtain the VID, PID, or serial number (it isn't enumerated that way), so the fact that I use the serial primarily to label the board, rather than to label the desired mode of operation, is still ok. I'll grant that I can change the VID:PID, but I haven't, as that's just more stuff that I could potentially get wrong every time I switch things around (the hardware on the two boards is identical), and also because I treat the PID as a hierarchy level above the serial(**). This is perhaps an obscure situation, and maybe not even applicable to people who wish to giveaway/sell devices to strangers, but Microsoft's decision helps me in this case. Could I have worked around it if they had done otherwise? Yes, but it'd have introduced an unwelcome complication in my host software. Michael (*) The exact model is unimportant to the example, but that's my specific case. Additionally, there are usually more than 2, but that's also irrelevant. (**) The host software assumes that all devices enumerated with that VID:PID are capable of some similar basic communication (primarily returning extra identification and status info), and yes I can enforce what devices are physically plugged in. ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ Libusb-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/libusb-devel |
|
In reply to this post by Pete Batard
On Mon, Jun 7, 2010 at 7:23 AM, Pete Batard <[hidden email]> wrote:
> Oh, and it doesn't look like Microsoft provides any official means of > temporarily disabling restore points, at least on Vista and later. Driver installation can potential go wrong, so I think it is good that Microsoft does not allow disabling restore points temporarily. > Moreover, disabling them manually on these platforms, while greatly > speeding up the driver installation process, does result in all restore > points to be deleted => not a viable option. > It looks unlikely that we'll be able to speed up the installation > process from libwdi then, as long as we use the official driver > installation API. I think this should not be a real issue for most users. For end users, it is not expected that they switch drivers from time to time. For manufacturers who are doing final testing (using same driver but with unique serial number), there is a trick to ignore the serial number (so that Windows does not prompt driver installation for every device). http://www.microchip.com/forums/fb.aspx?m=385062 Not so sure if this trick to ignore serial number work for this case. -- Xiaofan http://mcuee.blogspot.com ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ Libusb-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/libusb-devel |
|
On 2010.06.07 02:39, Xiaofan Chen wrote:
> http://www.microchip.com/forums/fb.aspx?m=385062 > > Not so sure if this trick to ignore serial number work for this case. Brilliant find once again! I'm pretty sure the IgnoreHWSerNumVVVVPPPP key was made just for this kind of situation (Microsoft probably knew their default choice was disputable), and I'm ready to bet the trick still works on recent platforms. If confirmed, I'll probably add an option to set the key in libwdi during driver installation, if people don't want to get prompted for devices that just differ from the serial. Do you have means to find out if that key works on Vista (or Windows 7) with your USBSTOR PIC device? Regards, /Pete ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ Libusb-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/libusb-devel |
|
^ Of course, this won't change the driver "election" system when plugged
on a different port, but that's a start. One has to wonder if there might exist registry entries to tweak the driver eligibility as well... /Pete ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ Libusb-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/libusb-devel |
|
Apparently, there's also a global flag to ignore all USB serial numbers:
(See the end of http://www.uwe-sieber.de/usbtrouble_e.html) [HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\UsbFlags] there is the value GlobalDisableSerNumGen: 0 disable all USB serials 1 enables USB serials Lakeview Research also mentions similar information: http://www.lvr.com/usbfaq.htm -> "How can I prevent Windows from asking to install a new driver every time a device is attached?" Some posts however seem to point out that this key is used to disable the generation of serials for devices that don't have one, rather than disable USB serial numbers globally, which would make sense with regards to the choice of key name: http://www.tech-archive.net/Archive/Development/microsoft.public.win32.programmer.kernel/2004-12/0109.html. It also appears that at least the USBSTOR driver has some provisions for tweaking, with usbstor.inf doing the following for iPod devices (05AC = "Apple Inc.", 12xx, 13xx = iPod, with 1300 being the PID of an iPod Shuffle for instance) [DeviceSpecific] HKLM,"System\CurrentControlSet\Control\usbstor\05AC12xx", "DeviceHackFlags", 0x00010003, 0x20 HKLM,"System\CurrentControlSet\Control\usbstor\05AC13xx", "DeviceHackFlags", 0x00010003, 0x20 Of course none of the DeviceHackFlags values are documented, and, while I have the default set to 0x20 on my machine, changing it to 0 and plugging an iPod didn't result in any noticeable change. Regards, /Pete ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ Libusb-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/libusb-devel |
|
In reply to this post by Pete Batard
On Mon, Jun 7, 2010 at 7:56 PM, Pete Batard <[hidden email]> wrote:
> I'm pretty sure the IgnoreHWSerNumVVVVPPPP key was made just for this > kind of situation (Microsoft probably knew their default choice was > disputable), and I'm ready to bet the trick still works on recent platforms. > If confirmed, I'll probably add an option to set the key in libwdi > during driver installation, if people don't want to get prompted for > devices that just differ from the serial. > > Do you have means to find out if that key works on Vista (or Windows 7) > with your USBSTOR PIC device? > Somehow it is not listed in the "usbflags" category. And I am not so sure what lists under "usbstor"means. -- Xiaofan http://mcuee.blogspot.com ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ Libusb-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/libusb-devel |
|
On Mon, Jun 7, 2010 at 10:29 PM, Xiaofan Chen <[hidden email]> wrote:
> On Mon, Jun 7, 2010 at 7:56 PM, Pete Batard <[hidden email]> wrote: >> I'm pretty sure the IgnoreHWSerNumVVVVPPPP key was made just for this >> kind of situation (Microsoft probably knew their default choice was >> disputable), and I'm ready to bet the trick still works on recent platforms. >> If confirmed, I'll probably add an option to set the key in libwdi >> during driver installation, if people don't want to get prompted for >> devices that just differ from the serial. >> >> Do you have means to find out if that key works on Vista (or Windows 7) >> with your USBSTOR PIC device? >> > > Somehow it is not listed in the "usbflags" category. And I am not > so sure what lists under "usbstor"means. > Hmm, it works. After creating a new string IgnoreHWSerNum04d80009 with value of 1 under usbflags, it does seem to ignore the serial number and continues to use the WinUSB driver even though I change the serial number of the firmware. This is under Win 7 X64 Home Premium. -- Xiaofan http://mcuee.blogspot.com ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ Libusb-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/libusb-devel |
|
On 2010.06.07 15:45, Xiaofan Chen wrote:
> Hmm, it works. After creating a new string > IgnoreHWSerNum04d80009 with value of 1 under usbflags, > it does seem to ignore the serial number and continues to use the > WinUSB driver even though I change the serial number > of the firmware. > > This is under Win 7 X64 Home Premium. Thanks a lot! This is behaving as expected, and I think it'll be worthwhile to provide the option to set this key automatically in libwdi. Regards, /Pete ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ Libusb-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/libusb-devel |
|
IgnoreHWSerNumVVVVPPPP does exactly what I needed.
Thank You Pete and Xiaofan for all you amazing help. In the near future I'll be doing some testing on windows 7 32 and 64 bit, so I'll be able to see if this works for those as well. Joseph, ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ Libusb-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/libusb-devel |
| Powered by Nabble | Edit this page |
