Ros-dev

ros-dev@reactos.org

5791 discussions

ping Alex regarding log2() for scheduler
by Royce Mitchell III 30 Mar '05

30 Mar '05

Alex, I managed to shave off another comparison in favor of a shift: int highest_bit ( int i ) { int ret = 0; if ( i > 0xffff ) i >>= 16, ret = 16; if ( i > 0xff ) i >>= 8, ret += 8; if ( i > 0xf ) i >>= 4, ret += 4; if ( i > 0x3 ) i >>= 2, ret += 2; return ret + (i>>1); } FWIW, rdtsc reports that compile for speed is 2x as fast as compile for size. I've thought about ways to possibly optimize pipeline throughput for this, but too many of the calculations depend on results of immediately previous calculations to get any real advantage. Here's updated micro-code for BSR, if you really think AMD & Intel will actually do it... IF r/m = 0 THEN ZF := 1; register := UNDEFINED; ELSE ZF := 0; register := 0; temp := r/m; IF OperandSize = 32 THEN IF (temp & 0xFFFF0000) != 0 THEN temp >>= 16; register |= 16; FI; FI; IF (temp & 0xFF00) != 0 THEN temp >>= 8; register |= 8; FI; IF (temp & 0xF0) != 0 THEN temp >>= 4; register |= 4; FI; IF (temp & 0xC) != 0 THEN temp >>= 2; register |= 2; FI; register |= (temp>>1); FI;

6 14

New proposed frontpage for reactos.com
by Jason Filby 29 Mar '05

29 Mar '05

Hi all A new proposed frontpage for reactos.com can be found here: http://reactos.com/newsite/reactos_index.html None of the links go anywhere yet. ezPublish will be phased out and most of the content hosted in a wiki (editable only by people trusted in the project). The wiki/forums/blogs would require a bit of a face-lift so as to look similar. Comments? Cheers Jason

10 13

Submit your changes fpr RC2
by Robert Köpferl 29 Mar '05

29 Mar '05

In ca. 16 h I will do SVN update again for the build 0.2.6 RC2 So think about, wether your patch should be submittet now or later. How about your feelings? Will RC2 be == final actually RC2 --------- So this is what I wanted to send yesterday. However my smtp-provider didn't work as I would have liked to. Thus I'm doing the LABEL now. __now

1 0

Re: [ros-dev] Service Control Manager, Plug and Pray, SMB and RPC
by Steven Edwards 29 Mar '05

29 Mar '05

--- Eric Kohl <eric.kohl(a)t-online.de> wrote: > IMHO, adding missing features to WIDL and fixing resulting bugs in > rpcrt4 is much more important right now. Access to remote machines, new > transports (tcp, udp, http, etc.) and a dynamic endpoint mapper (aka > rpcss) is not that important. Would it be possible to go ahead and port some of the Wine advapi32 SCM code? Its able to install and run services such as the Windows Installer so that might solve a few problems we are facing with the Msi service and CoLinux. Thanks Steven __________________________________ Do you Yahoo!? Yahoo! Small Business - Try our new resources site! http://smallbusiness.yahoo.com/resources/

7 14

RE: [ros-dev] ping Alex regarding log2() for scheduler
by Waldo Alvarez Cañizares 29 Mar '05

29 Mar '05

Can't that one be transformed to do a Bit Scan Reverse? I guess there are alogorithms for it too. Try near the link I provided. Or try the Angner Fog's optimization manual. If that does not works you could try SuperOptimizer. Is a program that finds algorithms in assembler for some arch (x86 included of course) by means of trying instructions using bruteforce (you can use several computers if you want), usually can be converted back to a HLL. This function is little enought to let superoptimizer work. Find it, is open source software, a little complex and MACROfied piece of code but somehow modifiable to suit your needs. Best Regards Waldo Alvarez ________________________________ From: ros-dev-bounces(a)reactos.com on behalf of Royce Mitchell III Sent: Fri 3/25/2005 4:19 PM To: ReactOS Development List Subject: Re: [ros-dev] ping Alex regarding log2() for scheduler Waldo Alvarez Cañizares wrote: >but if you are searching a fast BSF replacement... > > Unfortunately we have no need of BSF :( Is there a similar algo for BSR, which is what we need? _______________________________________________ Ros-dev mailing list Ros-dev(a)reactos.com http://reactos.com:8080/mailman/listinfo/ros-dev

1 0

call for an atapi tester
by Hartmut Birr 29 Mar '05

29 Mar '05

Hi, does anyone own a motherboard with an intel chipset and with an ICH5/6 io controller hub? I've modified atapi a little bit. Atapi may detect the second ide (SATA) controller of an ICH5/6. I'm interested on the debug output from atapi while booting. - Hartmut M:\Sandbox\ros_work\reactos>set SVN_EDITOR=notepad M:\Sandbox\ros_work\reactos>d:\programme\subversion\bin\svn.exe diff drivers\storage\atapi\atapi.c Index: drivers/storage/atapi/atapi.c =================================================================== --- drivers/storage/atapi/atapi.c (revision 14209) +++ drivers/storage/atapi/atapi.c (working copy) @@ -123,7 +123,15 @@ ULONG Dummy; } UNIT_EXTENSION, *PUNIT_EXTENSION; -PCI_SLOT_NUMBER LastSlotNumber; +#if 1 +/* + * FIXME: + * ScsiPortGetDeviceBase must check if a resource is already in use. + * This needs to implement IoReportResourceUsage or IoReportResourceForDetection. + */ +ULONG CommandPortArray[64]; +ULONG CommandPortCount = 0; +#endif #ifdef ENABLE_NATIVE_PCI typedef struct _PCI_NATIVE_CONTROLLER @@ -388,8 +396,6 @@ InitData.DeviceId = 0; InitData.DeviceIdLength = 0; - LastSlotNumber.u.AsULONG = 0xFFFFFFFF; - Status = ScsiPortInitialize(DriverObject, RegistryPath, &InitData, @@ -437,6 +443,27 @@ #ifdef ENABLE_DMA ULONG Length; #endif +#if 1 + /* + * FIXME: + * ScsiPortGetDeviceBase must check if a resource is already in use. + * This needs to implement IoReportResourceUsage or IoReportResourceForDetection. + */ + ULONG i; + + for (i = 0; i < CommandPortCount; i++) + { + if (CommandPortArray[i] == CommandPortBase) + { + break; + } + } + if (i < CommandPortCount || + CommandPortCount >= 64) + { + return FALSE; + } +#endif IoAddress = ScsiPortConvertUlongToPhysicalAddress(CommandPortBase); IoBase = ScsiPortGetDeviceBase((PVOID)DevExt, InterfaceType, @@ -534,6 +561,14 @@ { ConfigInfo->AtdiskSecondaryClaimed = TRUE; } +#if 1 + /* + * FIXME: + * ScsiPortGetDeviceBase must check if a resource is already in use. + * This needs to implement IoReportResourceUsage or IoReportResourceForDetection. + */ + CommandPortArray[CommandPortCount++] = CommandPortBase; +#endif return TRUE; } @@ -559,7 +594,7 @@ BOOLEAN DeviceFound; ULONG BusMasterBasePort = 0; - DPRINT("AtapiFindCompatiblePciController() Bus: %lu Slot: %lu\n", + DPRINT1("AtapiFindCompatiblePciController() Bus: %lu Slot: %lu\n", ConfigInfo->SystemIoBusNumber, ConfigInfo->SlotNumber); @@ -600,10 +635,10 @@ } } - DPRINT("%x %x\n", PciConfig.BaseClass, PciConfig.SubClass); - if (PciConfig.BaseClass == 0x01 && - PciConfig.SubClass == 0x01) // && -// (PciConfig.ProgIf & 0x05) == 0) + DPRINT("%x %x %x\n", PciConfig.BaseClass, PciConfig.SubClass, PciConfig.ProgIf); + if (PciConfig.BaseClass == PCI_CLASS_MASS_STORAGE_CTLR && + PciConfig.SubClass == PCI_SUBCLASS_MSC_IDE_CTLR && + (PciConfig.ProgIf & 0x05) == 0) { /* both channels are in compatibility mode */ DPRINT("Bus %1lu Device %2lu Func %1lu VenID 0x%04hx DevID 0x%04hx\n", @@ -692,7 +727,7 @@ BOOLEAN ChannelFound = FALSE; BOOLEAN DeviceFound = FALSE; - DPRINT("AtapiFindIsaBusController() called!\n"); + DPRINT1("AtapiFindIsaBusController() called!\n"); *Again = FALSE; @@ -770,7 +805,7 @@ ULONG Count; BOOLEAN ChannelFound; - DPRINT("AtapiFindNativePciController() called!\n"); + DPRINT1("AtapiFindNativePciController() called!\n"); SlotNumber.u.AsULONG = ConfigInfo->SlotNumber; StartDeviceNumber = SlotNumber.u.bits.DeviceNumber; @@ -799,9 +834,30 @@ break; } } - if (Count < sizeof(PciNativeController)/sizeof(PCI_NATIVE_CONTROLLER)) + if (PciConfig.BaseClass == 0x01) { - /* We have found a known native pci ide controller */ + DPRINT1("Bus %1lu Device %2lu Func %1lu VenID 0x%04hx DevID 0x%04hx\n", + ConfigInfo->SystemIoBusNumber, + SlotNumber.u.bits.DeviceNumber, + SlotNumber.u.bits.FunctionNumber, + PciConfig.VendorID, + PciConfig.DeviceID); + DPRINT1("BaseClass %02x, SubClass %02x, ProgIF %02x\n", PciConfig.BaseClass, PciConfig.SubClass, PciConfig.ProgIf); + DPRINT1("BaseAddresses[0] %08x\n", PciConfig.u.type0.BaseAddresses[0]); + DPRINT1("BaseAddresses[1] %08x\n", PciConfig.u.type0.BaseAddresses[1]); + DPRINT1("BaseAddresses[2] %08x\n", PciConfig.u.type0.BaseAddresses[2]); + DPRINT1("BaseAddresses[3] %08x\n", PciConfig.u.type0.BaseAddresses[3]); + DPRINT1("BaseAddresses[4] %08x\n", PciConfig.u.type0.BaseAddresses[4]); + DPRINT1("BaseAddresses[5] %08x\n", PciConfig.u.type0.BaseAddresses[5]); + DPRINT1("InterruptLine %d\n", PciConfig.u.type0.InterruptLine); + } + + if (Count < sizeof(PciNativeController)/sizeof(PCI_NATIVE_CONTROLLER) || + (PciConfig.BaseClass == PCI_CLASS_MASS_STORAGE_CTLR && + PciConfig.SubClass == PCI_SUBCLASS_MSC_IDE_CTLR && + (PciConfig.ProgIf & 0x05) == 0x05)) + { + /* We have found a known or a native pci ide controller */ if ((PciConfig.ProgIf & 0x80) && (PciConfig.u.type0.BaseAddresses[4] & PCI_ADDRESS_IO_SPACE)) { DPRINT("Found IDE Bus Master controller!\n"); @@ -813,40 +869,24 @@ BusMasterBasePort = 0; } - DPRINT("VendorID: %04x, DeviceID: %04x\n", PciConfig.VendorID, PciConfig.DeviceID); + DPRINT1("VendorID: %04x, DeviceID: %04x\n", PciConfig.VendorID, PciConfig.DeviceID); ConfigInfo->NumberOfBuses = 1; ConfigInfo->MaximumNumberOfTargets = 2; ConfigInfo->MaximumTransferLength = 0x10000; /* max 64Kbyte */ - /* FIXME: - We must not store and use the last tested slot number. If there is a recall - to the some device and we will claim the primary channel again than the call - to ScsiPortGetDeviceBase in AtapiClaimHwResource will fail and we can try to - claim the secondary channel. - */ ChannelFound = FALSE; - if (LastSlotNumber.u.AsULONG != SlotNumber.u.AsULONG) + /* try to claim primary channel */ + if ((PciConfig.u.type0.BaseAddresses[0] & PCI_ADDRESS_IO_SPACE) && + (PciConfig.u.type0.BaseAddresses[1] & PCI_ADDRESS_IO_SPACE)) { - /* try to claim primary channel */ - if ((PciConfig.u.type0.BaseAddresses[0] & PCI_ADDRESS_IO_SPACE) && - (PciConfig.u.type0.BaseAddresses[1] & PCI_ADDRESS_IO_SPACE)) - { - /* primary channel is enabled */ - ChannelFound = AtapiClaimHwResources(DevExt, - ConfigInfo, - PCIBus, - PciConfig.u.type0.BaseAddresses[0] & PCI_ADDRESS_IO_ADDRESS_MASK, - PciConfig.u.type0.BaseAddresses[1] & PCI_ADDRESS_IO_ADDRESS_MASK, - BusMasterBasePort, - PciConfig.u.type0.InterruptLine); - if (ChannelFound) - { - AtapiFindDevices(DevExt, ConfigInfo); - *Again = TRUE; - ConfigInfo->SlotNumber = LastSlotNumber.u.AsULONG = SlotNumber.u.AsULONG; - return SP_RETURN_FOUND; - } - } + /* primary channel is enabled */ + ChannelFound = AtapiClaimHwResources(DevExt, + ConfigInfo, + PCIBus, + PciConfig.u.type0.BaseAddresses[0] & PCI_ADDRESS_IO_ADDRESS_MASK, + PciConfig.u.type0.BaseAddresses[1] & PCI_ADDRESS_IO_ADDRESS_MASK, + BusMasterBasePort, + PciConfig.u.type0.InterruptLine); } if (!ChannelFound) { @@ -862,21 +902,21 @@ PciConfig.u.type0.BaseAddresses[3] & PCI_ADDRESS_IO_ADDRESS_MASK, BusMasterBasePort ? BusMasterBasePort + 8 : 0, PciConfig.u.type0.InterruptLine); - if (ChannelFound) - { - AtapiFindDevices(DevExt, ConfigInfo); - *Again = FALSE; - LastSlotNumber.u.AsULONG = 0xFFFFFFFF; - return SP_RETURN_FOUND; - } } } + if (ChannelFound) + { + CHECKPOINT1; + AtapiFindDevices(DevExt, ConfigInfo); + *Again = TRUE; + ConfigInfo->SlotNumber = SlotNumber.u.AsULONG; + return SP_RETURN_FOUND; + } } } StartFunctionNumber = 0; } *Again = FALSE; - LastSlotNumber.u.AsULONG = 0xFFFFFFFF; DPRINT("AtapiFindNativePciController() done!\n"); return(SP_RETURN_NOT_FOUND);

7 7

Changing syntax style of files to ms style...
by Gregor Anich 29 Mar '05

29 Mar '05

Hi! Sorry if I sound a little harsh, but I am upset by the fact that i don't understand what has driven someone to change the style of a file when adding one line of code from the old (whatever) style to ms style (where '{' and '}' are on the same line as the if/else keywords - that makes multiline if/else blocks unreadable to me, I esp. dislike it when the if-expression is also multiline - like if (some_very_long_expression_which_does && not_fit_onto_a_signle_line) { ... } ) I am not sure, but I think it was Alex. Would you please tell me why you changed the style of the file? I am pretty much used to the old style, esp. if i have been working on the file before it's strange for me to see the syntax changed in a way which i dislike as much as the ms style or whatever it is. :-/

1 1

Re: [ros-dev] STATEMENT
by \(\) Thomas Larsen \(\) 29 Mar '05

29 Mar '05

> > Taking credits for others work is not a nice thing to todo (Thx Alot).. > > but you made it worse anyway so no hard feelings :-) > What is it about? Can you point me to time/date/subject you refer to. Just Wierd that i make one implention of taskmgr their where better than the previous then some just take some of the work modify it and credit themself.. i where planning of release the version today with network status and user tab also an vc++ comp. but can see it dont matter.. :-) > > To talk about something else i start a new projekt called CubeOs > > and i would like to know if their are any special copyrights in reactos..? > Just curious: what is the project about? Its an Reactos Clone did alot of work my self and with some developers not to make a competition but to make an alternativ with some of my ideas and visions it has allways been my dream... hope its okay with the reactos team.. By the way thx for your german translation its a littel useless now on reactos but will be very happy to use it on CubeOS Greetz Thomas

3 2

Re: [ros-dev] Service Control Manager, Plug and Pray, SMB and RPC
by Steven Edwards 29 Mar '05

29 Mar '05

Hi, --- Royce Mitchell III <royce3(a)ev1.net> wrote: > Part of the reason he's bringing this up is because I was asking what > was necessary to get working before we could dynamically load drivers > through SCM. I'm working on learning driver development, but want to do > it in ReactOS. > > Anyways, these functions look most necessary... > > OpenSCManager > CreateService > ControlService > OpenService > CloseServiceHandle > DeleteService > QueryServiceStatus > EnumServicesStatus Yes those are most of the functions I need for CoLinux and the Windows Installer service. Wine has a good bit of that implemented using a pipe for communitcation but I assume on Windows its all LPC/RPC to support "net start servicename". Anyway if no one objects I will look at importing the hackish Wine code and see if I can make it work. Thanks Steven __________________________________ Do you Yahoo!? Yahoo! Small Business - Try our new resources site! http://smallbusiness.yahoo.com/resources/

2 1

Re: [ros-dev] STATEMENT
by \(\) Thomas Larsen \(\) 29 Mar '05

29 Mar '05

Just Mean that it is not fair when i uploade taskmgr with alot of bugfixes and translation and then someone just modify the code and take the credits but now ill just use my version on my new project on CubeOs so its okay but cant see why its done.. But thx about the license stuff Greetz Thomas

2 1

Jump to page:

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Ros-dev