In the wiki page related to Known problems for R..2.7 -->>
http://www.reactos.com/wiki/index.php?title=Known_Problems&curid=786&diff=0…
, it is mentionned that "The install CD wont install reactos in some
computers. LiveCD might boot in the same computers without problems"
I cannot Boot using Boocd based on svn 17600 and created using the
recommended "Built Environment" as indicated in the Wiki -->>
http://blight.reactos.at/reactos-be/ReactOS%20Build%20Environment%200.1-3.4…
The problem is reproductible on my 2 computers in real hardware , as per
debug messages below.
=> Is it the same problem ?
(ntoskrnl\ldr\loader.c:252) Could not open module file:
\SystemRoot\system32\drivers\sndblst.sys
(ntoskrnl\ldr\loader.c:252) Could not open module file:
\SystemRoot\system32\drivers\mpu401.sys
(registry.c:220) Can't read registry: c0000034
(registry.c:231) Manually set defaults
(ntoskrnl\ldr\loader.c:252) Could not open module file:
\SystemRoot\system32\drivers\xboxvmp.sys
Packet: DriverEntry
(ndis/protocol.c:678)(NdisRegisterProtocol) Called.
(ndis/protocol.c:685)(NdisRegisterProtocol) NDIS 3 protocol attempting
to register
(ndis/protocol.c:764)(NdisRegisterProtocol) Opening configuration key:
\Registry\Machine\System\CurrentControlSet\Services\PacketDriver\Linkage
(ndis/protocol.c:773)(NdisRegisterProtocol) Unable to open protocol
configuration
NPF: Failed to register protocol with NDIS
(ntoskrnl\ob\object.c:157) Invalid Size: 18 or Attributes: 1
(ntoskrnl\ob\object.c:243) Failed to capture, cleaning up
(ntoskrnl\ob\object.c:825) Capture failed
(init.c:108) SM: InitSessionManager: failed to create \SmApiPort
(Status=c000000d)
(ntoskrnl\ob\wait.c:246) Returning: 1
KeBugCheckEx at ntoskrnl\ex\init.c:716
A problem has been detected and ReactOS has been shut down to prevent
damage to your computer.
Technical information:
*** STOP: 0x00000071 (0x00000001,0x00000000,0x00000000,0x00000000)
Frames:
<ntoskrnl.exe:24c6 (ntoskrnl/ke/bug.c:498 (KeBugCheckEx))>
<ntoskrnl.exe:b8a5f (ntoskrnl/ex/init.c:716 (ExpInitializeExecutive))>
<ntoskrnl.exe:6706 (ntoskrnl/ke/main.c:100 (KiSystemStartup))>
<ntoskrnl.exe:b684e (ntoskrnl/ke/main.c:294 (_main))>
<ntoskrnl.exe:104b ({standard input}:47 (_section_alignment__))>
Regards
Gge
> From: gdalsnes(a)svn.reactos.com
>
> Updated files:
> trunk/reactos/subsys/win32k/include/win32k.h
Perhaps you forgot to commit a file?
[PCH] obj-i386/subsys/win32k/w32k.h.gch
In file included from subsys/win32k/w32k.h:44:
subsys/win32k/include/win32k.h:12:28: include/ntuser.h: No such file or
directory
make[1]: *** [obj-i386/subsys/win32k/w32k.h.gch] Error 1
> From: ion(a)svn.reactos.com
>
> - Fix kernel32 and ntoskrnl build issues.
> - Define public version of DEVOBJ_EXTENSION in DDK.
Ok, I have to wonder, did you bother to do ANY testing on your
DEVOBJ_EXTENSION/EXTENDED_DEVOBJ_EXTENSION changes at all? Like, uhm, just
trying to boot? 'cause it seems very unlikely to me that you were able to
successfully boot. IoCreateDevice allocates only enough room for
DEVOBJ_EXTENSION, but the allocated pointer is cast to
PEXTENDED_DEVOBJ_EXTENSION (a much larger struct) all over the place, so
we're reading/writing outside allocated memory in a zillion places.
Gé van Geldorp.
hpoussin(a)svn.reactos.com wrote:
>Implement SetComputerNameExA/W
>
>
>Updated files:
>trunk/reactos/lib/kernel32/misc/computername.c
>trunk/reactos/lib/kernel32/misc/stubs.c
>trunk/reactos/w32api/include/winbase.h
>
>_______________________________________________
>Ros-svn mailing list
>Ros-svn(a)reactos.com
>http://reactos.com:8080/mailman/listinfo/ros-svn
>
>
>
Hi,
this implementation breaks installing reactos. It does only accept 3 of
8 types to setting the computer name. The setup uses
ComputerNameNetBIOS, which isn't implemented.
- Hartmut
hpoussin(a)svn.reactos.com wrote:
>Allow compilation of fs_rec driver with MSVC
>
>
>
>
>
>+#ifdef _MSC_VER
>+#define STDCALL
>+#endif
>+
>
>
This isn't really correct...what are you trying to do?
Best regards,
Alex Ionescu
First, some stats:
In August there were:
* 676 Commits
* 49 bugs opened
* 139 bugs touched
* 44 bugs closed
Looking over the logs, here are some of the big and/or highly visible changes:
*sedwards bought in some winetests (in regtests/winetests)
This resulted in some immediate fixes
*navaraf replaced the pesum tool with pefixup
*sedwards brought dosfsck in late July, and enabled writing in August.
Still needs testing, but seems to work.
*brandonturner implemented windows style tab completion, replacing the
bash style we had before. I believe this can be toggled. Also, cmd
has really been getting attention from brandonturner and greatlord,
may little bugs have been fixed. Thanks guys!
*ion removed the third-party driver info from the hives and placed
them in the wiki.
I felt this was a good move, since it will motivate people to
fix/implement driver installation. The interim solution was going to
be creating .reg files to import or just re-add the info to your local
hives. This had the side affect of breaking our vmware video driver
installer. After some bugfixes, vmwinst now loads the driver, but
ReactOS seems unable to set the resolution.
*navaraf fixed the i8042prt detection, partially fixing the
AllocConsole bug. This was merged to 0.2.7 and we shipped it as a
known bug. We have received just a few reports of the AllocConsole
bug on IRC, so it seems to work for most.
*navaraf fixed the translation of extended keys, fixing my bug
hilighting text with Shift+Arrow keys. Merged to 0.2.7, thanks Filip!
*ion twiddled the priority code, fixing some winetests
* ea returned to work on subsystem code
* gdalsnes fixed ftol, progressing Abiword 2.0.3
* ion corrected the LPC Message Structures
* amunger fixed the freeldr.ini Timeout ;0)
* hbirr fixed several bugs that were blocking selfhosting (ros-on-ros
compilation)
I am able to compile ros on ros with the patch from the selfhosting
bug. Thanks Hartmut!
* We found that release builds of msi.dll allow the VMWare Tools
installer to run, several bugs were fixes in relation to this. Thanks
everyone!
*fireball got USB keyboards working on the XBOX, along with various USB fixes
* ion cleaned up the NDK headers
* gvg added the autogenerated DIB code for 32 and 8 bpp
* navaraf reimplemented the HAL DMA routines
* There was some work on the MSVC backend, thanks to sedwards and royce
* navaraf fixed some bugs in NDIS
* navaraf tweaked the explorer taskbar, fixing the misaligned buttons
under windows
This exposes a bug in ReactOS, so the taskbar looks a little off now.
17604 seems like a good way to start September:
* shutdown bug fixed
* Default Abiword 2.2.9 install works
* Abiword 2.2.9 runs and opens documents.
* 7-Zip 4.23 Installer works
* 7-Zip GUI has several issues
* 7za.exe (cli) works
* OO installer unpacks, but bombs a few pages in
* notepad opens and saves files properly
* winemine doesn't work
* solitare works
* ps works
* mc seems like it has issues (I'm not a user per se.)
* format still doesn't work
* calc doesn't even start
* cat works
* regedit can set values
* wallpaper stays if set in registry, not if using desk.cpl
Networking:
* qemu networking fixed
* ftp fixed
* ping works
* ipconfig has issues
* finger works
* "route print" works
* netstat doesn't work
* ncftp has issues
* hostname works
* arp doesn't work
* mIRC installer works, fails creating icons (or start menu I don't remember)
* mIRC has several issues, can't get it to connect
and the usual implementing new functions, syncing to wine, and bugfixes.
Overall 17604 "feels" good and stable. I was able to use the command
line without too much trouble, installed many packages, and only saw a
few BSODs, mainly in the network code.
I think that is a pretty good summary, and since it's been the day of
long emails, I won't go any further.
Please reply to this email if there is something I forgot, overlooked,
got wrong, etc..
WD
--
#irc.freenode.net #reactos
01:03PM <filip2307> i don't know about any bug
01:04PM <filip2307> none exist
01:04PM <filip2307> ReactOS is prefect
Hi all,
After doing some regression tests with SVN ISOs,
it appears some are missing/badly numbered.
Here is the updated list.
ReactOS-trunk-rxxxxx.iso -> real revision
ReactOS-trunk-r17465.iso -> r17560
ReactOS-trunk-r17466.iso -> r17539
ReactOS-trunk-r17468.iso -> r17559
ReactOS-trunk-r17471.iso -> r17567
ReactOS-trunk-r17472.iso -> r17553
ReactOS-trunk-r17473.iso -> r17565
ReactOS-trunk-r17478.iso -> r17552
ReactOS-trunk-r17479.iso -> r17569
ReactOS-trunk-r17490.iso -> r17572
ReactOS-trunk-r17495.iso -> r17553
These files did show the same version (md5sum proved they are different)
ReactOS-trunk-r17480.iso -> r17564
ReactOS-trunk-r17482.iso -> r17564
ReactOS-trunk-r17476.iso -> r17566
ReactOS-trunk-r17483.iso -> r17566
ReactOS-trunk-r17489.iso -> r17566
Kind regards,
Usurp (aka Sylvain Petreolle)
humans are like computers,
yesterday the BIOS was all
- today its just a word
Hey all,
I am thinking about starting, or helping write the SMSS and/or LSASS
implementation on ReactOS. I was wondering what I can help with, or
help someone to do the work...
Rick Langschultz
rlangschultz(a)cox.net (Home)
rlangschultz(a)ellemaespa.com (Work)
rlangschultz(a)email.uophx.edu (School)
navaraf(a)svn.reactos.com wrote:
>Check for Request == NULL.
>
>Modified: trunk/reactos/ntoskrnl/lpc/reply.c
>
>
> ------------------------------------------------------------------------
> *Modified: trunk/reactos/ntoskrnl/lpc/reply.c*
>
>--- trunk/reactos/ntoskrnl/lpc/reply.c 2005-09-02 11:29:40 UTC (rev 17603)
>+++ trunk/reactos/ntoskrnl/lpc/reply.c 2005-09-02 13:12:44 UTC (rev 17604)
>@@ -263,6 +263,12 @@
>
> Request = EiDequeueMessagePort(Port);
> KeReleaseSpinLock(&Port->Lock, oldIrql);
>
>
>
>+ if (Request == NULL)
>+ {
>+ ObDereferenceObject(Port);
>+ return STATUS_UNSUCCESSFUL;
>+ }
>+
>
>
> if (Request->Message.u2.s2.Type == LPC_CONNECTION_REQUEST)
> {
> PORT_MESSAGE Header;
>
>
Hi,
I think that this fix doesn't solve the real problem. The caller waits
on a semaphore and the semaphore is only signaled if a message is in the
queue. This means, the caller can't get a NULL pointer from
EiDequeueMessagePort(). The real problem is anywhere in
KeWaitForSingleObject, which does return STATUS_USER_APC for a non
alertable wait. That is wrong.
- Hartmut
