hbirr(a)svn.reactos.com wrote:
>- Reference the file object in IopSecurityFile.
>- Don't set FO_DIRECT_DEVICE_OPEN in IoCreateStreamFileObject.
>- Disabled the setting of IRP_NOCACHE, because vfat cannot handle cached and non cached requests for the same file.
>- Set the correct device object in NtWriteFile.
>
>
>
>
>
>
Hi,
This change breaks usetup. As I said, IoCreateStreamFileObject must be
left with the current hack intact.
Best regards,
Alex Ionescu
About a week or two ago usetup took about 20 seconds on my system under
qemu to copy over the installation files. I thought even that seemed a
bit slow, so I started rewriting the code to use memory mapped files to
see if I could speed it up. Now that I finished with it, I noticed that
the current code in the repository now takes just over 90 seconds to
install the files, which my changes reduce back down to 20 seconds.
Does anyone have any idea about what revision this huge slowdown was
introduced in? I'll try and figure out the problem tonight but it would
help if someone could point me towards the vicinity of it now so I don't
have to spend time looking when I get home.
I'd like to get the normal code back to it's former 20 second install
area so I can get a good comparison with my new code. I'd be thrilled
if it then could get the install times down to 10 seconds.
Just recently, (cd "Program Files") has stopped working. It works
without the quotes, but when you tab complete the directory name, it
quotes it for you automatically. Anyone touch this recently?
WD
--
"<arty> the tree was touched a lot recently"
Thanks to Alex and Arty! 15407 & 15408 fix the problems I was seeing
with ftp. Please check your networking apps again!
WD
--
"<arty> the tree was touched a lot recently"
The changes in this revision broke building, I get this:
ntoskrnl: [CC] io/pnpmgr.c
io/pnpmgr.c: In function `IoGetDeviceProperty':
io/pnpmgr.c:63: warning: 'Data' might be used uninitialized in this function
mingw32-make[1]: *** [io/pnpmgr.o] Error 1
mingw32-make: *** [ntoskrnl] Error 2