Re: [ros-dev] [ros-diffs] [dgorbachev] 42305: Add a hack in KiSystemStartupReal function until Better Times.

[Aleksey Bragin]

That function is not supposed to be called by old bootloader  
directly. Why is that needed? And why in this function?

WBR,
Aleksey Bragin.

On Jul 30, 2009, at 9:54 PM, [e-mail] wrote:
> Author: dgorbachev
> Date: Thu Jul 30 19:54:01 2009
> New Revision: 42305
>
> URL: http://svn.reactos.org/svn/reactos?rev=42305&view=rev
> Log:
> Add a hack in KiSystemStartupReal function until Better Times.
>
> Modified:
>     trunk/reactos/ntoskrnl/ke/i386/kiinit.c
>
> Modified: trunk/reactos/ntoskrnl/ke/i386/kiinit.c
> URL: http://svn.reactos.org/svn/reactos/trunk/reactos/ntoskrnl/ke/ 
> i386/kiinit.c?rev=42305&r1=42304&r2=42305&view=diff
> ====================================================================== 
> ========
> --- trunk/reactos/ntoskrnl/ke/i386/kiinit.c [iso-8859-1] (original)
> +++ trunk/reactos/ntoskrnl/ke/i386/kiinit.c [iso-8859-1] Thu Jul 30  
> 19:54:01 2009
> @@ -712,18 +712,34 @@
>      Ke386SetDs(KGDT_R3_DATA | RPL_MASK);
>      Ke386SetEs(KGDT_R3_DATA | RPL_MASK);
> /* Save NMI and double fault traps */
> RtlCopyMemory(&NmiEntry, &Idt[2], sizeof(KIDTENTRY));
> RtlCopyMemory(&DoubleFaultEntry, &Idt[8], sizeof(KIDTENTRY));
> /* Copy kernel's trap handlers */
> RtlCopyMemory(Idt,
>              (PVOID)KiIdtDescriptor.Base,
>              KiIdtDescriptor.Limit + 1);
> /* Restore NMI and double fault */
> RtlCopyMemory(&Idt[2], &NmiEntry, sizeof(KIDTENTRY));
> RtlCopyMemory(&Idt[8], &DoubleFaultEntry, sizeof(KIDTENTRY));
> /* HACK compensates for KiRosPrepareForSystemStartup hack */
> if ((ULONG)Idt == KiIdtDescriptor.Base)
> {
>    USHORT FlippedSelector;
>    FlippedSelector = KiIdt[2].Selector;
>    KiIdt[2].Selector = KiIdt[2].ExtendedOffset;
>    KiIdt[2].ExtendedOffset = FlippedSelector;
>    FlippedSelector = KiIdt[8].Selector;
>    KiIdt[8].Selector = KiIdt[8].ExtendedOffset;
>    KiIdt[8].ExtendedOffset = FlippedSelector;
> }
> else
> {
>    /* Save NMI and double fault traps */
>    RtlCopyMemory(&NmiEntry, &Idt[2], sizeof(KIDTENTRY));
>    RtlCopyMemory(&DoubleFaultEntry, &Idt[8], sizeof(KIDTENTRY));
>    /* Copy kernel's trap handlers */
>    RtlCopyMemory(Idt,
>                  (PVOID)KiIdtDescriptor.Base,
>                  KiIdtDescriptor.Limit + 1);
>    /* Restore NMI and double fault */
>    RtlCopyMemory(&Idt[2], &NmiEntry, sizeof(KIDTENTRY));
>    RtlCopyMemory(&Idt[8], &DoubleFaultEntry, sizeof(KIDTENTRY));
> }
>
> AppCpuInit:
>      /* Loop until we can release the freeze lock */

[Dmitry Gorbachev]

> That function is not supposed to be called by old bootloader directly.

It is called indirectly through KiRosPrepareForSystemStartup function,
after it applies a hack (ke/freeldr.c:1272). This new hack is to fix
that old hack. (Ugh!)
> Why is that needed?

Otherwise, the double fault and nmi descriptors become mutilated. When
these interrupts happen, it causes a (virtual or real) machine to
crash / reboot. Better to be able to examine situation from kdbg.
> And why in this function?

Simpler to do it here.

[Timo Kreuzer]

So you are saying that you are hacking correct code, to fix a bug in a
hack, cause it's easier?

Dmitry Gorbachev wrote:
> > That function is not supposed to be called by old bootloader directly.
>
> It is called indirectly through KiRosPrepareForSystemStartup function,
> after it applies a hack (ke/freeldr.c:1272). This new hack is to fix
> that old hack. (Ugh!)
> > Why is that needed?
>
> Otherwise, the double fault and nmi descriptors become mutilated. When
> these interrupts happen, it causes a (virtual or real) machine to
> crash / reboot. Better to be able to examine situation from kdbg.
> > And why in this function?
>
> Simpler to do it here.
>
> Ros-dev mailing list
> [e-mail]
> http://www.reactos.org/mailman/listinfo/ros-dev

[Dmitry Gorbachev]

> So you are saying that you are hacking correct code, to fix a bug in a
> hack, cause it's easier?

Not quite so; the loader should be fixed, then these hacks will be gone.

[Aleksey Bragin]

It shouldn't be that hard to fix freeldr, otherwise you're breaking  
ntldr boot style, I suppose.

Should I or would you?

WBR,
Aleksey Bragin.

On Aug 1, 2009, at 3:25 AM, Dmitry Gorbachev wrote:
> > So you are saying that you are hacking correct code, to fix a bug  
> > in a
> > hack, cause it's easier?
>
> Not quite so; the loader should be fixed, then these hacks will be  
> gone.

[Dmitry Gorbachev]

> otherwise you're breaking  ntldr boot style, I suppose.

No, the new hack is applied only if the old is present.
> Should I or would you?

Why to change that old style, if it is going to be replaced by the
new? There are probably more important things to do.

[Alex Ionescu]

Please apply this hack in ke/freeldr.c
Best regards,
Alex Ionescu

On Sat, Aug 1, 2009 at 3:55 AM, Dmitry Gorbachev [e-mail]:
> > otherwise you're breaking  ntldr boot style, I suppose.
>
> No, the new hack is applied only if the old is present.
> > Should I or would you?
>
> Why to change that old style, if it is going to be replaced by the
> new? There are probably more important things to do.
>
> Ros-dev mailing list
> [e-mail]
> http://www.reactos.org/mailman/listinfo/ros-dev

[Dmitry Gorbachev]

> Please apply this hack in ke/freeldr.c

Should I apply same sh*t to different file? :) I believe that now
everything is 'right' there, nothing to change.

[Alex Ionescu]

ke/freeldr.c was designed for sh*t. It really means, translated sh/itfile.c
and I was actually hoping to commit it that way.
The other files are designed to be 100% compatible and as much as possible
identical to 2003 SP1, so please remove that hack from there.

You can put anything you want in sh/itfile.c (oops, I mean ke/freeldr.c),
including GRUB code, Wine code and Linux_kernel_functions all day long.

Best regards,
Alex Ionescu

On Sat, Aug 1, 2009 at 2:27 PM, Dmitry Gorbachev [e-mail]:
> > Please apply this hack in ke/freeldr.c
>
> Should I apply same sh*t to different file? :) I believe that now
> everything is 'right' there, nothing to change.
>
> Ros-dev mailing list
> [e-mail]
> http://www.reactos.org/mailman/listinfo/ros-dev

[Alex Ionescu]

ASCII porn of RMS is also okay.
Best regards,
Alex Ionescu

On Sat, Aug 1, 2009 at 3:26 PM, Alex Ionescu [e-mail] wrote:
> ke/freeldr.c was designed for sh*t. It really means, translated sh/itfile.c
> and I was actually hoping to commit it that way.
> The other files are designed to be 100% compatible and as much as possible
> identical to 2003 SP1, so please remove that hack from there.
>
> You can put anything you want in sh/itfile.c (oops, I mean ke/freeldr.c),
> including GRUB code, Wine code and Linux_kernel_functions all day long.
>
> Best regards,
> Alex Ionescu
>
> On Sat, Aug 1, 2009 at 2:27 PM, Dmitry Gorbachev [e-mail]:
> > > Please apply this hack in ke/freeldr.c
> >
> > Should I apply same sh*t to different file? :) I believe that now
> > everything is 'right' there, nothing to change.
> >
> > Ros-dev mailing list
> > [e-mail]
> > http://www.reactos.org/mailman/listinfo/ros-dev

[Dmitry Gorbachev]

> The other files are designed to be 100% compatible and as much as possible
> identical to 2003 SP1, so please remove that hack from there.

All right, I will change it "because I want to stop the crying of a
baby." :) This small modification receives more attention then it
deserves. One fine day, these hacks will be removed, anyway.

Alex Ionescu wrote:
> ASCII porn of RMS is also okay.

[Alex Ionescu]

Thanks dad, baby Bragin and Ionescu love you.
Best regards,
Alex Ionescu

On Sat, Aug 1, 2009 at 5:15 PM, Dmitry Gorbachev [e-mail]:
> > The other files are designed to be 100% compatible and as much as
>
> possible
> > identical to 2003 SP1, so please remove that hack from there.
>
> All right, I will change it "because I want to stop the crying of a
> baby." :) This small modification receives more attention then it
> deserves. One fine day, these hacks will be removed, anyway.
>
> Alex Ionescu wrote:
> > ASCII porn of RMS is also okay.
>
> Ros-dev mailing list
> [e-mail]
> http://www.reactos.org/mailman/listinfo/ros-dev