reiser4 breaks vmware

reiser4 breaks vmware

[Shawn]

Forgive my line-wraps, but the following (among other do_mmap_pgoff
related snippets) break vmware.

Couple questions out of this:
1. Does anyone care enough to produce a patch for vmware's module?
2. What does this change accomplish for reiser4?

diff -ruN linux-2.6.0-test9/arch/i386/kernel/sys_i386.c
linux-2.6.0-test9-reiser4/arch/i386/kernel/sys_i386.c 
--- linux-2.6.0-test9/arch/i386/kernel/sys_i386.c       Sat Oct 25
22:44:51 2003 
+++ linux-2.6.0-test9-reiser4/arch/i386/kernel/sys_i386.c       Thu Nov
13 15:39:47 2003 
@@ -56,7 +56,7 @@ 
        } 

        down_write(&current->mm->mmap_sem); 
-       error = do_mmap_pgoff(file, addr, len, prot, flags, pgoff); 
+       error = do_mmap_pgoff(current->mm, file, addr, len, prot, flags,
pgoff); 
        up_write(&current->mm->mmap_sem); 

        if (file)

Re: reiser4 breaks vmware

[Hans Reiser]

Shawn wrote:

>Forgive my line-wraps, but the following (among other do_mmap_pgoff
>related snippets) break vmware.
>
>Couple questions out of this:
>1. Does anyone care enough to produce a patch for vmware's module?
>2. What does this change accomplish for reiser4?
>
>diff -ruN linux-2.6.0-test9/arch/i386/kernel/sys_i386.c
>linux-2.6.0-test9-reiser4/arch/i386/kernel/sys_i386.c 
>--- linux-2.6.0-test9/arch/i386/kernel/sys_i386.c       Sat Oct 25
>22:44:51 2003 
>+++ linux-2.6.0-test9-reiser4/arch/i386/kernel/sys_i386.c       Thu Nov
>13 15:39:47 2003 
>@@ -56,7 +56,7 @@ 
>        } 
>
>        down_write(&current->mm->mmap_sem); 
>-       error = do_mmap_pgoff(file, addr, len, prot, flags, pgoff); 
>+       error = do_mmap_pgoff(current->mm, file, addr, len, prot, flags,
>pgoff); 
>        up_write(&current->mm->mmap_sem); 
>
>        if (file)
>-
>To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
>the body of a message to majordomo@vger.kernel.org
>More majordomo info at  http://vger.kernel.org/majordomo-info.html
>Please read the FAQ at  http://www.tux.org/lkml/
>
>
>  
>
thanks Shawn.  Nikita will answer (tomorrow).

-- 
Hans

Re: reiser4 breaks vmware

[Nuno Silva]

Hi.

Shawn wrote:
> Forgive my line-wraps, but the following (among other do_mmap_pgoff
> related snippets) break vmware.
> 
> Couple questions out of this:
> 1. Does anyone care enough to produce a patch for vmware's module?
> 2. What does this change accomplish for reiser4?
> 
> diff -ruN linux-2.6.0-test9/arch/i386/kernel/sys_i386.c
> linux-2.6.0-test9-reiser4/arch/i386/kernel/sys_i386.c 
> --- linux-2.6.0-test9/arch/i386/kernel/sys_i386.c       Sat Oct 25
> 22:44:51 2003 
> +++ linux-2.6.0-test9-reiser4/arch/i386/kernel/sys_i386.c       Thu Nov
> 13 15:39:47 2003 
> @@ -56,7 +56,7 @@ 
>         } 
> 
>         down_write(&current->mm->mmap_sem); 
> -       error = do_mmap_pgoff(file, addr, len, prot, flags, pgoff); 
> +       error = do_mmap_pgoff(current->mm, file, addr, len, prot, flags,
> pgoff); 
>         up_write(&current->mm->mmap_sem); 
> 
>         if (file)

I'd say that namesys is using UML (user-mode-linux.sf.net) to develop 
raiser4 (smart guys, uml rocks).

These are probably left overs from UML's SKAS-host patch. If I'm correct 
you may try to reverse the SKAS patch from that tree (the patch is 
located at UML's site). It won't do any harm, anyway...

Regards,
Nuno Silva

Re: reiser4 breaks vmware

[Nikita Danilov]

Nuno Silva writes:
 > Hi.
 > 
 > Shawn wrote:
 > > Forgive my line-wraps, but the following (among other do_mmap_pgoff
 > > related snippets) break vmware.
 > > 
 > > Couple questions out of this:
 > > 1. Does anyone care enough to produce a patch for vmware's module?
 > > 2. What does this change accomplish for reiser4?
 > > 
 > > diff -ruN linux-2.6.0-test9/arch/i386/kernel/sys_i386.c
 > > linux-2.6.0-test9-reiser4/arch/i386/kernel/sys_i386.c 
 > > --- linux-2.6.0-test9/arch/i386/kernel/sys_i386.c       Sat Oct 25
 > > 22:44:51 2003 
 > > +++ linux-2.6.0-test9-reiser4/arch/i386/kernel/sys_i386.c       Thu Nov
 > > 13 15:39:47 2003 
 > > @@ -56,7 +56,7 @@ 
 > >         } 
 > > 
 > >         down_write(&current->mm->mmap_sem); 
 > > -       error = do_mmap_pgoff(file, addr, len, prot, flags, pgoff); 
 > > +       error = do_mmap_pgoff(current->mm, file, addr, len, prot, flags,
 > > pgoff); 
 > >         up_write(&current->mm->mmap_sem); 
 > > 
 > >         if (file)
 > 
 > I'd say that namesys is using UML (user-mode-linux.sf.net) to develop 
 > raiser4 (smart guys, uml rocks).
 > 
 > These are probably left overs from UML's SKAS-host patch. If I'm correct 
 > you may try to reverse the SKAS patch from that tree (the patch is 
 > located at UML's site). It won't do any harm, anyway...

Exactly. I included it into core.diff by mistake.
Revert it: http://www.namesys.com/snapshots/2003.12.23/broken-out/do_mmap2-fix.diff.patch

 > 
 > Regards,
 > Nuno Silva
 > 

Nikita.

Re: reiser4 breaks vmware

[Andi Kleen]

Nikita Danilov <Nikita@Namesys.COM> writes:

> Exactly. I included it into core.diff by mistake.
> Revert it: http://www.namesys.com/snapshots/2003.12.23/broken-out/do_mmap2-fix.diff.patch

There seem to be some other unnecessary patches in there, like
init_fixmap_vma.diff.patch. I cannot imagine why a file system should
need to change that. Same with spinlock-owner.diff.patch. Is that
really needed? If yes porting it to all architectures will be a lot of
work.

I would suggest separating your debug patches, like page-owner.diff.patch

And your webserver is misconfigured: I thinks READ.ME is a troff
document.

The other changes look reasonable, although a lot of the EXPORT_SYMBOLs
should be probably EXPORT_SYMBOL_GPL and carry some more comments about
their purpose.

-Andi

Re: reiser4 breaks vmware

[Nikita Danilov]

Andi Kleen writes:
 > Nikita Danilov <Nikita@Namesys.COM> writes:
 > 
 > > Exactly. I included it into core.diff by mistake.
 > > Revert it: http://www.namesys.com/snapshots/2003.12.23/broken-out/do_mmap2-fix.diff.patch
 > 
 > There seem to be some other unnecessary patches in there, like
 > init_fixmap_vma.diff.patch. I cannot imagine why a file system should

Yes, UML left-over also, thank you for noting.

 > need to change that. Same with spinlock-owner.diff.patch. Is that
 > really needed? If yes porting it to all architectures will be a lot of

No, it is debugging code, and I think it is reasonably to ship it
together with reiser4 while it is in the debugging stage. Debugging
patches are going to be removed eventually.

 > work.
 > 
 > I would suggest separating your debug patches, like page-owner.diff.patch
 > 
 > And your webserver is misconfigured: I thinks READ.ME is a troff
 > document.

Hmm. This was fixed long time ago.

$ telnet namesys.com 80
Trying 212.16.7.65...
Connected to thebsh.namesys.com (212.16.7.65).
Escape character is '^]'.
GET /snapshots/2003.12.23/READ.ME HTTP/1.0

HTTP/1.1 200 OK
Date: Thu, 25 Dec 2003 09:20:45 GMT
Server: Apache/1.3.23 (Unix)  (Red-Hat/Linux)
Last-Modified: Tue, 23 Dec 2003 11:31:22 GMT
ETag: "e3b05-fab-3fe8278a"
Accept-Ranges: bytes
Content-Length: 4011
Connection: close
Content-Type: text/plain

 > 
 > The other changes look reasonable, although a lot of the EXPORT_SYMBOLs
 > should be probably EXPORT_SYMBOL_GPL and carry some more comments about

What are the guidelines for using EXPORT_SYMBOL vs. EXPORT_SYMBOL_GPL?

 > their purpose.
 > 
 > -Andi

Nikita.