RE: Memory corruption when running VMware. (was File curruption when running VMware)

RE: Memory corruption when running VMware. (was File curruption when running VMware)

[Hong-Gunn Chew]

> Hi Petr, Andrea,
> 
> I have been trying different kernel configurations to 
> overcome this problem.  I found that turning on APIC seem to 
> work properly.  However turning on IO-APIC causes it to hang 
> just after X is started, which seems to be during the startup 
> of gdm. 
An added note is that with Local APIC on and IO-APIC off, 
I am able to start the system up, but swtiching from X to a virtual
console during a session, as well as during shutdown, causes the system
to hang.  There are no error messages that I can find.
Does anyone know the cause of the problem?

Recap of my system:
CPU:		P4 2.0A 2.0GHz
RAM:		4x256MB RDRAM PC800
MB:		ASUS P4-TE firmware:1005
		Intel i850
Disk:		IBM Deskstar 120GXP 80GB
Graphics:	ATI 7500 OEM

Distri:	RedHat 7.2
Kernel:	2.4.18
X:		Xfree 4.2.0
glibc:	2.2.4-19.3

Thanks,
Hong-Gunn

Re: File corruption when running VMware.

[Petr Vandrovec]

On 24 Apr 02 at 2:01, Rik van Riel wrote:
> On Wed, 24 Apr 2002, Hong-Gunn Chew wrote:
> 
> > I have a repeatable problem when running VMware workstation 3.00 and
> > 3.01.  The cause is still unknown, and could be VMware itself, the
> > hardware or the kernel.
> 
> If you can reproduce it without VMware or with only the
> open source part of VMware (ie without any of the binary
> only parts) we might have a chance of debugging it.

Hi again,
  one of 2.4.x kernel images available in SuSE's 8.0 has patched&enabled 
support for page tables in high memory, and this quickly revealed
incompatibility between VMware's vmmon page table handling and
ptes above directly mapped range.

  So if you have >890MB of RAM and your kernel is compiled with support
for pte in high memory, please stop using VMware, or reconfigure your 
kernel to not use pte in high memory (4GB config without pte-in-highmem
is OK). Using pte-in-highmem with vmmon will cause kernel oopses and/or 
memory corruption :-(

  If you do not have >890MB of memory, then reason for your memory corruption
is still unknown to me.
                                          Best regards,
                                                    Petr Vandrovec
                                                    vandrove@vc.cvut.cz
                                                    

RE: Memory corruption when running VMware. (was File curruption when running VMware)

[Hong-Gunn Chew]

Hi Petr, Andrea,

Petr wrote:
>   So if you have >890MB of RAM and your kernel is compiled 
> with support for pte in high memory, please stop using 
> VMware, or reconfigure your 
> kernel to not use pte in high memory (4GB config without 
> pte-in-highmem is OK). Using pte-in-highmem with vmmon will 
> cause kernel oopses and/or 
> memory corruption :-(

I have been trying different kernel configurations to overcome this
problem.  I found that turning on APIC seem to work properly.  However
turning on IO-APIC causes it to hang just after X is started, which
seems to be during the startup of gdm.
I do have a silly question to ask though.  How do I compile the kernel
NOT to use pte in high memory?

Andrea wrote:
> passing to the kernel mem=850M in lilo at boot will be enough.

This did not work as it causes vmware to seg fault with a kernel oops:
	Unable to handle kernel paging request at virtual address
ffffe350
	 printing eip:
	f61ff30a
	*pde = 00001063
	*pte = 00000000
	Oops: 0000
	CPU:    0
	EIP:
0010:[nls_iso8859-1:__insmod_nls_iso8859-1_S.data_L2336+139562/11435686]
Tainted: PF
	EIP:    0010:[<f61ff30a>]    Tainted: PF
	EFLAGS: 00013286
	eax: ffffe350   ebx: e7652600   ecx: ffffe350   edx: 000000d0
	esi: e7652600   edi: 00000069   ebp: e7625de8   esp: e7625de8
	ds: 0018   es: 0018   ss: 0018
	Process vmware (pid: 2595, stackpage=e7625000)
	Stack: e7625e08 f61ff34d ffffe350 c02f8ee0 e7625e2c f5095a00
000000d0 e6d21000  
	       e7625eb8 f61ff41f e7652600 e7625e8d e7625e8e e7625e8f
e7625e40 00003286  
	       00003286 efe78d80 000001f0 00000000 e7589a80 c01e330f
000000bc 000001f0
	Call Trace:
[nls_iso8859-1:__insmod_nls_iso8859-1_S.data_L2336+139629/11435619]
[nls_iso8859-1:__insmod_nls_iso8859-1_S.data_L2336+139839/11435409]
[sys_sendmsg+303/480] [wait_for_buffers+86/144]
[nls_iso8859-1:__insmod_nls_iso8859-1_S.data_L2336+137406/11437842]
	Call Trace: [<f61ff34d>] [<f61ff41f>] [<c01e330f>] [<c0134a36>]
[<f61fea9e>]
	
[nls_iso8859-1:__insmod_nls_iso8859-1_S.data_L2336+132081/11443167]
[generic_file_readahead+288/304] [bounce_end_io_read+164/288]
[isapnp_set_mem+23/272] [vfs_link+23/256] [system_call+51/56]
	   [<f61fd5d1>] [<c01272b0>] [<c0131e94>] [<c01e0527>]
[<c013e487>]  [<c0106f0b>]
	
	Code: 8b 11 89 d0 25 00 07 01 00 3d 00 04 00 00 75 12 81 ca 00
00

Cheers,
Hong-Gunn

Re: Memory corruption when running VMware. (was File curruption when running VMware)

[Andrea Arcangeli]

On Fri, May 03, 2002 at 12:23:35PM +0930, Hong-Gunn Chew wrote:
> Hi Petr, Andrea,
> 
> Petr wrote:
> >   So if you have >890MB of RAM and your kernel is compiled 
> > with support for pte in high memory, please stop using 
> > VMware, or reconfigure your 
> > kernel to not use pte in high memory (4GB config without 
> > pte-in-highmem is OK). Using pte-in-highmem with vmmon will 
> > cause kernel oopses and/or 
> > memory corruption :-(
> 
> I have been trying different kernel configurations to overcome this
> problem.  I found that turning on APIC seem to work properly.  However
> turning on IO-APIC causes it to hang just after X is started, which
> seems to be during the startup of gdm.
> I do have a silly question to ask though.  How do I compile the kernel
> NOT to use pte in high memory?

the simplest patch to avoid pte in highmem is this. It's against 19pre7aa3
but it will apply cleanly to the kernel you're using.

--- 2.4.19pre7aa3/mm/memory.c.~1~	Tue Apr 30 19:53:08 2002
+++ 2.4.19pre7aa3/mm/memory.c	Fri May  3 09:58:23 2002
@@ -1493,7 +1493,7 @@
 {
 	struct page * page;
 
-	page = alloc_page(GFP_KERNEL | __GFP_HIGHMEM);
+	page = alloc_page(GFP_KERNEL);
 	if (page)
 		clear_pagetable(page);
 	return page;

However if I would be in you I'd simply add mem=850M to the append line
in /etc/lilo.conf until the vmmon/vmnet update is released, that should
be the most confortable approch from your part.

> Andrea wrote:
> > passing to the kernel mem=850M in lilo at boot will be enough.
> 
> This did not work as it causes vmware to seg fault with a kernel oops:

Hmm looking at the bounce_end_io_read it looks like the kernel didn't
recognized you specified mem=850M via lilo, not idea why. You should
check with `free` that your kernel is using only 850M of ram before
starting vmware. Can you show the contents of /proc/meminfo right after
boot?

> [generic_file_readahead+288/304] [bounce_end_io_read+164/288]
> [isapnp_set_mem+23/272] [vfs_link+23/256] [system_call+51/56]
> 	   [<f61fd5d1>] [<c01272b0>] [<c0131e94>] [<c01e0527>]
> [<c013e487>]  [<c0106f0b>]
> 	
> 	Code: 8b 11 89 d0 25 00 07 01 00 3d 00 04 00 00 75 12 81 ca 00
> 00
> 
> Cheers,
> Hong-Gunn


Andrea