Re: hardware random number generator?

Re: hardware random number generator?

[Jeff Garzik]

Matthew Wilcox wrote:
> config HW_RANDOM
>         tristate "Intel/AMD/VIA HW Random Number Generator support"
>         depends on (X86 || IA64) && PCI
>         ---help---
>           This driver provides kernel-side support for the Random Number
>           Generator hardware found on Intel i8xx-based motherboards,
>           AMD 76x-based motherboards, and Via Nehemiah CPUs.
> 
> Why IA64 too?  This file contains the code:

No idea.  <guessing>  There are some Intel ICH-based motherboards w/ 
IA64 chips on them, maybe?  I certainly don't recall adding that IA64 
marker... :)

	Jeff

Re: hardware random number generator?

[Matthew Wilcox]

On Mon, Mar 22, 2004 at 09:00:14PM -0500, Jeff Garzik wrote:
> Matthew Wilcox wrote:
> >config HW_RANDOM
> >        tristate "Intel/AMD/VIA HW Random Number Generator support"
> >        depends on (X86 || IA64) && PCI
> >        ---help---
> >          This driver provides kernel-side support for the Random Number
> >          Generator hardware found on Intel i8xx-based motherboards,
> >          AMD 76x-based motherboards, and Via Nehemiah CPUs.
> >
> >Why IA64 too?  This file contains the code:
> 
> No idea.  <guessing>  There are some Intel ICH-based motherboards w/ 
> IA64 chips on them, maybe?  I certainly don't recall adding that IA64 
> marker... :)

Heh.  ICH is x86-only, as far as I know.  Intel did the 460GX chipset
for Merced and now have the E8870 chipset for McKinley/Madison.  Anyone
familiar with the Intel chipsets want to comment about the presence or
absence of a RNG?

-- 
"Next the statesmen will invent cheap lies, putting the blame upon 
the nation that is attacked, and every man will be glad of those
conscience-soothing falsities, and will diligently study them, and refuse
to examine any refutations of them; and thus he will by and by convince 
himself that the war is just, and will thank God for the better sleep 
he enjoys after this process of grotesque self-deception." -- Mark Twain

RE: hardware random number generator?

[Mallick, Asit K]

E8870 based McKinley/Madison platforms does have ICH4. It is used to
provide USB, IDE, On board LAN, LPC, SMBus, etc. funtionality. However,
the RNG does not work on current E8870 based platforms. Future platforms
may enable RNG.

Thanks,
Asit

>-----Original Message-----
>From: linux-ia64-owner@vger.kernel.org 
>[mailto:linux-ia64-owner@vger.kernel.org] On Behalf Of Matthew Wilcox
>Sent: Monday, March 22, 2004 7:05 PM
>To: Jeff Garzik
>Cc: Matthew Wilcox; linux-ia64@vger.kernel.org
>Subject: Re: hardware random number generator?
>
>On Mon, Mar 22, 2004 at 09:00:14PM -0500, Jeff Garzik wrote:
>> Matthew Wilcox wrote:
>> >config HW_RANDOM
>> >        tristate "Intel/AMD/VIA HW Random Number Generator support"
>> >        depends on (X86 || IA64) && PCI
>> >        ---help---
>> >          This driver provides kernel-side support for the 
>Random Number
>> >          Generator hardware found on Intel i8xx-based motherboards,
>> >          AMD 76x-based motherboards, and Via Nehemiah CPUs.
>> >
>> >Why IA64 too?  This file contains the code:
>> 
>> No idea.  <guessing>  There are some Intel ICH-based motherboards w/ 
>> IA64 chips on them, maybe?  I certainly don't recall adding 
>that IA64 
>> marker... :)
>
>Heh.  ICH is x86-only, as far as I know.  Intel did the 460GX chipset
>for Merced and now have the E8870 chipset for McKinley/Madison.  Anyone
>familiar with the Intel chipsets want to comment about the presence or
>absence of a RNG?
>
>-- 
>"Next the statesmen will invent cheap lies, putting the blame upon 
>the nation that is attacked, and every man will be glad of those
>conscience-soothing falsities, and will diligently study them, 
>and refuse
>to examine any refutations of them; and thus he will by and by 
>convince 
>himself that the war is just, and will thank God for the better sleep 
>he enjoys after this process of grotesque self-deception." -- 
>Mark Twain
>-
>To unsubscribe from this list: send the line "unsubscribe 
>linux-ia64" in
>the body of a message to majordomo@vger.kernel.org
>More majordomo info at  http://vger.kernel.org/majordomo-info.html
>

Re: hardware random number generator?

[Bjorn Helgaas]

On Wednesday 24 March 2004 12:23 am, Mallick, Asit K wrote:
> E8870 based McKinley/Madison platforms does have ICH4. It is used to
> provide USB, IDE, On board LAN, LPC, SMBus, etc. funtionality. However,
> the RNG does not work on current E8870 based platforms. Future platforms
> may enable RNG.

The following patch allows hw_random.c to build on ia64.  (The problem
was just that the VIA stuff has i386 assembly in it.  The current code
only probes for VIA on i386 anyway, so this patch just adds more ifdefs
so the VIA code is only built for i386.)

This boots fine on my Tiger, but reading it seems to give all ones,
which I suppose is what you mean about it being broken.

I don't know whether we should
  1) apply this patch, and ignore the fact that you don't
     get random data on current E8870-based platforms,
  2) put in some kind of check for known-buggy hardware (I don't
     know how to do that, of course), or
  3) change Kconfig so we don't offer it for ia64 (though
     we will presumably want to do so for future ia64 platforms).

I guess I vote for (1) at the moment, and if you think we should
be smart about identifying the broken versions of E8870, maybe
you can supply a patch, Asit.

=== drivers/char/hw_random.c 1.13 vs edited ==--- 1.13/drivers/char/hw_random.c	Fri Jan 30 21:25:15 2004
+++ edited/drivers/char/hw_random.c	Wed Mar 24 16:16:21 2004
@@ -92,10 +92,12 @@
 static unsigned int amd_data_present (void);
 static u32 amd_data_read (void);
 
+#ifdef __i386__
 static int __init via_init(struct pci_dev *dev);
 static void via_cleanup(void);
 static unsigned int via_data_present (void);
 static u32 via_data_read (void);
+#endif
 
 struct rng_operations {
 	int (*init) (struct pci_dev *dev);
@@ -137,8 +139,10 @@
 	/* rng_hw_amd */
 	{ amd_init, amd_cleanup, amd_data_present, amd_data_read, 4 },
 
+#ifdef __i386__
 	/* rng_hw_via */
 	{ via_init, via_cleanup, via_data_present, via_data_read, 1 },
+#endif
 };
 
 /*
@@ -341,6 +345,7 @@
 	/* FIXME: twiddle pmio, also? */
 }
 
+#ifdef __i386__
 /***********************************************************************
  *
  * VIA RNG operations
@@ -456,6 +461,7 @@
 {
 	/* do nothing */
 }
+#endif
 
 
 /***********************************************************************

Re: hardware random number generator?

[Bjorn Helgaas]

Jeff, can you apply this patch to hw_random.c?  Thanks.

The following patch allows hw_random.c to build on ia64.  (The problem
was just that the VIA stuff has i386 assembly in it.  The current code
only probes for VIA on i386 anyway, so this patch just adds more ifdefs
so the VIA code is only built for i386.)

=== drivers/char/hw_random.c 1.13 vs edited ==--- 1.13/drivers/char/hw_random.c	Fri Jan 30 21:25:15 2004
+++ edited/drivers/char/hw_random.c	Wed Mar 24 16:16:21 2004
@@ -92,10 +92,12 @@
 static unsigned int amd_data_present (void);
 static u32 amd_data_read (void);
 
+#ifdef __i386__
 static int __init via_init(struct pci_dev *dev);
 static void via_cleanup(void);
 static unsigned int via_data_present (void);
 static u32 via_data_read (void);
+#endif
 
 struct rng_operations {
 	int (*init) (struct pci_dev *dev);
@@ -137,8 +139,10 @@
 	/* rng_hw_amd */
 	{ amd_init, amd_cleanup, amd_data_present, amd_data_read, 4 },
 
+#ifdef __i386__
 	/* rng_hw_via */
 	{ via_init, via_cleanup, via_data_present, via_data_read, 1 },
+#endif
 };
 
 /*
@@ -341,6 +345,7 @@
 	/* FIXME: twiddle pmio, also? */
 }
 
+#ifdef __i386__
 /***********************************************************************
  *
  * VIA RNG operations
@@ -456,6 +461,7 @@
 {
 	/* do nothing */
 }
+#endif
 
 
 /***********************************************************************

Re: hardware random number generator?

[Randy.Dunlap]

On Tue, 23 Mar 2004 23:23:29 -0800 Mallick, Asit K wrote:

| E8870 based McKinley/Madison platforms does have ICH4. It is used to
| provide USB, IDE, On board LAN, LPC, SMBus, etc. funtionality. However,
| the RNG does not work on current E8870 based platforms. Future platforms
| may enable RNG.

Does this agree with or allow for this end-of-support announcement?
http://www.intel.com/design/security/rng/rng.htm


| Thanks,
| Asit
| 
| >-----Original Message-----
| >From: linux-ia64-owner@vger.kernel.org 
| >[mailto:linux-ia64-owner@vger.kernel.org] On Behalf Of Matthew Wilcox
| >Sent: Monday, March 22, 2004 7:05 PM
| >To: Jeff Garzik
| >Cc: Matthew Wilcox; linux-ia64@vger.kernel.org
| >Subject: Re: hardware random number generator?
| >
| >On Mon, Mar 22, 2004 at 09:00:14PM -0500, Jeff Garzik wrote:
| >> Matthew Wilcox wrote:
| >> >config HW_RANDOM
| >> >        tristate "Intel/AMD/VIA HW Random Number Generator support"
| >> >        depends on (X86 || IA64) && PCI
| >> >        ---help---
| >> >          This driver provides kernel-side support for the 
| >Random Number
| >> >          Generator hardware found on Intel i8xx-based motherboards,
| >> >          AMD 76x-based motherboards, and Via Nehemiah CPUs.
| >> >
| >> >Why IA64 too?  This file contains the code:
| >> 
| >> No idea.  <guessing>  There are some Intel ICH-based motherboards w/ 
| >> IA64 chips on them, maybe?  I certainly don't recall adding 
| >that IA64 
| >> marker... :)
| >
| >Heh.  ICH is x86-only, as far as I know.  Intel did the 460GX chipset
| >for Merced and now have the E8870 chipset for McKinley/Madison.  Anyone
| >familiar with the Intel chipsets want to comment about the presence or
| >absence of a RNG?
| >
| >-- 


--
~Randy
"We have met the enemy and he is us."  -- Pogo (by Walt Kelly)
(Again.  Sometimes I think ln -s /usr/src/linux/.config .signature)

Re: hardware random number generator?

[Jeff Garzik]

Bjorn Helgaas wrote:
> Jeff, can you apply this patch to hw_random.c?  Thanks.
> 
> The following patch allows hw_random.c to build on ia64.  (The problem
> was just that the VIA stuff has i386 assembly in it.  The current code
> only probes for VIA on i386 anyway, so this patch just adds more ifdefs
> so the VIA code is only built for i386.)


applied to 2.6

hardware random number generator?

[Matthew Wilcox]

config HW_RANDOM
        tristate "Intel/AMD/VIA HW Random Number Generator support"
        depends on (X86 || IA64) && PCI
        ---help---
          This driver provides kernel-side support for the Random Number
          Generator hardware found on Intel i8xx-based motherboards,
          AMD 76x-based motherboards, and Via Nehemiah CPUs.

Why IA64 too?  This file contains the code:

static inline u32 xstore(u32 *addr, u32 edx_in)
{
        u32 eax_out;

        asm(".byte 0x0F,0xA7,0xC0 /* xstore %%edi (addr=%0) */"
                :"=m"(*addr), "=a"(eax_out)
                :"D"(addr), "d"(edx_in));

        return eax_out;
}

which doesn't compile on ia64.  I think it should be unselectable.

-- 
"Next the statesmen will invent cheap lies, putting the blame upon 
the nation that is attacked, and every man will be glad of those
conscience-soothing falsities, and will diligently study them, and refuse
to examine any refutations of them; and thus he will by and by convince 
himself that the war is just, and will thank God for the better sleep 
he enjoys after this process of grotesque self-deception." -- Mark Twain