lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20121004131801.GD21028@aftab.osrc.amd.com>
Date:	Thu, 4 Oct 2012 15:18:02 +0200
From:	Borislav Petkov <bp@...64.org>
To:	Daniel J Blueman <daniel@...ascale-asia.com>
Cc:	Ingo Molnar <mingo@...hat.com>,
	Thomas Gleixner <tglx@...utronix.de>,
	"H. Peter Anvin" <hpa@...or.com>,
	Steffen Persvold <sp@...ascale.com>, x86@...nel.org,
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2] Fix AMD Northbridge-ID contiguity assumptions

On Wed, Oct 03, 2012 at 09:21:14PM +0800, Daniel J Blueman wrote:
> The AMD Northbridge initialisation code and EDAC assume the Northbridge IDs
> are contiguous, which no longer holds on federated systems with multiple
> HyperTransport fabrics and multiple PCI domains.
> 
> Address this assumption by searching the Northbridge ID array, rather than
> directly indexing it, using the upper bits for the PCI domain.
> 
> v2: Fix Northbridge entry initialisation
> 
> Tested on a single-socket system and 3-server federated system.
> 
> Signed-off-by: Daniel J Blueman <daniel@...ascale-asia.com>
> ---
>  arch/x86/include/asm/amd_nb.h |   23 +++++++++++++++++++++--
>  arch/x86/kernel/amd_nb.c      |   16 +++++++++-------
>  drivers/edac/amd64_edac.c     |   18 +++++++++---------
>  drivers/edac/amd64_edac.h     |    6 ------
>  4 files changed, 39 insertions(+), 24 deletions(-)
> 
> diff --git a/arch/x86/include/asm/amd_nb.h b/arch/x86/include/asm/amd_nb.h
> index b3341e9..0fd2f0c 100644
> --- a/arch/x86/include/asm/amd_nb.h
> +++ b/arch/x86/include/asm/amd_nb.h
> @@ -47,6 +47,7 @@ struct threshold_bank {
>  };
>  
>  struct amd_northbridge {
> +	u32 node;

Right, as pointed out before, are 4 bytes really needed for node index
or smaller is still ok?

>  	struct pci_dev *misc;
>  	struct pci_dev *link;
>  	struct amd_l3_cache l3_cache;
> @@ -76,15 +77,33 @@ static inline bool amd_nb_has_feature(unsigned feature)
>  	return ((amd_northbridges.flags & feature) == feature);
>  }
>  
> -static inline struct amd_northbridge *node_to_amd_nb(int node)
> +static inline int node_to_amd_index(u32 node)
>  {
> -	return (node < amd_northbridges.num) ? &amd_northbridges.nb[node] : NULL;
> +	int i;
> +
> +	for (i = 0; i < amd_northbridges.num; i++)
> +		if (amd_northbridges.nb[i].node == node)
> +			return i;
> +
> +	return 0;

Return -1 instead, in the failure case...

> +}
> +
> +static inline struct amd_northbridge *node_to_amd_nb(u32 node)
> +{
> +	return &amd_northbridges.nb[node_to_amd_index(node)];

.. which needs to be handled here.

> +}
> +
> +/* AMD sets the first MC device at device ID 0x18 */
> +static inline u32 get_node_id(struct pci_dev *pdev)
> +{
> +	return (pci_domain_nr(pdev->bus) << 8) | (PCI_SLOT(pdev->devfn) - 0x18);
>  }
>  
>  #else
>  
>  #define amd_nb_num(x)		0
>  #define amd_nb_has_feature(x)	false
> +#define node_to_amd_index(x)	0
>  #define node_to_amd_nb(x)	NULL

Those are both used in other files and thus in the global namespace, can we
change them to

amd_nb_node_index(x)

and

and_nb_get_desc(node)

?

>  #endif
> diff --git a/arch/x86/kernel/amd_nb.c b/arch/x86/kernel/amd_nb.c
> index aadf335..c29ce39 100644
> --- a/arch/x86/kernel/amd_nb.c
> +++ b/arch/x86/kernel/amd_nb.c
> @@ -75,10 +75,10 @@ int amd_cache_northbridges(void)
>  
>  	link = misc = NULL;
>  	for (i = 0; i != amd_nb_num(); i++) {
> -		node_to_amd_nb(i)->misc = misc =
> -			next_northbridge(misc, amd_nb_misc_ids);
> -		node_to_amd_nb(i)->link = link =
> -			next_northbridge(link, amd_nb_link_ids);
> +		nb->misc = misc = next_northbridge(misc, amd_nb_misc_ids);
> +		nb->node = get_node_id(misc);
> +		nb->link = link = next_northbridge(link, amd_nb_link_ids);
> +		nb++;

Pls reorder for better readability:

		nb->misc = misc = next_northbridge(misc, amd_nb_misc_ids);
		nb->link = link = next_northbridge(link, amd_nb_link_ids);
		nb->node = get_node_id(misc);
		nb++;

>          }
>  
>  	/* some CPU families (e.g. family 0x11) do not support GART */
> @@ -212,6 +212,7 @@ int amd_set_subcaches(int cpu, int mask)
>  static int amd_cache_gart(void)
>  {
>  	u16 i;
> +	struct amd_northbridge *nb = amd_northbridges.nb;
>  
>         if (!amd_nb_has_feature(AMD_NB_GART))
>                 return 0;
> @@ -222,9 +223,10 @@ static int amd_cache_gart(void)
>                 return -ENOMEM;
>         }
>  
> -       for (i = 0; i != amd_nb_num(); i++)
> -               pci_read_config_dword(node_to_amd_nb(i)->misc, 0x9c,
> -                                     &flush_words[i]);
> +	for (i = 0; i != amd_nb_num(); i++) {
> +		pci_read_config_dword(nb->misc, 0x9c, &flush_words[i]);
> +		nb++;
> +	}
>  
>         return 0;
>  }
> diff --git a/drivers/edac/amd64_edac.c b/drivers/edac/amd64_edac.c
> index 5a297a2..9c35565 100644
> --- a/drivers/edac/amd64_edac.c
> +++ b/drivers/edac/amd64_edac.c
> @@ -2549,7 +2549,7 @@ static int amd64_init_one_instance(struct pci_dev *F2)
>  	struct mem_ctl_info *mci = NULL;
>  	struct edac_mc_layer layers[2];
>  	int err = 0, ret;
> -	u8 nid = get_node_id(F2);
> +	u32 nid = get_node_id(F2);
>  
>  	ret = -ENOMEM;
>  	pvt = kzalloc(sizeof(struct amd64_pvt), GFP_KERNEL);
> @@ -2640,7 +2640,7 @@ err_ret:
>  static int __devinit amd64_probe_one_instance(struct pci_dev *pdev,
>  					     const struct pci_device_id *mc_type)
>  {
> -	u8 nid = get_node_id(pdev);
> +	u32 nid = get_node_id(pdev);

Let's add a

	node_idx = node_to_amd_index(nid);

and use it below instead, ok?

>  	struct pci_dev *F3 = node_to_amd_nb(nid)->misc;
>  	struct ecc_settings *s;
>  	int ret = 0;
> @@ -2656,7 +2656,7 @@ static int __devinit amd64_probe_one_instance(struct pci_dev *pdev,
>  	if (!s)
>  		goto err_out;
>  
> -	ecc_stngs[nid] = s;
> +	ecc_stngs[node_to_amd_index(nid)] = s;
>  
>  	if (!ecc_enabled(F3, nid)) {
>  		ret = -ENODEV;
> @@ -2680,7 +2680,7 @@ static int __devinit amd64_probe_one_instance(struct pci_dev *pdev,
>  
>  err_enable:
>  	kfree(s);
> -	ecc_stngs[nid] = NULL;
> +	ecc_stngs[node_to_amd_index(nid)] = NULL;
>  
>  err_out:
>  	return ret;
> @@ -2690,9 +2690,9 @@ static void __devexit amd64_remove_one_instance(struct pci_dev *pdev)
>  {
>  	struct mem_ctl_info *mci;
>  	struct amd64_pvt *pvt;
> -	u8 nid = get_node_id(pdev);
> +	u32 nid = get_node_id(pdev);
>  	struct pci_dev *F3 = node_to_amd_nb(nid)->misc;
> -	struct ecc_settings *s = ecc_stngs[nid];
> +	struct ecc_settings *s = ecc_stngs[node_to_amd_index(nid)];
>  
>  	mci = find_mci_by_dev(&pdev->dev);
>  	del_mc_sysfs_attrs(mci);
> @@ -2711,12 +2711,12 @@ static void __devexit amd64_remove_one_instance(struct pci_dev *pdev)
>  	amd_report_gart_errors(false);
>  	amd_unregister_ecc_decoder(amd64_decode_bus_error);
>  
> -	kfree(ecc_stngs[nid]);
> -	ecc_stngs[nid] = NULL;
> +	kfree(ecc_stngs[node_to_amd_index(nid)]);
> +	ecc_stngs[node_to_amd_index(nid)] = NULL;
>  
>  	/* Free the EDAC CORE resources */
>  	mci->pvt_info = NULL;
> -	mcis[nid] = NULL;
> +	mcis[node_to_amd_index(nid)] = NULL;
>  
>  	kfree(pvt);
>  	edac_mc_free(mci);
> diff --git a/drivers/edac/amd64_edac.h b/drivers/edac/amd64_edac.h
> index 8d48047..90cae61 100644
> --- a/drivers/edac/amd64_edac.h
> +++ b/drivers/edac/amd64_edac.h
> @@ -290,12 +290,6 @@
>  /* MSRs */
>  #define MSR_MCGCTL_NBE			BIT(4)
>  
> -/* AMD sets the first MC device at device ID 0x18. */
> -static inline u8 get_node_id(struct pci_dev *pdev)
> -{
> -	return PCI_SLOT(pdev->devfn) - 0x18;
> -}
> -
>  enum amd_families {
>  	K8_CPUS = 0,
>  	F10_CPUS,
> -- 
> 1.7.9.5

Thanks.

-- 
Regards/Gruss,
Boris.

Advanced Micro Devices GmbH
Einsteinring 24, 85609 Dornach
GM: Alberto Bozzo
Reg: Dornach, Landkreis Muenchen
HRB Nr. 43632 WEEE Registernr: 129 19551
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ