netdev - RE: [PATCH 2/7] vxge: fix crash of VF when unloading PF

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] [day] [month] [year] [list]

Message-ID: <FCA91A92EE52B041906A0358FC28FCC38EF8AF79FC@FRE1EXCH02.hq.exar.com>
Date:	Fri, 10 Dec 2010 17:35:00 -0800
From:	Ramkrishna Vepa <Ramkrishna.Vepa@...r.com>
To:	Chris Wright <chrisw@...s-sol.org>, Jon Mason <Jon.Mason@...r.com>
CC:	"David S. Miller" <davem@...emloft.net>,
	"netdev@...r.kernel.org" <netdev@...r.kernel.org>,
	Sivakumar Subramani <Sivakumar.Subramani@...r.com>,
	Sreenivasa Honnur <Sreenivasa.Honnur@...r.com>
Subject: RE: [PATCH 2/7] vxge: fix crash of VF when unloading PF

> > +
> >  /**
> >   * vxge_probe
> >   * @pdev : structure containing the PCI related information of the
> device.
> > @@ -4370,14 +4384,13 @@ vxge_probe(struct pci_dev *pdev, const struct
> pci_device_id *pre)
> >             num_vfs = vxge_get_num_vfs(function_mode) - 1;
> >
> >     /* Enable SRIOV mode, if firmware has SRIOV support and if it is a
> PF */
> > -   if (is_sriov(function_mode) && (max_config_dev > 1) &&
> > -           (ll_config->intr_type != INTA) &&
> > -           (is_privileged == VXGE_HW_OK)) {
> > -           ret = pci_enable_sriov(pdev, ((max_config_dev - 1) < num_vfs)
> > -                   ? (max_config_dev - 1) : num_vfs);
> > +   if (is_sriov(function_mode) && !is_sriov_initialized(pdev) &&
> > +      (ll_config->intr_type != INTA)) {
> > +           ret = pci_enable_sriov(pdev, num_vfs);
>
> This fundamentally changes the way VF's are allocated.  Now you cannot
> specifiy the number of vfs to allocate with max_config_dev module
> parameter.
The X3100 supports 11 different pci function modes where the user has the ability to choose the number of functions for each mode. This is more efficient usage of the hardware as the resources are carved out equally for the functions. By configuring the max_config_dev less than num_vfs, there's unnecessary wastage of resources.

>
> >             if (ret)
> >                     vxge_debug_ll_config(VXGE_ERR,
> >                             "Failed in enabling SRIOV mode: %d\n", ret);
> > +                   /* No need to fail out, as an error here is non-fatal */
> >     }
> >
> >     /*
> > @@ -4673,8 +4686,6 @@ static void __devexit vxge_remove(struct pci_dev
> *pdev)
> >
> >     iounmap(vdev->bar0);
> >
> > -   pci_disable_sriov(pdev);
> > -
>
> And you can never disable sriov.
If the device's pci function mode is changed, a power cycle is required in which case the functions are re-enumerated.

>
> This doesn't look like the right behaviour.
When the driver is loaded for the X3100 in SRIOV mode, it will be working in that mode even after it is unloaded and reloaded. As mentioned earlier, a change in the function mode requires and power cycle of the system.

The SRIOV feature is shipping in many distros and we need this fix back ported to prevent a possible crash when the PF is unloaded while the VFs are running in the guest OS in pass through mode.

If you have a better or simpler solution, that may take longer to implement, I would suggest that this solution be accepted in the interim.

Thanks,
Ram

The information and any attached documents contained in this message
may be confidential and/or legally privileged.  The message is
intended solely for the addressee(s).  If you are not the intended
recipient, you are hereby notified that any use, dissemination, or
reproduction is strictly prohibited and may be unlawful.  If you are
not the intended recipient, please contact the sender immediately by
return e-mail and destroy all copies of the original message.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html