linux-kernel - Re: [PATCH v3 04/10] PCI/RCEC: Add pcie_walk

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]

Date:   Wed, 16 Sep 2020 18:06:54 -0500
From:   Bjorn Helgaas <helgaas@...nel.org>
To:     Sean V Kelley <sean.v.kelley@...el.com>
Cc:     "Zhuo, Qiuxu" <qiuxu.zhuo@...el.com>, Jonathan.Cameron@...wei.com,
        rjw@...ysocki.net, sathyanarayanan.kuppuswamy@...ux.intel.com,
        "Raj, Ashok" <ashok.raj@...el.com>,
        "Luck, Tony" <tony.luck@...el.com>, linux-pci@...r.kernel.org,
        bhelgaas@...gle.com, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v3 04/10] PCI/RCEC: Add pcie_walk_rcec() to walk
 associated RCiEPs

On Tue, Sep 15, 2020 at 09:09:20AM -0700, Sean V Kelley wrote:

> Walking the bus with an RCEC as it is probed in the portdrv_pci.c can be
> done with both its own bus (bitmap) and with supported associated bus
> ranges.  In that walk I’m able to find all the associated endpoints via both
> bitmap on own bus and the bus ranges. No pci_get_slot() is needed as per
> your original suggestion.

OK.  That seems OK for now.

> The suggsted approach to the rcec_helper() seems to imply that we either
> walk it again or have cached all the associated RCiEP.  When we do the walk
> above, we are merely, finding the RCiEPs and linking them to the RCEC’s
> structure. There is no caching done of a list of RCiEPs per RCEC.

pcie_portdrv_init() is a device_initcall, so it should be called after
we enumerate all the PCI devices (at least those present at boot).  So
you don't have to worry about finding an RCiEP before its associated
RCEC -- we should have found them *all* by the time we get to
pcie_portdrv_probe().

Bjorn