linux-kernel - Re: [PATCH v2] x86/PCI: Ignore E820 reservations for bridge windows on newer systems

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:   Thu, 14 Oct 2021 10:15:13 +0200
From:   Hans de Goede <hdegoede@...hat.com>
To:     Krzysztof Wilczyński <kw@...ux.com>
Cc:     "Rafael J . Wysocki" <rjw@...ysocki.net>,
        Mika Westerberg <mika.westerberg@...ux.intel.com>,
        Bjorn Helgaas <bhelgaas@...gle.com>,
        Myron Stowe <myron.stowe@...hat.com>,
        Juha-Pekka Heikkila <juhapekka.heikkila@...il.com>,
        Thomas Gleixner <tglx@...utronix.de>,
        Ingo Molnar <mingo@...hat.com>, Borislav Petkov <bp@...en8.de>,
        "H . Peter Anvin" <hpa@...or.com>, linux-pci@...r.kernel.org,
        x86@...nel.org, linux-kernel@...r.kernel.org,
        Benoit Grégoire <benoitg@...us.ca>,
        Hui Wang <hui.wang@...onical.com>
Subject: Re: [PATCH v2] x86/PCI: Ignore E820 reservations for bridge windows
 on newer systems

Hi Krzysztof,

On 10/14/21 12:13 AM, Krzysztof Wilczyński wrote:
> Hi Hans,
> 
> Thank you for sending the patch over!
> 
> [...]
>> [    0.000000] BIOS-e820: [mem 0x000000004bc50000-0x00000000cfffffff] reserved
>> [    0.557473] pci_bus 0000:00: root bus resource [mem 0x65400000-0xbfffffff window]
> 
> A very small nitpick: we usually remove time/date stamps from kernel ring
> buffer outputs keeping only the relevant message parts left.

Ok, I'll do a v3 fixing this.

> 
> [...]
>> Old systems are defined here as BIOS year < 2018, this was chosen to
>> make sure that pci_use_e820 will not be set on the currently affected
>> systems, while at the same time also taking into account that the
>> systems for which the E820 checking was orignally added may have
> 
> A tiny typo of "originally" in the sentence above.

And this.

> [...]
>> @@ -232,3 +236,9 @@ static inline void mmio_config_writel(void __iomem *pos, u32 val)
>>  # define x86_default_pci_init_irq	NULL
>>  # define x86_default_pci_fixup_irqs	NULL
>>  #endif
>> +
>> +#if defined CONFIG_PCI && defined CONFIG_ACPI
> 
> I know that Mika already asked about this, and you responded, so I can only
> added: brackets, let's add brackets, most definitely. :)

I've no big preference either way, so I'll move to using
parentheses for the next version.

> 
> [...]
>> +/* Consumed in arch/x86/kernel/resource.c */
>> +bool pci_use_e820 = false;
> 
> A small nitpick: not sure if this comment is needed as probably most people
> working with this code would use "git grep" and likes to list occurrences
> where the variables is used.  But, this is highly subjective, thus there is
> probably nothing to change here.

I put it the comment there because the other use_foo flag directly above
it are all static, so it is there to explain why this one is not static.

At least that was my idea behind the comment :)

> 
>> +	printk(KERN_INFO "PCI: %s E820 reservations for host bridge windows\n",
>> +	       pci_use_e820 ? "Honoring" : "Ignoring");
> 
> I know you followed the existing style, which is very much appreciated, but
> if and where possible, we should move to newer API/style and replace the
> printk() above with pr_info().  New code should not be adding old style if
> it can be helped (checkpatch.pl would warn about this too).  What do yo you
> think?

Yes checkpatch complained about this, still I deliberately ignored that,
as you said I'm following the existing style here. I very much dislike
mixing styles in a single file.  If we want to change this for this file
then IMHO the right thing to do would be a follow up patch changing all
the printk-s at once.

Regards,

Hans