linux-kernel - Re: udldrm does not recover from powersave? Re: udldrmfb: causes WARN in i915 on X60 (x86-32)

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: <20210225170726.GA31110@duo.ucw.cz>
Date:   Thu, 25 Feb 2021 18:07:26 +0100
From:   Pavel Machek <pavel@....cz>
To:     Thomas Zimmermann <tzimmermann@...e.de>
Cc:     intel-gfx@...ts.freedesktop.org,
        kernel list <linux-kernel@...r.kernel.org>,
        dri-devel@...ts.freedesktop.org, rodrigo.vivi@...el.com,
        airlied@...hat.com, sean@...rly.run
Subject: Re: udldrm does not recover from powersave? Re: udldrmfb: causes
 WARN in i915 on X60 (x86-32)

Hi!

> > > > This is in -next, but I get same behaviour on 5.11; and no, udl does
> > > 
> > > Thanks for reporting. We are in the process of fixing the issue. The latest
> > > patch is at [1].
> > > 
> > 
> > Thank you, that fixes the DMA issue, and I can use the udl.
> > 
> > ...for a while. Then screensaver blanks laptop screen, udl screen
> > blanks too. Upon hitting a key, internal screen shows up, udl does
> > not.
> > 
> > I try rerunning xrandr ... --auto, but could not recover it.
> > 
> > Any ideas?
> 
> Did it work before the regression?
> 
> For testing, could you please remove the fix and then do
> 
>   git revert 6eb0233ec2d0
> 
> This would restore the old version. Please report back on the
> results.

Ok, I went to 7f206cf3ec2b with 6eb0233ec2d0 reverted. That fails to
build:

drivers/usb/core/message.c: In function ‘usb_set_configuration’:
drivers/usb/core/message.c:2100:12: error: ‘struct device’ has no member named ‘dma_pfn_offset’
 2100 |   intf->dev.dma_pfn_offset = dev->dev.dma_pfn_offset;
      |            ^
drivers/usb/core/message.c:2100:38: error: ‘struct device’ has no member named ‘dma_pfn_offset’
 2100 |   intf->dev.dma_pfn_offset = dev->dev.dma_pfn_offset;
      |                                      ^
  CC      drivers/net/ethernet/intel/e1000e/param.o
make[3]: *** [scripts/Makefile.build:271: drivers/usb/core/message.o]
Error 1

So I tried to go to bad commit's parent:

git checkout 6eb0233ec2d0^
git log
commit cf141ae989e2ff119cd320326da5923b480d1641
    ARM/keystone: move the DMA offset handling under ifdef CONFIG_ARM_LPAE

But that resulted in lockup soon after "--setprovidersource" command
was isued.

Best regards,
									Pavel

-- 
http://www.livejournal.com/~pavelmachek

Download attachment "signature.asc" of type "application/pgp-signature" (196 bytes)