| 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
| ||
|
Message-Id: <200705282021.30355.rjw@sisk.pl>
Date: Mon, 28 May 2007 20:21:29 +0200
From: "Rafael J. Wysocki" <rjw@...k.pl>
To: Maximilian Engelhardt <maxi@...monizer.de>
Cc: nigel@...el.suspend2.net, Herbert Xu <herbert@...dor.apana.org.au>,
linux-kernel <linux-kernel@...r.kernel.org>,
Pavel Machek <pavel@....cz>
Subject: Re: software suspend doesn't work with 2.6.22-rc3
On Monday, 28 May 2007 09:59, Rafael J. Wysocki wrote:
> On Monday, 28 May 2007 02:21, Maximilian Engelhardt wrote:
> > On Sunday 27 May 2007, Rafael J. Wysocki wrote:
> > > On Sunday, 27 May 2007 22:41, Maximilian Engelhardt wrote:
> > > > On Sunday 27 May 2007, Rafael J. Wysocki wrote:
> > > > > On Sunday, 27 May 2007 18:01, Maximilian Engelhardt wrote:
> > > > > > On Saturday 26 May 2007, Nigel Cunningham wrote:
> > > > > > > Hi.
> > > > > > >
> > > > > > > On Sat, 2007-05-26 at 14:49 +0200, Maximilian Engelhardt wrote:
> > > > > > > > On Saturday 26 May 2007, Nigel Cunningham wrote:
> > > > > > > > > Hi.
> > > > > > > > >
> > > > > > > > > On Sat, 2007-05-26 at 11:28 +0200, Maximilian Engelhardt wrote:
> > > > > > > > > > Hello,
> > > > > > > > > >
> > > > > > > > > > When I try software suspend on my laptop it always returns to
> > > > > > > > > > my running system after some time.
> > > > > > > > > > This is what's logged by the kernel:
> > > > > > > > > >
> > > > > > > > > > swsusp: Basic memory bitmaps created
> > > > > > > > > > Stopping tasks ...
> > > > > > > > > > Stopping kernel threads timed out after 20 seconds (1 tasks
> > > > > > > > > > refusing to freeze):
> > > > > > > > > > cryptd
> > > > > > > > > > Restarting tasks ... done.
> > > > > > > > > > swsusp: Basic memory bitmaps freed
> > > > > > > > > >
> > > > > > > > > > I have no idea what's the problem, but if you tell me what I
> > > > > > > > > > should do I can create debugging information and/or test
> > > > > > > > > > patches.
> > > > > > > > >
> > > > > > > > > Could you try this patch, please? It should help.
> > > > > > > > >
> > > > > > > > > Herbert, is this right? If cryptd is going to be used for block
> > > > > > > > > devs, the task should probably be PF_NOFREEZE (or whatever it
> > > > > > > > > is today) instead.
> > > > > > > > >
> > > > > > > > > Regards,
> > > > > > > > >
> > > > > > > > > Nigel
> > > > > > > > >
> > > > > > > > > crypto/cryptd.c | 1 +
> > > > > > > > > include/linux/freezer.h | 3 +++
> > > > > > > > > kernel/power/process.c | 2 +-
> > > > > > > > > 3 files changed, 5 insertions(+), 1 deletion(-)
> > > > > > > > > diff -ruNp 991-fix-cryptd.patch-old/crypto/cryptd.c
> > > > > > > > > 991-fix-cryptd.patch-new/crypto/cryptd.c ---
> > > > > > > > > 991-fix-cryptd.patch-old/crypto/cryptd.c 2007-05-19
> > > > > > > > > 18:16:47.000000000 +1000 +++
> > > > > > > > > 991-fix-cryptd.patch-new/crypto/cryptd.c 2007-05-26
> > > > > > > > > 19:45:42.000000000 +1000 @@ -341,6 +341,7 @@ static int
> > > > > > > > > cryptd_thread(void *data)
> > > > > > > > >
> > > > > > > > > mutex_unlock(&state->mutex);
> > > > > > > > >
> > > > > > > > > + try_to_freeze();
> > > > > > > > > schedule();
> > > > > > > > > } while (!stop);
> > > > > > > >
> > > > > > > > I tried your patch, but when I apply it my kernel doesn't compile
> > > > > > > > any more. I get these warnings/errors:
> > > > > > > >
> > > > > > > > [...]
> > > > > > > > CC crypto/cryptd.o
> > > > > > > > crypto/cryptd.c: In function ‘cryptd_thread’:
> > > > > > > > crypto/cryptd.c:344: warning: implicit declaration of function
> > > > > > > > ‘try_to_freeze’ [...]
> > > > > > > > LD init/built-in.o
> > > > > > > > LD .tmp_vmlinux1
> > > > > > > > crypto/built-in.o: In function `cryptd_thread':
> > > > > > > > cryptd.c:(.text+0xd7f5): undefined reference to `try_to_freeze'
> > > > > > > > make: *** [.tmp_vmlinux1] Error 1
> > > > > > >
> > > > > > > Ah. You'll need to add #include <linux/freezer.h> near that start
> > > > > > > of crypto/cryptd.c. Sorry for forgetting that.
> > > > > > >
> > > > > > > Nigel
> > > > > >
> > > > > > I added the include line and now I could compile the kernel, but
> > > > > > suspending still doesn't work.
> > > > > >
> > > > > > swsusp: Basic memory bitmaps created
> > > > > > Stopping tasks ...
> > > > > > Stopping kernel threads timed out after 20 seconds (1 tasks refusing
> > > > > > to freeze):
> > > > > > cryptd
> > > > > > Restarting tasks ... done.
> > > > > > swsusp: Basic memory bitmaps freed
> > > > >
> > > > > OK, this means that cryptd doesn't execute the try_to_freeze() for some
> > > > > reason.
> > > > >
> > > > > Please apply the appended patch on top of 2.6.22-rc3 and see if that
> > > > > helps.
> > > > >
> > > > > Greetings,
> > > > > Rafael
> > > > >
> > > > > ---
> > > > > crypto/cryptd.c | 1 +
> > > > > 1 file changed, 1 insertion(+)
> > > > >
> > > > > Index: linux-2.6.22-rc3/crypto/cryptd.c
> > > > > ===================================================================
> > > > > --- linux-2.6.22-rc3.orig/crypto/cryptd.c
> > > > > +++ linux-2.6.22-rc3/crypto/cryptd.c
> > > > > @@ -316,6 +316,7 @@ static int cryptd_thread(void *data)
> > > > > struct cryptd_state *state = data;
> > > > > int stop;
> > > > >
> > > > > + current->flags |= PF_NOFREEZE;
> > > > > do {
> > > > > struct crypto_async_request *req, *backlog;
> > > >
> > > > Even with this patch suspending doesn't work, dmesg shows the same error
> > > > message.
> > > > I also did build a kernel without cryptd and suspending does work there.
> > >
> > > Well, that's strange, because in that case the freezer shouldn't even wait
> > > for cryptd.
> > >
> > > Can you please try the patch at http://lkml.org/lkml/2007/5/26/169 ?
> >
> > With this patch applied suspend does work fine.
>
> Hmm. IMO the patch is too intrusive for 2.6.22, but OTOH it's going into the
> direction preferred by some prominent people. ;-)
>
> Let's try to combine the two threads and see what results from that.
Well, it looks like we have to fix this one separately.
Can you please tell me what to do to make cryptd run?
Rafael
-
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