From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Message-ID: <54CB9C6D.1080506@imgtec.com> Date: Fri, 30 Jan 2015 14:59:57 +0000 From: James Hogan MIME-Version: 1.0 To: "Maciej W. Rozycki" , Aaro Koskinen CC: Ralf Baechle , David Daney , , , Hemmo Nieminen , Subject: Re: [PATCH v2 2/2] MIPS: fix kernel lockup or crash after CPU offline/online References: <1421355719-17576-1-git-send-email-aaro.koskinen@iki.fi> <1421355719-17576-2-git-send-email-aaro.koskinen@iki.fi> <54CB5B59.5050203@imgtec.com> In-Reply-To: Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="DMNkN7kt0u2t2J4KlMio8mwJQ6mMt015t" Sender: linux-kernel-owner@vger.kernel.org List-ID: --DMNkN7kt0u2t2J4KlMio8mwJQ6mMt015t Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: quoted-printable On 30/01/15 12:47, Maciej W. Rozycki wrote: > On Fri, 30 Jan 2015, James Hogan wrote: >=20 >>> Hmm, why can a call to `printk' cause a TLB miss, what's so special = about=20 >>> this function? Does it use kernel mapped addresses for any purpose s= uch=20 >>> as `vmalloc'? >> >> It would be the fact netconsole (or whatever other console is in use) = is >> built as a kernel module, memory for which is allocated from the vmall= oc >> area. >=20 > Ah, I see, thanks for enlightening me. But in that case wouldn't it b= e=20 > possible to postpone console output from `printk' until it is safe to=20 > access the device? In a manner similar to how for example we handle ca= lls=20 > to `printk' made from the hardirq context. That would make things less= =20 > fragile. Hmm, kernel/printk/printk.c does have: static inline int can_use_console(unsigned int cpu) { return cpu_online(cpu) || have_callable_console(); } which should prevent it dumping printk buffer to console. CPU shouldn't be marked online that early, which suggests that the console has the CON_ANYTIME flag set, which it probably shouldn't if it depends on module code. call_console_drivers() seems to ensure the CPU is online or has CON_ANYTIME before calling the console write callback. A quick glance and I can't see any evidence of netconsole being able to get CON_ANYTIME. serial8250_console does appear to set that flag *and* is tristate though, which is slightly worrying. Aaro, what is the content of your /proc/consoles? Cheers James --DMNkN7kt0u2t2J4KlMio8mwJQ6mMt015t Content-Type: application/pgp-signature; name="signature.asc" Content-Description: OpenPGP digital signature Content-Disposition: attachment; filename="signature.asc" -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQIcBAEBAgAGBQJUy5x0AAoJEGwLaZPeOHZ6bqIP/A67+XDb6J0TVn65uvmsXvKD AwFMghX4pPMywmlqKRU4EV8QCRB0bOotA2HdE6fEhezhLDMzeC/dlHUVbqLKThOr PMynEUH/Tt5RKlZfpDfKzl0e7210HDCPB3HVfN7hcStOhao/G4kndrf36kiqmD4V DeM92x1/ng1Lx9UM5O9cK1KI7G13j8LeA16M/CsitUeDcxjIpyBZYNSDLuutRaCV ip/BR+x21q77Ad0s2a28mNl7e8cBwRZfKDjU/oU1Fbtg2pdJYI41muIWHaIhjQ3W 2j6dXj8AMhq8eRoMwGZaz8ZKP3JWzYCSIBemLkH36XOOrXoePwdgZyd9GIk4RQHh eXXkERVyW1bSrnA2XNq2TqvdesykV4ujKtDUgFwKGE/oZak+VjmtOHJlM5YoYKcY 4NgFQ62ihtGNdY02H/gFTLGzMCyxPeMLlKYhA6KgcdXW6oOuOmN+kCRj+6shspAr 2gIGRIt1cYUSv5aZxocQbwJcQrNHEqUqOApY80iIOYsTQ9kpiFPT306SDEX03dEb XDjG6KR1Ty/GGkPnGYMj/SRryFGeMDX6uXCB8ShCanHXBZjTrSTB/SkelyEG4w/D Qn2azTLsZfSD0m5bXnXG1FzYWp2hiJlEk8mXxv9D1XmyRSkqTpdzghMdmNWkRDGJ XYgSnqQe0rBmPQboIr8C =U6t3 -----END PGP SIGNATURE----- --DMNkN7kt0u2t2J4KlMio8mwJQ6mMt015t--