LOAD_FAULT_PTE macro is expected to set r2 with faulting vaddr.
However in case of CONFIG_ARC_DBG_TLB_MISS_COUNT, it was getting
clobbered with statistics collection code.
Fix latter by using a different register.
Note that only I-TLB Miss handler was potentially affected.
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
* No need to check for READ access in I-TLB Miss handler
* Redundant PAGE_PRESENT update in PTE
Post TLB entry installation, in updating PTE for software accessed/dity
bits, no need to update PAGE_PRESENT since it will already be set.
Infact the entry won't have installed if !PAGE_PRESENT.
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
* DWARF unwinder related
+ Force DWARF2 compliant .debug_frame (gcc 4.8 defaults to DWARF4
which kernel unwinder can't grok).
+ Discard the additional .eh_frame generated
+ Discard the dwarf4 debug info generated by -gdwarf-2 for normal
no debug case
* 4.8 already uses arc600 multilibs for -mno-mpy
* switch to using uclibc compiler (to get -mmedium-calls and -mno-sdata)
and also since buildroot can only use 1 toolchain
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Otherwise early boot exceptions such as instructions errors due to
configuration mismatch between kernel and hardware go off to la-la land,
as opposed to hitting the handler and panic()'ing properly.
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
With ECR now part of pt_regs
* No need to propagate from lowest asm handlers as arg
* No need to save it in tsk->thread.cause_code
* Avoid bit chopping to access the bit-fields
More code consolidation, cleanup
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
pt_regs->event was set with artificial values to identify the low level
system event (syscall trap / breakpoint trap / exceptions / interrupts)
With r8 saving out of the way, the full word can be used to save real
ECR (Exception Cause Register) which helps idenify the event naturally,
including additional info such as cause code, param.
Only for Interrupts, where ECR is not applicable, do we resort to
synthetic non ECR values.
SAVE_ALL_TRAP/EXCEPTIONS can now be merged as they both use ECR with
different runtime values.
The ptrace helpers now use the sub-fields of ECR to distinguish the
events (e.g. vector 0x25 is trap, param 0 is syscall...)
The following benefits will follow:
(1) This centralizes the location of where ECR is saved and will allow
the cleanup of task->thread.cause_code ECR placeholder which is set
in non-uniform way. Then ARC VM code can safely rely on it being
there for purpose of finer grained VM_EXEC dcache flush (based on
exec fault: I-TLB Miss)
(2) Further, ECR being passed around from low level handlers as arg can
be eliminated as it is part of standard reg-file in pt_regs
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Historically, pt_regs have had orig_r8, an overloaded container for
(1) backup copy of r8 (syscall number Trap Exceptions)
(2) additional system state: (syscall/Exception/Interrupt)
There is no point in keeping (1) since syscall number is never clobbered
in-place, in pt_regs, unlike r0 which duals as first syscall arg as well
as syscall return value and in case of syscall restart, the orig arg0
needs restoring (from orig_r0) after having been updated in-place with
syscall ret value.
This further paves way to convert (2) to contain ECR itself (rather than
current madeup values)
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
(This is a VERY IMP change for low level interrupt/exception handling)
-----------------------------------------------------------------------
WHAT
-----------------------------------------------------------------------
* User 25 now saved in pt_regs->user_r25 (vs. tsk->thread_info.user_r25)
* This allows Low level interrupt code to unconditionally save r25
(vs. the prev version which would only do it for U->K transition).
Ofcourse for nested interrupts, only the pt_regs->user_r25 of
bottom-most frame is useful.
* simplifies the interrupt prologue/epilogue
* Needed for ARCv2 ISA code and done here to keep design similar with
ARCompact event handling
-----------------------------------------------------------------------
WHY
-------------------------------------------------------------------------
With CONFIG_ARC_CURR_IN_REG, r25 is used to cache "current" task pointer
in kernel mode. So when entering kernel mode from User Mode
- user r25 is specially safe-kept (it being a callee reg is NOT part of
pt_regs which are saved by default on each interrupt/trap/exception)
- r25 loaded with current task pointer.
Further, if interrupt was taken in kernel mode, this is skipped since we
know that r25 already has valid "current" pointer.
With 2 level of interrupts in ARCompact ISA, detecting this is difficult
but still possible, since we could be in kernel mode but r25 not already saved
(in fact the stack itself might not have been switched).
A. User mode
B. L1 IRQ taken
C. L2 IRQ taken (while on 1st line of L1 ISR)
So in #C, although in kernel mode, r25 not saved (infact SP not
switched at all)
Given that ARcompact has manual stack switching, we could use a bit of
trickey - The low level code would make sure that SP is only set to kernel
mode value at the very end (after saving r25). So a non kernel mode SP,
even if in kernel mode, meant r25 was NOT saved.
The same paradigm won't work in ARCv2 ISA since SP is auto-switched so
it's setting can't be delayed/constrained.
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
This paves way for further simplifications.
There's an overhead of 1 insn for the non-common case of interrupt taken
from kernel mode.
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
* use artificial PUSH/POP contructs for CORE Reg save/restore to stack
* use artificial PUSHAX/POPAX contructs for Auxiliary Space regs
* macro'ize multiple copies of callee-reg-save/restore (SAVE_R13_TO_R24)
* use BIC insn for inverse-and operation
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
This is trickier than prev two:
* context switching code saves kernel mode callee regs in the format of
struct callee_regs thus needs adjustment. This also reduces the height
of topmost kernel stack frame by 1 word.
* Since kernel stack unwinder is sensitive to height of topmost kernel
stack frame, that needs a word of adjustment too.
ptrace needs a bit of updating since pt_regs now diverges from
user_regs_struct.
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Historically, pt_regs would end at offset of 1 word from end of stack
page.
----------------- -> START of page (task->stack)
| |
| thread_info |
-----------------
| |
^ ~ ~
| ~ ~
| | |
| | | <---- pt_regs used to END here
-----------------
| 1 word GUTTER |
----------------- -> End of page (START of kernel stack)
This required special "one-off" considerations in low level code.
The root cause is very likely assumption of "empty" SP by the original
ARC kernel hackers, despite ARC700 always been "full" SP.
So finally RIP one word gutter !
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
This can be ascertained within do_page_fault() since it gets the full
ECR (Exception Cause Register).
Further, for both the callers of do_page_fault(): Prot-V / D-TLB-Miss,
the cause sub-fields in ECR are same for same type of access, making the
code much more simpler.
D-TLB-Miss [LD] 0x00_21_01_00
Prot-V [LD] 0x00_23_01_00
^^
D-TLB-Miss [ST] 0x00_21_02_00
Prot-V [ST] 0x00_23_02_00
^^
D-TLB-Miss [EX] 0x00_21_03_00
Prot-V [EX] 0x00_23_03_00
^^
This helps code consolidation, which is even better when moving code from
assembler to "C".
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
1. For VM_EXEC based delayed dcache/icache flush, reduces the number of
flushes.
2. Makes this security feature ON by default rather than OFF before.
3. Applications can use mprotect() to selectively override this.
4. ELF binaries have a GNU_STACK segment which can easily override the
kernel default permissions.
For nested-functions/trampolines, gcc already auto-enables executable
stack in elf. Others needing this can use -Wl,-z,execstack option.
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Non-congruent SRC page in copy_user_page() is dcache clean in the end -
so record that fact, to avoid a subsequent extraneous flush.
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
* Number of (i|d)cache ways can be retrieved from BCRs and hence no need
to cross check with with built-in constants
* Use of IS_ENABLED() to check for a Kconfig option
* is_not_cache_aligned() not used anymore
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
* Move the various sub-system defines/types into relevant files/functions
(reduces compilation time)
* move CPU specific stuff out of asm/tlb.h into asm/mmu.h
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
This fixes the following:
- CONFIG_ARC_SERIAL_BAUD is only defined when CONFIG_SERIAL_ARC is defined.
Make sure that it isn't referenced otherwise.
- There is no use for initializing arc_uart_info[] when CONFIG_SERIAL_ARC is
not defined.
[vgupta: tweaked changelog title, used IS_ENABLED() kconfig helper]
Signed-off-by: Mischa Jonker <mjonker@synopsys.com>
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Another week, another batch of fixes for arm-soc platforms.
Again, nothing controversial. A few more than would be ideal, but all
are valid fixes. In particular the prima2 panic patch is critical since
it fixes a problem where multiplatform kernels panic on all but prima2
hardware.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
iQIcBAABAgAGBQJRvKO7AAoJEIwa5zzehBx3AE4P+wdAbXkBzt5L/MIlg2vgfS4O
5snAL6f/k1Yar7kJFL8o8jQ+DIPPo/Lq+kwJ+ZLOh4DkTIzd/JiJE5cGRlDwMF0R
KCbZsvxDf7kYapor7Fg8YHDDgDfZNCvudkBcmRHMEWCJVRjVbXxqrTIRTxNqLZVk
WyuZHKmLFj0/aY8K3Z+BhY8klWVmy6QsT8Cg+1la7qLXEoazXaEqLCSRhPSmvwCy
3ykPpmLdt/1MzVCgq+llNuT7C9N9buOw/bdb0JCYvGWiRxnLi2ee+pimGfOLIYDx
/eeMr091r7OOhFBsIPd8fzfzd5A2cEeR2USPhGEpFCKpPKv5gOF8m7ku2w62WDcN
PrMsYy2N4idxOBpsvLqxx9bkZ8nRqSj2anX6GLqHdL2v4k4pS/kufIihdgjd5Oja
xUvlzk0gVRc8wqENyluRYdrocm307p/URLu/s2EOTQ3ZOZOZ3PRjbEelxoKCp6Pd
RvADywebFMVL6Qyp4D996rvPgWYFnkSIkiSZQwlUHEMUHdE5JPbxcfBGTO5l8tKI
EehgZ6Xc28Jv/pGt5a5DbwNjUN4NPLe6jhMXEGEnOG0SZ+43Xfmlwj93Y96hReJq
+/8AMXx/NaPNrdSoqZBR4GJ6KBM/DATx23dr9VMskX5rBGxYzBVln9ck9boYnqyq
hg5z53tz0uTBtqot6W9v
=c555
-----END PGP SIGNATURE-----
Merge tag 'fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc
Pull ARM SoC fixes from Olof Johansson:
"These are a little later than I planned on since I got caught up with
handling merges for 3.11 most of the week.
Another week, another batch of fixes for arm-soc platforms.
Again, nothing controversial. A few more than would be ideal, but all
are valid fixes. In particular the prima2 panic patch is critical
since it fixes a problem where multiplatform kernels panic on all but
prima2 hardware."
* tag 'fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc:
ARM: SAMSUNG: pm: Adjust for pinctrl- and DT-enabled platforms
ARM: prima2: fix incorrect panic usage
arm: mvebu: armada-xp-{gp,openblocks-ax3-4}: specify PCIe range
ARM: Kirkwood: handle mv88f6282 cpu in __kirkwood_variant().
ARM: omap3: clock: fix wrong container_of in clock36xx.c
ARM: dts: OMAP5: Fix missing PWM capability to timer nodes
ARM: dts: omap4-panda|sdp: Fix mux for twl6030 IRQ pin and msecure line
ARM: dts: AM33xx: Fix properties on gpmc node
arm: omap2: fix AM33xx hwmod infos for UART2
ARM: OMAP3: Fix iva2_pwrdm settings for 3703
Pull networking fixes from David Miller:
1) Fix RTNL locking in batman-adv, from Matthias Schiffer.
2) Don't allow non-passthrough macvlan devices to set NOPROMISC via
netlink, otherwise we can end up with corrupted promisc counter
values on the device. From Michael S Tsirkin.
3) Fix stmmac driver build with debugging defines enabled, from Dinh
Nguyen.
4) Make sure name string we give in socket address in AF_PACKET is NULL
terminated, from Daniel Borkmann.
5) Fix leaking of two uninitialized bytes of memory to userspace in
l2tp, from Guillaume Nault.
6) Clear IPCB(skb) before tunneling otherwise we touch dangling IP
options state and crash. From Saurabh Mohan.
7) Fix suspend/resume for davinci_mdio by using suspend_late and
resume_early. From Mugunthan V N.
8) Don't tag ip_tunnel_init_net and ip_tunnel_delete_net with
__net_{init,exit}, they can be called outside of those contexts.
From Eric Dumazet.
9) Fix RX length error in sh_eth driver, from Yoshihiro Shimoda.
10) Fix missing sctp_outq initialization in some code paths of SCTP
stack, from Neil Horman.
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (21 commits)
sctp: fully initialize sctp_outq in sctp_outq_init
netiucv: Hold rtnl between name allocation and device registration.
tulip: Properly check dma mapping result
net: sh_eth: fix incorrect RX length error if R8A7740
ip_tunnel: remove __net_init/exit from exported functions
drivers: net: davinci_mdio: restore mdio clk divider in mdio resume
drivers: net: davinci_mdio: moving mdio resume earlier than cpsw ethernet driver
net/ipv4: ip_vti clear skb cb before tunneling.
tg3: Wait for boot code to finish after power on
l2tp: Fix sendmsg() return value
l2tp: Fix PPP header erasure and memory leak
bonding: fix igmp_retrans type and two related races
bonding: reset master mac on first enslave failure
packet: packet_getname_spkt: make sure string is always 0-terminated
net: ethernet: stmicro: stmmac: Fix compile error when STMMAC_XMIT_DEBUG used
be2net: Fix 32-bit DMA Mask handling
xen-netback: don't de-reference vif pointer after having called xenvif_put()
macvlan: don't touch promisc without passthrough
batman-adv: Don't handle address updates when bla is disabled
batman-adv: forward late OGMs from best next hop
...
Pull powerpc fixes from Benjamin Herrenschmidt:
"So here are 3 fixes still for 3.10. Fixes are simple, bugs are nasty
(though not recent regressions, nasty enough) and all targeted at
stable"
* 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc:
powerpc: Fix missing/delayed calls to irq_work
powerpc: Fix emulation of illegal instructions on PowerNV platform
powerpc: Fix stack overflow crash in resume_kernel when ftracing
Thanks to commit f91eb62f71 ("init: scream bloody murder if interrupts
are enabled too early"), "bloody murder" is now being screamed.
With a MIPS OCTEON config, we use on_each_cpu() in our
irq_chip.irq_bus_sync_unlock() function. This gets called in early as a
result of the time_init() call. Because the !SMP version of
on_each_cpu() unconditionally enables irqs, we get:
WARNING: at init/main.c:560 start_kernel+0x250/0x410()
Interrupts were enabled early
CPU: 0 PID: 0 Comm: swapper Not tainted 3.10.0-rc5-Cavium-Octeon+ #801
Call Trace:
show_stack+0x68/0x80
warn_slowpath_common+0x78/0xb0
warn_slowpath_fmt+0x38/0x48
start_kernel+0x250/0x410
Suggested fix: Do what we already do in the SMP version of
on_each_cpu(), and use local_irq_save/local_irq_restore. Because we
need a flags variable, make it a static inline to avoid name space
issues.
[ Change from v1: Convert on_each_cpu to a static inline function, add
#include <linux/irqflags.h> to avoid build breakage on some files.
on_each_cpu_mask() and on_each_cpu_cond() suffer the same problem as
on_each_cpu(), but they are not causing !SMP bugs for me, so I will
defer changing them to a less urgent patch. ]
Signed-off-by: David Daney <david.daney@cavium.com>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull VFS fixes from Al Viro:
"Several fixes + obvious cleanup (you've missed a couple of open-coded
can_lookup() back then)"
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
snd_pcm_link(): fix a leak...
use can_lookup() instead of direct checks of ->i_op->lookup
move exit_task_namespaces() outside of exit_notify()
fput: task_work_add() can fail if the caller has passed exit_task_work()
ncpfs: fix rmdir returns Device or resource busy
- Remove noisy warnings about experimental support which spams the logs
- Add padding to align directory and attr structures correctly
- Set block number on child buffer on a root btree split
- Disable verifiers during log recovery for non-CRC filesystems
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)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=bijX
-----END PGP SIGNATURE-----
Merge tag 'for-linus-v3.10-rc6' of git://oss.sgi.com/xfs/xfs
Pull xfs fixes from Ben Myers:
- Remove noisy warnings about experimental support which spams the logs
- Add padding to align directory and attr structures correctly
- Set block number on child buffer on a root btree split
- Disable verifiers during log recovery for non-CRC filesystems
* tag 'for-linus-v3.10-rc6' of git://oss.sgi.com/xfs/xfs:
xfs: don't shutdown log recovery on validation errors
xfs: ensure btree root split sets blkno correctly
xfs: fix implicit padding in directory and attr CRC formats
xfs: don't emit v5 superblock warnings on write
Here are some small mei driver fixes for 3.10-rc6 that fix some reported
problems.
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.19 (GNU/Linux)
iEYEABECAAYFAlG7RgwACgkQMUfUDdst+ynQjgCcCed/djDG6rEk8OHNwtH0qsGE
3o4AnjEW26lnses9dpudJOzhFGggCKJt
=wN5b
-----END PGP SIGNATURE-----
Merge tag 'char-misc-3.10-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc
Pull char / misc fixes from Greg Kroah-Hartman:
"Here are some small mei driver fixes for 3.10-rc6 that fix some
reported problems"
* tag 'char-misc-3.10-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc:
mei: me: clear interrupts on the resume path
mei: nfc: fix nfc device freeing
mei: init: Flush scheduled work before resetting the device
Here are some small USB driver fixes that resolve some reported problems
for 3.10-rc6
Nothing major, just 3 USB serial driver fixes, and two chipidea fixes.
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.19 (GNU/Linux)
iEYEABECAAYFAlG7Rq0ACgkQMUfUDdst+ykKmwCg0mta+HehUtBYrhLJGq9uADix
0YMAn1hEPP26BhVl/7a6GL+s8UoSVFxo
=9Vkq
-----END PGP SIGNATURE-----
Merge tag 'usb-3.10-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb
Pull USB fixes from Greg Kroah-Hartman:
"Here are some small USB driver fixes that resolve some reported
problems for 3.10-rc6
Nothing major, just 3 USB serial driver fixes, and two chipidea fixes"
* tag 'usb-3.10-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb:
usb: chipidea: fix id change handling
usb: chipidea: fix no transceiver case
USB: pl2303: fix device initialisation at open
USB: spcp8x5: fix device initialisation at open
USB: f81232: fix device initialisation at open
When replaying interrupts (as a result of the interrupt occurring
while soft-disabled), in the case of the decrementer, we are exclusively
testing for a pending timer target. However we also use decrementer
interrupts to trigger the new "irq_work", which in this case would
be missed.
This change the logic to force a replay in both cases of a timer
boundary reached and a decrementer interrupt having actually occurred
while disabled. The former test is still useful to catch cases where
a CPU having been hard-disabled for a long time completely misses the
interrupt due to a decrementer rollover.
CC: <stable@vger.kernel.org> [v3.4+]
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Tested-by: Steven Rostedt <rostedt@goodmis.org>
Normally, the kernel emulates a few instructions that are unimplemented
on some processors (e.g. the old dcba instruction), or privileged (e.g.
mfpvr). The emulation of unimplemented instructions is currently not
working on the PowerNV platform. The reason is that on these machines,
unimplemented and illegal instructions cause a hypervisor emulation
assist interrupt, rather than a program interrupt as on older CPUs.
Our vector for the emulation assist interrupt just calls
program_check_exception() directly, without setting the bit in SRR1
that indicates an illegal instruction interrupt. This fixes it by
making the emulation assist interrupt set that bit before calling
program_check_interrupt(). With this, old programs that use no-longer
implemented instructions such as dcba now work again.
CC: <stable@vger.kernel.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
It's possible for us to crash when running with ftrace enabled, eg:
Bad kernel stack pointer bffffd12 at c00000000000a454
cpu 0x3: Vector: 300 (Data Access) at [c00000000ffe3d40]
pc: c00000000000a454: resume_kernel+0x34/0x60
lr: c00000000000335c: performance_monitor_common+0x15c/0x180
sp: bffffd12
msr: 8000000000001032
dar: bffffd12
dsisr: 42000000
If we look at current's stack (paca->__current->stack) we see it is
equal to c0000002ecab0000. Our stack is 16K, and comparing to
paca->kstack (c0000002ecab3e30) we can see that we have overflowed our
kernel stack. This leads to us writing over our struct thread_info, and
in this case we have corrupted thread_info->flags and set
_TIF_EMULATE_STACK_STORE.
Dumping the stack we see:
3:mon> t c0000002ecab0000
[c0000002ecab0000] c00000000002131c .performance_monitor_exception+0x5c/0x70
[c0000002ecab0080] c00000000000335c performance_monitor_common+0x15c/0x180
--- Exception: f01 (Performance Monitor) at c0000000000fb2ec .trace_hardirqs_off+0x1c/0x30
[c0000002ecab0370] c00000000016fdb0 .trace_graph_entry+0xb0/0x280 (unreliable)
[c0000002ecab0410] c00000000003d038 .prepare_ftrace_return+0x98/0x130
[c0000002ecab04b0] c00000000000a920 .ftrace_graph_caller+0x14/0x28
[c0000002ecab0520] c0000000000d6b58 .idle_cpu+0x18/0x90
[c0000002ecab05a0] c00000000000a934 .return_to_handler+0x0/0x34
[c0000002ecab0620] c00000000001e660 .timer_interrupt+0x160/0x300
[c0000002ecab06d0] c0000000000025dc decrementer_common+0x15c/0x180
--- Exception: 901 (Decrementer) at c0000000000104d4 .arch_local_irq_restore+0x74/0xa0
[c0000002ecab09c0] c0000000000fe044 .trace_hardirqs_on+0x14/0x30 (unreliable)
[c0000002ecab0fb0] c00000000016fe3c .trace_graph_entry+0x13c/0x280
[c0000002ecab1050] c00000000003d038 .prepare_ftrace_return+0x98/0x130
[c0000002ecab10f0] c00000000000a920 .ftrace_graph_caller+0x14/0x28
[c0000002ecab1160] c0000000000161f0 .__ppc64_runlatch_on+0x10/0x40
[c0000002ecab11d0] c00000000000a934 .return_to_handler+0x0/0x34
--- Exception: 901 (Decrementer) at c0000000000104d4 .arch_local_irq_restore+0x74/0xa0
... and so on
__ppc64_runlatch_on() is called from RUNLATCH_ON in the exception entry
path. At that point the irq state is not consistent, ie. interrupts are
hard disabled (by the exception entry), but the paca soft-enabled flag
may be out of sync.
This leads to the local_irq_restore() in trace_graph_entry() actually
enabling interrupts, which we do not want. Because we have not yet
reprogrammed the decrementer we immediately take another decrementer
exception, and recurse.
The fix is twofold. Firstly make sure we call DISABLE_INTS before
calling RUNLATCH_ON. The badly named DISABLE_INTS actually reconciles
the irq state in the paca with the hardware, making it safe again to
call local_irq_save/restore().
Although that should be sufficient to fix the bug, we also mark the
runlatch routines as notrace. They are called very early in the
exception entry and we are asking for trouble tracing them. They are
also fairly uninteresting and tracing them just adds unnecessary
overhead.
[ This regression was introduced by fe1952fc0a
"powerpc: Rework runlatch code" by myself --BenH
]
CC: <stable@vger.kernel.org> [v3.4+]
Signed-off-by: Michael Ellerman <michael@ellerman.id.au>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
exit_notify() does exit_task_namespaces() after
forget_original_parent(). This was needed to ensure that ->nsproxy
can't be cleared prematurely, an exiting child we are going to
reparent can do do_notify_parent() and use the parent's (ours) pid_ns.
However, after 32084504 "pidns: use task_active_pid_ns in
do_notify_parent" ->nsproxy != NULL is no longer needed, we rely
on task_active_pid_ns().
Move exit_task_namespaces() from exit_notify() to do_exit(), after
exit_fs() and before exit_task_work().
This solves the problem reported by Andrey, free_ipc_ns()->shm_destroy()
does fput() which needs task_work_add().
Note: this particular problem can be fixed if we change fput(), and
that change makes sense anyway. But there is another reason to move
the callsite. The original reason for exit_task_namespaces() from
the middle of exit_notify() was subtle and it has already gone away,
now this looks confusing. And this allows us do simplify exit_notify(),
we can avoid unlock/lock(tasklist) and we can use ->exit_state instead
of PF_EXITING in forget_original_parent().
Reported-by: Andrey Vagin <avagin@openvz.org>
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Acked-by: "Eric W. Biederman" <ebiederm@xmission.com>
Acked-by: Andrey Vagin <avagin@openvz.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
fput() assumes that it can't be called after exit_task_work() but
this is not true, for example free_ipc_ns()->shm_destroy() can do
this. In this case fput() silently leaks the file.
Change it to fallback to delayed_fput_work if task_work_add() fails.
The patch looks complicated but it is not, it changes the code from
if (PF_KTHREAD) {
schedule_work(...);
return;
}
task_work_add(...)
to
if (!PF_KTHREAD) {
if (!task_work_add(...))
return;
/* fallback */
}
schedule_work(...);
As for shm_destroy() in particular, we could make another fix but I
think this change makes sense anyway. There could be another similar
user, it is not safe to assume that task_work_add() can't fail.
Reported-by: Andrey Vagin <avagin@openvz.org>
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Unfortunately, we cannot guarantee that items logged multiple times
and replayed by log recovery do not take objects back in time. When
they are taken back in time, the go into an intermediate state which
is corrupt, and hence verification that occurs on this intermediate
state causes log recovery to abort with a corruption shutdown.
Instead of causing a shutdown and unmountable filesystem, don't
verify post-recovery items before they are written to disk. This is
less than optimal, but there is no way to detect this issue for
non-CRC filesystems If log recovery successfully completes, this
will be undone and the object will be consistent by subsequent
transactions that are replayed, so in most cases we don't need to
take drastic action.
For CRC enabled filesystems, leave the verifiers in place - we need
to call them to recalculate the CRCs on the objects anyway. This
recovery problem can be solved for such filesystems - we have a LSN
stamped in all metadata at writeback time that we can to determine
whether the item should be replayed or not. This is a separate piece
of work, so is not addressed by this patch.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Ben Myers <bpm@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
(cherry picked from commit 9222a9cf86)
For CRC enabled filesystems, the BMBT is rooted in an inode, so it
passes through a different code path on root splits than the
freespace and inode btrees. This is much less traversed by xfstests
than the other trees. When testing on a 1k block size filesystem,
I've been seeing ASSERT failures in generic/234 like:
XFS: Assertion failed: cur->bc_btnum != XFS_BTNUM_BMAP || cur->bc_private.b.allocated == 0, file: fs/xfs/xfs_btree.c, line: 317
which are generally preceded by a lblock check failure. I noticed
this in the bmbt stats:
$ pminfo -f xfs.btree.block_map
xfs.btree.block_map.lookup
value 39135
xfs.btree.block_map.compare
value 268432
xfs.btree.block_map.insrec
value 15786
xfs.btree.block_map.delrec
value 13884
xfs.btree.block_map.newroot
value 2
xfs.btree.block_map.killroot
value 0
.....
Very little coverage of root splits and merges. Indeed, on a 4k
filesystem, block_map.newroot and block_map.killroot are both zero.
i.e. the code is not exercised at all, and it's the only generic
btree infrastructure operation that is not exercised by a default run
of xfstests.
Turns out that on a 1k filesystem, generic/234 accounts for one of
those two root splits, and that is somewhat of a smoking gun. In
fact, it's the same problem we saw in the directory/attr code where
headers are memcpy()d from one block to another without updating the
self describing metadata.
Simple fix - when copying the header out of the root block, make
sure the block number is updated correctly.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Ben Myers <bpm@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
(cherry picked from commit ade1335afe)
Michael L. Semon has been testing CRC patches on a 32 bit system and
been seeing assert failures in the directory code from xfs/080.
Thanks to Michael's heroic efforts with printk debugging, we found
that the problem was that the last free space being left in the
directory structure was too small to fit a unused tag structure and
it was being corrupted and attempting to log a region out of bounds.
Hence the assert failure looked something like:
.....
#5 calling xfs_dir2_data_log_unused() 36 32
#1 4092 4095 4096
#2 8182 8183 4096
XFS: Assertion failed: first <= last && last < BBTOB(bp->b_length), file: fs/xfs/xfs_trans_buf.c, line: 568
Where #1 showed the first region of the dup being logged (i.e. the
last 4 bytes of a directory buffer) and #2 shows the corrupt values
being calculated from the length of the dup entry which overflowed
the size of the buffer.
It turns out that the problem was not in the logging code, nor in
the freespace handling code. It is an initial condition bug that
only shows up on 32 bit systems. When a new buffer is initialised,
where's the freespace that is set up:
[ 172.316249] calling xfs_dir2_leaf_addname() from xfs_dir_createname()
[ 172.316346] #9 calling xfs_dir2_data_log_unused()
[ 172.316351] #1 calling xfs_trans_log_buf() 60 63 4096
[ 172.316353] #2 calling xfs_trans_log_buf() 4094 4095 4096
Note the offset of the first region being logged? It's 60 bytes into
the buffer. Once I saw that, I pretty much knew that the bug was
going to be caused by this.
Essentially, all direct entries are rounded to 8 bytes in length,
and all entries start with an 8 byte alignment. This means that we
can decode inplace as variables are naturally aligned. With the
directory data supposedly starting on a 8 byte boundary, and all
entries padded to 8 bytes, the minimum freespace in a directory
block is supposed to be 8 bytes, which is large enough to fit a
unused data entry structure (6 bytes in size). The fact we only have
4 bytes of free space indicates a directory data block alignment
problem.
And what do you know - there's an implicit hole in the directory
data block header for the CRC format, which means the header is 60
byte on 32 bit intel systems and 64 bytes on 64 bit systems. Needs
padding. And while looking at the structures, I found the same
problem in the attr leaf header. Fix them both.
Note that this only affects 32 bit systems with CRCs enabled.
Everything else is just fine. Note that CRC enabled filesystems created
before this fix on such systems will not be readable with this fix
applied.
Reported-by: Michael L. Semon <mlsemon35@gmail.com>
Debugged-by: Michael L. Semon <mlsemon35@gmail.com>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Ben Myers <bpm@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
(cherry picked from commit 8a1fd2950e)
We write the superblock every 30s or so which results in the
verifier being called. Right now that results in this output
every 30s:
XFS (vda): Version 5 superblock detected. This kernel has EXPERIMENTAL support enabled!
Use of these features in this kernel is at your own risk!
And spamming the logs.
We don't need to check for whether we support v5 superblocks or
whether there are feature bits we don't support set as these are
only relevant when we first mount the filesytem. i.e. on superblock
read. Hence for the write verification we can just skip all the
checks (and hence verbose output) altogether.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
(cherry picked from commit 34510185ab)
