It supposed to be safe to modify static branches after jump_label_init().
But, because static key modifying code eventually calls text_poke() it can
end up accessing a struct page which has not been initialized yet.
Here is how to quickly reproduce the problem. Insert code like this
into init/main.c:
| +static DEFINE_STATIC_KEY_FALSE(__test);
| asmlinkage __visible void __init start_kernel(void)
| {
| char *command_line;
|@@ -587,6 +609,10 @@ asmlinkage __visible void __init start_kernel(void)
| vfs_caches_init_early();
| sort_main_extable();
| trap_init();
|+ {
|+ static_branch_enable(&__test);
|+ WARN_ON(!static_branch_likely(&__test));
|+ }
| mm_init();
The following warnings show-up:
WARNING: CPU: 0 PID: 0 at arch/x86/kernel/alternative.c:701 text_poke+0x20d/0x230
RIP: 0010:text_poke+0x20d/0x230
Call Trace:
? text_poke_bp+0x50/0xda
? arch_jump_label_transform+0x89/0xe0
? __jump_label_update+0x78/0xb0
? static_key_enable_cpuslocked+0x4d/0x80
? static_key_enable+0x11/0x20
? start_kernel+0x23e/0x4c8
? secondary_startup_64+0xa5/0xb0
---[ end trace abdc99c031b8a90a ]---
If the code above is moved after mm_init(), no warning is shown, as struct
pages are initialized during handover from memblock.
Use text_poke_early() in static branching until early boot IRQs are enabled
and from there switch to text_poke. Also, ensure text_poke() is never
invoked when unitialized memory access may happen by using adding a
!after_bootmem assertion.
Signed-off-by: Pavel Tatashin <pasha.tatashin@oracle.com>
Cc: steven.sistare@oracle.com
Cc: daniel.m.jordan@oracle.com
Cc: linux@armlinux.org.uk
Cc: schwidefsky@de.ibm.com
Cc: heiko.carstens@de.ibm.com
Cc: john.stultz@linaro.org
Cc: sboyd@codeaurora.org
Cc: hpa@zytor.com
Cc: douly.fnst@cn.fujitsu.com
Cc: peterz@infradead.org
Cc: prarit@redhat.com
Cc: feng.tang@intel.com
Cc: pmladek@suse.com
Cc: gnomes@lxorguk.ukuu.org.uk
Cc: linux-s390@vger.kernel.org
Cc: boris.ostrovsky@oracle.com
Cc: jgross@suse.com
Cc: pbonzini@redhat.com
Link: https://lkml.kernel.org/r/20180719205545.16512-9-pasha.tatashin@oracle.com
...instead of kstrtol with a dirty cast.
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Miroslav Lichvar <mlichvar@redhat.com>
Cc: Richard Cochran <richardcochran@gmail.com>
Cc: Prarit Bhargava <prarit@redhat.com>
Cc: Stephen Boyd <sboyd@kernel.org>
Signed-off-by: Ondrej Mosnacek <omosnace@redhat.com>
Signed-off-by: John Stultz <john.stultz@linaro.org>
The 'ts' argument of process_adj_status() and process_adjtimex_modes()
is unused and can be safely removed.
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Miroslav Lichvar <mlichvar@redhat.com>
Cc: Richard Cochran <richardcochran@gmail.com>
Cc: Prarit Bhargava <prarit@redhat.com>
Cc: Stephen Boyd <sboyd@kernel.org>
Signed-off-by: Ondrej Mosnacek <omosnace@redhat.com>
Signed-off-by: John Stultz <john.stultz@linaro.org>
Simplify waiting for the contained link to become inactive, removing the
indirection to a unnecessary DPC-specific handler.
Signed-off-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Sinan Kaya <okaya@kernel.org>
Reviewed-by: Oza Pawandeep <poza@codeaurora.org>
Remove the work struct that was being used to handle a DPC event and use a
threaded IRQ instead.
Signed-off-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Sinan Kaya <okaya@kernel.org>
Reviewed-by: Oza Pawandeep <poza@codeaurora.org>
A DPC enabled device suppresses ERR_(NON)FATAL messages, preventing the AER
handler from reporting error details. If the DPC trigger reason says the
downstream port detected the error, collect the AER uncorrectable status
for logging, then clear the status.
Signed-off-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Sinan Kaya <okaya@kernel.org>
Reviewed-by: Oza Pawandeep <poza@codeaurora.org>
We don't need to save the rp pio status across multiple contexts as all
DPC event handling occurs in a single work queue context.
Signed-off-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Sinan Kaya <okaya@kernel.org>
Reviewed-by: Oza Pawandeep <poza@codeaurora.org>
Move all event handling to the existing work queue, which will
make it simpler to pass event information to the handler.
Signed-off-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Sinan Kaya <okaya@kernel.org>
Reviewed-by: Oza Pawandeep <poza@codeaurora.org>
Now that the DPC driver clears the interrupt status before exiting the
IRQ handler, we don't need to abuse the DPC control register to know if
a shared interrupt is for a new DPC event: a DPC port can not trigger
a second interrupt until the host clears the trigger status later in the
work queue handler.
Signed-off-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Sinan Kaya <okaya@kernel.org>
Reviewed-by: Oza Pawandeep <poza@codeaurora.org>
According to the documentation, "pcie_ports=native", linux should use
native AER and DPC services. While that is true for the _OSC method
parsing, this is not the only place that is checked. Should the HEST
list PCIe ports as firmware-first, linux will not use native services.
This happens because aer_acpi_firmware_first() doesn't take 'pcie_ports'
into account. This is wrong. DPC uses the same logic when it decides
whether to load or not, so fixing this also fixes DPC not loading.
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
[bhelgaas: return "false" from bool function (from kbuild robot)]
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Add sysfs attributes for rootport statistics (that are cumulative of all
the ERR_* messages seen on this PCI hierarchy).
Signed-off-by: Rajat Jain <rajatja@google.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Add sysfs attributes to provide total and breakdown of the AERs seen,
into different type of correctable, fatal and nonfatal errors:
/sys/bus/pci/devices/<dev>/aer_dev_correctable
/sys/bus/pci/devices/<dev>/aer_dev_fatal
/sys/bus/pci/devices/<dev>/aer_dev_nonfatal
Signed-off-by: Rajat Jain <rajatja@google.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Define a structure to hold the AER statistics. There are 2 groups of
statistics: dev_* counters that are to be collected for all AER capable
devices and rootport_* counters that are collected for all (AER capable)
rootports only. Allocate and free this structure when device is added or
released (thus counters survive the lifetime of the device).
Signed-off-by: Rajat Jain <rajatja@google.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Since pci_aer_init() and pci_no_aer() are used only internally, move their
declarations to the PCI internal header file. Also, no one cares about
return value of pci_aer_init(), so make it void.
Signed-off-by: Rajat Jain <rajatja@google.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
lspci uses abbreviated naming for AER error strings. Adopt the same naming
convention for the AER printing so they match.
Signed-off-by: Tyler Baicar <tbaicar@codeaurora.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Oza Pawandeep <poza@codeaurora.org>
Export some common AER functions and structures for other PCI core drivers
to use. Since this is making the function externally visible inside the
PCI core, prepend "aer_" to the function name.
Signed-off-by: Keith Busch <keith.busch@intel.com>
[bhelgaas: move AER declarations from linux/aer.h to drivers/pci/pci.h]
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Sinan Kaya <okaya@kernel.org>
Reviewed-by: Oza Pawandeep <poza@codeaurora.org>
We currently don't store the on-die ECC state (enabled/disabled) which
might force us to re-disable the engine even if it's already been
disabled after we've read the page in raw mode to count the actual
number of bitflips.
Add an "enabled" field to struct micron_on_die_ecc to keep track of the
ECC state.
Signed-off-by: Boris Brezillon <boris.brezillon@bootlin.com>
Tested-by: Chris Packham <chris.packham@alliedtelesis.co.nz>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Some chips have their on-die ECC forcibly enabled, there's no point in
trying to enable/disable the ECC engine in that case.
Signed-off-by: Boris Brezillon <boris.brezillon@bootlin.com>
Tested-by: Chris Packham <chris.packham@alliedtelesis.co.nz>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
The MT29F2Gxxx chips with 4bits/512byte on-die ECC let us know when
some bitflips were corrected by the on-die ECC, but they do not report
the actual number of bitflips that were present in the data+ECC chunk.
We initially decided to always return ecc->strength to avoid re-reading
the page in raw mode + comparing it to the corrected buffer to extract
the real number of bitflips, but this forces UBI to move data around as
soon as one bitflip is present in a page.
This not only wears the NAND out faster, but also degrades
performances, since reading a full PEB + writing it back to a different
PEB + erasing the old one is much more expensive than re-reading the
faulty page in raw mode and comparing it to the corrected buffer.
In most cases, the actual number of bitflips does not exceed the
bitflips threshold, and UBI won't have to move data around. Otherwise,
we can assume the time spent re-reading the page and doing the
comparison is negligible compared to the time UBI spends moving a full
PEB to another PEB.
Note that this logic is not applied to chips with 8bits/512byte on-die
ECC, because those chips provide fine-grained information (the maximum
error is 1 bit, and it will not force UBI to move blocks around at the
first bitflip).
Suggested-by: Bean Huo <beanhuo@micron.com>
Signed-off-by: Boris Brezillon <boris.brezillon@bootlin.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Some Micron NAND chips have on-die ECC forceably enabled. Allow such
chips to be used as long as the controller has set chip->ecc.mode to
NAND_ECC_ON_DIE.
Signed-off-by: Chris Packham <chris.packham@alliedtelesis.co.nz>
Reviewed-by: Boris Brezillon <boris.brezillon@bootlin.com>
Signed-off-by: Boris Brezillon <boris.brezillon@bootlin.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Some implementations simply can't disable their ECC engine. Expose
helpers returning -ENOTSUPP so that the caller knows that raw accesses
are not supported instead of silently falling back to non-raw
accessors.
Signed-off-by: Boris Brezillon <boris.brezillon@bootlin.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Micron MT29F1G08ABAFAWP-ITE:F supports an on-die ECC with 8 bits
per 512 bytes. Add support for this combination.
Signed-off-by: Chris Packham <chris.packham@alliedtelesis.co.nz>
Reviewed-by: Boris Brezillon <boris.brezillon@bootlin.com>
Signed-off-by: Boris Brezillon <boris.brezillon@bootlin.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Basing the "mandatory on-die" detection on ID byte 2 does not work,
because Micron has plenty of NANDs using the same device ID code, and
not all of them have forcibly enabled on-die ECC.
Since the "Array Operation" feature does not provide the "ECC
enabled/disabled" bit when the ECC can't be disabled, let's try to use
the "ECC enabled/disabled" bit in the READ_ID bytes.
It seems that this bit is dynamically updated on NANDs where on-die ECC
can freely be enabled/disabled, so let's hope it stays at one when we
have a NAND with on-die ECC forcibly enabled.
Fixes: 51f3b3970a8c ("mtd: rawnand: micron: detect forced on-die ECC")
Signed-off-by: Boris Brezillon <boris.brezillon@bootlin.com>
Tested-by: Chris Packham <chris.packham@alliedtelesis.co.nz>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
chip->read_buf is left unassigned since commit 4da712e702 ("mtd: nand:
fsmc: use ->exec_op()"), leading to a NULL pointer dereference when it's
called from fsmc_read_page_hwecc(). Fix that by using the appropriate
helper to read data out of the NAND.
Fixes: 4da712e702 ("mtd: nand: fsmc: use ->exec_op()")
Cc: <stable@vger.kernel.org>
Signed-off-by: Boris Brezillon <boris.brezillon@bootlin.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Compile-testing this driver on x86 caused a link error:
ERROR: "__phys_to_dma" [drivers/mtd/nand/raw/qcom_nandc.ko] undefined!
The problem here is that the driver attempts to convert the physical
address into the DMA controller as a dma_addr_t and calls phys_to_dma()
to do the conversion.
The correct way to do the conversion is using the dma mapping interfaces.
Fixes: c76b78d8ec ("mtd: nand: Qualcomm NAND controller driver")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Reviewed-by: Boris Brezillon <boris.brezillon@bootlin.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>