| Age | Commit message (Collapse) | Author |
|---|
|
This flag was used to support the PHB4 LSIs on P9 DD1 and we have
stopped supporting this CPU when DD2 came out. See skiboot commit:
https://github.com/open-power/skiboot/commit/0b0d15e3c170
Signed-off-by: Cédric Le Goater <clg@kaod.org>
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20201210171450.1933725-9-clg@kaod.org
|
|
PowerNV systems can handle up to 4K guests and 1M interrupt numbers
per chip. Increase the range of allowed interrupts to support a larger
number of guests.
Signed-off-by: Cédric Le Goater <clg@kaod.org>
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20201210171450.1933725-8-clg@kaod.org
|
|
Full state of the Linux interrupt descriptors can be dumped under
debugfs when compiled with CONFIG_GENERIC_IRQ_DEBUGFS. Add support for
the XIVE interrupt controller.
Signed-off-by: Cédric Le Goater <clg@kaod.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20201210171450.1933725-7-clg@kaod.org
|
|
We hope one day to handle multiple irq_domain in the XIVE driver.
Start simple by setting the name using the DT node.
Signed-off-by: Cédric Le Goater <clg@kaod.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20201210171450.1933725-6-clg@kaod.org
|
|
The XIVE driver deals with CPU IPIs in a peculiar way. Each CPU has
its own XIVE IPI interrupt allocated at the HW level, for PowerNV, or
at the hypervisor level for pSeries. In practice, these interrupts are
not always used. pSeries/PowerVM prefers local doorbells for local
threads since they are faster. On PowerNV, global doorbells are also
preferred for the same reason.
The mapping in the Linux is reduced to a single interrupt using HW
interrupt number 0 and a custom irq_chip to handle EOI. This can cause
performance issues in some benchmark (ipistorm) on multichip systems.
Clarify the use of the 0 value, it will help in improving multichip
support.
Signed-off-by: Cédric Le Goater <clg@kaod.org>
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20201210171450.1933725-4-clg@kaod.org
|
|
This is a simple cleanup to identify easily all flags of the XIVE
interrupt structure. The interrupts flagged with XIVE_IRQ_FLAG_NO_EOI
are the escalations used to wake up vCPUs in KVM. They are handled
very differently from the rest.
Signed-off-by: Cédric Le Goater <clg@kaod.org>
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20201210171450.1933725-3-clg@kaod.org
|
|
This is useful to track allocation of the HW resources on per guest
basis. Making sure IPIs are local to the chip of the vCPUs reduces
rerouting between interrupt controllers and gives better performance
in case of pinning. Checking the distribution of VP structures on the
chips also helps in reducing PowerBUS traffic.
[ clg: resurrected show_sources and reworked ouput ]
Signed-off-by: Greg Kurz <groug@kaod.org>
Signed-off-by: Cédric Le Goater <clg@kaod.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20201210171450.1933725-2-clg@kaod.org
|
|
Alexei Starovoitov says:
====================
pull-request: bpf 2020-12-10
The following pull-request contains BPF updates for your *net* tree.
We've added 21 non-merge commits during the last 12 day(s) which contain
a total of 21 files changed, 163 insertions(+), 88 deletions(-).
The main changes are:
1) Fix propagation of 32-bit signed bounds from 64-bit bounds, from Alexei.
2) Fix ring_buffer__poll() return value, from Andrii.
3) Fix race in lwt_bpf, from Cong.
4) Fix test_offload, from Toke.
5) Various xsk fixes.
Please consider pulling these changes from:
git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf.git
Thanks a lot!
Also thanks to reporters, reviewers and testers of commits in this pull-request:
Cong Wang, Hulk Robot, Jakub Kicinski, Jean-Philippe Brucker, John
Fastabend, Magnus Karlsson, Maxim Mikityanskiy, Yonghong Song
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Fix a possible hang in xdpsock that can occur when using multiple
threads. In this case, one or more of the threads might get stuck in
the while-loop in tx_only after the user has signaled the main thread
to stop execution. In this case, no more Tx packets will be sent, so a
thread might get stuck in the aforementioned while-loop. Fix this by
introducing a test inside the while-loop to check if the benchmark has
been terminated. If so, return from the function.
Fixes: cd9e72b6f210 ("samples/bpf: xdpsock: Add option to specify batch size")
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20201210163407.22066-1-magnus.karlsson@gmail.com
|
|
Mark tripped over the creative irqflags handling in the IO-APIC timer
delivery check which ends up doing:
local_irq_save(flags);
local_irq_enable();
local_irq_restore(flags);
which triggered a new consistency check he's working on required for
replacing the POPF based restore with a conditional STI.
That code is a historical mess and none of this is needed. Make it
straightforward use local_irq_disable()/enable() as that's all what is
required. It is invoked from interrupt enabled code nowadays.
Reported-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Mark Rutland <mark.rutland@arm.com>
Link: https://lore.kernel.org/r/87k0tpju47.fsf@nanos.tec.linutronix.de
|
|
Prarit reported that depending on the affinity setting the
' irq $N: Affinity broken due to vector space exhaustion.'
message is showing up in dmesg, but the vector space on the CPUs in the
affinity mask is definitely not exhausted.
Shung-Hsi provided traces and analysis which pinpoints the problem:
The ordering of trying to assign an interrupt vector in
assign_irq_vector_any_locked() is simply wrong if the interrupt data has a
valid node assigned. It does:
1) Try the intersection of affinity mask and node mask
2) Try the node mask
3) Try the full affinity mask
4) Try the full online mask
Obviously #2 and #3 are in the wrong order as the requested affinity
mask has to take precedence.
In the observed cases #1 failed because the affinity mask did not contain
CPUs from node 0. That made it allocate a vector from node 0, thereby
breaking affinity and emitting the misleading message.
Revert the order of #2 and #3 so the full affinity mask without the node
intersection is tried before actually affinity is broken.
If no node is assigned then only the full affinity mask and if that fails
the full online mask is tried.
Fixes: d6ffc6ac83b1 ("x86/vector: Respect affinity mask in irq descriptor")
Reported-by: Prarit Bhargava <prarit@redhat.com>
Reported-by: Shung-Hsi Yu <shung-hsi.yu@suse.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Shung-Hsi Yu <shung-hsi.yu@suse.com>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/87ft4djtyp.fsf@nanos.tec.linutronix.de
|
|
Tom Parkin says:
====================
add ppp_generic ioctl(s) to bridge channels
Following on from my previous RFC[1], this series adds two ioctl calls
to the ppp code to implement "channel bridging".
When two ppp channels are bridged, frames presented to ppp_input() on
one channel are passed to the other channel's ->start_xmit function for
transmission.
The primary use-case for this functionality is in an L2TP Access
Concentrator where PPP frames are typically presented in a PPPoE session
(e.g. from a home broadband user) and are forwarded to the ISP network in
a PPPoL2TP session.
The two new ioctls, PPPIOCBRIDGECHAN and PPPIOCUNBRIDGECHAN form a
symmetric pair.
Userspace code testing and illustrating use of the ioctl calls is
available in the go-l2tp[2] and l2tp-ktest[3] repositories.
[1]. Previous RFC series:
https://lore.kernel.org/netdev/20201106181647.16358-1-tparkin@katalix.com/
[2]. go-l2tp: a Go library for building L2TP applications on Linux
systems. Support for the PPPIOCBRIDGECHAN ioctl is on a branch:
https://github.com/katalix/go-l2tp/tree/tp_002_pppoe_2
[3]. l2tp-ktest: a test suite for the Linux Kernel L2TP subsystem.
Support for the PPPIOCBRIDGECHAN ioctl is on a branch:
https://github.com/katalix/l2tp-ktest/tree/tp_ac_pppoe_tests_2
Changelog:
v4:
* Fix NULL-pointer access in PPPIOCBRIDGECHAN in the case that the
ID of the channel to be bridged wasn't found.
* Add comment in ppp_unbridge_channels to better document the
unbridge process.
v3:
* Use rcu_dereference_protected for accessing struct channel
'bridge' field during updates with lock 'upl' held.
* Avoid race in ppp_unbridge_channels by ensuring that each channel
in the bridge points to it's peer before decrementing refcounts.
v2:
* Add missing __rcu annotation to struct channel 'bridge' field in
order to squash a sparse warning from a C=1 build
* Integrate review comments from gnault@redhat.com
* Have ppp_unbridge_channels return -EINVAL if the channel isn't
part of a bridge: this better aligns with the return code from
ppp_disconnect_channel.
* Improve docs update by including information on ioctl arguments
and error return codes.
====================
Reviewed-by: Guillaume Nault <gnault@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Add documentation of the newly-added PPPIOCBRIDGECHAN and
PPPIOCUNBRIDGECHAN ioctls.
Signed-off-by: Tom Parkin <tparkin@katalix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
This new ioctl pair allows two ppp channels to be bridged together:
frames arriving in one channel are transmitted in the other channel
and vice versa.
The practical use for this is primarily to support the L2TP Access
Concentrator use-case. The end-user session is presented as a ppp
channel (typically PPPoE, although it could be e.g. PPPoA, or even PPP
over a serial link) and is switched into a PPPoL2TP session for
transmission to the LNS. At the LNS the PPP session is terminated in
the ISP's network.
When a PPP channel is bridged to another it takes a reference on the
other's struct ppp_file. This reference is dropped when the channels
are unbridged, which can occur either explicitly on userspace calling
the PPPIOCUNBRIDGECHAN ioctl, or implicitly when either channel in the
bridge is unregistered.
In order to implement the channel bridge, struct channel is extended
with a new field, 'bridge', which points to the other struct channel
making up the bridge.
This pointer is RCU protected to avoid adding another lock to the data
path.
To guard against concurrent writes to the pointer, the existing struct
channel lock 'upl' coverage is extended rather than adding a new lock.
The 'upl' lock is used to protect the existing unit pointer. Since the
bridge effectively replaces the unit (they're mutually exclusive for a
channel) it makes coding easier to use the same lock to cover them
both.
Signed-off-by: Tom Parkin <tparkin@katalix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Change "Burusty" to "bursty".
Reported-by: Michael Larabel <Michael@phoronix.com>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Link: https://lore.kernel.org/r/20201210213324.2113041-1-srinivas.pandruvada@linux.intel.com
|
|
We've been seeing failures with xfstests generic/091 and generic/263
when using READ_PLUS. I've made some progress on these issues, and the
tests fail later on but still don't pass. Let's disable READ_PLUS by
default until we can work out what is going on.
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
|
|
Since commit b4868b44c5628 ("NFSv4: Wait for stateid updates after
CLOSE/OPEN_DOWNGRADE"), every inter server copy operation suffers 5
seconds delay regardless of the size of the copy. The delay is from
nfs_set_open_stateid_locked when the check by nfs_stateid_is_sequential
fails because the seqid in both nfs4_state and nfs4_stateid are 0.
Fix __nfs42_ssc_open to delay setting of NFS_OPEN_STATE in nfs4_state,
until after the call to update_open_stateid, to indicate this is the 1st
open. This fix is part of a 2 patches, the other patch is the fix in the
source server to return the stateid for COPY_NOTIFY request with seqid 1
instead of 0.
Fixes: ce0887ac96d3 ("NFSD add nfs4 inter ssc to nfsd4_copy")
Signed-off-by: Dai Ngo <dai.ngo@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
|
|
By switching to an XFS-backed export, I am able to reproduce the
ibcomp worker crash on my client with xfstests generic/013.
For the failing LISTXATTRS operation, xdr_inline_pages() is called
with page_len=12 and buflen=128.
- When ->send_request() is called, rpcrdma_marshal_req() does not
set up a Reply chunk because buflen is smaller than the inline
threshold. Thus rpcrdma_convert_iovs() does not get invoked at
all and the transport's XDRBUF_SPARSE_PAGES logic is not invoked
on the receive buffer.
- During reply processing, rpcrdma_inline_fixup() tries to copy
received data into rq_rcv_buf->pages because page_len is positive.
But there are no receive pages because rpcrdma_marshal_req() never
allocated them.
The result is that the ibcomp worker faults and dies. Sometimes that
causes a visible crash, and sometimes it results in a transport hang
without other symptoms.
RPC/RDMA's XDRBUF_SPARSE_PAGES support is not entirely correct, and
should eventually be fixed or replaced. However, my preference is
that upper-layer operations should explicitly allocate their receive
buffers (using GFP_KERNEL) when possible, rather than relying on
XDRBUF_SPARSE_PAGES.
Reported-by: Olga kornievskaia <kolga@netapp.com>
Suggested-by: Olga kornievskaia <kolga@netapp.com>
Fixes: c10a75145feb ("NFSv4.2: add the extended attribute proc functions.")
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Reviewed-by: Olga kornievskaia <kolga@netapp.com>
Reviewed-by: Frank van der Linden <fllinden@amazon.com>
Tested-by: Olga kornievskaia <kolga@netapp.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
|
|
Add the API for getting the domain from a vfio group. This could be used
by the physical device drivers which rely on the vfio/mdev framework for
mediated device user level access. The typical use case like below:
unsigned int pasid;
struct vfio_group *vfio_group;
struct iommu_domain *iommu_domain;
struct device *dev = mdev_dev(mdev);
struct device *iommu_device = mdev_get_iommu_device(dev);
if (!iommu_device ||
!iommu_dev_feature_enabled(iommu_device, IOMMU_DEV_FEAT_AUX))
return -EINVAL;
vfio_group = vfio_group_get_external_user_from_dev(dev);
if (IS_ERR_OR_NULL(vfio_group))
return -EFAULT;
iommu_domain = vfio_group_iommu_domain(vfio_group);
if (IS_ERR_OR_NULL(iommu_domain)) {
vfio_group_put_external_user(vfio_group);
return -EFAULT;
}
pasid = iommu_aux_get_pasid(iommu_domain, iommu_device);
if (pasid < 0) {
vfio_group_put_external_user(vfio_group);
return -EFAULT;
}
/* Program device context with pasid value. */
...
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
|
|
The Macronix MX35LF1G24AD(/2G24AD/4G24AD) are 3V, 1G/2G/4Gbit serial
SLC NAND flash device (without on-die ECC).
Validated by read, erase, read back, write, read back on Xilinx Zynq
PicoZed FPGA board which included Macronix SPI Host(drivers/spi/spi-mxic.c)
& S/W BCH ecc(drivers/mtd/nand/ecc-sw-bch.c) with bug fixing patch
(mtd: nand: ecc-bch: Fix the size of calc_buf/code_buf of the BCH).
Signed-off-by: YouChing Lin <ycllin@mxic.com.tw>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/1607570529-22341-3-git-send-email-ycllin@mxic.com.tw
|
|
This driver supports Rockchip NFC (NAND Flash Controller) found on RK3308,
RK2928, RKPX30, RV1108 and other SOCs. The driver has been tested using
8-bit NAND interface on the ARM based RK3308 platform.
Support Rockchip SoCs and NFC versions:
- PX30 and RK3326(NFCv900).
ECC: 16/40/60/70 bits/1KB.
CLOCK: ahb and nfc.
- RK3308 and RV1108(NFCv800).
ECC: 16 bits/1KB.
CLOCK: ahb and nfc.
- RK3036 and RK3128(NFCv622).
ECC: 16/24/40/60 bits/1KB.
CLOCK: ahb and nfc.
- RK3066, RK3188 and RK2928(NFCv600).
ECC: 16/24/40/60 bits/1KB.
CLOCK: ahb.
Supported features:
- Read full page data by DMA.
- Support HW ECC(one step is 1KB).
- Support 2 - 32K page size.
- Support 8 CS(depend on SoCs)
Limitations:
- No support for the ecc step size is 512.
- Untested on some SoCs.
- No support for subpages.
- No support for the builtin randomizer.
- The original bad block mask is not supported. It is recommended to use
the BBT(bad block table).
Suggested-by: Johan Jonker <jbx6244@gmail.com>
Signed-off-by: Yifeng Zhao <yifeng.zhao@rock-chips.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20201210002134.5686-3-yifeng.zhao@rock-chips.com
|
|
Documentation support for Rockchip RK3xxx NAND flash controllers
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Yifeng Zhao <yifeng.zhao@rock-chips.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20201210002134.5686-2-yifeng.zhao@rock-chips.com
|
|
The .compatible and .data pairs can be stored in a single line, which
makes the code more concise.
Signed-off-by: Fabio Estevam <festevam@gmail.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20201208221243.3255-1-festevam@gmail.com
|
|
To get better performance, current gpmi driver collected and chained all
small DMA transfers in gpmi_nfc_exec_op, the whole chain triggered and
wait for complete at the end.
But some random DMA timeout found in this new driver, with the help of
ftrace, we found the root cause is as follows:
Take gpmi_ecc_read_page() as an example, gpmi_nfc_exec_op collected 6
DMA transfers and the DMA chain triggered at the end. It waits for bch
completion and check jiffies if it's timeout. The typical function graph
shown below,
63.216351 | 1) | gpmi_ecc_read_page() {
63.216352 | 1) 0.750 us | gpmi_bch_layout_std();
63.216354 | 1) | gpmi_nfc_exec_op() {
63.216355 | 1) | gpmi_chain_command() {
63.216356 | 1) | mxs_dma_prep_slave_sg() {
63.216357 | 1) | /* mxs chan ccw idx: 0 */
63.216358 | 1) 1.750 us | }
63.216359 | 1) | mxs_dma_prep_slave_sg() {
63.216360 | 1) | /* mxs chan ccw idx: 1 */
63.216361 | 1) 2.000 us | }
63.216361 | 1) 6.500 us | }
63.216362 | 1) | gpmi_chain_command() {
63.216363 | 1) | mxs_dma_prep_slave_sg() {
63.216364 | 1) | /* mxs chan ccw idx: 2 */
63.216365 | 1) 1.750 us | }
63.216366 | 1) | mxs_dma_prep_slave_sg() {
63.216367 | 1) | /* mxs chan ccw idx: 3 */
63.216367 | 1) 1.750 us | }
63.216368 | 1) 5.875 us | }
63.216369 | 1) | /* gpmi_chain_wait_ready */
63.216370 | 1) | mxs_dma_prep_slave_sg() {
63.216372 | 1) | /* mxs chan ccw idx: 4 */
63.216373 | 1) 3.000 us | }
63.216374 | 1) | /* gpmi_chain_data_read */
63.216376 | 1) | mxs_dma_prep_slave_sg() {
63.216377 | 1) | /* mxs chan ccw idx: 5 */
63.216378 | 1) 2.000 us | }
63.216379 | 1) 1.125 us | mxs_dma_tx_submit();
63.216381 | 1) 1.000 us | mxs_dma_enable_chan();
63.216712 | 0) 2.625 us | mxs_dma_int_handler();
63.216717 | 0) 4.250 us | bch_irq();
63.216723 | 0) 1.250 us | mxs_dma_tasklet();
63.216723 | 1) | /* jiffies left 250 */
63.216725 | 1) ! 372.000 us | }
63.216726 | 1) 2.625 us | gpmi_count_bitflips();
63.216730 | 1) ! 379.125 us | }
but it's not gurantee that bch irq handled always after dma irq handled,
sometimes bch_irq comes first and gpmi_nfc_exec_op won't wait anymore,
another gpmi_nfc_exec_op may get invoked before last DMA chain IRQ
handled, this messed up the next DMA chain and causes DMA timeout. Check
the trace log when issue happened.
63.218923 | 1) | gpmi_ecc_read_page() {
63.218924 | 1) 0.625 us | gpmi_bch_layout_std();
63.218926 | 1) | gpmi_nfc_exec_op() {
63.218927 | 1) | gpmi_chain_command() {
63.218928 | 1) | mxs_dma_prep_slave_sg() {
63.218929 | 1) | /* mxs chan ccw idx: 0 */
63.218929 | 1) 1.625 us | }
63.218931 | 1) | mxs_dma_prep_slave_sg() {
63.218931 | 1) | /* mxs chan ccw idx: 1 */
63.218932 | 1) 1.750 us | }
63.218933 | 1) 5.875 us | }
63.218934 | 1) | gpmi_chain_command() {
63.218934 | 1) | mxs_dma_prep_slave_sg() {
63.218935 | 1) | /* mxs chan ccw idx: 2 */
63.218936 | 1) 1.875 us | }
63.218937 | 1) | mxs_dma_prep_slave_sg() {
63.218938 | 1) | /* mxs chan ccw idx: 3 */
63.218939 | 1) 1.625 us | }
63.218939 | 1) 5.875 us | }
63.218940 | 1) | /* gpmi_chain_wait_ready */
63.218941 | 1) | mxs_dma_prep_slave_sg() {
63.218942 | 1) | /* mxs chan ccw idx: 4 */
63.218942 | 1) 1.625 us | }
63.218943 | 1) | /* gpmi_chain_data_read */
63.218944 | 1) | mxs_dma_prep_slave_sg() {
63.218945 | 1) | /* mxs chan ccw idx: 5 */
63.218947 | 1) 2.375 us | }
63.218948 | 1) 0.625 us | mxs_dma_tx_submit();
63.218949 | 1) 1.000 us | mxs_dma_enable_chan();
63.219276 | 0) 5.125 us | bch_irq(); <----
63.219283 | 1) | /* jiffies left 250 */
63.219285 | 1) ! 358.625 us | }
63.219286 | 1) 2.750 us | gpmi_count_bitflips();
63.219289 | 1) ! 366.000 us | }
63.219290 | 1) | gpmi_ecc_read_page() {
63.219291 | 1) 0.750 us | gpmi_bch_layout_std();
63.219293 | 1) | gpmi_nfc_exec_op() {
63.219294 | 1) | gpmi_chain_command() {
63.219295 | 1) | mxs_dma_prep_slave_sg() {
63.219295 | 0) 1.875 us | mxs_dma_int_handler(); <----
63.219296 | 1) | /* mxs chan ccw idx: 6 */
63.219297 | 1) 2.250 us | }
63.219298 | 1) | mxs_dma_prep_slave_sg() {
63.219298 | 0) 1.000 us | mxs_dma_tasklet();
63.219299 | 1) | /* mxs chan ccw idx: 0 */
63.219300 | 1) 1.625 us | }
63.219300 | 1) 6.375 us | }
63.219301 | 1) | gpmi_chain_command() {
63.219302 | 1) | mxs_dma_prep_slave_sg() {
63.219303 | 1) | /* mxs chan ccw idx: 1 */
63.219304 | 1) 1.625 us | }
63.219305 | 1) | mxs_dma_prep_slave_sg() {
63.219306 | 1) | /* mxs chan ccw idx: 2 */
63.219306 | 1) 1.875 us | }
63.219307 | 1) 6.000 us | }
63.219308 | 1) | /* gpmi_chain_wait_ready */
63.219308 | 1) | mxs_dma_prep_slave_sg() {
63.219309 | 1) | /* mxs chan ccw idx: 3 */
63.219310 | 1) 2.000 us | }
63.219311 | 1) | /* gpmi_chain_data_read */
63.219312 | 1) | mxs_dma_prep_slave_sg() {
63.219313 | 1) | /* mxs chan ccw idx: 4 */
63.219314 | 1) 1.750 us | }
63.219315 | 1) 0.625 us | mxs_dma_tx_submit();
63.219316 | 1) 0.875 us | mxs_dma_enable_chan();
64.224227 | 1) | /* jiffies left 0 */
In the first gpmi_nfc_exec_op, bch_irq comes first and gpmi_nfc_exec_op
exits, but DMA IRQ still not happened yet until the middle of following
gpmi_nfc_exec_op, the first DMA transfer index get messed and DMA get
timeout.
To fix the issue, when there is bch ops in DMA chain, the
gpmi_nfc_exec_op should wait for both completions rather than bch
completion only.
Fixes: ef347c0cfd61 ("mtd: rawnand: gpmi: Implement exec_op")
Signed-off-by: Han Xu <han.xu@nxp.com>
Reviewed-by: Sascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20201209035104.22679-3-han.xu@nxp.com
|
|
Set the GPMI CTRL1 GANGED_RDYBUSY bit so driver can sense the R/B signal
from all CS.
For the NAND chip MT29F64G08AFAAAWP, only the first chip detected
without the patch.
[ 3.764118] nand: device found, Manufacturer ID: 0x2c, Chip ID: 0x68
[ 3.770613] nand: Micron MT29F64G08AFAAAWP
[ 3.774752] nand: 4096 MiB, SLC, erase size: 1024 KiB, page size: 8192, OOB size: 448
[ 3.786421] Bad block table found at page 524160, version 0x01
[ 3.792730] Bad block table found at page 524032, version 0x01
After applying the patch
[ 3.764445] nand: device found, Manufacturer ID: 0x2c, Chip ID: 0x68
[ 3.770941] nand: Micron MT29F64G08AFAAAWP
[ 3.775080] nand: 4096 MiB, SLC, erase size: 1024 KiB, page size: 8192, OOB size: 448
[ 3.784390] nand: 2 chips detected
[ 3.790900] Bad block table found at page 524160, version 0x01
[ 3.796776] Bad block table found at page 1048448, version 0x01
Signed-off-by: Han Xu <han.xu@nxp.com>
Acked-by: Sascha Hauer <s.hauer@pengutronix.de>
Reviewed-by: Fabio Estevam <festevam@gmail.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20201209035104.22679-2-han.xu@nxp.com
|
|
SDX55 uses QPIC version 2.0.0 IP for the NAND controller support.
In this version, DEV_CMD_* registers are moved to operational state,
hence CPU access in BAM mode is restricted. So, skip accessing these
registers and also use a different config for reading ONFI parameters.
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20201126085705.48399-3-manivannan.sadhasivam@linaro.org
|
|
Qualcomm SDX55 uses QPIC NAND controller version 2.0.0 with BAM DMA
Engine.
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
|
|
The .compatible and .data pairs can be stored in a single line, which
makes the code more concise.
Signed-off-by: Fabio Estevam <festevam@gmail.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20201204145818.32739-2-festevam@gmail.com
|
|
The retrieval of driver data can be a bit simplified by using
device_get_match_data(), so switch to it.
Signed-off-by: Fabio Estevam <festevam@gmail.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20201204145818.32739-1-festevam@gmail.com
|
|
Call clk_disable_unprepare(nfc->phase_rx) if the clk_set_rate() function
fails to avoid a resource leak.
Fixes: 8fae856c5350 ("mtd: rawnand: meson: add support for Amlogic NAND flash controller")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/X8ikVCnUsfTpffFB@mwanda
|
|
The retrieval of driver data via of_device_get_match_data() can make
the code simpler.
Use of_device_get_match_data() to simplify the code.
Signed-off-by: Fabio Estevam <festevam@gmail.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20201126030946.2058-1-festevam@gmail.com
|
|
This patch adds the new IP of Nand Flash Controller(NFC) support
on Intel's Lightning Mountain(LGM) SoC.
DMA is used for burst data transfer operation, also DMA HW supports
aligned 32bit memory address and aligned data access by default.
DMA burst of 8 supported. Data register used to support the read/write
operation from/to device.
Signed-off-by: Ramuthevar Vadivel Murugan <vadivel.muruganx.ramuthevar@linux.intel.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20201110012333.18647-3-vadivel.muruganx.ramuthevar@linux.intel.com
|
|
Add YAML file for dt-bindings to support NAND Flash Controller
on Intel's Lightning Mountain SoC.
Signed-off-by: Ramuthevar Vadivel Murugan <vadivel.muruganx.ramuthevar@linux.intel.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20201110012333.18647-2-vadivel.muruganx.ramuthevar@linux.intel.com
|
|
The MT29F2G01AAAED is a single die, 2Gb Micron SPI NAND Flash with 4-bit
ECC
Signed-off-by: Thirumalesha Narasimhappa <nthirumalesha7@gmail.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20201108113735.2533-3-nthirumalesha7@gmail.com
|
|
Rename the read/write/update of SPINAND_OP_VARIANTS() to more
specialized names.
Signed-off-by: Thirumalesha Narasimhappa <nthirumalesha7@gmail.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20201108113735.2533-2-nthirumalesha7@gmail.com
|
|
pm_runtime_get_sync() will increment pm usage at first and it
will resume the device later. If runtime of the device has
error or device is in inaccessible state(or other error state),
resume operation will fail. If we do not call put operation to
decrease the reference, it will result in reference leak in
the two functions(gpmi_init and gpmi_nfc_exec_op). Moreover,
this device cannot enter the idle state and always stay busy or
other non-idle state later. So we fixed it through adding
pm_runtime_put_noidle.
Fixes: 5bc6bb603b4d0 ("mtd: rawnand: gpmi: Fix suspend/resume problem")
Signed-off-by: Zhang Qilong <zhangqilong3@huawei.com>
Acked-by: Han Xu <han.xu@nxp.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20201107110552.1568742-1-zhangqilong3@huawei.com
|
|
Driver requires different amount of clocks for different SoCs. Describe
these requirements properly to fix dtbs_check warnings like:
arch/arm64/boot/dts/freescale/imx8mm-beacon-kit.dt.yaml: nand-controller@33002000: clock-names:1: 'gpmi_apb' was expected
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20201029162021.80839-1-krzk@kernel.org
|
|
The Macronix MX35LF2GE4AD / MX35LF4GE4AD are 3V, 2G / 4Gbit serial
SLC NAND flash device (with on-die ECC).
Validated by read, erase, read back, write, read back and nandtest
on Xilinx Zynq PicoZed FPGA board which included Macronix SPI Host
(drivers/spi/spi-mxic.c).
Signed-off-by: YouChing Lin <ycllin@mxic.com.tw>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/1604561020-13499-1-git-send-email-ycllin@mxic.com.tw
|
|
If an error happens in mtd_device_parse_register or mtd_device_register,
memory allocated for struct platram_info is leaked.
Make platram_probe() call platram_remove() on all error paths
after struct platram_info allocation to correctly free resources.
Found by Linux Driver Verification project (linuxtesting.org).
Signed-off-by: Baskov Evgeiny <baskov@ispras.ru>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20201113160537.899-1-baskov@ispras.ru
|
|
struct mtd_info has a flag oops_panic_write which is set when the write
operation is issued via the panic_write() callback. That allows controller
drivers to distinguish the panic write from a regular write.
Replace the open coded 'in_interrupt() | oops_in_progress' checks with a
check for that flag. in_interrupt() is an unrealiable indicator anyway as
it covers all sorts of atomic contexts not only hard and soft interrupt
service routines.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Vignesh Raghavendra <vigneshr@ti.com>
Cc: Tudor Ambarus <tudor.ambarus@microchip.com>
Cc: Richard Weinberger <richard@nod.at>
Cc: Kyungmin Park <kyungmin.park@samsung.com>
Cc: Boris Brezillon <boris.brezillon@collabora.com>
Cc: linux-mtd@lists.infradead.org
Cc: Miquel Raynal <miquel.raynal@bootlin.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20201113141422.2214771-1-bigeasy@linutronix.de
|
|
The raw NAND core now declares the on host ECC engine being the
default if none is provided in the DT. Drop this line doing exactly
the same from the Marvell driver.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20201113124114.449-1-miquel.raynal@bootlin.com
|
|
When the nand_chip structure is already available, there is no need to
dereference it through the info pointer. Use the chip pointer directly
in this case.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20201113124045.32743-1-miquel.raynal@bootlin.com
|
|
While working a bit on this driver I dropped the platform includes and
commented a few lines just to verify the correctness of my changes. It
appeared the following:
drivers/mtd/nand/raw/au1550nd.c: In function ‘au1550nd_waitrdy’:
drivers/mtd/nand/raw/au1550nd.c:130:3: error: implicit declaration of function ‘usleep_range’ [-Werror=implicit-function-declaration]
usleep_range(10, 100);
^~~~~~~~~~~~
drivers/mtd/nand/raw/au1550nd.c: In function ‘au1550nd_exec_instr’:
drivers/mtd/nand/raw/au1550nd.c:188:3: error: implicit declaration of function ‘ndelay’ [-Werror=implicit-function-declaration]
ndelay(instr->delay_ns);
^~~~~~
I think the delay.h header should be included in this file and not
come from one of its platform includes, so let's add it here.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20201113124021.32675-1-miquel.raynal@bootlin.com
|
|
The NAND ECC core is included in the generic NAND core when it is
compiled in.
Different software ECC engines drivers will select the NAND ECC core
and thus also have a dependency on the NAND core. Using a "depends on"
between the two leads to possible cases (not real cases, but created
by robots) where one is still unselected because of the "select does
not verifies depends on" game:
WARNING: unmet direct dependencies detected for MTD_NAND_ECC
Depends on [n]: MTD [=m] && MTD_NAND_CORE [=n]
Selected by [m]:
- MTD_NAND_ECC_SW_HAMMING [=y] && MTD [=m]
- MTD_NAND_ECC_SW_BCH [=y] && MTD [=m]
Fix this by using a select instead.
Reported-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Acked-by: Randy Dunlap <rdunlap@infradead.org> # build-tested
Link: https://lore.kernel.org/linux-mtd/20201113123945.32592-1-miquel.raynal@bootlin.com
|
|
Moving files around produced the following warnings:
Error: Cannot open file drivers/mtd/nand/raw/nand_ecc.c
Error: Cannot open file drivers/mtd/nand/raw/nand_ecc.c
Fix one by just dropping the reference because it is not relevant, the
other by using a better noun instead of a file name.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20201113123831.32429-1-miquel.raynal@bootlin.com
|
|
i.MX is a devicetree-only platform now and the existing platform data
support in this driver was only useful for old non-devicetree platforms.
Get rid of the platform data support since it is no longer used.
Signed-off-by: Fabio Estevam <festevam@gmail.com>
Reviewed-by: Sascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20201110121908.19400-1-festevam@gmail.com
|
|
of_find_device_by_node() already takes a reference to the device, and
ingenic_ecc_release() will drop the reference. So, the get_device() in
ingenic_ecc_get() is redundand.
Fixes: 15de8c6efd0e("mtd: rawnand: ingenic: Separate top-level and SoC specific code")
Signed-off-by: Yu Kuai <yukuai3@huawei.com>
Acked-by: Paul Cercueil <paul@crapouillou.net>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20201031105439.2304211-1-yukuai3@huawei.com
|
|
This code has been written in 2008 and is fine, but in order to keep
robots happy, I think it's time to change a little bit this code just
to clarify the different possible values of eccsize_mult. Indeed, this
variable may only take the value 1 or 2 because step_size, in the case
of the software Hamming ECC engine may only be 256 or 512. Depending
on the value of eccsize_mult, an extra rp17 variable is set, or not
and triggers the following warning:
smatch warnings:
ecc_sw_hamming_calculate() error: uninitialized symbol 'rp17'.
As highlighted by Dan Carpenter, if the only possible values for
eccsize_mult are 1 and 2, then the code is fine, but "it's hard to
tell just from looking".
So instead of shifting step_size, let's use a ternary condition to
assign to eccsize_mult the only two possible values and clarify the
driver's logic.
Now that the situation is clarified for humans, ensure rp17 is
initialized to 0 to keep compilers and robots silent as well.
Reported-by: kernel test robot <lkp@intel.com>
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20201030172333.28390-1-miquel.raynal@bootlin.com
|
|
This patch enables NAND MDMA (MBUS DMA) mode for
the Allwinner SoCs A23/A33/H3.
The DMA transfer method gets sets now to MBUS DMA as default for
the sun8i-a23-nand-controller (till now DMA transfer was executed
via the shared DMA engine).
The main advantage is more bandwidth for the users of the shared DMA
engine and also that the MBUS DMA setup requires less configuration
effort. For example you don't need to define a dedicated DMA channel
in the device-tree any more.
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Signed-off-by: Manuel Dipolt <manuel.dipolt@robart.cc>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/154840787.280672.1602517282173.JavaMail.zimbra@robart.cc
|
|
Arguments 'infolen' and 'datalen' to meson_nfc_dma_buffer_release() were mixed up.
Fixes: 8fae856c53500 ("mtd: rawnand: meson: add support for Amlogic NAND flash controller")
Cc: stable@vger.kernel.org
Signed-off-by: Sergei Antonov <saproj@gmail.com>
Acked-by: Liang Yang <liang.yang@amlogic.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20201028094940.11765-1-saproj@gmail.com
|
