euleros/rtos inclusion
category: bugfix
bugzilla: https://gitee.com/openeuler/kernel/issues/I4LOJ6
CVE: NA

--------------------------------

This patch set introduce many conflicts while backporting mainline
bcache patches, revert it temporarily.

This reverts commit a460ae11.
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>
Acked-by: NXie XiuQi <xiexiuqi@huawei.com>

euleros/rtos inclusion
category: bugfix
bugzilla: https://gitee.com/openeuler/kernel/issues/I4LOJ6
CVE: NA

--------------------------------

This patch set introduce many conflicts while backporting mainline
bcache patches, revert it temporarily.

This reverts commit 30dc9d9c.
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>
Acked-by: NXie XiuQi <xiexiuqi@huawei.com>

euleros/rtos inclusion
category: bugfix
bugzilla: https://gitee.com/openeuler/kernel/issues/I4LOJ6
CVE: NA

--------------------------------

This patch set introduce many conflicts while backporting mainline
bcache patches, revert it temporarily.

This reverts commit 08a3ac0e.
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>
Acked-by: NXie XiuQi <xiexiuqi@huawei.com>

euleros/rtos inclusion
category: bugfix
bugzilla: https://gitee.com/openeuler/kernel/issues/I4LOJ6
CVE: NA

--------------------------------

This patch set introduce many conflicts while backporting mainline
bcache patches, revert it temporarily.

This reverts commit 1751c6ad.
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>
Acked-by: NXie XiuQi <xiexiuqi@huawei.com>

euleros/rtos inclusion
category: bugfix
bugzilla: https://gitee.com/openeuler/kernel/issues/I4LOJ6
CVE: NA

--------------------------------

This patch set introduce many conflicts while backporting mainline
bcache patches, revert it temporarily.

This reverts commit 87f258f0.
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>
Acked-by: NXie XiuQi <xiexiuqi@huawei.com>

euleros/rtos inclusion
category: bugfix
bugzilla: https://gitee.com/openeuler/kernel/issues/I4LOJ6
CVE: NA

--------------------------------

This patch set introduce many conflicts while backporting mainline
bcache patches, revert it temporarily.

This reverts commit e8c75ee9.
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>
Acked-by: NXie XiuQi <xiexiuqi@huawei.com>

euleros/rtos inclusion
category: bugfix
bugzilla: https://gitee.com/openeuler/kernel/issues/I4LOJ6
CVE: NA

--------------------------------

This patch set introduce many conflicts while backporting mainline
bcache patches, revert it temporarily.

This reverts commit 6214e257.
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>
Acked-by: NXie XiuQi <xiexiuqi@huawei.com>

euleros/rtos inclusion
category: bugfix
bugzilla: https://gitee.com/openeuler/kernel/issues/I4LOJ6
CVE: NA

--------------------------------

This patch set introduce many conflicts while backporting mainline
bcache patches, revert it temporarily.

This reverts commit 4c16a439.
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>
Acked-by: NXie XiuQi <xiexiuqi@huawei.com>

mainline inclusion
from mainline-v5.16-rc7
commit ebb966d3
category: bugfix
bugzilla: https://gitee.com/openeuler/kernel/issues/I4PVZR?from=project-issue
CVE: NA

Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?h=v5.16-rc7&id=ebb966d3bdfed581ecccbb4a7432341baf7619b4

--------------------------------

In commit 5648b5e1 ("netfilter: nfnetlink_queue: fix OOB when mac
header was cleared"), the test for non-empty MAC header introduced in
commit 2c38de4c ("netfilter: fix looped (broad|multi)cast's MAC
handling") has been replaced with a test for a set MAC header.

This breaks the case when the MAC header has been reset (using
skb_reset_mac_header), as is the case with looped-back multicast
packets.  As a result, the packets ending up in NFQUEUE get a bogus
hwaddr interpreted from the first bytes of the IP header.

This patch adds a test for a non-empty MAC header in addition to the
test for a set MAC header.  The same two tests are also implemented in
nfnetlink_log.c, where the initial code of commit 2c38de4c
("netfilter: fix looped (broad|multi)cast's MAC handling") has not been
touched, but where supposedly the same situation may happen.

Fixes: 5648b5e1 ("netfilter: nfnetlink_queue: fix OOB when mac header was cleared")
Signed-off-by: NIgnacy Gawędzki <ignacy.gawedzki@green-communications.fr>
Reviewed-by: NFlorian Westphal <fw@strlen.de>
Signed-off-by: NPablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: NHuang Guobin <huangguobin4@huawei.com>
Reviewed-by: NWei Yongjun <weiyongjun1@huawei.com>
Reviewed-by: NYue Haibing <yuehaibing@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: bugfix
bugzilla: https://gitee.com/openeuler/kernel/issues/I4HYY4
CVE: NA

-------------------------------------------------

Per-CPU variables cpc_desc_ptr are initialized in
acpi_cppc_processor_probe() when the processor devices are present and
added into the system. But when cpu_possible_mask and cpu_present_mask
is not equal, only cpc_desc_ptr in cpu_present_mask are initialized,
this will cause acpi_get_psd_map() failed in cppc_cpufreq_init().

To fix this issue, we parse the _PSD method for all possible CPUs to get
the P-State topology and modify acpi_get_psd_map() to rely on this
information.
Signed-off-by: NXiongfeng Wang <wangxiongfeng@huawei.com>
Reviewed-by: NKeqian Zhu <zhukeqian1@huawei.com>
Reviewed-by: NHanjun Guo <guohanjun@huawei.com>
Reviewed-by: NCheng Jian <cj.chengjian@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>

 Conflicts:
	drivers/acpi/cppc_acpi.c
Signed-off-by: NXiongfeng Wang <wangxiongfeng@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

mainline inclusion
from mainline-v5.16-rc3
category: bugfix
bugzilla: https://gitee.com/openeuler/kernel/issues/I4HYY4
CVE: NA

-------------------------------

When I hot added a CPU, I found 'cpufreq' directory was not created
below /sys/devices/system/cpu/cpuX/.

It is because get_cpu_device() failed in add_cpu_dev_symlink().

cpufreq_add_dev() is the .add_dev callback of a CPU subsys interface.
It will be called when the CPU device registered into the system.
The call chain is as follows:

  register_cpu()
  ->device_register()
   ->device_add()
    ->bus_probe_device()
     ->cpufreq_add_dev()

But only after the CPU device has been registered, we can get the
CPU device by get_cpu_device(), otherwise it will return NULL.

Since we already have the CPU device in cpufreq_add_dev(), pass
it to add_cpu_dev_symlink().

I noticed that the 'kobj' of the CPU device has been added into
the system before cpufreq_add_dev().

Fixes: 2f0ba790 ("cpufreq: Fix creation of symbolic links to policy directories")
Signed-off-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Acked-by: NViresh Kumar <viresh.kumar@linaro.org>
Cc: All applicable <stable@vger.kernel.org>
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: NXiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: NHanjun Guo <guohanjun@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

serial: amba-pl011: Fix serial port discard interrupt when interrupt signal line of serial port is connected to mbigen.

ascend inclusion
category: bugfix
Bugzilla: https://gitee.com/openeuler/kernel/issues/I4K2U5
CVE: N/A

---------------------------------------

Hisi when designing ascend chip, connect the serial port interrupt
signal lines to mbigen equipment, mbigen write GICD_SETSPI_NSR
register trigger the SPI interrupt. This can result in serial
port drop interrupts.
Signed-off-by: NXu Qiang <xuqiang36@huawei.com>
Signed-off-by: NWang Wensheng <wangwensheng4@huawei.com>
Reviewed-by: NHanjun Guo <guohanjun@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

ascend inclusion
category: feature
bugzilla: https://gitee.com/openeuler/kernel/issues/I4L735
CVE: NA

---------------------------------------------------

Add devicetree bindings document for arm64/mpam, this add basic
description for the main properties of mpam devicetree definition.
Signed-off-by: NXingang Wang <wangxingang5@huawei.com>
Reviewed-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

ascend inclusion
category: feature
bugzilla: https://gitee.com/openeuler/kernel/issues/I4L735
CVE: NA

---------------------------------------------------

For now, only ACPI boot is supported for the arm64 mpam init.
This introduce device tree support, treat the mpam device as
a platform device, add a platform driver and use of interface
to parse the dts node. Add a common init function to call the
device tree or ACPI init procedure, according to whether ACPI
is disabled and the boot arguments.
Signed-off-by: NXingang Wang <wangxingang5@huawei.com>
Reviewed-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

ascend inclusion
category: feature
bugzilla: https://gitee.com/openeuler/kernel/issues/I4L735
CVE: NA

---------------------------------------------------

To support device tree boot for arm64 mpam, the __init macro might be
used by the dts driver. This remove the necessary __init macro for the
related functions.
Signed-off-by: NXingang Wang <wangxingang5@huawei.com>
Reviewed-by: NWang ShaoBo <bobo.shaobowang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: bugfix
bugzilla: https://gitee.com/openeuler/kernel/issues/I4OAFT
CVE: NA

-------------------------------------------------

Perf script was failed to print the phys_addr for SPE profiling.
One 'dummy' event is added by SPE profiling but it doesn't have PHYS_ADDR
attribute set, perf script then exits with error.

Now referring to 'addr', use evsel__do_check_stype() to check the type.

Before:

  # perf record -e arm_spe_0/branch_filter=0,ts_enable=1,pa_enable=1,load_filter=1,jitter=0,store_filter=0,min_latency=0,event_filter=2/ -p 4064384 -- sleep 3
  # perf script -F pid,tid,addr,phys_addr
  Samples for 'dummy:u' event do not have PHYS_ADDR attribute set. Cannot print 'phys_addr' field.

After:

  # perf record -e arm_spe_0/branch_filter=0,ts_enable=1,pa_enable=1,load_filter=1,jitter=0,store_filter=0,min_latency=0,event_filter=2/ -p 4064384 -- sleep 3
  # perf script -F pid,tid,addr,phys_addr
  4064384/4064384 ffff802f921be0d0      2f921be0d0
  4064384/4064384 ffff802f921be0d0      2f921be0d0
Signed-off-by: NYao Jin <jinyao5@huawei.com>
Signed-off-by: NWei Li <liwei391@huawei.com>
Reviewed-by: NYang Jihong <yangjihong1@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

euleros/rtos inclusion
category: bugfix
bugzilla: https://gitee.com/openeuler/kernel/issues/I4LOJ6
CVE: NA

--------------------------------

Before this patch bypassed I/O request will not record start time,
therefore when calculating latency data, start time will be calculated
wrongly. This patch makes start time always be recorded to fix this.
Signed-off-by: NLi Ruilin <liruilin4@huawei.com>
Reviewed-by: NSong Chao <chao.song@huawei.com>
Acked-by: NXie XiuQi <xiexiuqi@huawei.com>
Signed-off-by: NCheng Jian <cj.chengjian@huawei.com>
Reviewed-by: NGuangxing Deng <dengguangxing@huawei.com>
Reviewed-by: Nchao song <chao.song@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

euleros/rtos inclusion
category: bugfix
bugzilla: https://gitee.com/openeuler/kernel/issues/I4LOJ6
CVE: NA

--------------------------------

commit 6947676c374("bcache: add a framework to perform prefetch")
collects data insert info which includes device info got from bio.
However, bio created by write_moving here has no device info, causing
a null pointer dereference.

[ 1497.991768] BUG: unable to handle kernel NULL pointer dereference at 0000000000000000
[ 1497.991869] PGD 0 P4D 0
[ 1497.991912] Oops: 0000 [#1] SMP PTI
[ 1497.991962] CPU: 2 PID: 733 Comm: kworker/2:3 Not tainted 4.19.90+ #33
[ 1497.992030] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.10.2-1ubuntu1 04/01/2014
[ 1497.992137] Workqueue: bcache_gc write_moving [bcache]
[ 1497.992219] RIP: 0010:bch_data_insert+0x4c/0x140 [bcache]
...
[ 1497.993367] Call Trace:
[ 1497.993427]  ? cached_dev_read_error+0x140/0x140 [bcache]
[ 1497.993526]  write_moving+0x19e/0x1b0 [bcache]
[ 1497.993621]  process_one_work+0x1fd/0x440
[ 1497.993742]  worker_thread+0x34/0x410
[ 1497.993811]  kthread+0x121/0x140
[ 1497.993873]  ? process_one_work+0x440/0x440
[ 1497.993946]  ? kthread_create_worker_on_cpu+0x70/0x70
[ 1497.994043]  ret_from_fork+0x35/0x40
Signed-off-by: NLi Ruilin <liruilin4@huawei.com>
Review-by: NSong Chao <chao.song@huawei.com>
Review-by: NXu Wei <xuwei56@huawei.com>
Signed-off-by: NLi Ruilin <liruilin4@huawei.com>
Acked-by: NXie XiuQi <xiexiuqi@huawei.com>
Signed-off-by: NCheng Jian <cj.chengjian@huawei.com>
Reviewed-by: NGuangxing Deng <dengguangxing@huawei.com>
Reviewed-by: Nchao song <chao.song@huawei.com>
Reviewed-by: Nchao song <chao.song@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

euleros inclusion
category: feature
bugzilla: https://gitee.com/openeuler/kernel/issues/I4LOJ6
CVE: NA

------------------------------

The recently pushd bugfix patch "Delay to invalidate cache data
in writearound write" has a stupid copy&paste bug, which causes
bypassed write request will never invalidate data in cache device,
causing a data corruption. This patch fixes this corruption.
This patch also ensures that the writeback lock is released after
data insert.

Fixes: 6a1d9c41b367 ("bcache: Delay to invalidate cache data in writearound write")
Signed-off-by: NLi Ruilin <liruilin4@huawei.com>
Signed-off-by: NSong Chao <chao.song@huawei.com>
Reviewed-by: NPeng Junyi <pengjunyi1@huawei.com>
Acked-by: NXie Xiuqi <xiexiuqi@huawei.com>
Signed-off-by: NCheng Jian <cj.chengjian@huawei.com>
Reviewed-by: NGuangxing Deng <dengguangxing@huawei.com>
Reviewed-by: Nchao song <chao.song@huawei.com>
Reviewed-by: Nchao song <chao.song@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

euleros inclusion
category: feature
bugzilla: https://gitee.com/openeuler/kernel/issues/I4LOJ6
CVE: NA

------------------------------

Add a sample for new prefetch frame on bcache. As a sample, the
program just reads every read request received by bcache device
and send a prefetch request for each read request. The length of
the prefetch request is equal as the read request and the position
of the prefetch request is follow the read request.
Signed-off-by: NLi Ruilin <liruilin4@huawei.com>
Reviewed-by: NLuan Jianhai <luanjianhai@huawei.com>
Reviewed-by: NPeng Junyi <pengjunyi1@huawei.com>
Acked-by: NXie Xiuqi <xiexiuqi@huawei.com>
Signed-off-by: NCheng Jian <cj.chengjian@huawei.com>
Reviewed-by: NGuangxing Deng <dengguangxing@huawei.com>
Reviewed-by: Nchao song <chao.song@huawei.com>
Reviewed-by: Nchao song <chao.song@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

euleros inclusion
category: feature
bugzilla: https://gitee.com/openeuler/kernel/issues/I4LOJ6
CVE: NA

------------------------------

In writearound cache mode, read request quickly followed by write
request may overwrite the invalidate bkey inserted by the write
request.

The function bch_data_insert() is invoked asynchronously as the bio
subbmited to backend block device, therefore there may be a read
request subbmited after the bch_data_insert() done and ended before
the backend bio is end. This read request will read data from the
backend block device, and insert dirty data to cache device. However
by writearound cache mode, bcache will not invalidate data again,
so that read request after will read dirty data from the cache,
causing a data corruption.

By this patch we delay the invalidation to end of backend bio to
avoid this corruption.
Signed-off-by: NLi Ruilin <liruilin4@huawei.com>
Reviewed-by: NLuan Jianhai <luanjianhai@huawei.com>
Reviewed-by: NPeng Junyi <pengjunyi1@huawei.com>
Acked-by: NXie Xiuqi <xiexiuqi@huawei.com>
Signed-off-by: NCheng Jian <cj.chengjian@huawei.com>
Reviewed-by: NGuangxing Deng <dengguangxing@huawei.com>
Reviewed-by: Nchao song <chao.song@huawei.com>
Reviewed-by: Nchao song <chao.song@huawei.com>
Reviewed-by: Nchao song <chao.song@huawei.com>
Reviewed-by: Nchao song <chao.song@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

euleros inclusion
category: feature
bugzilla: https://gitee.com/openeuler/kernel/issues/I4LOJ6
CVE: NA

------------------------------

Add a list to save all prefetch requests. When an IO request comes,
check if the request has overlap with some of prefetch requests. If
it das have, block the request until the prefetch request is end.

Add a switch to control whether to enable this. If not enabled, count
the overlapped IO request as a fake hit for performance analysis.
Signed-off-by: NLi Ruilin <liruilin4@huawei.com>
Reviewed-by: NLuan Jianhai <luanjianhai@huawei.com>
Reviewed-by: NPeng Junyi <pengjunyi1@huawei.com>
Acked-by: NXie Xiuqi <xiexiuqi@huawei.com>
Signed-off-by: NCheng Jian <cj.chengjian@huawei.com>
Reviewed-by: NGuangxing Deng <dengguangxing@huawei.com>
Reviewed-by: Nchao song <chao.song@huawei.com>
Reviewed-by: Nchao song <chao.song@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

euleros inclusion
category: feature
bugzilla: https://gitee.com/openeuler/kernel/issues/I4LOJ6
CVE: NA

------------------------------

provide a switch named read_bypass. If enbale, all IO requests will
bypass the cache. This option could be useful when we enable userspace
prefetch and the cache device is low capacity.
Signed-off-by: NLi Ruilin <liruilin4@huawei.com>
Reviewed-by: NLuan Jianhai <luanjianhai@huawei.com>
Reviewed-by: NPeng Junyi <pengjunyi1@huawei.com>
Acked-by: NXie Xiuqi <xiexiuqi@huawei.com>
Signed-off-by: NCheng Jian <cj.chengjian@huawei.com>
Reviewed-by: NGuangxing Deng <dengguangxing@huawei.com>
Reviewed-by: Nchao song <chao.song@huawei.com>
Reviewed-by: Nchao song <chao.song@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

euleros inclusion
category: feature
bugzilla: https://gitee.com/openeuler/kernel/issues/I4LOJ6
CVE: NA

------------------------------

Add a framwork to transform io informations to userspace client and
process prefetch request sent by userspace client. Create a char device
namede "acache" for connecting between kernelspace and userspace.
Save informations of all io requests into a buffer and pass them to
client when client reads from the device.

The prefetch request could be treated as normal io request. As deference,
those requests have no need return data back to userspace, and they
should not append readahead part.

Add two parameters. acache_dev_size is for controlling size of buffer
to save io informations. acache_prefetch_workers is for controlling
max threads to process prefetch requests.
Signed-off-by: NLi Ruilin <liruilin4@huawei.com>
Reviewed-by: NLuan Jianhai <luanjianhai@huawei.com>
Reviewed-by: NPeng Junyi <pengjunyi1@huawei.com>
Acked-by: NXie Xiuqi <xiexiuqi@huawei.com>
Signed-off-by: NCheng Jian <cj.chengjian@huawei.com>
Reviewed-by: NGuangxing Deng <dengguangxing@huawei.com>
Reviewed-by: Nchao song <chao.song@huawei.com>
Reviewed-by: Nchao song <chao.song@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

hulk inclusion
category: feature
bugzilla: https://gitee.com/openeuler/kernel/issues/I4PY1Q
CVE: NA

--------------------------------

We add stub info in some structures to maintain the consistency of KABI
Signed-off-by: NBaisong Zhong <zhongbaisong@huawei.com>
Reviewed-by: NWei Yongjun <weiyongjun1@huawei.com>
Reviewed-by: NYue Haibing <yuehaibing@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

driver inclusion
category: feature
bugzilla: https://gitee.com/openeuler/kernel/issues/I4ETXO
CVE: NA

-----------------------------------------

Fix following build warnings in arm32 builds:

drivers/net/ethernet/huawei/bma/edma_drv/bma_devintf.c: In function ‘bma_cdev_add_msg’:
drivers/net/ethernet/huawei/bma/edma_drv/bma_pci.h:92:20: warning: format ‘%ld’ expects argument of type ‘long int’, but argument 5 has type ‘size_t {aka unsigned int}’ [-Wformat=]

drivers/net/ethernet/huawei/bma/veth_drv/veth_hb.c: In function ‘veth_recv_pkt’:
drivers/net/ethernet/huawei/bma/veth_drv/veth_hb.c:74:37: warning: format ‘%llx’ expects argument of type ‘long long unsigned int’, but argument 7 has type ‘dma_addr_t {aka unsigned int}’ [-Wformat=]
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>
Reviewed-by: NKefeng Wang <wangkefeng.wang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

euleros inclusion
category: feature
feature: etmem
bugzilla: https://gitee.com/openeuler/kernel/issues/I4OODH?from=project-issue
CVE: NA

-------------------------------------------------

Add /proc/sys/kernel/hugepage_pmem_allocall switch. Set 1 to allowed all
memory in pmem could alloc for hugepage. Set 0(default) hugepage alloc is
limited by zone watermark as usual.
Add /proc/sys/kernel/hugepage_mig_noalloc switch. Set 1 to forbid new
hugepage alloc in hugepage migration when hugepage in dest node runs
out. Set 0(default) to allow hugepage alloc in hugepage migration as
usual.
Signed-off-by: NKemeng Shi <shikemeng@huawei.com>
Reviewed-by: Nlouhongxiang <louhongxiang@huawei.com>
Reviewed-by: NKefeng Wang <wangkefeng.wang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

euleros inclusion
category: feature
feature: etmem
bugzilla: https://gitee.com/openeuler/kernel/issues/I4OODH?from=project-issue
CVE: NA

-------------------------------------------------

User space migration daemon could check
/sys/bus/node/devices/nodeX/type for node type.

Software can interrogate node type for node memory type and distance
to get desirable target node in migration.

grep -r . /sys/devices/system/node/*/type
/sys/devices/system/node/node0/type:dram
/sys/devices/system/node/node1/type:dram
/sys/devices/system/node/node2/type:pmem
/sys/devices/system/node/node3/type:pmem

Along with next patch which export `peer_node`, migration daemon
could easily find the memory type of current node, and the target
node in case of migration.

grep -r . /sys/devices/system/node/*/peer_node
/sys/devices/system/node/node0/peer_node:2
/sys/devices/system/node/node1/peer_node:3
/sys/devices/system/node/node2/peer_node:0
/sys/devices/system/node/node3/peer_node:1
Signed-off-by: NFan Du <fan.du@intel.com>
Signed-off-by: NFengguang Wu <fengguang.wu@intel.com>
Signed-off-by: NKemeng Shi <shikemeng@huawei.com>
Reviewed-by: Nlouhongxiang <louhongxiang@huawei.com>
Reviewed-by: NKefeng Wang <wangkefeng.wang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

euleros inclusion
category: feature
feature: etmem
bugzilla: https://gitee.com/openeuler/kernel/issues/I4OODH?from=project-issue
CVE: NA

-------------------------------------------------

Driver dax_kmem will export pmem as a NUMA node. This patch will
record node consists of persistent memory for futher use.
Signed-off-by: NKemeng Shi <shikemeng@huawei.com>
Reviewed-by: Nlouhongxiang <louhongxiang@huawei.com>
Reviewed-by: NKefeng Wang <wangkefeng.wang@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

euleros inclusion
category: feature
feature: etmem
bugzilla: https://gitee.com/openeuler/kernel/issues/I4OODH?from=project-issue
CVE: NA

-------------------------------------------------

add a callback in pte_hole during walk_page_range for user to scan page
without page table.
Signed-off-by: NKemeng Shi <shikemeng@huawei.com>
Reviewed-by: Nlouhongxiang <louhongxiang@huawei.com>
Reviewed-by: NChen Wandun <chenwandun@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

euleros inclusion
category: feature
feature: etmem
bugzilla: https://gitee.com/openeuler/kernel/issues/I4OODH?from=project-issue
CVE: NA

-------------------------------------------------

Now, we will call cond_resched after scan a full memslot. If we scan a huge
memslot, it will take long time before cond_resched. So call cond_resched
after scan walk_step size memory.
Signed-off-by: NKemeng Shi <shikemeng@huawei.com>
Reviewed-by: Nlouhongxiang <louhongxiang@huawei.com>
Reviewed-by: NChen Wandun <chenwandun@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

euleros inclusion
category: feature
feature: etmem
bugzilla: https://gitee.com/openeuler/kernel/issues/I4OODH?from=project-issue
CVE: NA

-------------------------------------------------

Kvm shadow page may be freed when etmem_scan is walking ept page table.
Hold mmu_lock when walking ept page table to avoid UAF.
To avoid holding mmu_lock for too long time, walk step module parameter
is added to control lock holding time.
Signed-off-by: NKemeng Shi <shikemeng@huawei.com>
Reviewed-by: Nlouhongxiang <louhongxiang@huawei.com>
Reviewed-by: NChen Wandun <chenwandun@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

euleros inclusion
category: feature
feature: etmem
bugzilla: https://gitee.com/openeuler/kernel/issues/I4OODH?from=project-issue
CVE: NA

-------------------------------------------------

Module scan/swap and etmem access export file operations without
protection.
Kernel crash can be triggered by following:
1.insert scan/swap module.
2.etmem check if exported file operations are set.
3.remove scan/swap module.
4.etmem call checked file operation.
5.kernel crash happens.

Fix this as following:
Module scan/swap set and clear operations with lock held.
Etmem in kernel calls try_module_get to with lock held.
Etmem call read/open/release/ioctl callback without lock held with module
get.

Another concurrent access situaction is that open for idles_pages and
swap_pages will success without scan/swap module inserted. If scan/swap
module is inserteds after open, subsequent call of open/read/close will
call exported file operations set by scan/swap. This also may trigger
kernel crash as following:
1.open idle_pages or swap_pages
2.modprobe scan/swap module
3.close idle_pages or swap_pages(module_put is called without
try_module_get)
4.modprobe -r scan/swap module found invalid module reference count in
trace delete_module syscall->try_stop_module->try_release_module_ref
and report a BUG_ON for ret < 0.

Fix this by only return file successfully with scan/swap module inserted.
Signed-off-by: NKemeng Shi <shikemeng@huawei.com>
Reviewed-by: Nlouhongxiang <louhongxiang@huawei.com>
Reviewed-by: NChen Wandun <chenwandun@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

euleros inclusion
category: feature
feature: etmem
bugzilla: https://gitee.com/openeuler/kernel/issues/I4OODH?from=project-issue
CVE: NA

-------------------------------------------------

free pic before return from vm_idle_read in etmem scan
Signed-off-by: NKemeng Shi <shikemeng@huawei.com>
Reviewed-by: Nlouhongxiang <louhongxiang@huawei.com>
Reviewed-by: NChen Wandun <chenwandun@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

euleros inclusion
category: feature
feature: etmem
bugzilla: https://gitee.com/openeuler/kernel/issues/I4OODH?from=project-issue
CVE: NA

-------------------------------------------------

Before this patch, etmem_scan is failed if vm and host has different
page level. This patch supports scan 4 level ept while 5 level page
is enabled in host.
Signed-off-by: NKemeng Shi <shikemeng@huawei.com>
Reviewed-by: Nlouhongxiang <louhongxiang@huawei.com>
Reviewed-by: NChen Wandun <chenwandun@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

euleros inclusion
category: feature
feature: etmem
bugzilla: https://gitee.com/openeuler/kernel/issues/I4OODH?from=project-issue
CVE: NA

-------------------------------------------------

1. add hugetlb_entry callback to report hugetlb page.
2. try to walk host page table when ept entry is not present.
3. add SCAN_AS_HUGE to report ept page in pmd level as host
hugetlb page may be splited into 4k ept page in vm.
4. add SCAN_IGN_HOST for user to ignore access from host.
Signed-off-by: NKemeng Shi <shikemeng@huawei.com>
Reviewed-by: Nlouhongxiang <louhongxiang@huawei.com>
Reviewed-by: NChen Wandun <chenwandun@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

euleros inclusion
category: feature
feature: etmem
bugzilla: https://gitee.com/openeuler/kernel/issues/I4OODH?from=project-issue
CVE: NA

-------------------------------------------------

support ioctl for etmem scan to set scan flag
Signed-off-by: NKemeng Shi <shikemeng@huawei.com>
Reviewed-by: Nlouhongxiang <louhongxiang@huawei.com>
Reviewed-by: NChen Wandun <chenwandun@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

mainline inclusion
from mainline-v5.16-rc5
commit 38d22088
category: feature
bugzilla: https://gitee.com/openeuler/kernel/issues/I4O23M
CVE: NA

---------------------------------------------------------------------

HIP06 is no longer supported. In order to reduce unnecessary maintenance,
the code of HIP06 is removed.

Link: https://lore.kernel.org/r/20211220130558.61585-1-liangwenpeng@huawei.comSigned-off-by: NChengchang Tang <tangchengchang@huawei.com>
Signed-off-by: NWenpeng Liang <liangwenpeng@huawei.com>
Reviewed-by: NLeon Romanovsky <leonro@nvidia.com>
Signed-off-by: NJason Gunthorpe <jgg@nvidia.com>
sigend-off-by: NGuofeng Yue <yueguofeng@hisilicon.com>
Reviewed-by: NYangyang Li <liyangyang20@huawei.com>
Acked-by: NXie XiuQi <xiexiuqi@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

mainline inclusion
from mainline-v5.15-rc1
commit ab5cbb9d
category: feature
bugzilla: https://gitee.com/openeuler/kernel/issues/I4O23M
CVE: NA

---------------------------------------------------------------------

There is no need to prints error for hw_v1.

Link: https://lore.kernel.org/r/1629985056-57004-5-git-send-email-liangwenpeng@huawei.comSigned-off-by: NWeihang Li <liweihang@huawei.com>
Signed-off-by: NWenpeng Liang <liangwenpeng@huawei.com>
Signed-off-by: NJason Gunthorpe <jgg@nvidia.com>
sigend-off-by: NGuofeng Yue <yueguofeng@hisilicon.com>
Reviewed-by: NYangyang Li <liyangyang20@huawei.com>
Acked-by: NXie XiuQi <xiexiuqi@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>

mainline inclusion
from mainline-v5.15-rc1
commit 0045e0d3
category: feature
bugzilla: https://gitee.com/openeuler/kernel/issues/I4O23M
CVE: NA

Reference: https://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma.git/commit/?id=0045e0d3f42ed7d05434bb5bc16acfc793ea4891

---------------------------------------------------------------------

The current write wqe mechanism is to write to DDR first, and then notify
the hardware through doorbell to read the data. Direct wqe is a mechanism
to fill wqe directly into the hardware. In the case of light load, the wqe
will be filled into pcie bar space of the hardware, this will reduce one
memory access operation and therefore reduce the latency. SIMD
instructions allows cpu to write the 512 bits at one time to device
memory, thus it can be used for posting direct wqe.

Add direct wqe enable switch and address mapping.

Link: https://lore.kernel.org/r/20211207124901.42123-2-liangwenpeng@huawei.comSigned-off-by: NYixing Liu <liuyixing1@huawei.com>
Signed-off-by: NWenpeng Liang <liangwenpeng@huawei.com>
Signed-off-by: NJason Gunthorpe <jgg@nvidia.com>
sigend-off-by: NGuofeng Yue <yueguofeng@hisilicon.com>
Reviewed-by: NYangyang Li <liyangyang20@huawei.com>
Acked-by: NXie XiuQi <xiexiuqi@huawei.com>
Signed-off-by: NZheng Zengkai <zhengzengkai@huawei.com>