This bug was introduced in v4.8-rc1.
Signed-off-by: NEric Biggers <ebiggers@google.com>
Signed-off-by: NTheodore Ts'o <tytso@mit.edu>
Cc: stable@vger.kernel.org

There are some repetitive code in jbd2_journal_init_dev() and
jbd2_journal_init_inode(). So this patch moves the common code into
journal_init_common() helper to simplify the code. And fix the coding
style warnings reported by checkpatch.pl by the way.
Signed-off-by: NGeliang Tang <geliangtang@gmail.com>
Signed-off-by: NTheodore Ts'o <tytso@mit.edu>
Reviewed-by: NJan Kara <jack@suse.cz>

MAX_32_NUM isn't used in ext4
Signed-off-by: NFabian Frederick <fabf@skynet.be>
Signed-off-by: NTheodore Ts'o <tytso@mit.edu>

Create a macro to calculate length + offset -> maximum blocks
This adds more readability.
Signed-off-by: NFabian Frederick <fabf@skynet.be>
Signed-off-by: NTheodore Ts'o <tytso@mit.edu>

ext4_alloc_file_blocks() is called from ext4_zero_range() and
ext4_fallocate() both already testing EXT4_INODE_EXTENTS
We can call ext_depth(inode) unconditionnally.

[ Added BUG_ON check to make sure ext4_alloc_file_blocks() won't get
  called for a indirect-mapped inode in the future.  -- tytso ]
Signed-off-by: NFabian Frederick <fabf@skynet.be>
Signed-off-by: NTheodore Ts'o <tytso@mit.edu>

Running xfstests generic/013 with kmemleak gives the following:

unreferenced object 0xffff8801d3d27de0 (size 96):
  comm "fsstress", pid 4941, jiffies 4294860168 (age 53.485s)
  hex dump (first 32 bytes):
    00 00 00 00 00 00 00 00 00 00 01 00 00 00 00 00  ................
    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
  backtrace:
    [<ffffffff818eaaf3>] kmemleak_alloc+0x23/0x40
    [<ffffffff81179805>] __kmalloc+0xf5/0x1d0
    [<ffffffff8122ef5c>] ext4_find_extent+0x1ec/0x2f0
    [<ffffffff8123530c>] ext4_insert_range+0x34c/0x4a0
    [<ffffffff81235942>] ext4_fallocate+0x4e2/0x8b0
    [<ffffffff81181334>] vfs_fallocate+0x134/0x210
    [<ffffffff8118203f>] SyS_fallocate+0x3f/0x60
    [<ffffffff818efa9b>] entry_SYSCALL_64_fastpath+0x13/0x8f
    [<ffffffffffffffff>] 0xffffffffffffffff

Problem seems mitigated by dropping refs and freeing path
when there's no path[depth].p_ext

Cc: stable@vger.kernel.org
Signed-off-by: NFabian Frederick <fabf@skynet.be>
Signed-off-by: NTheodore Ts'o <tytso@mit.edu>

Pages clear buffers after ext4 delayed block allocation failed,
However, it does not clean its pte_dirty flag.
if the pages unmap ,in cording to the pte_dirty ,
unmap_page_range may try to call __set_page_dirty,

which may lead to the bugon at 
mpage_prepare_extent_to_map:head = page_buffers(page);.

This patch just call clear_page_dirty_for_io to clean pte_dirty 
at mpage_release_unused_pages for pages mmaped. 

Steps to reproduce the bug:

（1） mmap a file in ext4
	addr = (char *)mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_SHARED,
	       	            fd, 0);
	memset(addr, 'i', 4096);

（2） return EIO at 

	ext4_writepages->mpage_map_and_submit_extent->mpage_map_one_extent 

which causes this log message to be print:

                ext4_msg(sb, KERN_CRIT,
                        "Delayed block allocation failed for "
                        "inode %lu at logical offset %llu with"
                        " max blocks %u with error %d",
                        inode->i_ino,
                        (unsigned long long)map->m_lblk,
                        (unsigned)map->m_len, -err);

(3）Unmap the addr cause warning at

	__set_page_dirty:WARN_ON_ONCE(warn && !PageUptodate(page));

(4) wait for a minute,then bugon happen.

Cc: stable@vger.kernel.org
Signed-off-by: Nwangguang <wangguang03@zte.com>
Signed-off-by: NTheodore Ts'o <tytso@mit.edu>

ext4lazyinit is a global thread. This thread performs itable
initalization under li_list_mtx mutex.

It basically does the following:
ext4_lazyinit_thread
  ->mutex_lock(&eli->li_list_mtx);
  ->ext4_run_li_request(elr)
    ->ext4_init_inode_table-> Do a lot of IO if the list is large

And when new mount/umount arrive they have to block on ->li_list_mtx
because  lazy_thread holds it during full walk procedure.
ext4_fill_super
 ->ext4_register_li_request
   ->mutex_lock(&ext4_li_info->li_list_mtx);
   ->list_add(&elr->lr_request, &ext4_li_info >li_request_list);
In my case mount takes 40minutes on server with 36 * 4Tb HDD.
Common user may face this in case of very slow dev ( /dev/mmcblkXXX)
Even more. If one of filesystems was frozen lazyinit_thread will simply
block on sb_start_write() so other mount/umount will be stuck forever.

This patch changes logic like follows:
- grab ->s_umount read sem before processing new li_request.
  After that it is safe to drop li_list_mtx because all callers of
  li_remove_request are holding ->s_umount for write.
- li_thread skips frozen SB's

Locking order:
Mh KOrder is asserted by umount path like follows: s_umount ->li_list_mtx so
the only way to to grab ->s_mount inside li_thread is via down_read_trylock

xfstests:ext4/023
#PSBM-49658
Signed-off-by: NDmitry Monakhov <dmonakhov@openvz.org>
Signed-off-by: NTheodore Ts'o <tytso@mit.edu>

A condition !hlist_empty(&inode->i_dentry) is always true for open file.
Just remove it. Also ext4_sync_parent() could use some explanation why
races with rmdir() are not an issue - add a comment explaining that.
Reported-by: NAl Viro <viro@ZenIV.linux.org.uk>
Signed-off-by: NJan Kara <jack@suse.cz>
Signed-off-by: NTheodore Ts'o <tytso@mit.edu>

Use the ext4_{has,set,clear}_feature_* helpers to replace the old
feature helpers.
Signed-off-by: NKaho Ng <ngkaho1234@gmail.com>
Signed-off-by: NTheodore Ts'o <tytso@mit.edu>
Reviewed-by: NJan Kara <jack@suse.cz>
Reviewed-by: NDarrick J. Wong <darrick.wong@oracle.com>

When quota information is stored in quota files, we enable only quota
accounting on mount and enforcement is enabled only in response to
Q_QUOTAON quotactl. To make ext4 behavior consistent with XFS, we add a
possibility to enable quota enforcement on mount by specifying
corresponding quota mount option (usrquota, grpquota, prjquota).
Signed-off-by: NJan Kara <jack@suse.cz>
Signed-off-by: NTheodore Ts'o <tytso@mit.edu>

Now, ext4_do_update_inode() clears high 16-bit fields of uid/gid
of deleted and evicted inode to fix up interoperability with old
kernels. However, it checks only i_dtime of an inode to determine
whether the inode was deleted and evicted, and this is very risky,
because i_dtime can be used for the pointer maintaining orphan inode
list, too. We need to further check whether the i_dtime is being
used for the orphan inode list even if the i_dtime is not NULL.

We found that high 16-bit fields of uid/gid of inode are unintentionally
and permanently cleared when the inode truncation is just triggered,
but not finished, and the inode metadata, whose high uid/gid bits are
cleared, is written on disk, and the sudden power-off follows that
in order.

Cc: stable@vger.kernel.org
Signed-off-by: NDaeho Jeong <daeho.jeong@samsung.com>
Signed-off-by: NHobin Woo <hobin.woo@samsung.com>
Signed-off-by: NTheodore Ts'o <tytso@mit.edu>

register_shrinker in mb_cache_create may fail due to no memory. This
patch fixes to do the check of return value of register_shrinker and
handle the error case, otherwise mb_cache_create may return with no
error, but losing the inner shrinker.
Signed-off-by: NChao Yu <yuchao0@huawei.com>
Reviewed-by: NJan Kara <jack@suse.cz>
Signed-off-by: NTheodore Ts'o <tytso@mit.edu>

Online defragging of encrypted files is not currently implemented.
However, the move extent ioctl can still return successfully when
called.  For example, this occurs when xfstest ext4/020 is run on an
encrypted file system, resulting in a corrupted test file and a
corresponding test failure.

Until the proper functionality is implemented, fail the move extent
ioctl if either the original or donor file is encrypted.

Cc: stable@vger.kernel.org
Signed-off-by: NEric Whitney <enwlinux@gmail.com>
Signed-off-by: NTheodore Ts'o <tytso@mit.edu>

Move loop to make enough space in the inode from
ext4_expand_extra_isize_ea() into a separate function to make that
function smaller and better readable and also to avoid delaration of
variables inside a loop block.
Signed-off-by: NJan Kara <jack@suse.cz>
Signed-off-by: NTheodore Ts'o <tytso@mit.edu>

'start' variable is completely unused in ext4_expand_extra_isize_ea().
Variable 'first' is used only once in one place. So just remove them.
Variables 'entry' and 'last' are only really used later in the function
inside a loop. Move their declarations there.
Signed-off-by: NJan Kara <jack@suse.cz>
Signed-off-by: NTheodore Ts'o <tytso@mit.edu>

Factor out function for moving xattrs from inode into external xattr
block from ext4_expand_extra_isize_ea(). That function is already quite
long and factoring out this rather standalone functionality helps
readability.
Signed-off-by: NJan Kara <jack@suse.cz>
Signed-off-by: NTheodore Ts'o <tytso@mit.edu>

We were checking whether computed offsets do not exceed end of block in
ext4_xattr_shift_entries(). However this does not make sense since we
always only decrease offsets. So replace that assertion with a check
whether we really decrease xattrs value offsets.
Signed-off-by: NJan Kara <jack@suse.cz>
Signed-off-by: NTheodore Ts'o <tytso@mit.edu>

Currently we don't support xattrs with e_value_block set. We don't allow
them to pass initial xattr check so there's no point for checking for
this later. Since these tests were untested, bugs were creeping in and
not all places which should have checked were checking e_value_block
anyway.
Signed-off-by: NJan Kara <jack@suse.cz>
Signed-off-by: NTheodore Ts'o <tytso@mit.edu>

Currently we don't support xattrs with values stored out of line. Check
for that in ext4_xattr_check_names() to make sure we never work with
such xattrs since not all the code counts with that resulting is possible
weird corruption issues.
Signed-off-by: NJan Kara <jack@suse.cz>
Signed-off-by: NTheodore Ts'o <tytso@mit.edu>

Conditions checking whether there is enough free space in an xattr block
and when xattr is large enough to make enough space in the inode forgot
to account for the fact that inode need not be completely filled up with
xattrs. Thus we could move unnecessarily many xattrs out of inode or
even falsely claim there is not enough space to expand the inode. We
also forgot to update the amount of free space in xattr block when moving
more xattrs and thus could decide to move too big xattr resulting in
unexpected failure.

Fix these problems by properly updating free space in the inode and
xattr block as we move xattrs. To simplify the math, avoid shifting
xattrs after removing each one xattr and instead just shift xattrs only
once there is enough free space in the inode.
Signed-off-by: NJan Kara <jack@suse.cz>
Signed-off-by: NTheodore Ts'o <tytso@mit.edu>

Pull ext4 fixes from Ted Ts'o:
 "Fix bugs that could cause kernel deadlocks or file system corruption
  while moving xattrs to expand the extended inode.

  Also add some sanity checks to the block group descriptors to make
  sure we don't end up overwriting the superblock"

* tag 'ext4_for_linus_stable' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4:
  ext4: avoid deadlock when expanding inode size
  ext4: properly align shifted xattrs when expanding inodes
  ext4: fix xattr shifting when expanding inodes part 2
  ext4: fix xattr shifting when expanding inodes
  ext4: validate that metadata blocks do not overlap superblock
  ext4: reserve xattr index for the Hurd

Pull networking fixes from David Miller:

 1) Segregate namespaces properly in conntrack dumps, from Liping Zhang.

 2) tcp listener refcount fix in netfilter tproxy, from Eric Dumazet.

 3) Fix timeouts in qed driver due to xmit_more, from Yuval Mintz.

 4) Fix use-after-free in tcp_xmit_retransmit_queue().

 5) Userspace header fixups (use of __u32, missing includes, etc.) from
    Mikko Rapeli.

 6) Further refinements to fragmentation wrt gso and tunnels, from
    Shmulik Ladkani.

 7) Trigger poll correctly for zero length UDP packets, from Eric
    Dumazet.

 8) TCP window scaling fix, also from Eric Dumazet.

 9) SLAB_DESTROY_BY_RCU is not relevant any more for UDP sockets.

10) Module refcount leak in qdisc_create_dflt(), from Eric Dumazet.

11) Fix deadlock in cp_rx_poll() of 8139cp driver, from Gao Feng.

12) Memory leak in rhashtable's alloc_bucket_locks(), from Eric Dumazet.

13) Add new device ID to alx driver, from Owen Lin.

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (83 commits)
  Add Killer E2500 device ID in alx driver.
  net: smc91x: fix SMC accesses
  Documentation: networking: dsa: Remove platform device TODO
  net/mlx5: Increase number of ethtool steering priorities
  net/mlx5: Add error prints when validate ETS failed
  net/mlx5e: Fix memory leak if refreshing TIRs fails
  net/mlx5e: Add ethtool counter for TX xmit_more
  net/mlx5e: Fix ethtool -g/G rx ring parameter report with striding RQ
  net/mlx5e: Don't wait for SQ completions on close
  net/mlx5e: Don't post fragmented MPWQE when RQ is disabled
  net/mlx5e: Don't wait for RQ completions on close
  net/mlx5e: Limit UMR length to the device's limitation
  rhashtable: fix a memory leak in alloc_bucket_locks()
  sfc: fix potential stack corruption from running past stat bitmask
  team: loadbalance: push lacpdus to exact delivery
  net: hns: dereference ppe_cb->ppe_common_cb if it is non-null
  8139cp: Fix one possible deadloop in cp_rx_poll
  i40e: Change some init flow for the client
  Revert "phy: IRQ cannot be shared"
  net: dsa: bcm_sf2: Fix race condition while unmasking interrupts
  ...

Merge tag 'platform-drivers-x86-v4.8-4' of git://git.infradead.org/users/dvhart/linux-platform-drivers-x86

Pull x86 platform driver fixes from Darren Hart:
 "Remove module related code from two drivers that are only configurable
  as built-in: intel_pmic_gpio and platform/olpc"

* tag 'platform-drivers-x86-v4.8-4' of git://git.infradead.org/users/dvhart/linux-platform-drivers-x86:
  intel_pmic_gpio: Make explicitly non-modular
  platform/olpc: Make ec explicitly non-modular

Pull powerpc fixes from Ben Herrenschmidt:
 "This was meant to be sent early last week, but I has a change pending
  on one of the fixes and other things made me forget all about.  Ugh.

  We have some misc fixes for powerpc 4.8.  Some trivial bits and some
  regressions, and a trivial cleanup or two that I saw no point in
  letting rot in patchwork"

* tag 'powerpc-4.8-4' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:
  powerpc: signals: Discard transaction state from signal frames
  powerpc/powernv : Drop reference added by kset_find_obj()
  powerpc/tm: do not use r13 for tabort_syscall
  powerpc: move hmi.c to arch/powerpc/kvm/
  powerpc: sysdev: cpm: fix gpio save_regs functions
  powerpc/pseries: PACA save area fix for MCE vs MCE
  powerpc/pseries: PACA save area fix for general exception vs MCE
  powerpc/prom: Fix sub-processor option passed to ibm, client-architecture-support
  powerpc, hotplug: Avoid to touch non-existent cpumasks.
  powerpc: migrate exception table users off module.h and onto extable.h
  powerpc/powernv/pci: fix iterator signedness
  powerpc/pseries: use pci_host_bridge.release_fn() to kfree(phb)
  cxl: use pcibios_free_controller_deferred() when removing vPHBs
  powerpc: mpc8349emitx: Delete unnecessary assignment for the field "owner"
  powerpc/512x: Delete unnecessary assignment for the field "owner"
  drivers/macintosh: Delete owner assignment
  powerpc: cputhreads: Add missing include file

The Kconfig entry controlling compilation of this code is:

drivers/platform/x86/Kconfig:config GPIO_INTEL_PMIC
drivers/platform/x86/Kconfig:   bool "Intel PMIC GPIO support"

...meaning that it currently is not being built as a module by anyone.

Lets remove the couple traces of modular infrastructure use, so that
when reading the driver there is no doubt it is builtin-only.

We delete the MODULE_LICENSE tag etc. since all that information
was (or is now) contained at the top of the file in the comments.

We don't replace module.h with init.h since the file already has that.

Cc: Alek Du <alek.du@intel.com>
Cc: platform-driver-x86@vger.kernel.org
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: NDarren Hart <dvhart@linux.intel.com>

The Kconfig entry controlling compilation of this code is:

arch/x86/Kconfig:config OLPC
arch/x86/Kconfig:       bool "One Laptop Per Child support"

...meaning that it currently is not being built as a module by anyone.

Lets remove the couple traces of modular infrastructure use, so that
when reading the driver there is no doubt it is builtin-only.

We delete the MODULE_LICENSE tag etc. since all that information
was (or is now) contained at the top of the file in the comments.

Cc: platform-driver-x86@vger.kernel.org
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>
Acked-by: NAndres Salomon <dilinger@queued.net>
Signed-off-by: NDarren Hart <dvhart@linux.intel.com>

Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Commit b70661c7 ("net: smc91x: use run-time configuration on all ARM
machines") broke some ARM platforms through several mistakes.  Firstly,
the access size must correspond to the following rule:

(a) at least one of 16-bit or 8-bit access size must be supported
(b) 32-bit accesses are optional, and may be enabled in addition to
    the above.

Secondly, it provides no emulation of 16-bit accesses, instead blindly
making 16-bit accesses even when the platform specifies that only 8-bit
is supported.

Reorganise smc91x.h so we can make use of the existing 16-bit access
emulation already provided - if 16-bit accesses are supported, use
16-bit accesses directly, otherwise if 8-bit accesses are supported,
use the provided 16-bit access emulation.  If neither, BUG().  This
exactly reflects the driver behaviour prior to the commit being fixed.

Since the conversion incorrectly cut down the available access sizes on
several platforms, we also need to go through every platform and fix up
the overly-restrictive access size: Arnd assumed that if a platform can
perform 32-bit, 16-bit and 8-bit accesses, then only a 32-bit access
size needed to be specified - not so, all available access sizes must
be specified.

This likely fixes some performance regressions in doing this: if a
platform does not support 8-bit accesses, 8-bit accesses have been
emulated by performing a 16-bit read-modify-write access.

Tested on the Intel Assabet/Neponset platform, which supports only 8-bit
accesses, which was broken by the original commit.

Fixes: b70661c7 ("net: smc91x: use run-time configuration on all ARM machines")
Signed-off-by: NRussell King <rmk+kernel@armlinux.org.uk>
Tested-by: NRobert Jarzmik <robert.jarzmik@free.fr>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Since commit 83c0afae ("net: dsa: Add new binding implementation"),
the shortcomings of the dsa platform device have been addressed, remove
that TODO item.
Signed-off-by: NFlorian Fainelli <f.fainelli@gmail.com>
Acked-by: NAndrew Lunn <andrew@lunn.ch>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Saeed Mahameed says:

====================
Mellanox 100G mlx5 fixes 2016-08-29

This series contains some bug fixes for the mlx5 core and mlx5
ethernet driver.

From Saeed, Fix UMR to consider hardware translation table field
size limitation when calculating the maximum number of MTTs required
by the driver.  Three patches to speed-up netdevice close time by
serializing channel (SQs & RQs) destruction rather than issuing and
waiting for hardware interrupts to free them.

From Eran, Fix ethtool ring parameter reporting for striding RQ layout.
Add error prints on ETS validation failure.

From Kamal, Fix memory leak on error flow.

From Maor, Fix ethtool steering priorities number.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Ethtool has 11 flow tables, each flow table has its own priority.
Increase the number of priorities to be aligned with the number of flow
tables.

Fixes: 1174fce8 ('net/mlx5e: Support l3/l4 flow type specs in ethtool flow steering')
Signed-off-by: NMaor Gottlieb <maorg@mellanox.com>
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Upon set ETS failure due to user invalid input, add error prints to
specify the exact error to the user.

Fixes: cdcf1121 ('net/mlx5e: Validate BW weight values of ETS')
Signed-off-by: NEran Ben Elisha <eranbe@mellanox.com>
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Free 'in' command object also when mlx5_core_modify_tir fails.

Fixes: 724b2aa1 ("net/mlx5e: TIRs management refactoring")
Signed-off-by: NKamal Heib <kamalh@mellanox.com>
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add a counter in ethtool for the number of times that
TX xmit_more was used.
Signed-off-by: NTariq Toukan <tariqt@mellanox.com>
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The driver RQ has two possible configurations: striding RQ and
non-striding RQ.  Until this patch, the driver always reported the
number of hardware WQEs (ring descriptors). For non striding RQ
configuration, this was OK since we have one WQE per pending packet
For striding RQ, multiple packets can fit into one WQE. For better
user experience we normalize the rx_pending parameter (size of wqe/mtu)
as the average ring size in case of striding RQ.

Fixes: 461017cb ('net/mlx5e: Support RX multi-packet WQE ...')
Signed-off-by: NEran Ben Elisha <eranbe@mellanox.com>
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Instead of asking the firmware to flush the SQ (Send Queue) via
asynchronous completions when moved to error, we handle SQ flush
manually (mlx5e_free_tx_descs) same as we did when SQ flush got
timed out or on tx_timeout.

This will reduce SQs flush time and speedup interface down procedure.

Moved mlx5e_free_tx_descs to the end of en_tx.c for tx
critical code locality.

Fixes: 29429f33 ('net/mlx5e: Timeout if SQ doesn't flush during close')
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

ICO (Internal control operations) SQ (Send Queue) is closed/disabled
after RQ (Receive Queue).  After RQ is closed an ICO SQ completion
might post a fragmented MPWQE (Multi Packet Work Queue Element) into
that RQ.

As on regular RQ post, check if we are allowed to post to that
RQ (RQ is enabled). Cleanup in-progress UMR MPWQE on mlx5e_free_rx_descs
if needed.

Fixes: bc77b240 ('net/mlx5e: Add fragmented memory support for RX multi packet WQE')
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This will significantly reduce receive queue flush time on interface
down.

Instead of asking the firmware to flush the RQ (Receive Queue) via
asynchronous completions when moved to error, we handle RQ flush
manually (mlx5e_free_rx_descs) same as we did when RQ flush got timed
out.

This will reduce RQs flush time and speedup interface down procedure
(ifconfig down) from 6 sec to 0.3 sec on a 48 cores system.

Moved mlx5e_free_rx_descs en_main.c where it is needed, to keep en_rx.c
free form non critical data path code for better code locality.

Fixes: 6cd392a0 ('net/mlx5e: Handle RQ flush in error cases')
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

ConnectX-4 UMR (User Memory Region) MTT translation table offset in WQE
is limited to U16_MAX, before this patch we ignored that limitation and
requested the maximum possible UMR translation length that the netdev
might need (MAX channels * MAX pages per channel).
In case of a system with #cores > 32 and when linear WQE allocation fails,
falling back to using UMR WQEs will cause the RQ (Receive Queue) to get
stuck.

Here we limit UMR length to min(U16_MAX, max required pages) (while
considering the required alignments) on driver load, by default U16_MAX is
sufficient since the default RX rings value guarantees that we are in
range, dynamically (on set_ringparam/set_channels) we will check if the
new required UMR length (num mtts) is still in range, if not, fail the
request.

Fixes: bc77b240 ('net/mlx5e: Add fragmented memory support for RX multi packet WQE')
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>