Newer firmware has a new call HWRM_FUNC_QSTATS_EXT to retrieve the
masks of all ring counters.  Make this call when supported to
initialize the hardware masks of all ring counters.  If the call
is not available, assume 48-bit ring counter masks on P5 chips.
Reviewed-by: Vasundhara Volam <vasundhara-v.volam@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Some of these DMAed hardware counters are not full 64-bit counters and
so we need to accumulate them as they overflow.  Allocate copies of these
DMA statistics memory blocks with the same size for accumulation.  The
hardware counter widths are also counter specific so we allocate
memory for masks that correspond to each counter.
Reviewed-by: Vasundhara Volam <vasundhara-v.volam@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The driver manages multiple statistics structures of different sizes.
They are all allocated, freed, and handled practically the same.  Define
a new bnxt_stats_mem structure and common allocation and free functions
for all staistics memory blocks.
Reviewed-by: Vasundhara Volam <vasundhara-v.volam@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The port statistics structures have hard coded padding and offset.
Define macros to make this look cleaner.
Reviewed-by: Pavan Chebbi <pavan.chebbi@broadcom.com>
Reviewed-by: Vasundhara Volam <vasundhara-v.volam@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Main changes are 200G support and fixing the definitions of discard and
error counters to match the hardware definitions.

Because the HWRM_PORT_PHY_QCFG message size has now exceeded the max.
encapsulated response message size of 96 bytes from the PF to the VF,
we now need to cap this message to 96 bytes for forwarding.  The forwarded
response only needs to contain the basic link status and speed information
and can be capped without adding the new information.

v2: Fix bnxt_re compile error.

Cc: Selvin Xavier <selvin.xavier@broadcom.com>
Reviewed-by: Vasundhara Volam <vasundhara-v.volam@broadcom.com>
Reviewed-by: Edwin Peer <edwin.peer@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Remove PCIe non-counters display from ethtool statistics, as
they are not simple counters but register dump.  The next few
patches will add logic to detect counter roll-over and it won't
work with these PCIe non-counters.

There will be a follow up patch to get PCIe information via
ethtool register dump.
Signed-off-by: Vasundhara Volam <vasundhara-v.volam@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

MPLS has no dependency with the device type of underlying devices.
Hence the device type check to add mpls support for devices can be
avoided.
Signed-off-by: Martin Varghese <martin.varghese@nokia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Cited commit mistakenly copied provided option to 'val' instead of to
'mfc':

```
-               if (copy_from_user(&mfc, optval, sizeof(mfc))) {
+               if (copy_from_sockptr(&val, optval, sizeof(val))) {
```

Fix this by copying the option to 'mfc'.

selftest router_multicast.sh before:

$ ./router_multicast.sh
smcroutectl: Unknown or malformed IPC message 'a' from client.
smcroutectl: failed removing multicast route, does not exist.
TEST: mcast IPv4                                                    [FAIL]
        Multicast not received on first host
TEST: mcast IPv6                                                    [ OK ]
smcroutectl: Unknown or malformed IPC message 'a' from client.
smcroutectl: failed removing multicast route, does not exist.
TEST: RPF IPv4                                                      [FAIL]
        Multicast not received on first host
TEST: RPF IPv6                                                      [ OK ]

selftest router_multicast.sh after:

$ ./router_multicast.sh
TEST: mcast IPv4                                                    [ OK ]
TEST: mcast IPv6                                                    [ OK ]
TEST: RPF IPv4                                                      [ OK ]
TEST: RPF IPv6                                                      [ OK ]

Fixes: 01ccb5b4 ("net/ipv4: switch ip_mroute_setsockopt to sockptr_t")
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

Karsten Graul says:

====================
net/smc: show unique rsn code for exceeded max dmb count

Resolve some confusion at the user side when the reason code shows
out-of-memory but actually there is enough memory left.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

When the maximum dmb buffer limit for an ism device is reached no more
dmb buffers can be registered. When this happens the reason code is set
to SMC_CLC_DECL_MEM indicating out-of-memory. This is the same reason
code that is used when no memory could be allocated for the new dmb
buffer.
This is confusing for users when they see this error but there is more
memory available. To solve this set a separate new reason code when the
maximum dmb limit exceeded.
Reviewed-by: Ursula Braun <ubraun@linux.ibm.com>
Signed-off-by: Karsten Graul <kgraul@linux.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

When the ism driver allocates a new dmb in ism_alloc_dmb() it must
first check for and reserve a slot in the sba bitmap. When
find_next_zero_bit() finds no free slot then the return code is -ENOMEM.
This code conflicts with the error when the alloc() fails later in the
code. As a result of that the caller can not differentiate
between out-of-memory conditions and sba-bitmap-full conditions.
Fix that by using the return code -ENOSPC when the sba slot
reservation failed.
Reviewed-by: Ursula Braun <ubraun@linux.ibm.com>
Signed-off-by: Karsten Graul <kgraul@linux.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

list_for_each_safe is able to handle an empty list.
The only effect of avoiding the loop is not initializing the
index variable.
Drop list_empty tests in cases where these variables are not
used.

The semantic patch that makes this change is as follows:
(http://coccinelle.lip6.fr/)

<smpl>
@@
expression x,e;
iterator name list_for_each_safe;
statement S;
identifier i,j;
@@

-if (!(list_empty(x))) {
   list_for_each_safe(i,j,x) S
- }
 ... when != i
     when != j
(
  i = e;
|
? j = e;
)
</smpl>
Signed-off-by: Julia Lawall <Julia.Lawall@inria.fr>
Acked-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The UDP reuseport conflict was a little bit tricky.

The net-next code, via bpf-next, extracted the reuseport handling
into a helper so that the BPF sk lookup code could invoke it.

At the same time, the logic for reuseport handling of unconnected
sockets changed via commit efc6b6f6
which changed the logic to carry on the reuseport result into the
rest of the lookup loop if we do not return immediately.

This requires moving the reuseport_has_conns() logic into the callers.

While we are here, get rid of inline directives as they do not belong
in foo.c files.

The other changes were cases of more straightforward overlapping
modifications.
Signed-off-by: David S. Miller <davem@davemloft.net>

Merge tag 'riscv-for-linus-5.8-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/riscv/linux into master

Pull RISC-V fixes from Palmer Dabbelt:
 "A few more fixes this week:

   - A fix to avoid using SBI calls during kasan initialization, as the
     SBI calls themselves have not been probed yet.

   - Three fixes related to systems with multiple memory regions"

* tag 'riscv-for-linus-5.8-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/riscv/linux:
  riscv: Parse all memory blocks to remove unusable memory
  RISC-V: Do not rely on initrd_start/end computed during early dt parsing
  RISC-V: Set maximum number of mapped pages correctly
  riscv: kasan: use local_tlb_flush_all() to avoid uninitialized __sbi_rfence

Merge tag 'x86-urgent-2020-07-25' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip into master

Pull x86 fixes from Ingo Molnar:
 "Misc fixes:

   - Fix a section end page alignment assumption that was causing
     crashes

   - Fix ORC unwinding on freshly forked tasks which haven't executed
     yet and which have empty user task stacks

   - Fix the debug.exception-trace=1 sysctl dumping of user stacks,
     which was broken by recent maccess changes"

* tag 'x86-urgent-2020-07-25' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86/dumpstack: Dump user space code correctly again
  x86/stacktrace: Fix reliable check for empty user task stacks
  x86/unwind/orc: Fix ORC for newly forked tasks
  x86, vmlinux.lds: Page-align end of ..page_aligned sections

Merge tag 'perf-urgent-2020-07-25' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip into master

Pull uprobe fix from Ingo Molnar:
 "Fix an interaction/regression between uprobes based shared library
  tracing & GDB"

* tag 'perf-urgent-2020-07-25' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  uprobes: Change handle_swbp() to send SIGTRAP with si_code=SI_KERNEL, to fix GDB regression

Merge tag 'timers-urgent-2020-07-25' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip into master

Pull timer fix from Ingo Molnar:
 "Fix a suspend/resume regression (crash) on TI AM3/AM4 SoC's"

* tag 'timers-urgent-2020-07-25' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  clocksource/drivers/timer-ti-dm: Fix suspend and resume for am3 and am4

Merge tag 'sched-urgent-2020-07-25' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip into master

Pull scheduler fixes from Ingo Molnar:
 "Fix a race introduced by the recent loadavg race fix, plus add a debug
  check for a hard to debug case of bogus wakeup function flags"

* tag 'sched-urgent-2020-07-25' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  sched: Warn if garbage is passed to default_wake_function()
  sched: Fix race against ptrace_freeze_trace()

Merge tag 'efi-urgent-2020-07-25' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip into master

Pull EFI fixes from Ingo Molnar:
 "Various EFI fixes:

   - Fix the layering violation in the use of the EFI runtime services
     availability mask in users of the 'efivars' abstraction

   - Revert build fix for GCC v4.8 which is no longer supported

   - Clean up some x86 EFI stub details, some of which are borderline
     bugs that copy around garbage into padding fields - let's fix these
     out of caution.

   - Fix build issues while working on RISC-V support

   - Avoid --whole-archive when linking the stub on arm64"

* tag 'efi-urgent-2020-07-25' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  efi: Revert "efi/x86: Fix build with gcc 4"
  efi/efivars: Expose RT service availability via efivars abstraction
  efi/libstub: Move the function prototypes to header file
  efi/libstub: Fix gcc error around __umoddi3 for 32 bit builds
  efi/libstub/arm64: link stub lib.a conditionally
  efi/x86: Only copy upto the end of setup_header
  efi/x86: Remove unused variables

Pull cifs fix from Steve French:
 "A fix for a recently discovered regression in rename to older servers
  caused by a recent patch"

* tag '5.8-rc6-cifs-fix' of git://git.samba.org/sfrench/cifs-2.6:
  Revert "cifs: Fix the target file was deleted when rename failed."

Pull networking fixes from David Miller:

 1) Fix RCU locaking in iwlwifi, from Johannes Berg.

 2) mt76 can access uninitialized NAPI struct, from Felix Fietkau.

 3) Fix race in updating pause settings in bnxt_en, from Vasundhara
    Volam.

 4) Propagate error return properly during unbind failures in ax88172a,
    from George Kennedy.

 5) Fix memleak in adf7242_probe, from Liu Jian.

 6) smc_drv_probe() can leak, from Wang Hai.

 7) Don't muck with the carrier state if register_netdevice() fails in
    the bonding driver, from Taehee Yoo.

 8) Fix memleak in dpaa_eth_probe, from Liu Jian.

 9) Need to check skb_put_padto() return value in hsr_fill_tag(), from
    Murali Karicheri.

10) Don't lose ionic RSS hash settings across FW update, from Shannon
    Nelson.

11) Fix clobbered SKB control block in act_ct, from Wen Xu.

12) Missing newlink in "tx_timeout" sysfs output, from Xiongfeng Wang.

13) IS_UDPLITE cleanup a long time ago, incorrectly handled
    transformations involving UDPLITE_RECV_CC. From Miaohe Lin.

14) Unbalanced locking in netdevsim, from Taehee Yoo.

15) Suppress false-positive error messages in qed driver, from Alexander
    Lobakin.

16) Out of bounds read in ax25_connect and ax25_sendmsg, from Peilin Ye.

17) Missing SKB release in cxgb4's uld_send(), from Navid Emamdoost.

18) Uninitialized value in geneve_changelink(), from Cong Wang.

19) Fix deadlock in xen-netfront, from Andera Righi.

19) flush_backlog() frees skbs with IRQs disabled, so should use
    dev_kfree_skb_irq() instead of kfree_skb(). From Subash Abhinov
    Kasiviswanathan.

* git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (111 commits)
  drivers/net/wan: lapb: Corrected the usage of skb_cow
  dev: Defer free of skbs in flush_backlog
  qrtr: orphan socket in qrtr_release()
  xen-netfront: fix potential deadlock in xennet_remove()
  flow_offload: Move rhashtable inclusion to the source file
  geneve: fix an uninitialized value in geneve_changelink()
  bonding: check return value of register_netdevice() in bond_newlink()
  tcp: allow at most one TLP probe per flight
  AX.25: Prevent integer overflows in connect and sendmsg
  cxgb4: add missing release on skb in uld_send()
  net: atlantic: fix PTP on AQC10X
  AX.25: Prevent out-of-bounds read in ax25_sendmsg()
  sctp: shrink stream outq when fails to do addstream reconf
  sctp: shrink stream outq only when new outcnt < old outcnt
  AX.25: Fix out-of-bounds read in ax25_connect()
  enetc: Remove the mdio bus on PF probe bailout
  net: ethernet: ti: add NETIF_F_HW_TC hw feature flag for taprio offload
  net: ethernet: ave: Fix error returns in ave_init
  drivers/net/wan/x25_asy: Fix to make it work
  ipvs: fix the connection sync failed in some cases
  ...

Currently, maximum physical memory allowed is equal to -PAGE_OFFSET.
That's why we remove any memory blocks spanning beyond that size. However,
it is done only for memblock containing linux kernel which will not work
if there are multiple memblocks.

Process all memory blocks to figure out how much memory needs to be removed
and remove at the end instead of updating the memblock list in place.
Signed-off-by: Atish Patra <atish.patra@wdc.com>
Signed-off-by: Palmer Dabbelt <palmerdabbelt@google.com>

Currently, initrd_start/end are computed during early_init_dt_scan
but used during arch_setup. We will get the following panic if initrd is used
and CONFIG_DEBUG_VIRTUAL is turned on.

[    0.000000] ------------[ cut here ]------------
[    0.000000] kernel BUG at arch/riscv/mm/physaddr.c:33!
[    0.000000] Kernel BUG [#1]
[    0.000000] Modules linked in:
[    0.000000] CPU: 0 PID: 0 Comm: swapper Not tainted 5.8.0-rc4-00015-ged0b226fed02 #886
[    0.000000] epc: ffffffe0002058d2 ra : ffffffe0000053f0 sp : ffffffe001001f40
[    0.000000]  gp : ffffffe00106e250 tp : ffffffe001009d40 t0 : ffffffe00107ee28
[    0.000000]  t1 : 0000000000000000 t2 : ffffffe000a2e880 s0 : ffffffe001001f50
[    0.000000]  s1 : ffffffe0001383e8 a0 : ffffffe00c087e00 a1 : 0000000080200000
[    0.000000]  a2 : 00000000010bf000 a3 : ffffffe00106f3c8 a4 : ffffffe0010bf000
[    0.000000]  a5 : ffffffe000000000 a6 : 0000000000000006 a7 : 0000000000000001
[    0.000000]  s2 : ffffffe00106f068 s3 : ffffffe00106f070 s4 : 0000000080200000
[    0.000000]  s5 : 0000000082200000 s6 : 0000000000000000 s7 : 0000000000000000
[    0.000000]  s8 : 0000000080011010 s9 : 0000000080012700 s10: 0000000000000000
[    0.000000]  s11: 0000000000000000 t3 : 000000000001fe30 t4 : 000000000001fe30
[    0.000000]  t5 : 0000000000000000 t6 : ffffffe00107c471
[    0.000000] status: 0000000000000100 badaddr: 0000000000000000 cause: 0000000000000003
[    0.000000] random: get_random_bytes called from print_oops_end_marker+0x22/0x46 with crng_init=0

To avoid the error, initrd_start/end can be computed from phys_initrd_start/size
in setup itself. It also improves the initrd placement by aligning the start
and size with the page size.

Fixes: 76d2a049 ("RISC-V: Init and Halt Code")
Signed-off-by: Atish Patra <atish.patra@wdc.com>
Signed-off-by: Palmer Dabbelt <palmerdabbelt@google.com>

This patch fixed 2 issues with the usage of skb_cow in LAPB drivers
"lapbether" and "hdlc_x25":

1) After skb_cow fails, kfree_skb should be called to drop a reference
to the skb. But in both drivers, kfree_skb is not called.

2) skb_cow should be called before skb_push so that is can ensure the
safety of skb_push. But in "lapbether", it is incorrectly called after
skb_push.

More details about these 2 issues:

1) The behavior of calling kfree_skb on failure is also the behavior of
netif_rx, which is called by this function with "return netif_rx(skb);".
So this function should follow this behavior, too.

2) In "lapbether", skb_cow is called after skb_push. This results in 2
logical issues:
   a) skb_push is not protected by skb_cow;
   b) An extra headroom of 1 byte is ensured after skb_push. This extra
      headroom has no use in this function. It also has no use in the
      upper-layer function that this function passes the skb to
      (x25_lapb_receive_frame in net/x25/x25_dev.c).
So logically skb_cow should instead be called before skb_push.

Cc: Eric Dumazet <edumazet@google.com>
Cc: Martin Schiller <ms@dev.tdt.de>
Signed-off-by: Xie He <xie.he.0141@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Chris Packham says:

====================
net: dsa: mv88e6xxx: port mtu support

This series connects up the mv88e6xxx switches to the dsa infrastructure for
configuring the port MTU. The first patch is also a bug fix which might be a
candiatate for stable.

I've rebased this series on top of net-next/master to pick up Andrew's change
for the gigabit switches. Patch 1 and 2 are unchanged (aside from adding
Andrew's Reviewed-by). Patch 3 is reworked to make use of the existing mtu
support.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Some of the chips in the mv88e6xxx family don't support jumbo
configuration per port. But they do have a chip-wide max frame size that
can be used. Use this to approximate the behaviour of configuring a port
based MTU.
Signed-off-by: Chris Packham <chris.packham@alliedtelesis.co.nz>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

The MV88E6190 and MV88E6190X both support per port jumbo configuration
just like the other GE switches. Install the appropriate ops.
Signed-off-by: Chris Packham <chris.packham@alliedtelesis.co.nz>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

The MV88E6097 chip does not support configuring jumbo frames. Prior to
commit 5f436666 only the 6352, 6351, 6165 and 6320 chips configured
jumbo mode. The refactor accidentally added the function for the 6097.
Remove the erroneous function pointer assignment.

Fixes: 5f436666 ("net: dsa: mv88e6xxx: Refactor setting of jumbo frames")
Signed-off-by: Chris Packham <chris.packham@alliedtelesis.co.nz>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

IRQs are disabled when freeing skbs in input queue.
Use the IRQ safe variant to free skbs here.

Fixes: 145dd5f9 ("net: flush the softnet backlog in process context")
Signed-off-by: Subash Abhinov Kasiviswanathan <subashab@codeaurora.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

Currently, maximum number of mapper pages are set to the pfn calculated
from the memblock size of the memblock containing kernel. This will work
until that memblock spans the entire memory. However, it will be set to
a wrong value if there are multiple memblocks defined in kernel
(e.g. with efi runtime services).

Set the the maximum value to the pfn calculated from dram size.
Signed-off-by: Atish Patra <atish.patra@wdc.com>
Signed-off-by: Palmer Dabbelt <palmerdabbelt@google.com>

Merge tag 'pci-v5.8-fixes-2' of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci into master

Pull PCI fixes from Bjorn Helgaas:

 - Reject invalid IRQ 0 command line argument for virtio_mmio because
   IRQ 0 now generates warnings (Bjorn Helgaas)

 - Revert "PCI/PM: Assume ports without DLL Link Active train links in
   100 ms", which broke nouveau (Bjorn Helgaas)

* tag 'pci-v5.8-fixes-2' of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci:
  Revert "PCI/PM: Assume ports without DLL Link Active train links in 100 ms"
  virtio-mmio: Reject invalid IRQ 0 command line argument

We have to detach sock from socket in qrtr_release(),
otherwise skb->sk may still reference to this socket
when the skb is released in tun->queue, particularly
sk->sk_wq still points to &sock->wq, which leads to
a UAF.

Reported-and-tested-by: syzbot+6720d64f31c081c2f708@syzkaller.appspotmail.com
Fixes: 28fb4e59 ("net: qrtr: Expose tunneling endpoint to user space")
Cc: Bjorn Andersson <bjorn.andersson@linaro.org>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Remove casting the values returned by memory allocation function.

Coccinelle emits WARNING:

./drivers/net/ethernet/hisilicon/hix5hd2_gmac.c:1027:9-23: WARNING:
 casting value returned by memory allocation function to (struct sg_desc *) is useless.

This issue was detected by using the Coccinelle software.
Signed-off-by: Wang Hai <wanghai38@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Merge tag 'wireless-drivers-2020-07-24' of git://git.kernel.org/pub/scm/linux/kernel/git/kvalo/wireless-drivers

Kalle Valo says:

====================
wireless-drivers fixes for v5.8

Second set of fixes for v5.8, and hopefully also the last. Three
important regressions fixed.

ath9k

* fix a regression which broke support for all ath9k usb devices

ath10k

* fix a regression which broke support for all QCA4019 AHB devices

iwlwifi

* fix a regression which broke support for some Killer Wireless-AC 1550 cards
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Tom Parkin says:

====================
l2tp: avoid multiple assignment, remove BUG_ON

l2tp hasn't been kept up to date with the static analysis checks offered
by checkpatch.pl.

This patchset builds on the series: "l2tp: cleanup checkpatch.pl
warnings" and "l2tp: further checkpatch.pl cleanups" to resolve some of
the remaining checkpatch warnings in l2tp.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

l2tp_session_free called BUG_ON if the tunnel magic feather value wasn't
correct.  The intent of this was to catch lifetime bugs; for example
early tunnel free due to incorrect use of reference counts.

Since the tunnel magic feather being wrong indicates either early free
or structure corruption, we can avoid doing more damage by simply
leaving the tunnel structure alone.  If the tunnel refcount isn't
dropped when it should be, the tunnel instance will remain in the
kernel, resulting in the tunnel structure and socket leaking.
Signed-off-by: Tom Parkin <tparkin@katalix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

l2tp_session_free is only called by l2tp_session_dec_refcount when the
reference count reaches zero, so it's of limited value to validate the
reference count value in l2tp_session_free itself.
Signed-off-by: Tom Parkin <tparkin@katalix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

l2tp_session_queue_purge is used during session shutdown to drop any
skbs queued for reordering purposes according to L2TP dataplane rules.

The BUG_ON in this function checks the session magic feather in an
attempt to catch lifetime bugs.

Rather than crashing the kernel with a BUG_ON, we can simply WARN_ON and
refuse to do anything more -- in the worst case this could result in a
leak.  However this is highly unlikely given that the session purge only
occurs from codepaths which have obtained the session by means of a lookup
via. the parent tunnel and which check the session "dead" flag to
protect against shutdown races.

While we're here, have l2tp_session_queue_purge return void rather than
an integer, since neither of the callsites checked the return value.
Signed-off-by: Tom Parkin <tparkin@katalix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

checkpatch advises that WARN_ON and recovery code are preferred over
BUG_ON which crashes the kernel.

l2tp_ppp has a BUG_ON check of struct seq_file's private pointer in
pppol2tp_seq_start prior to accessing data through that pointer.

Rather than crashing, we can simply bail out early and return NULL in
order to terminate the seq file processing in much the same way as we do
when reaching the end of tunnel/session instances to render.

Retain a WARN_ON to help trace possible bugs in this area.
Signed-off-by: Tom Parkin <tparkin@katalix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

checkpatch advises that WARN_ON and recovery code are preferred over
BUG_ON which crashes the kernel.

l2tp_ppp.c's BUG_ON checks of the l2tp session structure's "magic" field
occur in code paths where it's reasonably easy to recover:

 * In the case of pppol2tp_sock_to_session, we can return NULL and the
   caller will bail out appropriately.  There is no change required to
   any of the callsites of this function since they already handle
   pppol2tp_sock_to_session returning NULL.

 * In the case of pppol2tp_session_destruct we can just avoid
   decrementing the reference count on the suspect session structure.
   In the worst case scenario this results in a memory leak, which is
   preferable to a crash.

Convert these uses of BUG_ON to WARN_ON accordingly.
Signed-off-by: Tom Parkin <tparkin@katalix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>