linux-stable.git/drivers/staging, branch v5.4.84

speakup: Reject setting the speakup line discipline outside of speakup

2020-12-11T12:23:31+00:00

commit f0992098cadb4c9c6a00703b66cafe604e178fea upstream.

Speakup exposing a line discipline allows userland to try to use it,
while it is deemed to be useless, and thus uselessly exposes potential
bugs. One of them is simply that in such a case if the line sends data,
spk_ttyio_receive_buf2 is called and crashes since spk_ttyio_synth
is NULL.

This change restricts the use of the speakup line discipline to
speakup drivers, thus avoiding such kind of issues altogether.

Cc: stable@vger.kernel.org
Reported-by: Shisong Qin 
Signed-off-by: Samuel Thibault 
Tested-by: Shisong Qin 
Link: https://lore.kernel.org/r/20201129193523.hm3f6n5xrn6fiyyc@function
Signed-off-by: Greg Kroah-Hartman

staging/octeon: fix up merge error

2020-12-08T09:40:24+00:00

commit 673b41e04a035d760bc0aff83fa9ee24fd9c2779 upstream.

There's a semantic conflict in the Octeon staging network driver, which
used the skb_reset_tc() function to reset skb state when re-using an
skb.  But that inline helper function was removed in mainline by commit
2c64605b590e ("net: Fix CONFIG_NET_CLS_ACT=n and
CONFIG_NFT_FWD_NETDEV={y, m} build").

Fix it by using skb_reset_redirect() instead.  Also move it out of the

This code path only ends up triggering if REUSE_SKBUFFS_WITHOUT_FREE is
enabled, which in turn only happens if you don't have CONFIG_NETFILTER
configured.  Which was how this wasn't caught by the usual allmodconfig
builds.

Signed-off-by: Randy Dunlap 
Reported-by: Stephen Rothwell 
Signed-off-by: Linus Torvalds 
Cc: Pablo Neira Ayuso 
Signed-off-by: Greg Kroah-Hartman

staging: ralink-gdma: fix kconfig dependency bug for DMA_RALINK

2020-12-02T07:49:47+00:00

[ Upstream commit 06ea594051707c6b8834ef5b24e9b0730edd391b ]

When DMA_RALINK is enabled and DMADEVICES is disabled, it results in the
following Kbuild warnings:

WARNING: unmet direct dependencies detected for DMA_ENGINE
  Depends on [n]: DMADEVICES [=n]
  Selected by [y]:
  - DMA_RALINK [=y] && STAGING [=y] && RALINK [=y] && !SOC_RT288X [=n]

WARNING: unmet direct dependencies detected for DMA_VIRTUAL_CHANNELS
  Depends on [n]: DMADEVICES [=n]
  Selected by [y]:
  - DMA_RALINK [=y] && STAGING [=y] && RALINK [=y] && !SOC_RT288X [=n]

The reason is that DMA_RALINK selects DMA_ENGINE and DMA_VIRTUAL_CHANNELS
without depending on or selecting DMADEVICES while DMA_ENGINE and
DMA_VIRTUAL_CHANNELS are subordinate to DMADEVICES. This can also fail
building the kernel as demonstrated in a bug report.

Honor the kconfig dependency to remove unmet direct dependency warnings
and avoid any potential build failures.

Link: https://bugzilla.kernel.org/show_bug.cgi?id=210055
Signed-off-by: Necip Fazil Yildiran 
Link: https://lore.kernel.org/r/20201104181522.43567-1-fazilyildiran@gmail.com
Signed-off-by: Greg Kroah-Hartman 
Signed-off-by: Sasha Levin

staging: rtl8723bs: Add 024c:0627 to the list of SDIO device-ids

2020-11-24T12:29:21+00:00

commit aee9dccc5b64e878cf1b18207436e73f66d74157 upstream.

Add 024c:0627 to the list of SDIO device-ids, based on hardware found in
the wild. This hardware exists on at least some Acer SW1-011 tablets.

Signed-off-by: Brian O'Keefe 
Reviewed-by: Hans de Goede 
Link: https://lore.kernel.org/r/b9e1523f-2ba7-fb82-646a-37f095b4440e@alum.wpi.edu
Cc: stable 
Signed-off-by: Greg Kroah-Hartman

speakup: Do not let the line discipline be used several times

2020-11-24T12:29:20+00:00

commit d4122754442799187d5d537a9c039a49a67e57f1 upstream.

Speakup has only one speakup_tty variable to store the tty it is managing. This
makes sense since its codebase currently assumes that there is only one user who
controls the screen reading.

That however means that we have to forbid using the line discipline several
times, otherwise the second closure would try to free a NULL ldisc_data, leading to

general protection fault: 0000 [#1] SMP KASAN PTI
RIP: 0010:spk_ttyio_ldisc_close+0x2c/0x60
Call Trace:
 tty_ldisc_release+0xa2/0x340
 tty_release_struct+0x17/0xd0
 tty_release+0x9d9/0xcc0
 __fput+0x231/0x740
 task_work_run+0x12c/0x1a0
 do_exit+0x9b5/0x2230
 ? release_task+0x1240/0x1240
 ? __do_page_fault+0x562/0xa30
 do_group_exit+0xd5/0x2a0
 __x64_sys_exit_group+0x35/0x40
 do_syscall_64+0x89/0x2b0
 ? page_fault+0x8/0x30
 entry_SYSCALL_64_after_hwframe+0x44/0xa9

Cc: stable@vger.kernel.org
Reported-by: 秦世松 
Signed-off-by: Samuel Thibault 
Tested-by: Shisong Qin 
Link: https://lore.kernel.org/r/20201110183541.fzgnlwhjpgqzjeth@function
Signed-off-by: Greg Kroah-Hartman

staging: octeon: Drop on uncorrectable alignment or FCS error

2020-11-05T10:43:38+00:00

commit 49d28ebdf1e30d806410eefc7de0a7a1ca5d747c upstream.

Currently in case of alignment or FCS error if the packet cannot be
corrected it's still not dropped. Report the error properly and drop the
packet while making the code around a little bit more readable.

Fixes: 80ff0fd3ab64 ("Staging: Add octeon-ethernet driver files.")
Signed-off-by: Alexander Sverdlin 
Cc: stable 
Link: https://lore.kernel.org/r/20201016145630.41852-1-alexander.sverdlin@nokia.com
Signed-off-by: Greg Kroah-Hartman

staging: octeon: repair "fixed-link" support

2020-11-05T10:43:37+00:00

commit 179f5dc36b0a1aa31538d7d8823deb65c39847b3 upstream.

The PHYs must be registered once in device probe function, not in device
open callback because it's only possible to register them once.

Fixes: a25e278020bf ("staging: octeon: support fixed-link phys")
Signed-off-by: Alexander Sverdlin 
Cc: stable 
Link: https://lore.kernel.org/r/20201016101858.11374-1-alexander.sverdlin@nokia.com
Signed-off-by: Greg Kroah-Hartman

staging: comedi: cb_pcidas: Allow 2-channel commands for AO subdevice

2020-11-05T10:43:37+00:00

commit 647a6002cb41d358d9ac5de101a8a6dc74748a59 upstream.

The "cb_pcidas" driver supports asynchronous commands on the analog
output (AO) subdevice for those boards that have an AO FIFO.  The code
(in `cb_pcidas_ao_check_chanlist()` and `cb_pcidas_ao_cmd()`) to
validate and set up the command supports output to a single channel or
to two channels simultaneously (the boards have two AO channels).
However, the code in `cb_pcidas_auto_attach()` that initializes the
subdevices neglects to initialize the AO subdevice's `len_chanlist`
member, leaving it set to 0, but the Comedi core will "correct" it to 1
if the driver neglected to set it.  This limits commands to use a single
channel (either channel 0 or 1), but the limit should be two channels.
Set the AO subdevice's `len_chanlist` member to be the same value as the
`n_chan` member, which will be 2.

Cc: 
Signed-off-by: Ian Abbott 
Link: https://lore.kernel.org/r/20201021122142.81628-1-abbotti@mev.co.uk
Signed-off-by: Greg Kroah-Hartman

staging: fieldbus: anybuss: jump to correct label in an error path

2020-11-05T10:43:37+00:00

commit 7e97e4cbf30026b49b0145c3bfe06087958382c5 upstream.

In current code, controller_probe() misses to call ida_simple_remove()
in an error path. Jump to correct label to fix it.

Fixes: 17614978ed34 ("staging: fieldbus: anybus-s: support the Arcx anybus controller")
Reviewed-by: Sven Van Asbroeck 
Signed-off-by: Jing Xiangfeng 
Cc: stable 
Link: https://lore.kernel.org/r/20201012132404.113031-1-jingxiangfeng@huawei.com
Signed-off-by: Greg Kroah-Hartman 
Signed-off-by: Greg Kroah-Hartman

net: wilc1000: clean up resource in error path of init mon interface

2020-10-29T08:57:40+00:00

[ Upstream commit 55bd149978679742374c800e56e8f6bc74378bbe ]

The wilc_wfi_init_mon_int() forgets to clean up resource when
register_netdevice() failed. Add the missed call to fix it.
And the return value of netdev_priv can't be NULL, so remove
the unnecessary error handling.

Fixes: 588713006ea4 ("staging: wilc1000: avoid the use of 'wilc_wfi_mon' static variable")
Reported-by: Hulk Robot 
Signed-off-by: Huang Guobin 
Signed-off-by: Kalle Valo 
Link: https://lore.kernel.org/r/20200917123019.206382-1-huangguobin4@huawei.com
Signed-off-by: Sasha Levin