Documentation/networking/ip-sysctl.txt:46 says:
ip_forward_use_pmtu - BOOLEAN
By default we don't trust protocol path MTUs while forwarding
because they could be easily forged and can lead to unwanted
fragmentation by the router.
You only need to enable this if you have user-space software
which tries to discover path mtus by itself and depends on the
kernel honoring this information. This is normally not the case.
Default: 0 (disabled)
Possible values:
0 - disabled
1 - enabled
Which makes it pretty clear that setting it to 1 is a potential
security/safety/DoS issue, and yet it is entirely reasonable to want
forwarded traffic to honour explicitly administrator configured
route mtus (instead of defaulting to device mtu).
Indeed, I can't think of a single reason why you wouldn't want to.
Since you configured a route mtu you probably know better...
It is pretty common to have a higher device mtu to allow receiving
large (jumbo) frames, while having some routes via that interface
(potentially including the default route to the internet) specify
a lower mtu.
Note that ipv6 forwarding uses device mtu unless the route is locked
(in which case it will use the route mtu).
This approach is not usable for IPv4 where an 'mtu lock' on a route
also has the side effect of disabling TCP path mtu discovery via
disabling the IPv4 DF (don't frag) bit on all outgoing frames.
I'm not aware of a way to lock a route from an IPv6 RA, so that also
potentially seems wrong.
Signed-off-by: Maciej Żenczykowski <maze@google.com>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Willem de Bruijn <willemb@google.com>
Cc: Lorenzo Colitti <lorenzo@google.com>
Cc: Sunmeet Gill (Sunny) <sgill@qti.qualcomm.com>
Cc: Vinay Paradkar <vparadka@qti.qualcomm.com>
Cc: Tyler Wear <twear@qti.qualcomm.com>
Cc: David Ahern <dsahern@kernel.org>
Reviewed-by: Eric Dumazet <edumazet@google.com>
(Backported from commit 02a1b175b0e92d9e0fa5df3957ade8d733ceb6a0).
Change-Id: I26f336f891711f0149b2835d2c4a78fc8407b5ba
Git-Commit: 02a1b175b0e92d9e0fa5df3957ade8d733ceb6a0
Git-repo: https://kernel.googlesource.com/pub/scm/linux/kernel/git/stable/linux
Signed-off-by: Sauvik Saha <ssaha@codeaurora.org>
Before copying dump to userspace, memory is mapped and then the
content get copied to userspace and later that memory should be
unmapped.
Let's unmap the memory after the content get copied to
userspace.
Change-Id: I03d53ebd1cdb743e1820f226a49d46136eca4874
Signed-off-by: Mukesh Ojha <mojha@codeaurora.org>
Send Calibration mode to FW only at the time of cold
boot. At the time of SSR Calibration mode will not be
sent.
Change-Id: Idf7016384937433480f62d0912e5e3dcf76f4e44
Signed-off-by: Naman Padhiar <npadhiar@codeaurora.org>
Normally clk_vote_vdd_level() and clk_unvote_vdd_level() are called
while holding the clock framework's prepare_lock since they are called
in the prepare paths. However, we're also calling clk_unvote_vdd_level()
from qcom_cc_sync_state() when removing proxy votes. This happens
outside of the clock framework, so the prepare_lock isn't protecting
this case. If the sync_state callback fires while another thread is
voting/unvoting voltage from a normal clock call, then there's a race
condition that can result in regulators being set to the incorrect
voltage or being pre-maturely disabled.
Add a mutex to protect this case.
Change-Id: Ib979e2edaaa5824dc80f7d4f065d72ca0b1a4444
Signed-off-by: Mike Tipton <mdtipton@codeaurora.org>
The target platforms use PM_SUSPEND_TO_IDLE as its default
suspend target. Since syscore_ops's suspend/resume callbacks
are not called in this path, the IPCC driver will never be
able to log the client's details that caused the wakeup in
it's resume handler. Hence, move the suspend/resume handlers
to dev_pm_ops for better visibility.
Change-Id: I095f02217aeea5ed7c5036017a7e960e2b98757a
Signed-off-by: Raghavendra Rao Ananta <rananta@codeaurora.org>
According to the comment in arch/arm/include/asm/delay.h,
__bad_udelay is specifically designed on ARM to produce a
build failure when udelay is called with a value > 2000.
Fix the issue by using mdelay for value > 2000.
Change-Id: Ia5ee988eeef68ce60b637232e0dd3989b8cbfe64
Signed-off-by: Hemant Kumar <hemantk@codeaurora.org>
Driver is directly performing a 64 bit read on time sync
register. In order to avoid compilation error on a 32 bit
platform perform two 32 bit reads instead of 64 bit read.
Change-Id: I477647c88a0bb2b4302ce2564bdc58fc4d68e806
Signed-off-by: Hemant Kumar <hemantk@codeaurora.org>
Both UFS3.0 and higher version of ufs device are capable of running
at lower level (2.5v) of VCC rail. If both host and device supports
2.5v level then we do runtime switch of UFS VCC voltage level.
Change-Id: I6fea6c354cde5f7bd9d2a8a275c66e842726d358
Signed-off-by: Bao D. Nguyen <nguyenb@codeaurora.org>
Signed-off-by: Subhash Jadavani <subhashj@codeaurora.org>
Signed-off-by: Nitin Rawat <nitirawa@codeaurora.org>
Initialize cpu freq table pointer to NULL while defining the same.
It avoids a possible use of uninitialized value of this stack pointer
in error path.
Change-Id: If9fb6bfeb836d1f3c81ecd5d9910cf0c76b4a0c5
Signed-off-by: Manaf Meethalavalappu Pallikunhi <manafm@codeaurora.org>
This reverts commit 410f0c2b19.
Additional enum value in include header is causing CRC differences
with GKI kernel.
Change-Id: Ia27f4d217a6bb2e234f49cb93c4163f372f4bcd7
Signed-off-by: Jeevan Shriram <jshriram@codeaurora.org>
In PM suspend of WCN6750, host send WOW Enable to FW
and FW allows target to go to low power mode along
with PCIe link. In Host DP, even after sending WOW
enable it tries to access target for flushing pending
transactions. This will cause NOC error as link is
down.
Send enter power save mode to FW after driver suspend
returns successfully to platform driver and all pending
transactions are flushed. FW will go to low power mode
only after receiveing enter power save mode QMI.
Change-Id: I5dfb257fe3db4dc453bae456c40d19f841e3c5c8
Signed-off-by: Naman Padhiar <npadhiar@codeaurora.org>
Enable qfprom_sys as module in holi gki defconfig.
Change-Id: I5deb36ee66418bdb3252b62368d75132bce26124
Signed-off-by: Naina Mehta <nainmeht@codeaurora.org>
Add support for QMI beamer ADC and RF sensors to apps.
It enables these sensors as thermal zone sensors so
that we can read and set trip thresholds via
thermal zone.
Change-Id: I0f56e59a319d0d6ea5b18cadc5eb675a62c485e1
Signed-off-by: Gopala Krishna Nuthaki <gnuthaki@codeaurora.org>
