- 04 Nov, 2018 17 commits
-
-
Michal Simek authored
[ Upstream commit e1e5d8a9 ] Clear ADDR64 dma bit in DMACFG register in case that HW_DMA_CAP_64B is not detected on 64bit system. The issue was observed when bootloader(u-boot) does not check macb feature at DCFG6 register (DAW64_OFFSET) and enabling 64bit dma support by default. Then macb driver is reading DMACFG register back and only adding 64bit dma configuration but not cleaning it out. Signed-off-by:
Michal Simek <michal.simek@xilinx.com> Acked-by:
Nicolas Ferre <nicolas.ferre@microchip.com> Signed-off-by:
David S. Miller <davem@davemloft.net> Signed-off-by:
Sasha Levin <sashal@kernel.org>
-
Florian Fainelli authored
[ Upstream commit 3ab97942 ] A number of our interrupts were incorrectly specified, fix both the PPI and SPI interrupts to be correct. Fixes: b5762cac ("ARM: bcm63138: add NAND DT support") Fixes: 46d4bca0 ("ARM: BCM63XX: add BCM63138 minimal Device Tree") Signed-off-by:
Florian Fainelli <f.fainelli@gmail.com> Signed-off-by:
-
Steve Capper authored
[ Upstream commit 469ed9d8 ] In the contiguous bit hugetlb break-before-make code we assume that all hugetlb pages are young. In fact, remove_migration_pte is able to place an old hugetlb pte so this assumption is not valid. This patch fixes the contiguous hugetlb scanning code to preserve young ptes. Fixes: d8bdcff2 ("arm64: hugetlb: Add break-before-make logic for contiguous entries") Signed-off-by:
Steve Capper <steve.capper@arm.com> Signed-off-by:
Will Deacon <will.deacon@arm.com> Signed-off-by:
-
David Ahern authored
[ Upstream commit a173f066 ] For starters, the bridge netfilter code registers operations that are invoked any time nh_hook is called. Specifically, ip_sabotage_in watches for nested calls for NF_INET_PRE_ROUTING when a bridge is in the stack. Packet wise, the bridge netfilter hook runs first. br_nf_pre_routing allocates nf_bridge, sets in_prerouting to 1 and calls NF_HOOK for NF_INET_PRE_ROUTING. It's finish function, br_nf_pre_routing_finish, then resets in_prerouting flag to 0 and the packet continues up the stack. The packet eventually makes it to the VRF driver and it invokes nf_hook for NF_INET_PRE_ROUTING in case any rules have been added against the vrf device. Because of the registered operations the call to nf_hook causes ip_sabotage_in to be invoked. That function sees the nf_bridge on the skb and that in_prerouting is not set. Thinking it is an invalid nested call it steals (drops) the packet. Update ip_sabotage_in to recognize that the bridge or one of its upper devices (e.g., vlan) can be enslaved to a VRF (L3 master device) and allow the packet to go through the nf_hook a second time. Fixes: 73e20b76 ("net: vrf: Add support for PREROUTING rules on vrf device") Reported-by:
D'Souza, Nelson <ndsouza@ciena.com> Signed-off-by:
David Ahern <dsahern@gmail.com> Signed-off-by:
Pablo Neira Ayuso <pablo@netfilter.org> Signed-off-by:
-
Sean Tranchetti authored
[ Upstream commit 32bf94fb ] XFRM mode parameters passed as part of the user templates in the IP_XFRM_POLICY are never properly validated. Passing values other than valid XFRM modes can cause stack-out-of-bounds reads to occur later in the XFRM processing: [ 140.535608] ================================================================ [ 140.543058] BUG: KASAN: stack-out-of-bounds in xfrm_state_find+0x17e4/0x1cc4 [ 140.550306] Read of size 4 at addr ffffffc0238a7a58 by task repro/5148 [ 140.557369] [ 140.558927] Call trace: [ 140.558936] dump_backtrace+0x0/0x388 [ 140.558940] show_stack+0x24/0x30 [ 140.558946] __dump_stack+0x24/0x2c [ 140.558949] dump_stack+0x8c/0xd0 [ 140.558956] print_address_description+0x74/0x234 [ 140.558960] kasan_report+0x240/0x264 [ 140.558963] __asan_report_load4_noabort+0x2c/0x38 [ 140.558967] xfrm_state_find+0x17e4/0x1cc4 [ 140.558971] xfrm_resolve_and_create_bundle+0x40c/0x1fb8 [ 140.558975] xfrm_lookup+0x238/0x1444 [ 140.558977] xfrm_lookup_route+0x48/0x11c [ 140.558984] ip_route_output_flow+0x88/0xc4 [ 140.558991] raw_sendmsg+0xa74/0x266c [ 140.558996] inet_sendmsg+0x258/0x3b0 [ 140.559002] sock_sendmsg+0xbc/0xec [ 140.559005] SyS_sendto+0x3a8/0x5a8 [ 140.559008] el0_svc_naked+0x34/0x38 [ 140.559009] [ 140.592245] page dumped because: kasan: bad access detected [ 140.597981] page_owner info is not active (free page?) [ 140.603267] [ 140.653503] ================================================================ Signed-off-by:
Sean Tranchetti <stranche@codeaurora.org> Signed-off-by:
Steffen Klassert <steffen.klassert@secunet.com> Signed-off-by:
-
Thomas Petazzoni authored
[ Upstream commit 3a58ac65 ] IO_SPACE_LIMIT is the ending address of the PCI IO space, i.e something like 0xfffff (and not 0x100000). Therefore, when offset = 0xf0000 is passed as argument, this function fails even though the offset + SZ_64K fits below the IO_SPACE_LIMIT. This makes the last chunk of 64 KB of the I/O space not usable as it cannot be mapped. This patch fixes that by substracing 1 to offset + SZ_64K, so that we compare the addrss of the last byte of the I/O space against IO_SPACE_LIMIT instead of the address of the first byte of what is after the I/O space. Fixes: c2794437 ("ARM: Add fixed PCI i/o mapping") Signed-off-by:
Thomas Petazzoni <thomas.petazzoni@bootlin.com> Acked-by:
Nicolas Pitre <nico@linaro.org> Signed-off-by:
Russell King <rmk+kernel@armlinux.org.uk> Signed-off-by:
-
Steffen Klassert authored
[ Upstream commit 9e143793 ] Since commit 222d7dbd ("net: prevent dst uses after free") skb_dst_force() might clear the dst_entry attached to the skb. The xfrm code don't expect this to happen, so we crash with a NULL pointer dereference in this case. Fix it by checking skb_dst(skb) for NULL after skb_dst_force() and drop the packet in cast the dst_entry was cleared. Fixes: 222d7dbd ("net: prevent dst uses after free") Reported-by:
Tobias Hommel <netdev-list@genoetigt.de> Reported-by:
Kristian Evensen <kristian.evensen@gmail.com> Reported-by:
Wolfgang Walter <linux@stwm.de> Signed-off-by:
-
Yuan-Chi Pang authored
[ Upstream commit c4205510 ] TX status reporting to ieee80211s is through ieee80211s_update_metric. There are two problems about ieee80211s_update_metric: 1. The purpose is to estimate the fail probability to a specific link. No need to restrict to data frame. 2. Current implementation does not work if wireless driver does not pass tx_status with skb. Fix this by removing ieee80211_is_data condition, passing ieee80211_tx_status directly to ieee80211s_update_metric, and putting it in both __ieee80211_tx_status and ieee80211_tx_status_ext. Signed-off-by:
Yuan-Chi Pang <fu3mo6goo@gmail.com> Signed-off-by:
Johannes Berg <johannes.berg@intel.com> Signed-off-by:
-
Johannes Berg authored
[ Upstream commit cb59bc14 ] If the TDLS setup happens over a connection to an AP that doesn't have QoS, we nevertheless assign a non-zero TID (skb->priority) and queue mapping, which may confuse us or drivers later. Fix it by just assigning the special skb->priority and then using ieee80211_select_queue() just like other data frames would go through. Signed-off-by:
-
Jouni Malinen authored
[ Upstream commit 119f94a6 ] cfg80211_get_bss_channel() is used to update the RX channel based on the available frame payload information (channel number from DSSS Parameter Set element or HT Operation element). This is needed on 2.4 GHz channels where frames may be received on neighboring channels due to overlapping frequency range. This might of some use on the 5 GHz band in some corner cases, but things are more complex there since there is no n:1 or 1:n mapping between channel numbers and frequencies due to multiple different starting frequencies in different operating classes. This could result in ieee80211_channel_to_frequency() returning incorrect frequency and ieee80211_get_channel() returning incorrect channel information (or indication of no match). In the previous implementation, this could result in some scan results being dropped completely, e.g., for the 4.9 GHz channels. That prevented connection to such BSSs. Fix this by using the driver-provided channel pointer if ieee80211_get_channel() does not find matching channel data for the channel number in the frame payload and if the scan is done with 5 MHz or 10 MHz channel bandwidth. While doing this, also add comments describing what the function is trying to achieve to make it easier to understand what happens here and why. Signed-off-by:
Jouni Malinen <jouni@codeaurora.org> Signed-off-by:
-
Bob Copeland authored
[ Upstream commit 6eae4a6c ] In our environment running lots of mesh nodes, we are seeing the pending queue hang periodically, with the debugfs queues file showing lines such as: 00: 0x00000000/348 i.e. there are a large number of frames but no stop reason set. One way this could happen is if queue processing from the pending tasklet exited early without processing all frames, and without having some future event (incoming frame, stop reason flag, ...) to reschedule it. Exactly this can occur today if ieee80211_tx() returns false due to packet drops or power-save buffering in the tx handlers. In the past, this function would return true in such cases, and the change to false doesn't seem to be intentional. Fix this case by reverting to the previous behavior. Fixes: bb42f2d1 ("mac80211: Move reorder-sensitive TX handlers to after TXQ dequeue") Signed-off-by:
Bob Copeland <bobcopeland@fb.com> Acked-by:
Toke Høiland-Jørgensen <toke@toke.dk> Signed-off-by:
-
Andrei Otcheretianski authored
[ Upstream commit 24f33e64 ] Core regulatory hints didn't set wiphy_idx to WIPHY_IDX_INVALID. Since the regulatory request is zeroed, wiphy_idx was always implicitly set to 0. This resulted in updating only phy #0. Fix that. Fixes: 806a9e39 ("cfg80211: make regulatory_request use wiphy_idx instead of wiphy") Signed-off-by:
Andrei Otcheretianski <andrei.otcheretianski@intel.com> Signed-off-by:
Luca Coelho <luciano.coelho@intel.com> [add fixes tag] Signed-off-by:
-
Andrei Otcheretianski authored
[ Upstream commit 8682250b ] If a frame is dropped for any reason, mac80211 wouldn't report the TX status back to user space. As the user space may rely on the TX_STATUS to kick its state machines, resends etc, it's better to just report this frame as not acked instead. Signed-off-by:
-
Sowmini Varadhan authored
[ Upstream commit 782710e3 ] We only support one offloaded xfrm (we do not have devices that can handle more than one offload), so reset crypto_done in xfrm_input() when iterating over multiple transforms in xfrm_input, so that we can invoke the appropriate x->type->input for the non-offloaded transforms Fixes: d77e38e6 ("xfrm: Add an IPsec hardware offloading API") Signed-off-by:
Sowmini Varadhan <sowmini.varadhan@oracle.com> Signed-off-by:
-
Sowmini Varadhan authored
[ Upstream commit bfc0698b ] A policy may have been set up with multiple transforms (e.g., ESP and ipcomp). In this situation, the ingress IPsec processing iterates in xfrm_input() and applies each transform in turn, processing the nexthdr to find any additional xfrm that may apply. This patch resets the transport header back to network header only after the last transformation so that subsequent xfrms can find the correct transport header. Fixes: 7785bba2 ("esp: Add a software GRO codepath") Suggested-by:
-
Thadeu Lima de Souza Cascardo authored
[ Upstream commit 215ab0f0 ] After commit d6990976 ("vti6: fix PMTU caching and reporting on xmit"), some too big skbs might be potentially passed down to __xfrm6_output, causing it to fail to transmit but not free the skb, causing a leak of skb, and consequentially a leak of dst references. After running pmtu.sh, that shows as failure to unregister devices in a namespace: [ 311.397671] unregister_netdevice: waiting for veth_b to become free. Usage count = 1 The fix is to call kfree_skb in case of transmit failures. Fixes: dd767856 ("xfrm6: Don't call icmpv6_send on local error") Signed-off-by:
Thadeu Lima de Souza Cascardo <cascardo@canonical.com> Reviewed-by:
Sabrina Dubroca <sd@queasysnail.net> Signed-off-by:
-
Steffen Klassert authored
[ Upstream commit 07bf7908 ] We don't validate the address prefix lengths in the xfrm selector we got from userspace. This can lead to undefined behaviour in the address matching functions if the prefix is too big for the given address family. Fix this by checking the prefixes and refuse SA/policy insertation when a prefix is invalid. Fixes: 1da177e4 ("Linux-2.6.12-rc2") Reported-by:
Air Icy <icytxw@gmail.com> Signed-off-by:
-
- 20 Oct, 2018 23 commits
-
-
Greg Kroah-Hartman authored
-
Michael J. Ruhl authored
commit b4a4957d upstream. rvt_destroy_qp() cannot complete until all in process packets have been released from the underlying hardware. If a link down event occurs, an application can hang with a kernel stack similar to: cat /proc/<app PID>/stack quiesce_qp+0x178/0x250 [hfi1] rvt_reset_qp+0x23d/0x400 [rdmavt] rvt_destroy_qp+0x69/0x210 [rdmavt] ib_destroy_qp+0xba/0x1c0 [ib_core] nvme_rdma_destroy_queue_ib+0x46/0x80 [nvme_rdma] nvme_rdma_free_queue+0x3c/0xd0 [nvme_rdma] nvme_rdma_destroy_io_queues+0x88/0xd0 [nvme_rdma] nvme_rdma_error_recovery_work+0x52/0xf0 [nvme_rdma] process_one_work+0x17a/0x440 worker_thread+0x126/0x3c0 kthread+0xcf/0xe0 ret_from_fork+0x58/0x90 0xffffffffffffffff quiesce_qp() waits until all outstanding packets have been freed. This wait should be momentary. During a link down event, the cleanup handling does not ensure that all packets caught by the link down are flushed properly. This is caused by the fact that the freeze path and the link down event is handled the same. This is not correct. The freeze path waits until the HFI is unfrozen and then restarts PIO. A link down is not a freeze event. The link down path cannot restart the PIO until link is restored. If the PIO path is restarted before the link comes up, the application (QP) using the PIO path will hang (until link is restored). Fix by separating the linkdown path from the freeze path and use the link down path for link down events. Close a race condition sc_disable() by acquiring both the progress and release locks. Close a race condition in sc_stop() by moving the setting of the flag bits under the alloc lock. Cc: <stable@vger.kernel.org> # 4.9.x+ Fixes: 77241056 ("IB/hfi1: add driver files") Reviewed-by:
Mike Marciniszyn <mike.marciniszyn@intel.com> Signed-off-by:
Michael J. Ruhl <michael.j.ruhl@intel.com> Signed-off-by:
Dennis Dalessandro <dennis.dalessandro@intel.com> Signed-off-by:
Jason Gunthorpe <jgg@mellanox.com> Signed-off-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Wolfram Sang authored
commit 2b16fd63 upstream. On Gen3, we can only do RXDMA once per transfer reliably. For that, we must reset the device, then we can have RXDMA once. This patch implements this. When there is no reset controller or the reset fails, RXDMA will be blocked completely. Otherwise, it will be disabled after the first RXDMA transfer. Based on a commit from the BSP by Hiromitsu Yamasaki, yet completely refactored to handle multiple read messages within one transfer. Signed-off-by:
Wolfram Sang <wsa+renesas@sang-engineering.com> Reviewed-by:
Geert Uytterhoeven <geert+renesas@glider.be> Signed-off-by:
Wolfram Sang <wsa@the-dreams.de> Cc: stable@kernel.org Signed-off-by:
Sudip Mukherjee <sudipm.mukherjee@gmail.com> Signed-off-by:
-
Clint Taylor authored
commit 0ca94881 upstream. On GLK NUC platforms the HDMI retiming buffer needs additional disabled time to correctly sync to a faster incoming signal. When measured on a scope the highspeed lines of the HDMI clock turn off for ~400uS during a normal resolution change. The HDMI retimer on the GLK NUC appears to require at least a full frame of quiet time before a new faster clock can be correctly sync'd. Wait 100ms due to msleep inaccuracies while waiting for a completed frame. Add a quirk to the driver for GLK boards that use ITE66317 HDMI retimers. V2: Add more devices to the quirk list V3: Delay increased to 100ms, check to confirm crtc type is HDMI. V4: crtc type check extended to include _DDI and whitespace fixes v5: Fix white spaces, remove the macro for delay. Revert the crtc type check introduced in v4. Cc: Imre Deak <imre.deak@intel.com> Cc: <stable@vger.kernel.org> # v4.14+ Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=105887Signed-off-by:
Clint Taylor <clinton.a.taylor@intel.com> Tested-by:
Daniel Scheller <d.scheller.oss@gmail.com> Signed-off-by:
Radhakrishna Sripada <radhakrishna.sripada@intel.com> Signed-off-by:
Imre Deak <imre.deak@intel.com> Reviewed-by:
Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by:
-
Dave Jiang authored
commit 15d36fec upstream. When pmem namespaces created are smaller than section size, this can cause an issue during removal and gpf was observed: general protection fault: 0000 1 SMP PTI CPU: 36 PID: 3941 Comm: ndctl Tainted: G W 4.14.28-1.el7uek.x86_64 #2 task: ffff88acda150000 task.stack: ffffc900233a4000 RIP: 0010:__put_page+0x56/0x79 Call Trace: devm_memremap_pages_release+0x155/0x23a release_nodes+0x21e/0x260 devres_release_all+0x3c/0x48 device_release_driver_internal+0x15c/0x207 device_release_driver+0x12/0x14 unbind_store+0xba/0xd8 drv_attr_store+0x27/0x31 sysfs_kf_write+0x3f/0x46 kernfs_fop_write+0x10f/0x18b __vfs_write+0x3a/0x16d vfs_write+0xb2/0x1a1 SyS_write+0x55/0xb9 do_syscall_64+0x79/0x1ae entry_SYSCALL_64_after_hwframe+0x3d/0x0 Add code to check whether we have a mapping already in the same section and prevent additional mappings from being created if that is the case. Link: http://lkml.kernel.org/r/152909478401.50143.312364396244072931.stgit@djiang5-desk3.ch.intel.comSigned-off-by:
Dave Jiang <dave.jiang@intel.com> Cc: Dan Williams <dan.j.williams@intel.com> Cc: Robert Elliott <elliott@hpe.com> Cc: Jeff Moyer <jmoyer@redhat.com> Cc: Matthew Wilcox <willy@infradead.org> Cc: <stable@vger.kernel.org> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by:
-
Gilad Ben-Yossef authored
commit 2f7caf6b upstream. If we ran out of DMA pool buffers, we get into the unmap code path with a NULL before. Deal with this by checking the virtual mapping is not NULL. Cc: stable@vger.kernel.org Signed-off-by:
Gilad Ben-Yossef <gilad@benyossef.com> Signed-off-by:
-
Ville Syrjälä authored
commit 05c72e77 upstream. We broke the LVDS notifier resume thing in (presumably) commit e2c8b870 ("drm/i915: Use atomic helpers for suspend, v2.") as we no longer duplicate the current state in the LVDS notifier and thus we never resume it properly either. Instead of trying to fix it again let's just kill off the lid notifier entirely. None of the machines tested thus far have apparently needed it. Originally the lid notifier was added to work around cases where the VBIOS was clobbering some of the hardware state behind the driver's back, mostly on Thinkpads. We now have a few report of Thinkpads working just fine without the notifier. So maybe it was misdiagnosed originally, or something else has changed (ACPI video stuff perhaps?). If we do end up finding a machine where the VBIOS is still causing problems I would suggest that we first try setting various bits in the VBIOS scratch registers. There are several to choose from that may instruct the VBIOS to steer clear. With the notifier gone we'll also stop looking at the panel status in ->detect(). v2: Nuke enum modeset_restore (Rodrigo) Cc: stable@vger.kernel.org Cc: Wolfgang Draxinger <wdraxinger.maillist@draxit.de> Cc: Vito Caputo <vcaputo@pengaru.com> Cc: kitsunyan <kitsunyan@airmail.cc> Cc: Joonas Saarinen <jza@saunalahti.fi> Tested-by: Vito Caputo <vcaputo@pengaru.com> # Thinkapd X61s Tested-by: kitsunyan <kitsunyan@airmail.cc> # ThinkPad X200 Tested-by: Joonas Saarinen <jza@saunalahti.fi> # Fujitsu Siemens U9210 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=105902 References: https://lists.freedesktop.org/archives/intel-gfx/2018-June/169315.html References: https://bugs.freedesktop.org/show_bug.cgi?id=21230 Fixes: e2c8b870 ("drm/i915: Use atomic helpers for suspend, v2.") Signed-off-by:
Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20180717174216.22252-1-ville.syrjala@linux.intel.comReviewed-by:
-
Natanael Copa authored
Commit b6cc0ba2 (HID: add support for Apple Magic Keyboards) backported support for the Magic Keyboard over Bluetooth, but did not add the BT_VENDOR_ID_APPLE to hid_have_special_driver[] so the hid-apple driver is never loaded and Fn key does not work at all. Adding BT_VENDOR_ID_APPLE to hid_have_special_driver[] is not needed after commit e04a0442 (HID: core: remove the absolute need of hid_have_special_driver[]), so 4.16 kernels and newer does not need it. Fixes: b6cc0ba2 (HID: add support for Apple Magic Keyboards) Bugzilla-id: https://bugzilla.kernel.org/show_bug.cgi?id=99881Signed-off-by:
Natanael Copa <ncopa@alpinelinux.org> Acked-by:
Benjamin Tissoires <benjamin.tissoires@redhat.com> Signed-off-by:
-
Alexey Brodkin authored
commit 40660f1f upstream. There's not much sense in doing that because if user or his build-system didn't set CROSS_COMPILE we still may very well make incorrect guess. But as it turned out setting CROSS_COMPILE is not as harmless as one may think: with recent changes that implemented automatic discovery of __host__ gcc features unconditional setup of CROSS_COMPILE leads to failures on execution of "make xxx_defconfig" with absent cross-compiler, for more info see [1]. Set CROSS_COMPILE as well gets in the way if we want only to build .dtb's (again with absent cross-compiler which is not really needed for building .dtb's), see [2]. Note, we had to change LIBGCC assignment type from ":=" to "=" so that is is resolved on its usage, otherwise if it is resolved at declaration time with missing CROSS_COMPILE we're getting this error message from host GCC: | gcc: error: unrecognized command line option -mmedium-calls | gcc: error: unrecognized command line option -mno-sdata [1] http://lists.infradead.org/pipermail/linux-snps-arc/2018-September/004308.html [2] http://lists.infradead.org/pipermail/linux-snps-arc/2018-September/004320.htmlSigned-off-by:
Alexey Brodkin <abrodkin@synopsys.com> Cc: Masahiro Yamada <yamada.masahiro@socionext.com> Cc: Rob Herring <robh@kernel.org> Signed-off-by:
Vineet Gupta <vgupta@synopsys.com> Signed-off-by:
-
Alexey Brodkin authored
commit 615f6445 upstream. This check is very naive: we simply test if GCC invoked without "-mcpu=XXX" has ARC700 define set. In that case we think that GCC was built with "--with-cpu=arc700" and has libgcc built for ARC700. Otherwise if ARC700 is not defined we think that everythng was built for ARCv2. But in reality our life is much more interesting. 1. Regardless of GCC configuration (i.e. what we pass in "--with-cpu" it may generate code for any ARC core). 2. libgcc might be built with explicitly specified "--mcpu=YYY" That's exactly what happens in case of multilibbed toolchains: - GCC is configured with default settings - All the libs built for many different CPU flavors I.e. that check gets in the way of usage of multilibbed toolchains. And even non-multilibbed toolchains are affected. OpenEmbedded also builds GCC without "--with-cpu" because each and every target component later is compiled with explicitly set "-mcpu=ZZZ". Acked-by:
Rob Herring <robh@kernel.org> Signed-off-by:
-
Linus Torvalds authored
commit eb66ae03 upstream. Jann Horn points out that our TLB flushing was subtly wrong for the mremap() case. What makes mremap() special is that we don't follow the usual "add page to list of pages to be freed, then flush tlb, and then free pages". No, mremap() obviously just _moves_ the page from one page table location to another. That matters, because mremap() thus doesn't directly control the lifetime of the moved page with a freelist: instead, the lifetime of the page is controlled by the page table locking, that serializes access to the entry. As a result, we need to flush the TLB not just before releasing the lock for the source location (to avoid any concurrent accesses to the entry), but also before we release the destination page table lock (to avoid the TLB being flushed after somebody else has already done something to that page). This also makes the whole "need_flush" logic unnecessary, since we now always end up flushing the TLB for every valid entry. Reported-and-tested-by:
Jann Horn <jannh@google.com> Acked-by:
Ingo Molnar <mingo@kernel.org> Acked-by:
Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by:
-
Greg Kroah-Hartman authored
This reverts commit 4f4374a9 which was commit a6795a58 upstream. Turns out this causes problems and was to fix a patch only in the 4.19 and newer tree. Reported-by:
Amir Goldstein <amir73il@gmail.com> Cc: Miklos Szeredi <mszeredi@redhat.com> Cc: Christoph Hellwig <hch@lst.de> Cc: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by:
-
Christophe Leroy authored
commit 8183d99f upstream. feature fixups need to use patch_instruction() early in the boot, even before the code is relocated to its final address, requiring patch_instruction() to use PTRRELOC() in order to address data. But feature fixups applies on code before it is set to read only, even for modules. Therefore, feature fixups can use raw_patch_instruction() instead. Signed-off-by:
Christophe Leroy <christophe.leroy@c-s.fr> Signed-off-by:
Michael Ellerman <mpe@ellerman.id.au> Reported-by:
David Gounaris <david.gounaris@infinera.com> Tested-by:
-
Arindam Nath authored
[ Upstream commit 5ebb1bc2 ] ACPI HID devices do not actually have an alias for them in the IVRS. But dev_data->alias is still used for indexing into the IOMMU device table for devices being handled by the IOMMU. So for ACPI HID devices, we simply return the corresponding devid as an alias, as parsed from IVRS table. Signed-off-by:
Arindam Nath <arindam.nath@amd.com> Fixes: 2bf9a0a1 ('iommu/amd: Add iommu support for ACPI HID devices') Signed-off-by:
Joerg Roedel <jroedel@suse.de> Signed-off-by:
Sasha Levin <alexander.levin@microsoft.com> Signed-off-by:
-
Michael Neuling authored
[ Upstream commit 96dc89d5 ] Current we store the userspace r1 to PACATMSCRATCH before finally saving it to the thread struct. In theory an exception could be taken here (like a machine check or SLB miss) that could write PACATMSCRATCH and hence corrupt the userspace r1. The SLB fault currently doesn't touch PACATMSCRATCH, but others do. We've never actually seen this happen but it's theoretically possible. Either way, the code is fragile as it is. This patch saves r1 to the kernel stack (which can't fault) before we turn MSR[RI] back on. PACATMSCRATCH is still used but only with MSR[RI] off. We then copy r1 from the kernel stack to the thread struct once we have MSR[RI] back on. Suggested-by:
Breno Leitao <leitao@debian.org> Signed-off-by:
Michael Neuling <mikey@neuling.org> Signed-off-by:
-
Michael Neuling authored
[ Upstream commit cf13435b ] When we treclaim we store the userspace checkpointed r13 to a scratch SPR and then later save the scratch SPR to the user thread struct. Unfortunately, this doesn't work as accessing the user thread struct can take an SLB fault and the SLB fault handler will write the same scratch SPRG that now contains the userspace r13. To fix this, we store r13 to the kernel stack (which can't fault) before we access the user thread struct. Found by running P8 guest + powervm + disable_1tb_segments + TM. Seen as a random userspace segfault with r13 looking like a kernel address. Signed-off-by:
-
Tao Ren authored
[ Upstream commit 4451d3f5 ] Currently, the aspeed MATCH1 register is updated to <current_count - cycles> in set_next_event handler, with the assumption that COUNT register value is preserved when the timer is disabled and it continues decrementing after the timer is enabled. But the assumption is wrong: RELOAD register is loaded into COUNT register when the aspeed timer is enabled, which means the next event may be delayed because timer interrupt won't be generated until <0xFFFFFFFF - current_count + cycles>. The problem can be fixed by updating RELOAD register to <cycles>, and COUNT register will be re-loaded when the timer is enabled and interrupt is generated when COUNT register overflows. The test result on Facebook Backpack-CMM BMC hardware (AST2500) shows the issue is fixed: without the patch, usleep(100) suspends the process for several milliseconds (and sometimes even over 40 milliseconds); after applying the fix, usleep(100) takes averagely 240 microseconds to return under the same workload level. Signed-off-by:
Tao Ren <taoren@fb.com> Reviewed-by:
Linus Walleij <linus.walleij@linaro.org> Tested-by:
Lei YU <mine260309@gmail.com> Signed-off-by:
Daniel Lezcano <daniel.lezcano@linaro.org> Signed-off-by:
-
Nathan Chancellor authored
[ Upstream commit 8ac1ee6f ] Clang warns that the address of a pointer will always evaluated as true in a boolean context: drivers/net/ethernet/mellanox/mlx4/eq.c:243:11: warning: address of array 'eq->affinity_mask' will always evaluate to 'true' [-Wpointer-bool-conversion] if (!eq->affinity_mask || cpumask_empty(eq->affinity_mask)) ~~~~~^~~~~~~~~~~~~ 1 warning generated. Use cpumask_available, introduced in commit f7e30f01 ("cpumask: Add helper cpumask_available()"), which does the proper checking and avoids this warning. Link: https://github.com/ClangBuiltLinux/linux/issues/86Signed-off-by:
Nathan Chancellor <natechancellor@gmail.com> Signed-off-by:
-
Johannes Thumshirn authored
[ Upstream commit f1f1fada ] When sd_init_command() get's a command with a unknown req_op() it crashes the system via BUG(). This makes debugging the actual reason for the broken request cmd_flags pretty hard as the system is down before it's able to write out debugging data on the serial console or the trace buffer. Change the BUG() to a WARN_ON() and return BLKPREP_KILL to fail gracefully and return an I/O error to the producer of the request. Signed-off-by:
Johannes Thumshirn <jthumshirn@suse.de> Cc: Hannes Reinecke <hare@suse.de> Cc: Bart Van Assche <bvanassche@acm.org> Cc: Christoph Hellwig <hch@lst.de> Reviewed-by:
Christoph Hellwig <hch@lst.de> Reviewed-by:
Bart Van Assche <bvanassche@acm.org> Signed-off-by:
Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by:
-
Wen Xiong authored
[ Upstream commit 318ddb34 ] While dlpar adding primary ipr adapter back, driver goes through adapter initialization then schedule ipr_worker_thread to start te disk scan by dropping the host lock, calling scsi_add_device. Then get the adapter reset request again, so driver does scsi_block_requests, this will cause the scsi_add_device get hung until we unblock. But we can't run ipr_worker_thread to do the unblock because its stuck in scsi_add_device. This patch fixes the issue. [mkp: typo and whitespace fixes] Signed-off-by:
Wen Xiong <wenxiong@linux.vnet.ibm.com> Acked-by:
Brian King <brking@linux.vnet.ibm.com> Signed-off-by:
-
Alexandru Gheorghe authored
[ Upstream commit 69be1984 ] Currently, if userspace calls drm_wait_vblank before the crtc is activated the crtc vblank_enable hook is called, which in case of malidp driver triggers some warninngs. This happens because on device init we don't inform the drm core about the vblank state by calling drm_crtc_vblank_on/off/reset which together with drm_vblank_get have some magic that prevents calling drm_vblank_enable when crtc is off. Signed-off-by:
Alexandru Gheorghe <alexandru-cosmin.gheorghe@arm.com> Acked-by:
Liviu Dudau <liviu.dudau@arm.com> Signed-off-by:
-
Jisheng Zhang authored
[ Upstream commit 9024143e ] When programming the inbound/outbound ATUs, we call usleep_range() after each checking PCIE_ATU_ENABLE bit. Unfortunately, the ATU programming can be executed in atomic context: inbound ATU programming could be called through pci_epc_write_header() =>dw_pcie_ep_write_header() =>dw_pcie_prog_inbound_atu() outbound ATU programming could be called through pci_bus_read_config_dword() =>dw_pcie_rd_conf() =>dw_pcie_prog_outbound_atu() Fix this issue by calling mdelay() instead. Fixes: f8aed6ec ("PCI: dwc: designware: Add EP mode support") Fixes: d8bbeb39 ("PCI: designware: Wait for iATU enable") Signed-off-by:
Jisheng Zhang <Jisheng.Zhang@synaptics.com> [lorenzo.pieralisi@arm.com: commit log update] Signed-off-by:
Lorenzo Pieralisi <lorenzo.pieralisi@arm.com> Signed-off-by:
Bjorn Helgaas <bhelgaas@google.com> Acked-by:
Gustavo Pimentel <gustavo.pimentel@synopsys.com> Signed-off-by:
-
Kazuya Mizuguchi authored
[ Upstream commit 2fe397a3 ] EtherAVB hardware requires 0 to be written to status register bits in order to clear them, however, care must be taken not to: 1. Clear other bits, by writing zero to them 2. Write one to reserved bits This patch corrects the ravb driver with respect to the second point above. This is done by defining reserved bit masks for the affected registers and, after auditing the code, ensure all sites that may write a one to a reserved bit use are suitably masked. Signed-off-by:
Kazuya Mizuguchi <kazuya.mizuguchi.ks@renesas.com> Signed-off-by:
Simon Horman <horms+renesas@verge.net.au> Reviewed-by:
Sergei Shtylyov <sergei.shtylyov@cogentembedded.com> Signed-off-by:
-
