- 09 May, 2022 40 commits
-
-
Daniel Starke authored
commit 1adf6fee upstream. gsm_control_modem() informs the virtual tty that more data can be written after receiving a control signal octet via modem status command (MSC). However, gsm_dlci_data() fails to do the same after receiving a control signal octet from the convergence layer type 2 header. Add tty_wakeup() in gsm_dlci_data() for convergence layer type 2 to fix this. Fixes: e1eaea46 ("tty: n_gsm line discipline") Cc: stable@vger.kernel.org Signed-off-by:
Daniel Starke <daniel.starke@siemens.com> Link: https://lore.kernel.org/r/20220414094225.4527-14-daniel.starke@siemens.com Signed-off-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Daniel Starke authored
commit 317f86af upstream. n_gsm is based on the 3GPP 07.010 and its newer version is the 3GPP 27.010. See https://portal.3gpp.org/desktopmodules/Specifications/SpecificationDetails.aspx?specificationId=1516 The changes from 07.010 to 27.010 are non-functional. Therefore, I refer to the newer 27.010 here. The value of the modem status command (MSC) frame contains an address field, control signal and optional break signal octet. The address field is encoded as described in chapter 5.2.1.2 with only one octet (may be extended to more in future versions of the standard). Whereas the control signal and break signal octet are always one byte each. This is strange at first glance as it makes the EA bit redundant. However, the same two octets are also encoded as header in convergence layer type 2 as described in chapter 5.5.2. No header length field is given and the only way to test if there is an optional break signal octet is via the EA flag which extends the control signal octet with a break signal octet. Now it becomes obvious how the EA bit for those two octets shall be encoded in the MSC frame. The current implementation treats the signal octet different for MSC frame and convergence layer type 2 header even though the standard describes it for both in the same way. Use the EA bit to encode the signal octets not only in the convergence layer type 2 header but also in the MSC frame in the same way with either 1 or 2 bytes in case of an optional break signal. Adjust the receiving path accordingly in gsm_control_modem(). Fixes: 3ac06b90 ("tty: n_gsm: Fix for modems with brk in modem status control") Cc: stable@vger.kernel.org Signed-off-by:
-
Daniel Starke authored
commit 398867f5 upstream. n_gsm is based on the 3GPP 07.010 and its newer version is the 3GPP 27.010. See https://portal.3gpp.org/desktopmodules/Specifications/SpecificationDetails.aspx?specificationId=1516 The changes from 07.010 to 27.010 are non-functional. Therefore, I refer to the newer 27.010 here. Chapter 5.4.6.1 states that each command frame shall be made up from type, length and value. Looking for example in chapter 5.4.6.3.5 at the description for the encoding of a flow control on command it becomes obvious, that the type and length field is always present whereas the value may be zero bytes long. The current implementation omits the length field if the value is not present. This is wrong. Correct this by always sending the length in gsm_control_transmit(). So far only the modem status command (MSC) has included a value and encoded its length directly. Therefore, also change gsmtty_modem_update(). Fixes: e1eaea46 ("tty: n_gsm line discipline") Cc: stable@vger.kernel.org Signed-off-by:
-
Daniel Starke authored
commit d0bcdffc upstream. n_gsm is based on the 3GPP 07.010 and its newer version is the 3GPP 27.010. See https://portal.3gpp.org/desktopmodules/Specifications/SpecificationDetails.aspx?specificationId=1516 The changes from 07.010 to 27.010 are non-functional. Therefore, I refer to the newer 27.010 here. Chapter 5.7.3 states that the valid range for the maximum number of retransmissions (N2) is from 0 to 255 (both including). gsm_config() fails to limit this range correctly. Furthermore, gsm_control_retransmit() handles this number incorrectly by performing N2 - 1 retransmission attempts. Setting N2 to zero results in more than 255 retransmission attempts. Fix the range check in gsm_config() and the value handling in gsm_control_send() and gsm_control_retransmit() to comply with 3GPP 27.010. Fixes: e1eaea46 ("tty: n_gsm line discipline") Cc: stable@vger.kernel.org Signed-off-by:
-
Daniel Starke authored
commit 17eac652 upstream. In gsm_cleanup_mux() the muxer is closed down and all queues are removed. However, removing the queues is done without explicit control of the underlying buffers. Flush those before freeing up our queues to ensure that all outgoing queues are cleared consistently. Otherwise, a new mux connection establishment attempt may time out while the underlying tty is still busy sending out the remaining data from the previous connection. Fixes: e1eaea46 ("tty: n_gsm line discipline") Cc: stable@vger.kernel.org Signed-off-by:
-
Daniel Starke authored
commit deefc58b upstream. The current DLCI release order starts with the control channel followed by the user channels. Reverse this order to keep the control channel open until all user channels have been released. Fixes: e1eaea46 ("tty: n_gsm line discipline") Cc: stable@vger.kernel.org Signed-off-by:
-
Daniel Starke authored
commit 535bf600 upstream. n_gsm is based on the 3GPP 07.010 and its newer version is the 3GPP 27.010. See https://portal.3gpp.org/desktopmodules/Specifications/SpecificationDetails.aspx?specificationId=1516 The changes from 07.010 to 27.010 are non-functional. Therefore, I refer to the newer 27.010 here. Chapter 5.7.2 states that the maximum frame size (N1) refers to the length of the information field (i.e. user payload). However, 'txframe' stores the whole frame including frame header, checksum and start/end flags. We also need to consider the byte stuffing overhead. Define constant for the protocol overhead and adjust the 'txframe' size calculation accordingly to reserve enough space for a complete mux frame including byte stuffing for advanced option mode. Note that no byte stuffing is applied to the start and end flag. Also use MAX_MTU instead of MAX_MRU as this buffer is used for data transmission. Fixes: e1eaea46 ("tty: n_gsm line discipline") Cc: stable@vger.kernel.org Signed-off-by:
-
Florian Westphal authored
commit 743b83f1 upstream. Check if the incoming interface is available and NFT_BREAK in case neither skb->sk nor input device are set. Because nf_sk_lookup_slow*() assume packet headers are in the 'in' direction, use in postrouting is not going to yield a meaningful result. Same is true for the forward chain, so restrict the use to prerouting, input and output. Use in output work if a socket is already attached to the skb. Fixes: 554ced0a ("netfilter: nf_tables: add support for native socket matching") Reported-and-tested-by:
Topi Miettinen <toiwoton@gmail.com> Signed-off-by:
Florian Westphal <fw@strlen.de> Signed-off-by:
Pablo Neira Ayuso <pablo@netfilter.org> Signed-off-by:
-
Daniel Starke authored
commit a24b4b2f upstream. The gsm_mux field 'malformed' represents the number of malformed frames received. However, gsm1_receive() also increases this counter for any out of frame byte. Fix this by ignoring out of frame data for the malformed counter. Fixes: e1eaea46 ("tty: n_gsm line discipline") Cc: stable@vger.kernel.org Signed-off-by:
-
Daniel Starke authored
commit 06d5afd4 upstream. n_gsm is based on the 3GPP 07.010 and its newer version is the 3GPP 27.010. See https://portal.3gpp.org/desktopmodules/Specifications/SpecificationDetails.aspx?specificationId=1516 The changes from 07.010 to 27.010 are non-functional. Therefore, I refer to the newer 27.010 here. Chapter 5.5.2 describes that the signal octet in convergence layer type 2 can be either one or two bytes. The length is encoded in the EA bit. This is set 1 for the last byte in the sequence. gsmtty_modem_update() handles this correctly but gsm_dlci_data_output() fails to set EA to 1. There is no case in which we encode two signal octets as there is no case in which we send out a break signal. Therefore, always set the EA bit to 1 for the signal octet to fix this. Fixes: e1eaea46 ("tty: n_gsm line discipline") Cc: stable@vger.kernel.org Signed-off-by:
-
Daniel Starke authored
commit 284260f2 upstream. Internally, we manage the alive state of the mux channels and mux itself with the field member 'dead'. This makes it possible to notify the user if the accessed underlying link is already gone. On the other hand, however, removing the virtual ttys before terminating the channels may result in peer messages being received without any internal target. Move the mux cleanup procedure from gsmld_detach_gsm() to gsmld_close() to fix this by keeping the virtual ttys open until the mux has been cleaned up. Fixes: e1eaea46 ("tty: n_gsm line discipline") Cc: stable@vger.kernel.org Signed-off-by:
-
Daniel Starke authored
commit 1ec92e97 upstream. The active mux instances are managed in the gsm_mux array and via mux_get() and mux_put() functions separately. This gives a very loose coupling between the actual instance and the gsm_mux array which manages it. It also results in unnecessary lockings which makes it prone to failures. And it creates a race condition if more than the maximum number of mux instances are requested while the user changes the parameters of an active instance. The user may loose ownership of the current mux instance in this case. Fix this by moving the gsm_mux array handling to the mux allocation and deallocation functions. Fixes: e1eaea46 ("tty: n_gsm line discipline") Cc: stable@vger.kernel.org Signed-off-by:
-
Daniel Starke authored
commit aa371e96 upstream. n_gsm is based on the 3GPP 07.010 and its newer version is the 3GPP 27.010. See https://portal.3gpp.org/desktopmodules/Specifications/SpecificationDetails.aspx?specificationId=1516 The changes from 07.010 to 27.010 are non-functional. Therefore, I refer to the newer 27.010 here. Chapter 5.8.2 states that both sides will revert to the non-multiplexed mode via a close-down message (CLD). The usual program flow is as following: - start multiplex mode by sending AT+CMUX to the mobile - establish the control channel (DLCI 0) - establish user channels (DLCI >0) - terminate user channels - send close-down message (CLD) - revert to AT protocol (i.e. leave multiplexed mode) The AT protocol is out of scope of the n_gsm driver. However, gsm_disconnect() sends CLD if gsm_config() detects that the requested parameters require the mux protocol to restart. The next immediate action is to start the mux protocol by opening DLCI 0 again. Any responder side which handles CLD commands correctly forces us to fail at this point because AT+CMUX needs to be sent to the mobile to start the mux again. Therefore, remove the CLD command in this phase and keep both sides in multiplexed mode. Remove the gsm_disconnect() function as it become unnecessary and merge the remaining parts into gsm_cleanup_mux() to handle the termination order and locking correctly. Fixes: 71e07791 ("tty: n_gsm: do not send/receive in ldisc close path") Cc: stable@vger.kernel.org Signed-off-by:
-
Namhyung Kim authored
commit 8799ebce upstream. Now arch-specific functions all do the same thing. When it fixes the symbol address it needs to check the boundary between the kernel image and modules. For the last symbol in the previous region, it cannot know the exact size as it's discarded already. Thus it just uses a small page size (4096) and rounds it up like the last symbol. Fixes: 3cf6a32f ("perf symbols: Fix symbol size calculation condition") Signed-off-by:
Namhyung Kim <namhyung@kernel.org> Acked-by:
Ian Rogers <irogers@google.com> Cc: Heiko Carstens <hca@linux.ibm.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@kernel.org> Cc: John Garry <john.garry@huawei.com> Cc: Leo Yan <leo.yan@linaro.org> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Cc: Michael Ellerman <mpe@ellerman.id.au> Cc: Michael Petlan <mpetlan@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Song Liu <songliubraving@fb.com> Cc: Will Deacon <will@kernel.org> Cc: linux-s390@vger.kernel.org Cc: linuxppc-dev@lists.ozlabs.org Link: https://lore.kernel.org/r/20220416004048.1514900-3-namhyung@kernel.org Signed-off-by:
Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by:
-
Namhyung Kim authored
commit 838425f2 upstream. The symbol fixup is necessary for symbols in kallsyms since they don't have size info. So we use the next symbol's address to calculate the size. Now it's also used for user binaries because sometimes they miss size for hand-written asm functions. There's a arch-specific function to handle kallsyms differently but currently it cannot distinguish kallsyms from others. Pass this information explicitly to handle it properly. Note that those arch functions will be moved to the generic function so I didn't added it to the arch-functions. Fixes: 3cf6a32f ("perf symbols: Fix symbol size calculation condition") Signed-off-by:
-
Borislav Petkov authored
commit f9e14dbb upstream. When resuming from system sleep state, restore_processor_state() restores the boot CPU MSRs. These MSRs could be emulated by microcode. If microcode is not loaded yet, writing to emulated MSRs leads to unchecked MSR access error: ... PM: Calling lapic_suspend+0x0/0x210 unchecked MSR access error: WRMSR to 0x10f (tried to write 0x0...0) at rIP: ... (native_write_msr) Call Trace: <TASK> ? restore_processor_state x86_acpi_suspend_lowlevel acpi_suspend_enter suspend_devices_and_enter pm_suspend.cold state_store kobj_attr_store sysfs_kf_write kernfs_fop_write_iter new_sync_write vfs_write ksys_write __x64_sys_write do_syscall_64 entry_SYSCALL_64_after_hwframe RIP: 0033:0x7fda13c260a7 To ensure microcode emulated MSRs are available for restoration, load the microcode on the boot CPU before restoring these MSRs. [ Pawan: write commit message and productize it. ] Fixes: e2a1256b ("x86/speculation: Restore speculation related MSRs during S3 resume") Reported-by:
Kyle D. Pelton <kyle.d.pelton@intel.com> Signed-off-by:
Borislav Petkov <bp@suse.de> Signed-off-by:
Pawan Gupta <pawan.kumar.gupta@linux.intel.com> Tested-by:
-
Tim Harvey authored
commit 4c79865f upstream. The GW71xx, GW72xx and GW73xx boards have USB1 routed to a USB OTG connectors and USB2 routed to a USB hub. The OTG connector has a over-currently protection with an active-low pin and the USB1 to HUB connection has no over-current protection (as the HUB itself implements this for its downstream ports). Add proper dt nodes to specify the over-current pin polarity for USB1 and disable over-current protection for USB2. Fixes: 6f30b27c ("arm64: dts: imx8mm: Add Gateworks i.MX 8M Mini Development Kits") Cc: stable@vger.kernel.org Signed-off-by:
Tim Harvey <tharvey@gateworks.com> Signed-off-by:
Shawn Guo <shawnguo@kernel.org> Signed-off-by:
-
Eugen Hristev authored
commit 3f7ce6d7 upstream. Flexcom3 is used as board console serial. There are no pull-ups on these lines on the board. This means that if a cable is not connected (that has pull-ups included), stray characters could appear on the console as the floating pins voltage levels are interpreted as incoming characters. To avoid this problem, enable the internal pull-ups on these lines. Fixes: 7540629e ("ARM: dts: at91: add sama7g5 SoC DT and sama7g5-ek") Cc: stable@vger.kernel.org # v5.15+ Signed-off-by:
Eugen Hristev <eugen.hristev@microchip.com> Reviewed-by:
Tudor Ambarus <tudor.ambarus@microchip.com> Signed-off-by:
Nicolas Ferre <nicolas.ferre@microchip.com> Link: https://lore.kernel.org/r/20220307113827.2419331-1-eugen.hristev@microchip.com Signed-off-by:
-
Filipe Manana authored
commit 50ff5788 upstream. On a zoned filesystem, if we fail to allocate the root node for the log root tree while syncing the log, we end up returning without finishing the IO plug we started before, resulting in leaking resources as we have started writeback for extent buffers of a log tree before. That allocation failure, which typically is either -ENOMEM or -ENOSPC, is not fatal and the fsync can safely fallback to a full transaction commit. So release the IO plug if we fail to allocate the extent buffer for the root of the log root tree when syncing the log on a zoned filesystem. Fixes: 3ddebf27 ("btrfs: zoned: reorder log node allocation on zoned filesystem") CC: stable@vger.kernel.org # 5.15+ Reviewed-by:
Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by:
Filipe Manana <fdmanana@suse.com> Reviewed-by:
David Sterba <dsterba@suse.com> Signed-off-by:
-
Kees Cook authored
commit d0f6cfb2 upstream. Control Flow Integrity (CFI) instrumentation of the kernel noticed that the caller, dev_attr_show(), and the callback, odvp_show(), did not have matching function prototypes, which would cause a CFI exception to be raised. Correct the prototype by using struct device_attribute instead of struct kobj_attribute. Reported-and-tested-by:
Joao Moreira <joao@overdrivepizza.com> Link: https://lore.kernel.org/lkml/067ce8bd4c3968054509831fa2347f4f@overdrivepizza.com/ Fixes: 006f006f ("thermal/int340x_thermal: Export OEM vendor variables") Cc: 5.8+ <stable@vger.kernel.org> # 5.8+ Signed-off-by:
Kees Cook <keescook@chromium.org> Signed-off-by:
Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by:
-
Ville Syrjälä authored
commit fc45e55e upstream. The "safe state" index is used by acpi_idle_enter_bm() to avoid entering a C-state that may require bus mastering to be disabled on entry in the cases when this is not going to happen. For this reason, it should not be set to point to C3 type of C-states, because they may require bus mastering to be disabled on entry in principle. This was broken by commit d6b88ce2 ("ACPI: processor idle: Allow playing dead in C3 state") which inadvertently allowed the "safe state" index to point to C3 type of C-states. This results in a machine that won't boot past the point when it first enters C3. Restore the correct behaviour (either demote to C1/C2, or use C3 but also set ARB_DIS=1). I hit this on a Fujitsu Siemens Lifebook S6010 (P3) machine. Fixes: d6b88ce2 ("ACPI: processor idle: Allow playing dead in C3 state") Cc: 5.16+ <stable@vger.kernel.org> # 5.16+ Signed-off-by:
Ville Syrjälä <ville.syrjala@linux.intel.com> Tested-by:
Woody Suwalski <wsuwalski@gmail.com> [ rjw: Subject and changelog adjustments ] Signed-off-by:
-
Dinh Nguyen authored
commit 5fd1fe48 upstream. I made a mistake with the commit a6aaa003 ("net: ethernet: stmmac: fix altr_tse_pcs function when using a fixed-link"). I should have tested against both scenario of having a SGMII interface and one without. Without the SGMII PCS TSE adpater, the sgmii_adapter_base address is NULL, thus a write to this address will fail. Cc: stable@vger.kernel.org Fixes: a6aaa003 ("net: ethernet: stmmac: fix altr_tse_pcs function when using a fixed-link") Signed-off-by:
Dinh Nguyen <dinguyen@kernel.org> Link: https://lore.kernel.org/r/20220420152345.27415-1-dinguyen@kernel.org Signed-off-by:
Jakub Kicinski <kuba@kernel.org> Signed-off-by:
-
Imre Deak authored
commit 4ae4dd2e upstream. Fix typo in the _SEL_FETCH_PLANE_BASE_1_B register base address. Fixes: a5523e2f ("drm/i915: Add PSR2 selective fetch registers") References: https://gitlab.freedesktop.org/drm/intel/-/issues/5400 Cc: José Roberto de Souza <jose.souza@intel.com> Cc: <stable@vger.kernel.org> # v5.9+ Signed-off-by:
Imre Deak <imre.deak@intel.com> Reviewed-by:
José Roberto de Souza <jose.souza@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20220421162221.2261895-1-imre.deak@intel.com (cherry picked from commit af2cbc6e ) Signed-off-by:
Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Signed-off-by:
-
Jouni Högander authored
commit c05d8332 upstream. We have now seen panel (XMG Core 15 e21 laptop) advertizing support for Intel proprietary eDP backlight control via DPCD registers, but actually working only with legacy pwm control. This patch adds panel EDID check for possible HDR static metadata and Intel proprietary eDP backlight control is used only if that exists. Missing HDR static metadata is ignored if user specifically asks for Intel proprietary eDP backlight control via enable_dpcd_backlight parameter. v2 : - Ignore missing HDR static metadata if Intel proprietary eDP backlight control is forced via i915.enable_dpcd_backlight - Printout info message if panel is missing HDR static metadata and support for Intel proprietary eDP backlight control is detected Fixes: 4a8d7990 ("drm/i915/dp: Enable Intel's HDR backlight interface (only SDR for now)") Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/5284 Cc: Lyude Paul <lyude@redhat.com> Cc: Mika Kahola <mika.kahola@intel.com> Cc: Jani Nikula <jani.nikula@intel.com> Cc: Filippo Falezza <filippo.falezza@outlook.it> Cc: stable@vger.kernel.org Signed-off-by:
Jouni Högander <jouni.hogander@intel.com> Signed-off-by:
Lyude Paul <lyude@redhat.com> (cherry picked from commit b4b15757 ) Signed-off-by:
-
Martin Willi authored
commit 8ddffdb9 upstream. The commit referenced below fixed packet re-routing if Netfilter mangles a routing key property of a packet and the packet is routed in a VRF L3 domain. The fix, however, addressed IPv4 re-routing, only. This commit applies the same behavior for IPv6. While at it, untangle the nested ternary operator to make the code more readable. Fixes: 6d8b49c3 ("netfilter: Update ip_route_me_harder to consider L3 domain") Cc: stable@vger.kernel.org Signed-off-by:
Martin Willi <martin@strongswan.org> Reviewed-by:
David Ahern <dsahern@kernel.org> Signed-off-by:
-
Md Sadre Alam authored
commit ba7542eb upstream. This patch fixes a memory corruption that occurred in the nand_scan() path for Hynix nand device. On boot, for Hynix nand device will panic at a weird place: | Unable to handle kernel NULL pointer dereference at virtual address 00000070 | [00000070] *pgd=00000000 | Internal error: Oops: 5 [#1] PREEMPT SMP ARM | Modules linked in: | CPU: 0 PID: 1 Comm: swapper/0 Not tainted 5.17.0-01473-g13ae1769cfb0 #38 | Hardware name: Generic DT based system | PC is at nandc_set_reg+0x8/0x1c | LR is at qcom_nandc_command+0x20c/0x5d0 | pc : [<c088b74c>] lr : [<c088d9c8>] psr: 00000113 | sp : c14adc50 ip : c14ee208 fp : c0cc970c | r10: 000000a3 r9 : 00000000 r8 : 00000040 | r7 : c16f6a00 r6 : 00000090 r5 : 00000004 r4 :c14ee040 | r3 : 00000000 r2 : 0000000b r1 : 00000000 r0 :c14ee040 | Flags: nzcv IRQs on FIQs on Mode SVC_32 ISA ARM Segment none | Control: 10c5387d Table: 8020406a DAC: 00000051 | Register r0 information: slab kmalloc-2k start c14ee000 pointer offset 64 size 2048 | Process swapper/0 (pid: 1, stack limit = 0x(ptrval)) | nandc_set_reg from qcom_nandc_command+0x20c/0x5d0 | qcom_nandc_command from nand_readid_op+0x198/0x1e8 | nand_readid_op from hynix_nand_has_valid_jedecid+0x30/0x78 | hynix_nand_has_valid_jedecid from hynix_nand_init+0xb8/0x454 | hynix_nand_init from nand_scan_with_ids+0xa30/0x14a8 | nand_scan_with_ids from qcom_nandc_probe+0x648/0x7b0 | qcom_nandc_probe from platform_probe+0x58/0xac The problem is that the nand_scan()'s qcom_nand_attach_chip callback is updating the nandc->max_cwperpage from 1 to 4 or 8 based on page size. This causes the sg_init_table of clear_bam_transaction() in the driver's qcom_nandc_command() to memset much more than what was initially allocated by alloc_bam_transaction(). This patch will update nandc->max_cwperpage 1 to 4 or 8 based on page size in qcom_nand_attach_chip call back after freeing the previously allocated memory for bam txn as per nandc->max_cwperpage = 1 and then again allocating bam txn as per nandc->max_cwperpage = 4 or 8 based on page size in qcom_nand_attach_chip call back itself. Cc: stable@vger.kernel.org Fixes: 6a3cec64 ("mtd: rawnand: qcom: convert driver to nand_scan()") Reported-by:
Konrad Dybcio <konrad.dybcio@somainline.org> Reviewed-by:
Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org> Co-developed-by:
Sricharan R <quic_srichara@quicinc.com> Signed-off-by:
Md Sadre Alam <quic_mdalam@quicinc.com> Signed-off-by:
Miquel Raynal <miquel.raynal@bootlin.com> Link: https://lore.kernel.org/linux-mtd/1650268107-5363-1-git-send-email-quic_mdalam@quicinc.com Signed-off-by:
-
Zqiang authored
commit 31fa985b upstream. kasan_quarantine_remove_cache() is called in kmem_cache_shrink()/ destroy(). The kasan_quarantine_remove_cache() call is protected by cpuslock in kmem_cache_destroy() to ensure serialization with kasan_cpu_offline(). However the kasan_quarantine_remove_cache() call is not protected by cpuslock in kmem_cache_shrink(). When a CPU is going offline and cache shrink occurs at same time, the cpu_quarantine may be corrupted by interrupt (per_cpu_remove_cache operation). So add a cpu_quarantine offline flags check in per_cpu_remove_cache(). [akpm@linux-foundation.org: add comment, per Zqiang] Link: https://lkml.kernel.org/r/20220414025925.2423818-1-qiang1.zhang@intel.com Signed-off-by:
Zqiang <qiang1.zhang@intel.com> Reviewed-by:
Dmitry Vyukov <dvyukov@google.com> Cc: Andrey Ryabinin <ryabinin.a.a@gmail.com> Cc: Alexander Potapenko <glider@google.com> Cc: Andrey Konovalov <andreyknvl@gmail.com> 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:
-
Damien Le Moal authored
commit 694852ea upstream. Ensure that the i_flags field of struct zonefs_inode_info is cleared to 0 when initializing a zone file inode, avoiding seeing the flag ZONEFS_ZONE_OPEN being incorrectly set. Fixes: b5c00e97 ("zonefs: open/close zone on file open/close") Cc: <stable@vger.kernel.org> Signed-off-by:
Damien Le Moal <damien.lemoal@opensource.wdc.com> Reviewed-by:
Chaitanya Kulkarni <kch@nvidia.com> Reviewed-by:
Hans Holmberg <hans.holmberg@wdc.com> Signed-off-by:
-
Damien Le Moal authored
commit 1da18a29 upstream. The mount option "explicit_open" manages the device open zone resources to ensure that if an application opens a sequential file for writing, the file zone can always be written by explicitly opening the zone and accounting for that state with the s_open_zones counter. However, if some zones are already open when mounting, the device open zone resource usage status will be larger than the initial s_open_zones value of 0. Ensure that this inconsistency does not happen by closing any sequential zone that is open when mounting. Furthermore, with ZNS drives, closing an explicitly open zone that has not been written will change the zone state to "closed", that is, the zone will remain in an active state. Since this can then cause failures of explicit open operations on other zones if the drive active zone resources are exceeded, we need to make sure that the zone is not active anymore by resetting it instead of closing it. To address this, zonefs_zone_mgmt() is modified to change a REQ_OP_ZONE_CLOSE request into a REQ_OP_ZONE_RESET for sequential zones that have not been written. Fixes: b5c00e97 ("zonefs: open/close zone on file open/close") Cc: <stable@vger.kernel.org> Signed-off-by:
-
Ville Syrjälä authored
commit 20e582e1 upstream. This reverts commit bfe55a1f . This was presumably misdiagnosed as an inability to use C3 at all when I suspect the real problem is just misconfiguration of C3 vs. ARB_DIS. Signed-off-by:
-
Sidhartha Kumar authored
[ Upstream commit 18d609da ] Because mremap does not have a MAP_FIXED_NOREPLACE flag, it can destroy existing mappings. This causes a segfault when regions such as text are remapped and the permissions are changed. Verify the requested mremap destination address does not overlap any existing mappings by using mmap's MAP_FIXED_NOREPLACE flag. Keep incrementing the destination address until a valid mapping is found or fail the current test once the max address is reached. Link: https://lkml.kernel.org/r/20220420215721.4868-2-sidhartha.kumar@oracle.com Signed-off-by:
Sidhartha Kumar <sidhartha.kumar@oracle.com> Reviewed-by:
Shuah Khan <skhan@linuxfoundation.org> Signed-off-by:
Sasha Levin <sashal@kernel.org>
-
Sidhartha Kumar authored
[ Upstream commit 9c85a9ba ] Avoid calling mmap with requested addresses that are less than the system's mmap_min_addr. When run as root, mmap returns EACCES when trying to map addresses < mmap_min_addr. This is not one of the error codes for the condition to retry the mmap in the test. Rather than arbitrarily retrying on EACCES, don't attempt an mmap until addr > vm.mmap_min_addr. Add a munmap call after an alignment check as the mappings are retained after the retry and can reach the vm.max_map_count sysctl. Link: https://lkml.kernel.org/r/20220420215721.4868-1-sidhartha.kumar@oracle.com Signed-off-by:
-
Alexey Kardashevskiy authored
[ Upstream commit bb82c574 ] The "read_bhrb" global symbol is only called under CONFIG_PPC64 of arch/powerpc/perf/core-book3s.c but it is compiled for both 32 and 64 bit anyway (and LLVM fails to link this on 32bit). This fixes it by moving bhrb.o to obj64 targets. Signed-off-by:
Alexey Kardashevskiy <aik@ozlabs.ru> Signed-off-by:
Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/20220421025756.571995-1-aik@ozlabs.ru Signed-off-by:
-
Duoming Zhou authored
[ Upstream commit bc6de287 ] There is a deadlock in rr_close(), which is shown below: (Thread 1) | (Thread 2) | rr_open() rr_close() | add_timer() spin_lock_irqsave() //(1) | (wait a time) ... | rr_timer() del_timer_sync() | spin_lock_irqsave() //(2) (wait timer to stop) | ... We hold rrpriv->lock in position (1) of thread 1 and use del_timer_sync() to wait timer to stop, but timer handler also need rrpriv->lock in position (2) of thread 2. As a result, rr_close() will block forever. This patch extracts del_timer_sync() from the protection of spin_lock_irqsave(), which could let timer handler to obtain the needed lock. Signed-off-by:
Duoming Zhou <duoming@zju.edu.cn> Link: https://lore.kernel.org/r/20220417125519.82618-1-duoming@zju.edu.cn Signed-off-by:
Paolo Abeni <pabeni@redhat.com> Signed-off-by:
-
Ronnie Sahlberg authored
[ Upstream commit f5d0f921 ] because the copychunk_write might cover a region of the file that has not yet been sent to the server and thus fail. A simple way to reproduce this is: truncate -s 0 /mnt/testfile; strace -f -o x -ttT xfs_io -i -f -c 'pwrite 0k 128k' -c 'fcollapse 16k 24k' /mnt/testfile the issue is that the 'pwrite 0k 128k' becomes rearranged on the wire with the 'fcollapse 16k 24k' due to write-back caching. fcollapse is implemented in cifs.ko as a SMB2 IOCTL(COPYCHUNK_WRITE) call and it will fail serverside since the file is still 0b in size serverside until the writes have been destaged. To avoid this we must ensure that we destage any unwritten data to the server before calling COPYCHUNK_WRITE. Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=1997373 Reported-by:
Xiaoli Feng <xifeng@redhat.com> Signed-off-by:
Ronnie Sahlberg <lsahlber@redhat.com> Signed-off-by:
Steve French <stfrench@microsoft.com> Signed-off-by:
-
Mikulas Patocka authored
[ Upstream commit a6823e4e ] The first "if" condition in __memcpy_flushcache is supposed to align the "dest" variable to 8 bytes and copy data up to this alignment. However, this condition may misbehave if "size" is greater than 4GiB. The statement min_t(unsigned, size, ALIGN(dest, 8) - dest); casts both arguments to unsigned int and selects the smaller one. However, the cast truncates high bits in "size" and it results in misbehavior. For example: suppose that size == 0x100000001, dest == 0x200000002 min_t(unsigned, size, ALIGN(dest, 8) - dest) == min_t(0x1, 0xe) == 0x1; ... dest += 0x1; so we copy just one byte "and" dest remains unaligned. This patch fixes the bug by replacing unsigned with size_t. Signed-off-by:
Mikulas Patocka <mpatocka@redhat.com> Signed-off-by:
-
suresh kumar authored
[ Upstream commit 49aefd13 ] Commit b5f86218 was introduced to discard lowest hash bit for layer3+4 hashing but it also removes last bit from non layer3+4 hashing Below script shows layer2+3 hashing will result in same slave to be used with above commit. $ cat hash.py #/usr/bin/python3.6 h_dests=[0xa0, 0xa1] h_source=0xe3 hproto=0x8 saddr=0x1e7aa8c0 daddr=0x17aa8c0 for h_dest in h_dests: hash = (h_dest ^ h_source ^ hproto ^ saddr ^ daddr) hash ^= hash >> 16 hash ^= hash >> 8 print(hash) print("with last bit removed") for h_dest in h_dests: hash = (h_dest ^ h_source ^ hproto ^ saddr ^ daddr) hash ^= hash >> 16 hash ^= hash >> 8 hash = hash >> 1 print(hash) Output: $ python3.6 hash.py 522133332 522133333 <-------------- will result in both slaves being used with last bit removed 261066666 261066666 <-------------- only single slave used Signed-off-by:
suresh kumar <suresh2514@gmail.com> Signed-off-by:
David S. Miller <davem@davemloft.net> Signed-off-by:
-
Namjae Jeon authored
[ Upstream commit 02655a70 ] Currently ksmbd is using ->f_bsize from vfs_statfs() as sector size. If fat/exfat is a local share, ->f_bsize is a cluster size that is too large to be used as a sector size. Sector sizes larger than 4K cause problem occurs when mounting an iso file through windows client. The error message can be obtained using Mount-DiskImage command, the error is: "Mount-DiskImage : The sector size of the physical disk on which the virtual disk resides is not supported." This patch reports fixed 4KB sector size if ->s_blocksize is bigger than 4KB. Signed-off-by:
Namjae Jeon <linkinjeon@kernel.org> Signed-off-by:
-
Namjae Jeon authored
[ Upstream commit 8510a043 ] Add missing increment reference count of parent fp in ksmbd_lookup_fd_inode(). Signed-off-by:
Hyunchul Lee <hyc.lee@gmail.com> Signed-off-by:
-
Duoming Zhou authored
[ Upstream commit eb5adc70 ] There is a deadlock in rs_close(), which is shown below: (Thread 1) | (Thread 2) | rs_open() rs_close() | mod_timer() spin_lock_bh() //(1) | (wait a time) ... | rs_poll() del_timer_sync() | spin_lock() //(2) (wait timer to stop) | ... We hold timer_lock in position (1) of thread 1 and use del_timer_sync() to wait timer to stop, but timer handler also need timer_lock in position (2) of thread 2. As a result, rs_close() will block forever. This patch deletes the redundant timer_lock in order to prevent the deadlock. Because there is no race condition between rs_close, rs_open and rs_poll. Signed-off-by:
Max Filippov <jcmvbkbc@gmail.com> Signed-off-by:
-
