- 12 May, 2021 1 commit
-
-
Paolo Abeni authored
If the owing socket is shutting down - e.g. the sock reference count already dropped to 0 and only sk_wmem_alloc is keeping the sock alive, skb_orphan_partial() becomes a no-op. When forwarding packets over veth with GRO enabled, the above causes refcount errors. This change addresses the issue with a plain skb_orphan() call in the critical scenario. Fixes: 9adc89af ("net: let skb_orphan_partial wake-up waiters.") Signed-off-by:
Paolo Abeni <pabeni@redhat.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
- 11 May, 2021 34 commits
-
-
Alex Elder authored
IPA configuration data includes an array of memory region descriptors. That was a fixed-size array at one time, but at some point we started defining it such that it was only as big as required for a given platform. The actual number of entries in the array is recorded in the configuration data along with the array. A loop in ipa_mem_config() still assumes the array has entries for all defined memory region IDs. As a result, this loop can go past the end of the actual array and attempt to write "canary" values based on nonsensical data. Fix this, by stashing the number of entries in the array, and using that rather than IPA_MEM_COUNT in the initialization loop found in ipa_mem_config(). The only remaining use of IPA_MEM_COUNT is in a validation check to ensure configuration data doesn't have too many entries. That's fine for now. Fixes: 3128aae8 ("net: ipa: redefine struct ipa_mem_data") Signed-off-by:
Alex Elder <elder@linaro.org> Signed-off-by:
-
Shannon Nelson authored
When IONIC=y and PTP_1588_CLOCK=m were set in the .config file the driver link failed with undefined references. We add the dependancy depends on PTP_1588_CLOCK || !PTP_1588_CLOCK to clear this up. If PTP_1588_CLOCK=m, the depends limits IONIC to =m (or disabled). If PTP_1588_CLOCK is disabled, IONIC can be any of y/m/n. Fixes: 61db421d ("ionic: link in the new hw timestamp code") Reported-by:
kernel test robot <lkp@intel.com> Cc: Jakub Kicinski <kuba@kernel.org> Cc: Randy Dunlap <rdunlap@infradead.org> Cc: Allen Hubbe <allenbh@pensando.io> Signed-off-by:
Shannon Nelson <snelson@pensando.io> Acked-by:
Randy Dunlap <rdunlap@infradead.org> Signed-off-by:
-
Paolo Abeni authored
Maxim reported several issues when forcing a TCP transparent proxy to use the MPTCP protocol for the inbound connections. He also provided a clean reproducer. The problem boils down to 'mptcp_frag_can_collapse_to()' assuming that only MPTCP will use the given page_frag. If others - e.g. the plain TCP protocol - allocate page fragments, we can end-up re-using already allocated memory for mptcp_data_frag. Fix the issue ensuring that the to-be-expanded data fragment is located at the current page frag end. v1 -> v2: - added missing fixes tag (Mat) Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/178Reported-and-tested-by:
Maxim Galaganov <max@internet.ru> Fixes: 18b683bf ("mptcp: queue data for mptcp level retransmission") Signed-off-by:
-
git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpfDavid S. Miller authored
Daniel Borkmann says: ==================== pull-request: bpf 2021-05-11 The following pull-request contains BPF updates for your *net* tree. We've added 13 non-merge commits during the last 8 day(s) which contain a total of 21 files changed, 817 insertions(+), 382 deletions(-). The main changes are: 1) Fix multiple ringbuf bugs in particular to prevent writable mmap of read-only pages, from Andrii Nakryiko & Thadeu Lima de Souza Cascardo. 2) Fix verifier alu32 known-const subregister bound tracking for bitwise operations and/or/xor, from Daniel Borkmann. 3) Reject trampoline attachment for functions with variable arguments, and also add a deny list of other forbidden functions, from Jiri Olsa. 4) Fix nested bpf_bprintf_prepare() calls used by various helpers by switching to per-CPU buffers, from Florent Revest. 5) Fix kernel compilation with BTF debug info on ppc64 due to pahole missing TCP-CC functions like cubictcp_init, from Martin KaFai Lau. 6) Add a kconfig entry to provide an option to disallow unprivileged BPF by default, from Daniel Borkmann. 7) Fix libbpf compilation for older libelf when GELF_ST_VISIBILITY() macro is not available, from Arnaldo Carvalho de Melo. 8) Migrate test_tc_redirect to test_progs framework as prep work for upcoming skb_change_head() fix & selftest, from Jussi Maki. 9) Fix a libbpf segfault in add_dummy_ksym_var() if BTF is not present, from Ian Rogers. 10) Fix tx_only micro-benchmark in xdpsock BPF sample with proper frame size, from Magnus Karlsson. ==================== Signed-off-by: -
David S. Miller authored
Merge tag 'mac80211-for-net-2021-05-11' of git://git.kernel.org/pub/scm/linux/kernel/git/jberg/mac80211 Johannes Berg says: ==================== pull-request: mac80211 2021-05-11 So exciting times, for the first pull request for fixes I have a bunch of security things that have been under embargo for a while - see more details in the tag below, and at the patch posting message I linked to. I organized with Kalle to just have a single set of fixes for mac80211 and ath10k/ath11k, we don't know about any of the other vendors (the mac80211 + already released firmware is sufficient to fix iwlwifi.) Please pull and let me know if there's any problem. Several security issues in the 802.11 implementations were found by Mathy Vanhoef (New York University Abu Dhabi), and this contains the fixes developed for mac80211 and specifically Qualcomm drivers, I'm sending this together (as agreed with Kalle) to have just a single set of patches for now. We don't know about other vendors though. More details in the patch posting: https://lore.kernel.org/r/20210511180259.159598-1-johannes@sipsolutions.net ==================== Signed-off-by:
-
Joakim Zhang authored
Both get and set WoL will check device_can_wakeup(), if MAC supports PMT, it will set device wakeup capability. After commit 1d8e5b0f ("net: stmmac: Support WOL with phy"), device wakeup capability will be overwrite in stmmac_init_phy() according to phy's Wol feature. If phy doesn't support WoL, then MAC will lose wakeup capability. To fix this issue, only overwrite device wakeup capability when MAC doesn't support PMT. For STMMAC now driver checks MAC's WoL capability if MAC supports PMT, if not support, driver will check PHY's WoL capability. Fixes: 1d8e5b0f ("net: stmmac: Support WOL with phy") Reviewed-by:
Jisheng Zhang <Jisheng.Zhang@synaptics.com> Signed-off-by:
Joakim Zhang <qiangqing.zhang@nxp.com> Signed-off-by:
-
Martin KaFai Lau authored
During the discussion in [0]. It was pointed out that static functions in ppc64 is prefixed with ".". For example, the 'readelf -s vmlinux.ppc': 89326: c000000001383280 24 NOTYPE LOCAL DEFAULT 31 cubictcp_init 89327: c000000000c97c50 168 FUNC LOCAL DEFAULT 2 .cubictcp_init The one with FUNC type is ".cubictcp_init" instead of "cubictcp_init". The "." seems to be done by arch/powerpc/include/asm/ppc_asm.h. This caused that pahole cannot generate the BTF for these tcp-cc kernel functions because pahole only captures the FUNC type and "cubictcp_init" is not. It then failed the kernel compilation in ppc64. This behavior is only reported in ppc64 so far. I tried arm64, s390, and sparc64 and did not observe this "." prefix and NOTYPE behavior. Since the kfunc call is only supported in the x86_64 and x86_32 JIT, this patch limits those tcp-cc functions to x86 only to avoid unnecessary compilation issue in other ARCHs. In the future, we can examine if it is better to change all those functions from static to extern. [0] https://lore.kernel.org/bpf/4e051459-8532-7b61-c815-f3435767f8a0@kernel.org/ Fixes: e78aea8b ("bpf: tcp: Put some tcp cong functions in allowlist for bpf-tcp-cc") Signed-off-by:
Martin KaFai Lau <kafai@fb.com> Signed-off-by:
Daniel Borkmann <daniel@iogearbox.net> Cc: Michal Suchánek <msuchanek@suse.de> Cc: Jiri Slaby <jslaby@suse.com> Cc: Jiri Olsa <jolsa@redhat.com> Link: https://lore.kernel.org/bpf/20210508005011.3863757-1-kafai@fb.com
-
Jussi Maki authored
As discussed in [0], this ports test_tc_redirect.sh to the test_progs framework and removes the old test. This makes it more in line with rest of the tests and makes it possible to run this test case with vmtest.sh and under the bpf CI. The upcoming skb_change_head() helper fix in [0] is depending on it and extending the test case to redirect a packet from L3 device to veth. [0] https://lore.kernel.org/bpf/20210427135550.807355-1-joamaki@gmail.comSigned-off-by:
Jussi Maki <joamaki@gmail.com> Signed-off-by:
-
Arnaldo Carvalho de Melo authored
Where that macro isn't available. Signed-off-by:
Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by:
-
Florent Revest authored
The bpf_seq_printf, bpf_trace_printk and bpf_snprintf helpers share one per-cpu buffer that they use to store temporary data (arguments to bprintf). They "get" that buffer with try_get_fmt_tmp_buf and "put" it by the end of their scope with bpf_bprintf_cleanup. If one of these helpers gets called within the scope of one of these helpers, for example: a first bpf program gets called, uses bpf_trace_printk which calls raw_spin_lock_irqsave which is traced by another bpf program that calls bpf_snprintf, then the second "get" fails. Essentially, these helpers are not re-entrant. They would return -EBUSY and print a warning message once. This patch triples the number of bprintf buffers to allow three levels of nesting. This is very similar to what was done for tracepoints in "9594dc3c bpf: fix nested bpf tracepoints with per-cpu data" Fixes: d9c9e4db ("bpf: Factorize bpf_trace_printk and bpf_seq_printf") Reported-by: syzbot+63122d0bc347f18c1884@syzkaller.appspotmail.com Signed-off-by:
Florent Revest <revest@chromium.org> Signed-off-by:
Alexei Starovoitov <ast@kernel.org> Link: https://lore.kernel.org/bpf/20210511081054.2125874-1-revest@chromium.org
-
Jiri Olsa authored
The recursion check in __bpf_prog_enter and __bpf_prog_exit leaves some (not inlined) functions unprotected: In __bpf_prog_enter: - migrate_disable is called before prog->active is checked In __bpf_prog_exit: - migrate_enable,rcu_read_unlock_strict are called after prog->active is decreased When attaching trampoline to them we get panic like: traps: PANIC: double fault, error_code: 0x0 double fault: 0000 [#1] SMP PTI RIP: 0010:__bpf_prog_enter+0x4/0x50 ... Call Trace: <IRQ> bpf_trampoline_6442466513_0+0x18/0x1000 migrate_disable+0x5/0x50 __bpf_prog_enter+0x9/0x50 bpf_trampoline_6442466513_0+0x18/0x1000 migrate_disable+0x5/0x50 __bpf_prog_enter+0x9/0x50 bpf_trampoline_6442466513_0+0x18/0x1000 migrate_disable+0x5/0x50 __bpf_prog_enter+0x9/0x50 bpf_trampoline_6442466513_0+0x18/0x1000 migrate_disable+0x5/0x50 ... Fixing this by adding deny list of btf ids for tracing programs and checking btf id during program verification. Adding above functions to this list. Suggested-by:Jiri Olsa <jolsa@kernel.org> Signed-off-by:
-
Daniel Borkmann authored
Add a kconfig knob which allows for unprivileged bpf to be disabled by default. If set, the knob sets /proc/sys/kernel/unprivileged_bpf_disabled to value of 2. This still allows a transition of 2 -> {0,1} through an admin. Similarly, this also still keeps 1 -> {1} behavior intact, so that once set to permanently disabled, it cannot be undone aside from a reboot. We've also added extra2 with max of 2 for the procfs handler, so that an admin still has a chance to toggle between 0 <-> 2. Either way, as an additional alternative, applications can make use of CAP_BPF that we added a while ago. Signed-off-by: -
Daniel Borkmann authored
Right now, all core BPF related options are scattered in different Kconfig locations mainly due to historic reasons. Moving forward, lets add a proper subsystem entry under ... General setup ---> BPF subsystem ---> ... in order to have all knobs in a single location and thus ease BPF related configuration. Networking related bits such as sockmap are out of scope for the general setup and therefore better suited to remain in net/Kconfig. Signed-off-by: -
Sriram R authored
Fragmentation is used only with unicast frames. Drop multicast fragments to avoid any undesired behavior. Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01734-QCAHKSWPL_SILICONZ-1 v2 Cc: stable@vger.kernel.org Signed-off-by:
Sriram R <srirrama@codeaurora.org> Signed-off-by:
Jouni Malinen <jouni@codeaurora.org> Link: https://lore.kernel.org/r/20210511200110.1d53bfd20a8b.Ibb63283051bb5e2c45951932c6e1f351d5a73dc3@changeidSigned-off-by:
Johannes Berg <johannes.berg@intel.com>
-
Sriram R authored
Currently the fragment cache setup during peer assoc is cleared only during peer delete. In case a key reinstallation happens with the same peer, the same fragment cache with old fragments added before key installation could be clubbed with fragments received after. This might be exploited to mix fragments of different data resulting in a proper unintended reassembled packet to be passed up the stack. Hence flush the fragment cache on every key installation to prevent potential attacks (CVE-2020-24587). Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01734-QCAHKSWPL_SILICONZ-1 v2 Cc: stable@vger.kernel.org Signed-off-by:
-
Sriram R authored
In certain scenarios a normal MSDU can be received as an A-MSDU when the A-MSDU present bit of a QoS header gets flipped during reception. Since this bit is unauthenticated, the hardware crypto engine can pass the frame to the driver without any error indication. This could result in processing unintended subframes collected in the A-MSDU list. Hence, validate A-MSDU list by checking if the first frame has a valid subframe header. Comparing the non-aggregated MSDU and an A-MSDU, the fields of the first subframe DA matches the LLC/SNAP header fields of a normal MSDU. In order to avoid processing such frames, add a validation to filter such A-MSDU frames where the first subframe header DA matches with the LLC/SNAP header pattern. Tested-on: QCA9984 hw1.0 PCI 10.4-3.10-00047 Cc: stable@vger.kernel.org Signed-off-by:
-
Wen Gong authored
TKIP Michael MIC was not verified properly for PCIe cases since the validation steps in ieee80211_rx_h_michael_mic_verify() in mac80211 did not get fully executed due to unexpected flag values in ieee80211_rx_status. Fix this by setting the flags property to meet mac80211 expectations for performing Michael MIC validation there. This fixes CVE-2020-26141. It does the same as ath10k_htt_rx_proc_rx_ind_hl() for SDIO which passed MIC verification case. This applies only to QCA6174/QCA9377 PCIe. Tested-on: QCA6174 hw3.2 PCI WLAN.RM.4.4.1-00110-QCARMSWP-1 Cc: stable@vger.kernel.org Signed-off-by:
Wen Gong <wgong@codeaurora.org> Signed-off-by:
-
Wen Gong authored
When the discard flag is set by the firmware for an MPDU, it should be dropped. This allows a mitigation for CVE-2020-24588 to be implemented in the firmware. Tested-on: QCA6174 hw3.2 SDIO WLAN.RMH.4.4.1-00049 Cc: stable@vger.kernel.org Signed-off-by:
-
Wen Gong authored
Fragmentation is not used with multicast frames. Discard unexpected fragments with multicast DA. This fixes CVE-2020-26145. Tested-on: QCA6174 hw3.2 SDIO WLAN.RMH.4.4.1-00049 Cc: stable@vger.kernel.org Signed-off-by:
-
Wen Gong authored
Fragmentation is not used with multicast frames. Discard unexpected fragments with multicast DA. This fixes CVE-2020-26145. Tested-on: QCA6174 hw3.2 PCI WLAN.RM.4.4.1-00110-QCARMSWP-1 Cc: stable@vger.kernel.org Signed-off-by:
-
Wen Gong authored
PN replay check for not fragmented frames is finished in the firmware, but this was not done for fragmented frames when ath10k is used with QCA6174/QCA6377 PCIe. mac80211 has the function ieee80211_rx_h_defragment() for PN replay check for fragmented frames, but this does not get checked with QCA6174 due to the ieee80211_has_protected() condition not matching the cleared Protected bit case. Validate the PN of received fragmented frames within ath10k when CCMP is used and drop the fragment if the PN is not correct (incremented by exactly one from the previous fragment). This applies only for QCA6174/QCA6377 PCIe. Tested-on: QCA6174 hw3.2 PCI WLAN.RM.4.4.1-00110-QCARMSWP-1 Cc: stable@vger.kernel.org Signed-off-by:
-
Wen Gong authored
For some chips/drivers, e.g., QCA6174 with ath10k, the decryption is done by the hardware, and the Protected bit in the Frame Control field is cleared in the lower level driver before the frame is passed to mac80211. In such cases, the condition for ieee80211_has_protected() is not met in ieee80211_rx_h_defragment() of mac80211 and the new security validation steps are not executed. Extend mac80211 to cover the case where the Protected bit has been cleared, but the frame is indicated as having been decrypted by the hardware. This extends protection against mixed key and fragment cache attack for additional drivers/chips. This fixes CVE-2020-24586 and CVE-2020-24587 for such cases. Tested-on: QCA6174 hw3.2 PCI WLAN.RM.4.4.1-00110-QCARMSWP-1 Cc: stable@vger.kernel.org Signed-off-by:
-
Johannes Berg authored
EAPOL frames are used for authentication and key management between the AP and each individual STA associated in the BSS. Those frames are not supposed to be sent by one associated STA to another associated STA (either unicast for broadcast/multicast). Similarly, in 802.11 they're supposed to be sent to the authenticator (AP) address. Since it is possible for unexpected EAPOL frames to result in misbehavior in supplicant implementations, it is better for the AP to not allow such cases to be forwarded to other clients either directly, or indirectly if the AP interface is part of a bridge. Accept EAPOL (control port) frames only if they're transmitted to the own address, or, due to interoperability concerns, to the PAE group address. Disable forwarding of EAPOL (or well, the configured control port protocol) frames back to wireless medium in all cases. Previously, these frames were accepted from fully authenticated and authorized stations and also from unauthenticated stations for one of the cases. Additionally, to avoid forwarding by the bridge, rewrite the PAE group address case to the local MAC address. Cc: stable@vger.kernel.org Co-developed-by:
-
Johannes Berg authored
Similar to the issues fixed in previous patches, TKIP and WEP should be protected even if for TKIP we have the Michael MIC protecting it, and WEP is broken anyway. However, this also somewhat protects potential other algorithms that drivers might implement. Cc: stable@vger.kernel.org Link: https://lore.kernel.org/r/20210511200110.430e8c202313.Ia37e4e5b6b3eaab1a5ae050e015f6c92859dbe27@changeidSigned-off-by:
-
Johannes Berg authored
As pointed out by Mathy Vanhoef, we implement the RX PN check on fragmented frames incorrectly - we check against the last received PN prior to the new frame, rather than to the one in this frame itself. Prior patches addressed the security issue here, but in order to be able to reason better about the code, fix it to really compare against the current frame's PN, not the last stored one. Cc: stable@vger.kernel.org Link: https://lore.kernel.org/r/20210511200110.bfbc340ff071.Id0b690e581da7d03d76df90bb0e3fd55930bc8a0@changeidSigned-off-by:
-
Johannes Berg authored
Prior patches protected against fragmentation cache attacks by coloring keys, but this shows that it can lead to issues when multiple stations use the same sequence number. Add a fragment cache to struct sta_info (in addition to the one in the interface) to separate fragments for different stations properly. This then automatically clear most of the fragment cache when a station disconnects (or reassociates) from an AP, or when client interfaces disconnect from the network, etc. On the way, also fix the comment there since this brings us in line with the recommendation in 802.11-2016 ("An AP should support ..."). Additionally, remove a useless condition (since there's no problem purging an already empty list). Cc: stable@vger.kernel.org Link: https://lore.kernel.org/r/20210511200110.fc35046b0d52.I1ef101e3784d13e8f6600d83de7ec9a3a45bcd52@changeidSigned-off-by: -
Johannes Berg authored
With old ciphers (WEP and TKIP) we shouldn't be using A-MSDUs since A-MSDUs are only supported if we know that they are, and the only practical way for that is HT support which doesn't support old ciphers. However, we would normally accept them anyway. Since we check the MMIC before deaggregating A-MSDUs, and the A-MSDU bit in the QoS header is not protected in TKIP (or WEP), this enables attacks similar to CVE-2020-24588. To prevent that, drop A-MSDUs completely with old ciphers. Cc: stable@vger.kernel.org Link: https://lore.kernel.org/r/20210511200110.076543300172.I548e6e71f1ee9cad4b9a37bf212ae7db723587aa@changeidSigned-off-by:
-
Mathy Vanhoef authored
Mitigate A-MSDU injection attacks (CVE-2020-24588) by detecting if the destination address of a subframe equals an RFC1042 (i.e., LLC/SNAP) header, and if so dropping the complete A-MSDU frame. This mitigates known attacks, although new (unknown) aggregation-based attacks may remain possible. This defense works because in A-MSDU aggregation injection attacks, a normal encrypted Wi-Fi frame is turned into an A-MSDU frame. This means the first 6 bytes of the first A-MSDU subframe correspond to an RFC1042 header. In other words, the destination MAC address of the first A-MSDU subframe contains the start of an RFC1042 header during an aggregation attack. We can detect this and thereby prevent this specific attack. For details, see Section 7.2 of "Fragment and Forge: Breaking Wi-Fi Through Frame Aggregation and Fragmentation". Note that for kernel 4.9 and above this patch depends on "mac80211: properly handle A-MSDUs that start with a rfc1042 header". Otherwise this patch has no impact and attacks will remain possible. Cc: stable@vger.kernel.org Signed-off-by:
Mathy Vanhoef <Mathy.Vanhoef@kuleuven.be> Link: https://lore.kernel.org/r/20210511200110.25d93176ddaf.I9e265b597f2cd23eb44573f35b625947b386a9de@changeidSigned-off-by:
-
Mathy Vanhoef authored
Properly parse A-MSDUs whose first 6 bytes happen to equal a rfc1042 header. This can occur in practice when the destination MAC address equals AA:AA:03:00:00:00. More importantly, this simplifies the next patch to mitigate A-MSDU injection attacks. Cc: stable@vger.kernel.org Signed-off-by:
-
Mathy Vanhoef authored
Simultaneously prevent mixed key attacks (CVE-2020-24587) and fragment cache attacks (CVE-2020-24586). This is accomplished by assigning a unique color to every key (per interface) and using this to track which key was used to decrypt a fragment. When reassembling frames, it is now checked whether all fragments were decrypted using the same key. To assure that fragment cache attacks are also prevented, the ID that is assigned to keys is unique even over (re)associations and (re)connects. This means fragments separated by a (re)association or (re)connect will not be reassembled. Because mac80211 now also prevents the reassembly of mixed encrypted and plaintext fragments, all cache attacks are prevented. Cc: stable@vger.kernel.org Signed-off-by:
-
Mathy Vanhoef authored
Do not mix plaintext and encrypted fragments in protected Wi-Fi networks. This fixes CVE-2020-26147. Previously, an attacker was able to first forward a legitimate encrypted fragment towards a victim, followed by a plaintext fragment. The encrypted and plaintext fragment would then be reassembled. For further details see Section 6.3 and Appendix D in the paper "Fragment and Forge: Breaking Wi-Fi Through Frame Aggregation and Fragmentation". Because of this change there are now two equivalent conditions in the code to determine if a received fragment requires sequential PNs, so we also move this test to a separate function to make the code easier to maintain. Cc: stable@vger.kernel.org Signed-off-by:
-
Andrii Nakryiko authored
Only the very first page of BPF ringbuf that contains consumer position counter is supposed to be mapped as writeable by user-space. Producer position is read-only and can be modified only by the kernel code. BPF ringbuf data pages are read-only as well and are not meant to be modified by user-code to maintain integrity of per-record headers. This patch allows to map only consumer position page as writeable and everything else is restricted to be read-only. remap_vmalloc_range() internally adds VM_DONTEXPAND, so all the established memory mappings can't be extended, which prevents any future violations through mremap()'ing. Fixes: 457f4436 ("bpf: Implement BPF ring buffer and verifier support for it") Reported-by: Ryota Shiga (Flatt Security) Reported-by:
Thadeu Lima de Souza Cascardo <cascardo@canonical.com> Signed-off-by:
Andrii Nakryiko <andrii@kernel.org> Signed-off-by:
-
Thadeu Lima de Souza Cascardo authored
A BPF program might try to reserve a buffer larger than the ringbuf size. If the consumer pointer is way ahead of the producer, that would be successfully reserved, allowing the BPF program to read or write out of the ringbuf allocated area. Reported-by: Ryota Shiga (Flatt Security) Fixes: 457f4436 ("bpf: Implement BPF ring buffer and verifier support for it") Signed-off-by:
-
Daniel Borkmann authored
Fix a bug in the verifier's scalar32_min_max_*() functions which leads to incorrect tracking of 32 bit bounds for the simulation of and/or/xor bitops. When both the src & dst subreg is a known constant, then the assumption is that scalar_min_max_*() will take care to update bounds correctly. However, this is not the case, for example, consider a register R2 which has a tnum of 0xffffffff00000000, meaning, lower 32 bits are known constant and in this case of value 0x00000001. R2 is then and'ed with a register R3 which is a 64 bit known constant, here, 0x100000002. What can be seen in line '10:' is that 32 bit bounds reach an invalid state where {u,s}32_min_value > {u,s}32_max_value. The reason is scalar32_min_max_*() delegates 32 bit bounds updates to scalar_min_max_*(), however, that really only takes place when both the 64 bit src & dst register is a known constant. Given scalar32_min_max_*() is intended to be designed as closely as possible to scalar_min_max_*(), update the 32 bit bounds in this situation through __mark_reg32_known() which will set all {u,s}32_{min,max}_value to the correct constant, which is 0x00000000 after the fix (given 0x00000001 & 0x00000002 in 32 bit space). This is possible given var32_off already holds the final value as dst_reg->var_off is updated before calling scalar32_min_max_*(). Before fix, invalid tracking of R2: [...] 9: R0_w=inv1337 R1=ctx(id=0,off=0,imm=0) R2_w=inv(id=0,smin_value=-9223372036854775807 (0x8000000000000001),smax_value=9223372032559808513 (0x7fffffff00000001),umin_value=1,umax_value=0xffffffff00000001,var_off=(0x1; 0xffffffff00000000),s32_min_value=1,s32_max_value=1,u32_min_value=1,u32_max_value=1) R3_w=inv4294967298 R10=fp0 9: (5f) r2 &= r3 10: R0_w=inv1337 R1=ctx(id=0,off=0,imm=0) R2_w=inv(id=0,smin_value=0,smax_value=4294967296 (0x100000000),umin_value=0,umax_value=0x100000000,var_off=(0x0; 0x100000000),s32_min_value=1,s32_max_value=0,u32_min_value=1,u32_max_value=0) R3_w=inv4294967298 R10=fp0 [...] After fix, correct tracking of R2: [...] 9: R0_w=inv1337 R1=ctx(id=0,off=0,imm=0) R2_w=inv(id=0,smin_value=-9223372036854775807 (0x8000000000000001),smax_value=9223372032559808513 (0x7fffffff00000001),umin_value=1,umax_value=0xffffffff00000001,var_off=(0x1; 0xffffffff00000000),s32_min_value=1,s32_max_value=1,u32_min_value=1,u32_max_value=1) R3_w=inv4294967298 R10=fp0 9: (5f) r2 &= r3 10: R0_w=inv1337 R1=ctx(id=0,off=0,imm=0) R2_w=inv(id=0,smin_value=0,smax_value=4294967296 (0x100000000),umin_value=0,umax_value=0x100000000,var_off=(0x0; 0x100000000),s32_min_value=0,s32_max_value=0,u32_min_value=0,u32_max_value=0) R3_w=inv4294967298 R10=fp0 [...] Fixes: 3f50f132 ("bpf: Verifier, do explicit ALU32 bounds tracking") Fixes: 2921c90d ("bpf: Fix a verifier failure with xor") Reported-by: Manfred Paul (@_manfp) Reported-by:John Fastabend <john.fastabend@gmail.com> Acked-by:
-
- 10 May, 2021 5 commits
-
-
Michael Walle authored
Commit 316bcffe ("net: dsa: felix: disable always guard band bit for TAS config") disabled the guard band and broke 802.3Qbv compliance. There are two issues here: (1) Without the guard band the end of the scheduling window could be overrun by a frame in transit. (2) Frames that don't fit into a configured window will still be sent. The reason for both issues is that the switch will schedule the _start_ of a frame transmission inside the predefined window without taking the length of the frame into account. Thus, we'll need the guard band which will close the gate early, so that a complete frame can still be sent. Revert the commit and add a note. For a lengthy discussion see [1]. [1] https://lore.kernel.org/netdev/c7618025da6723418c56a54fe4683bd7@walle.cc/ Fixes: 316bcffe ("net: dsa: felix: disable always guard band bit for TAS config") Signed-off-by:
Michael Walle <michael@walle.cc> Reviewed-by:
Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by:
-
Hoang Le authored
The using of the node address and node link identity are not thread safe, meaning that two publications may be published the same values, as result one of them will get failure because of already existing in the name table. To avoid this we have to use the node address and node link identity values from inside the node item's write lock protection. Fixes: 50a3499a ("tipc: simplify signature of tipc_namtbl_publish()") Acked-by:
Jon Maloy <jmaloy@redhat.com> Signed-off-by:
Hoang Le <hoang.h.le@dektech.com.au> Signed-off-by:
-
Vladimir Oltean authored
DSA implements a bunch of 'standardized' ethtool statistics counters, namely tx_packets, tx_bytes, rx_packets, rx_bytes. So whatever the hardware driver returns in .get_sset_count(), we need to add 4 to that. That is ok, except that .get_sset_count() can return a negative error code, for example: b53_get_sset_count -> phy_ethtool_get_sset_count -> return -EIO -EIO is -5, and with 4 added to it, it becomes -1, aka -EPERM. One can imagine that certain error codes may even become positive, although based on code inspection I did not see instances of that. Check the error code first, if it is negative return it as-is. Based on a similar patch for dsa_master_get_strings from Dan Carpenter: https://patchwork.kernel.org/project/netdevbpf/patch/YJaSe3RPgn7gKxZv@mwanda/ Fixes: 91da11f8 ("net: Distributed Switch Architecture protocol support") Signed-off-by:
Florian Fainelli <f.fainelli@gmail.com> Reviewed-by:
Andrew Lunn <andrew@lunn.ch> Signed-off-by:
-
Vladyslav Tarasiuk authored
Fix SFP and QSFP* EEPROM queries by setting i2c_address, offset and page number correctly. For SFP set the following params: - I2C address for offsets 0-255 is 0x50. For 256-511 - 0x51. - Page number is zero. - Offset is 0-255. At the same time, QSFP* parameters are different: - I2C address is always 0x50. - Page number is not limited to zero. - Offset is 0-255 for page zero and 128-255 for others. To set parameters accordingly to cable used, implement function to query module ID and implement respective helper functions to set parameters correctly. Fixes: 135dd959 ("net/mlx4_en: ethtool, Remove unsupported SFP EEPROM high pages query") Signed-off-by:
Vladyslav Tarasiuk <vladyslavt@nvidia.com> Signed-off-by:
Tariq Toukan <tariqt@nvidia.com> Signed-off-by:
-
Dan Carpenter authored
If ds->ops->get_sset_count() fails then it "count" is a negative error code such as -EOPNOTSUPP. Because "i" is an unsigned int, the negative error code is type promoted to a very high value and the loop will corrupt memory until the system crashes. Fix this by checking for error codes and changing the type of "i" to just int. Fixes: badf3ada ("net: dsa: Provide CPU port statistics to master netdev") Signed-off-by:
Dan Carpenter <dan.carpenter@oracle.com> Reviewed-by:
Vladimir Oltean <olteanv@gmail.com> Signed-off-by:
-
