| CVE |
Vendors |
Products |
Updated |
CVSS v3.1 |
| In the Linux kernel, the following vulnerability has been resolved:
ionic: use dev_consume_skb_any outside of napi
If we're not in a NAPI softirq context, we need to be careful
about how we call napi_consume_skb(), specifically we need to
call it with budget==0 to signal to it that we're not in a
safe context.
This was found while running some configuration stress testing
of traffic and a change queue config loop running, and this
curious note popped out:
[ 4371.402645] BUG: using smp_processor_id() in preemptible [00000000] code: ethtool/20545
[ 4371.402897] caller is napi_skb_cache_put+0x16/0x80
[ 4371.403120] CPU: 25 PID: 20545 Comm: ethtool Kdump: loaded Tainted: G OE 6.10.0-rc3-netnext+ #8
[ 4371.403302] Hardware name: HPE ProLiant DL360 Gen10/ProLiant DL360 Gen10, BIOS U32 01/23/2021
[ 4371.403460] Call Trace:
[ 4371.403613] <TASK>
[ 4371.403758] dump_stack_lvl+0x4f/0x70
[ 4371.403904] check_preemption_disabled+0xc1/0xe0
[ 4371.404051] napi_skb_cache_put+0x16/0x80
[ 4371.404199] ionic_tx_clean+0x18a/0x240 [ionic]
[ 4371.404354] ionic_tx_cq_service+0xc4/0x200 [ionic]
[ 4371.404505] ionic_tx_flush+0x15/0x70 [ionic]
[ 4371.404653] ? ionic_lif_qcq_deinit.isra.23+0x5b/0x70 [ionic]
[ 4371.404805] ionic_txrx_deinit+0x71/0x190 [ionic]
[ 4371.404956] ionic_reconfigure_queues+0x5f5/0xff0 [ionic]
[ 4371.405111] ionic_set_ringparam+0x2e8/0x3e0 [ionic]
[ 4371.405265] ethnl_set_rings+0x1f1/0x300
[ 4371.405418] ethnl_default_set_doit+0xbb/0x160
[ 4371.405571] genl_family_rcv_msg_doit+0xff/0x130
[...]
I found that ionic_tx_clean() calls napi_consume_skb() which calls
napi_skb_cache_put(), but before that last call is the note
/* Zero budget indicate non-NAPI context called us, like netpoll */
and
DEBUG_NET_WARN_ON_ONCE(!in_softirq());
Those are pretty big hints that we're doing it wrong. We can pass a
context hint down through the calls to let ionic_tx_clean() know what
we're doing so it can call napi_consume_skb() correctly. |
| In the Linux kernel, the following vulnerability has been resolved:
bpf: Add schedule points in batch ops
syzbot reported various soft lockups caused by bpf batch operations.
INFO: task kworker/1:1:27 blocked for more than 140 seconds.
INFO: task hung in rcu_barrier
Nothing prevents batch ops to process huge amount of data,
we need to add schedule points in them.
Note that maybe_wait_bpf_programs(map) calls from
generic_map_delete_batch() can be factorized by moving
the call after the loop.
This will be done later in -next tree once we get this fix merged,
unless there is strong opinion doing this optimization sooner. |
| JOSE is "JSON Web Almost Everything" - JWA, JWS, JWE, JWT, JWK, JWKS with no dependencies using runtime's native crypto in Node.js, Browser, Cloudflare Workers, Electron, and Deno. The PBKDF2-based JWE key management algorithms expect a JOSE Header Parameter named `p2c` PBES2 Count, which determines how many PBKDF2 iterations must be executed in order to derive a CEK wrapping key. The purpose of this parameter is to intentionally slow down the key derivation function in order to make password brute-force and dictionary attacks more expensive. This makes the PBES2 algorithms unsuitable for situations where the JWE is coming from an untrusted source: an adversary can intentionally pick an extremely high PBES2 Count value, that will initiate a CPU-bound computation that may take an unreasonable amount of time to finish. Under certain conditions, it is possible to have the user's environment consume unreasonable amount of CPU time. The impact is limited only to users utilizing the JWE decryption APIs with symmetric secrets to decrypt JWEs from untrusted parties who do not limit the accepted JWE Key Management Algorithms (`alg` Header Parameter) using the `keyManagementAlgorithms` (or `algorithms` in v1.x) decryption option or through other means. The `v1.28.2`, `v2.0.6`, `v3.20.4`, and `v4.9.2` releases limit the maximum PBKDF2 iteration count to `10000` by default. It is possible to adjust this limit with a newly introduced `maxPBES2Count` decryption option. If users are unable to upgrade their required library version, they have two options depending on whether they expect to receive JWEs using any of the three PBKDF2-based JWE key management algorithms. They can use the `keyManagementAlgorithms` decryption option to disable accepting PBKDF2 altogether, or they can inspect the JOSE Header prior to using the decryption API and limit the PBKDF2 iteration count (`p2c` Header Parameter). |
| The mp4ff_read_stts function in common/mp4ff/mp4atom.c in Freeware Advanced Audio Decoder 2 (FAAD2) 2.7 allows remote attackers to cause a denial of service (large loop and CPU consumption) via a crafted mp4 file. |
| The mp4ff_read_stsc function in common/mp4ff/mp4atom.c in Freeware Advanced Audio Decoder 2 (FAAD2) 2.7 allows remote attackers to cause a denial of service (large loop and CPU consumption) via a crafted mp4 file. |
| The mp4ff_read_stco function in common/mp4ff/mp4atom.c in Freeware Advanced Audio Decoder 2 (FAAD2) 2.7 allows remote attackers to cause a denial of service (large loop and CPU consumption) via a crafted mp4 file. |
| GraphicsMagick 1.3.26 has a denial of service issue in ReadXBMImage() in a coders/xbm.c "Read hex image data" version!=10 case that results in the reader not returning; it would cause large amounts of CPU and memory consumption although the crafted file itself does not request it. |
| GraphicsMagick 1.3.26 has a denial of service issue in ReadXBMImage() in a coders/xbm.c "Read hex image data" version==10 case that results in the reader not returning; it would cause large amounts of CPU and memory consumption although the crafted file itself does not request it. |
| The ReadRLEImage function in coders\rle.c in ImageMagick 7.0.6-1 has a large loop vulnerability via a crafted rle file that triggers a huge number_pixels value. |
| In Wireshark 2.0.0 to 2.0.13, the GPRS LLC dissector could go into a large loop. This was addressed in epan/dissectors/packet-gprs-llc.c by using a different integer data type. |
| The mp4ff_read_ctts function in common/mp4ff/mp4atom.c in Freeware Advanced Audio Decoder 2 (FAAD2) 2.7 allows remote attackers to cause a denial of service (large loop and CPU consumption) via a crafted mp4 file. |
| The mp4ff_read_stsd function in common/mp4ff/mp4atom.c in Freeware Advanced Audio Decoder 2 (FAAD2) 2.7 allows remote attackers to cause a denial of service (large loop and CPU consumption) via a crafted mp4 file. |
| In libavformat/rl2.c in FFmpeg 3.3.3, a DoS in rl2_read_header() due to lack of an EOF (End of File) check might cause huge CPU and memory consumption. When a crafted RL2 file, which claims a large "frame_count" field in the header but does not contain sufficient backing data, is provided, the loops (for offset and size tables) would consume huge CPU and memory resources, since there is no EOF check inside these loops. |
| In FFmpeg 3.3.3, a DoS in asf_read_marker() due to lack of an EOF (End of File) check might cause huge CPU and memory consumption. When a crafted ASF file, which claims a large "name_len" or "count" field in the header but does not contain sufficient backing data, is provided, the loops over the name and markers would consume huge CPU and memory resources, since there is no EOF check inside these loops. |
| In FFmpeg 3.3.3, a DoS in cine_read_header() due to lack of an EOF check might cause huge CPU and memory consumption. When a crafted CINE file, which claims a large "duration" field in the header but does not contain sufficient backing data, is provided, the image-offset parsing loop would consume huge CPU and memory resources, since there is no EOF check inside the loop. |
| In libavformat/mxfdec.c in FFmpeg 3.3.3 -> 2.4, a DoS in mxf_read_index_entry_array() due to lack of an EOF (End of File) check might cause huge CPU consumption. When a crafted MXF file, which claims a large "nb_index_entries" field in the header but does not contain sufficient backing data, is provided, the loop would consume huge CPU resources, since there is no EOF check inside the loop. Moreover, this big loop can be invoked multiple times if there is more than one applicable data segment in the crafted MXF file. |
| In libavformat/nsvdec.c in FFmpeg 2.4 and 3.3.3, a DoS in nsv_parse_NSVf_header() due to lack of an EOF (End of File) check might cause huge CPU consumption. When a crafted NSV file, which claims a large "table_entries_used" field in the header but does not contain sufficient backing data, is provided, the loop over 'table_entries_used' would consume huge CPU resources, since there is no EOF check inside the loop. |
| In coders/ps.c in ImageMagick 7.0.7-0 Q16, a DoS in ReadPSImage() due to lack of an EOF (End of File) check might cause huge CPU consumption. When a crafted PSD file, which claims a large "extent" field in the header but does not contain sufficient backing data, is provided, the loop over "length" would consume huge CPU resources, since there is no EOF check inside the loop. |
| In coders/psd.c in ImageMagick 7.0.7-0 Q16, a DoS in ReadPSDLayersInternal() due to lack of an EOF (End of File) check might cause huge CPU consumption. When a crafted PSD file, which claims a large "length" field in the header but does not contain sufficient backing data, is provided, the loop over "length" would consume huge CPU resources, since there is no EOF check inside the loop. |
| In coders/xbm.c in ImageMagick 7.0.6-1 Q16, a DoS in ReadXBMImage() due to lack of an EOF (End of File) check might cause huge CPU consumption. When a crafted XBM file, which claims large rows and columns fields in the header but does not contain sufficient backing data, is provided, the loop over the rows would consume huge CPU resources, since there is no EOF check inside the loop. |
