Instead of changing write_pnet() and read_pnet() and potentially
hurt performance, add the needed READ_ONCE()/WRITE_ONCE()
in tm_net() and tcpm_new().
Fixes: 849e8a0ca8d5 ("tcp_metrics: Add a field tcpm_net and verify it matches on lookup") Signed-off-by: Eric Dumazet <edumazet@google.com> Reviewed-by: David Ahern <dsahern@kernel.org> Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com> Link: https://lore.kernel.org/r/20230802131500.1478140-6-edumazet@google.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
Because v4 and v6 families use separate inetpeer trees (respectively
net->ipv4.peers and net->ipv6.peers), inetpeer_addr_cmp(a, b) assumes
a & b share the same family.
tcp_metrics use a common hash table, where entries can have different
families.
We must therefore make sure to not call inetpeer_addr_cmp()
if the families do not match.
Fixes: d39d14ffa24c ("net: Add helper function to compare inetpeer addresses") Signed-off-by: Eric Dumazet <edumazet@google.com> Reviewed-by: David Ahern <dsahern@kernel.org> Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com> Link: https://lore.kernel.org/r/20230802131500.1478140-2-edumazet@google.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
When both supported and previous version have the same major version,
and the firmwares are missing, the driver ends in a loop requesting the
same (previous) version over and over again:
[ 76.327413] Prestera DX 0000:01:00.0: missing latest mrvl/prestera/mvsw_prestera_fw-v4.1.img firmware, fall-back to previous 4.0 version
[ 76.339802] Prestera DX 0000:01:00.0: missing latest mrvl/prestera/mvsw_prestera_fw-v4.0.img firmware, fall-back to previous 4.0 version
[ 76.352162] Prestera DX 0000:01:00.0: missing latest mrvl/prestera/mvsw_prestera_fw-v4.0.img firmware, fall-back to previous 4.0 version
[ 76.364502] Prestera DX 0000:01:00.0: missing latest mrvl/prestera/mvsw_prestera_fw-v4.0.img firmware, fall-back to previous 4.0 version
[ 76.376848] Prestera DX 0000:01:00.0: missing latest mrvl/prestera/mvsw_prestera_fw-v4.0.img firmware, fall-back to previous 4.0 version
[ 76.389183] Prestera DX 0000:01:00.0: missing latest mrvl/prestera/mvsw_prestera_fw-v4.0.img firmware, fall-back to previous 4.0 version
[ 76.401522] Prestera DX 0000:01:00.0: missing latest mrvl/prestera/mvsw_prestera_fw-v4.0.img firmware, fall-back to previous 4.0 version
[ 76.413860] Prestera DX 0000:01:00.0: missing latest mrvl/prestera/mvsw_prestera_fw-v4.0.img firmware, fall-back to previous 4.0 version
[ 76.426199] Prestera DX 0000:01:00.0: missing latest mrvl/prestera/mvsw_prestera_fw-v4.0.img firmware, fall-back to previous 4.0 version
...
Fix this by inverting the check to that we aren't yet at the previous
version, and also check the minor version.
This also catches the case where both versions are the same, as it was
after commit bb5dbf2cc64d ("net: marvell: prestera: add firmware v4.0
support").
With this fix applied:
[ 88.499622] Prestera DX 0000:01:00.0: missing latest mrvl/prestera/mvsw_prestera_fw-v4.1.img firmware, fall-back to previous 4.0 version
[ 88.511995] Prestera DX 0000:01:00.0: failed to request previous firmware: mrvl/prestera/mvsw_prestera_fw-v4.0.img
[ 88.522403] Prestera DX: probe of 0000:01:00.0 failed with error -2
In the cited commit, new type of FS_TYPE_PRIO_CHAINS fs_prio was added
to support multiple parallel namespaces for multi-chains. And we skip
all the flow tables under the fs_node of this type unconditionally,
when searching for the next or previous flow table to connect for a
new table.
As this search function is also used for find new root table when the
old one is being deleted, it will skip the entire FS_TYPE_PRIO_CHAINS
fs_node next to the old root. However, new root table should be chosen
from it if there is any table in it. Fix it by skipping only the flow
tables in the same FS_TYPE_PRIO_CHAINS fs_node when finding the
closest FT for a fs_node.
Besides, complete the connecting from FTs of previous priority of prio
because there should be multiple prevs after this fs_prio type is
introduced. And also the next FT should be chosen from the first flow
table next to the prio in the same FS_TYPE_PRIO_CHAINS fs_prio, if
this prio is the first child.
Fixes: 328edb499f99 ("net/mlx5: Split FDB fast path prio to multiple namespaces") Signed-off-by: Jianbo Liu <jianbol@nvidia.com> Reviewed-by: Paul Blakey <paulb@nvidia.com> Signed-off-by: Leon Romanovsky <leonro@nvidia.com> Link: https://lore.kernel.org/r/7a95754df479e722038996c97c97b062b372591f.1690803944.git.leonro@nvidia.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
As find_closest_ft_recursive is called to find the closest FT, the
first parameter of find_closest_ft can be changed from fs_prio to
fs_node. Thus this function is extended to find the closest FT for the
nodes of any type, not only prios, but also the sub namespaces.
The nexthop code expects a 31 bit hash, such as what is returned by
fib_multipath_hash() and rt6_multipath_hash(). Passing the 32 bit hash
returned by skb_get_hash() can lead to problems related to the fact that
'int hash' is a negative number when the MSB is set.
In the case of hash threshold nexthop groups, nexthop_select_path_hthr()
will disproportionately select the first nexthop group entry. In the case
of resilient nexthop groups, nexthop_select_path_res() may do an out of
bounds access in nh_buckets[], for example:
hash = -912054133
num_nh_buckets = 2
bucket_index = 65535
Fix this problem by ensuring the MSB of hash is 0 using a right shift - the
same approach used in fib_multipath_hash() and rt6_multipath_hash().
Fixes: 1274e1cc4226 ("vxlan: ecmp support for mac fdb entries") Signed-off-by: Benjamin Poirier <bpoirier@nvidia.com> Reviewed-by: Ido Schimmel <idosch@nvidia.com> Reviewed-by: Simon Horman <horms@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
When setup a vlan device on dev pim6reg, DAD ns packet may sent on reg_vif_xmit().
reg_vif_xmit()
ip6mr_cache_report()
skb_push(skb, -skb_network_offset(pkt));//skb_network_offset(pkt) is 4
And skb_push declared as:
void *skb_push(struct sk_buff *skb, unsigned int len);
skb->data -= len;
//0xffff88805f86a84c - 0xfffffffc = 0xffff887f5f86a850
skb->data is set to 0xffff887f5f86a850, which is invalid mem addr, lead to skb_push() fails.
Fixes: 14fb64e1f449 ("[IPV6] MROUTE: Support PIM-SM (SSM).") Signed-off-by: Yue Haibing <yuehaibing@huawei.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
dev_close() and dev_open() are issued to change the interface state to DOWN
or UP (dev->flags IFF_UP). When the netdev is set DOWN it loses e.g its
Ipv6 addresses and routes. We don't want this in cases of device recovery
(triggered by hardware or software) or when the qeth device is set
offline.
Setting a qeth device offline or online and device recovery actions call
netif_device_detach() and/or netif_device_attach(). That will reset or
set the LOWER_UP indication i.e. change the dev->state Bit
__LINK_STATE_PRESENT. That is enough to e.g. cause bond failovers, and
still preserves the interface settings that are handled by the network
stack.
Don't call dev_open() nor dev_close() from the qeth device driver. Let the
network stack handle this.
Fixes: d4560150cb47 ("s390/qeth: call dev_close() during recovery") Signed-off-by: Alexandra Winter <wintera@linux.ibm.com> Reviewed-by: Wenjia Zhang <wenjia@linux.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
The dcbnl_bcn_setcfg uses erroneous policy to parse tb[DCB_ATTR_BCN],
which is introduced in commit 859ee3c43812 ("DCB: Add support for DCB
BCN"). Please see the comment in below code
static int dcbnl_bcn_setcfg(...)
{
...
ret = nla_parse_nested_deprecated(..., dcbnl_pfc_up_nest, .. )
// !!! dcbnl_pfc_up_nest for attributes
// DCB_PFC_UP_ATTR_0 to DCB_PFC_UP_ATTR_ALL in enum dcbnl_pfc_up_attrs
...
for (i = DCB_BCN_ATTR_RP_0; i <= DCB_BCN_ATTR_RP_7; i++) {
// !!! DCB_BCN_ATTR_RP_0 to DCB_BCN_ATTR_RP_7 in enum dcbnl_bcn_attrs
...
value_byte = nla_get_u8(data[i]);
...
}
...
for (i = DCB_BCN_ATTR_BCNA_0; i <= DCB_BCN_ATTR_RI; i++) {
// !!! DCB_BCN_ATTR_BCNA_0 to DCB_BCN_ATTR_RI in enum dcbnl_bcn_attrs
...
value_int = nla_get_u32(data[i]);
...
}
...
}
That is, the nla_parse_nested_deprecated uses dcbnl_pfc_up_nest
attributes to parse nlattr defined in dcbnl_pfc_up_attrs. But the
following access code fetch each nlattr as dcbnl_bcn_attrs attributes.
By looking up the associated nla_policy for dcbnl_bcn_attrs. We can find
the beginning part of these two policies are "same".
Therefore, the current code is buggy and this
nla_parse_nested_deprecated could overflow the dcbnl_pfc_up_nest and use
the adjacent nla_policy to parse attributes from DCB_BCN_ATTR_BCNA_0.
Hence use the correct policy dcbnl_bcn_nest to parse the nested
tb[DCB_ATTR_BCN] TLV.
Fixes: 859ee3c43812 ("DCB: Add support for DCB BCN") Signed-off-by: Lin Ma <linma@zju.edu.cn> Reviewed-by: Simon Horman <horms@kernel.org> Link: https://lore.kernel.org/r/20230801013248.87240-1-linma@zju.edu.cn Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
The existing code does not allow the MTU to be set to the maximum even
after an XDP program supporting multiple buffers is attached. Fix it
to set the netdev->max_mtu to the maximum value if the attached XDP
program supports mutiple buffers, regardless of the current MTU value.
Also use a local variable dev instead of repeatedly using bp->dev.
Fixes: 1dc4c557bfed ("bnxt: adding bnxt_xdp_build_skb to build skb from multibuffer xdp_buff") Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com> Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Reviewed-by: Andy Gospodarek <andrew.gospodarek@broadcom.com> Signed-off-by: Michael Chan <michael.chan@broadcom.com> Link: https://lore.kernel.org/r/20230731142043.58855-3-michael.chan@broadcom.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
The RXBD length field on all bnxt chips is 16-bit and so we cannot
support a full page when the native page size is 64K or greater.
The non-XDP (non page pool) code path has logic to handle this but
the XDP page pool code path does not handle this. Add the missing
logic to use page_pool_dev_alloc_frag() to allocate 32K chunks if
the page size is 64K or greater.
As documented in acd7aaf51b20 ("netsec: ignore 'phy-mode' device
property on ACPI systems") the SocioNext SynQuacer platform ships with
firmware defining the PHY mode as RGMII even though the physical
configuration of the PHY is for TX and RX delays. Since bbc4d71d63549bc
("net: phy: realtek: fix rtl8211e rx/tx delay config") this has caused
misconfiguration of the PHY, rendering the network unusable.
This was worked around for ACPI by ignoring the phy-mode property but
the system is also used with DT. For DT instead if we're running on a
SynQuacer force a working PHY mode, as well as the standard EDK2
firmware with DT there are also some of these systems that use u-boot
and might not initialise the PHY if not netbooting. Newer firmware
imagaes for at least EDK2 are available from Linaro so print a warning
when doing this.
Fixes: 533dd11a12f6 ("net: socionext: Add Synquacer NetSec driver") Signed-off-by: Mark Brown <broonie@kernel.org> Acked-by: Ard Biesheuvel <ardb@kernel.org> Acked-by: Ilias Apalodimas <ilias.apalodimas@linaro.org> Reviewed-by: Andrew Lunn <andrew@lunn.ch> Link: https://lore.kernel.org/r/20230731-synquacer-net-v3-1-944be5f06428@kernel.org Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
in korina_probe(), the return value of clk_prepare_enable()
should be checked since it might fail. we can use
devm_clk_get_optional_enabled() instead of devm_clk_get_optional()
and clk_prepare_enable() to automatically handle the error.
Fixes: e4cd854ec487 ("net: korina: Get mdio input clock via common clock framework") Signed-off-by: Yuanjun Gong <ruc_gongyuanjun@163.com> Link: https://lore.kernel.org/r/20230731090535.21416-1-ruc_gongyuanjun@163.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
Most kernel functions return negative error codes but some irq functions
return zero on error. In this code irq_of_parse_and_map(), returns zero
and platform_get_irq() returns negative error codes. We need to handle
both cases appropriately.
Fixes: 8425c41d1ef7 ("net: ll_temac: Extend support to non-device-tree platforms") Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org> Acked-by: Esben Haabendal <esben@geanix.com> Reviewed-by: Yang Yingliang <yangyingliang@huawei.com> Reviewed-by: Harini Katakam <harini.katakam@amd.com> Link: https://lore.kernel.org/r/3d0aef75-06e0-45a5-a2a6-2cc4738d4143@moroto.mountain Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
Disabling preemption in sock_map_sk_acquire conflicts with GFP_ATOMIC
allocation later in sk_psock_init_link on PREEMPT_RT kernels, since
GFP_ATOMIC might sleep on RT (see bpf: Make BPF and PREEMPT_RT co-exist
patchset notes for details).
This causes calling bpf_map_update_elem on BPF_MAP_TYPE_SOCKMAP maps to
BUG (sleeping function called from invalid context) on RT kernels.
preempt_disable was introduced together with lock_sk and rcu_read_lock
in commit 99ba2b5aba24e ("bpf: sockhash, disallow bpf_tcp_close and update
in parallel"), probably to match disabled migration of BPF programs, and
is no longer necessary.
Remove preempt_disable to fix BUG in sock_map_update_common on RT.
When route4_change() is called on an existing filter, the whole
tcf_result struct is always copied into the new instance of the filter.
This causes a problem when updating a filter bound to a class,
as tcf_unbind_filter() is always called on the old instance in the
success path, decreasing filter_cnt of the still referenced class
and allowing it to be deleted, leading to a use-after-free.
Fix this by no longer copying the tcf_result struct from the old filter.
Fixes: 1109c00547fc ("net: sched: RCU cls_route") Reported-by: valis <sec@valis.email> Reported-by: Bing-Jhong Billy Jheng <billy@starlabs.sg> Signed-off-by: valis <sec@valis.email> Signed-off-by: Jamal Hadi Salim <jhs@mojatatu.com> Reviewed-by: Victor Nogueira <victor@mojatatu.com> Reviewed-by: Pedro Tammela <pctammela@mojatatu.com> Reviewed-by: M A Ramdhan <ramdhan@starlabs.sg> Link: https://lore.kernel.org/r/20230729123202.72406-4-jhs@mojatatu.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
When fw_change() is called on an existing filter, the whole
tcf_result struct is always copied into the new instance of the filter.
This causes a problem when updating a filter bound to a class,
as tcf_unbind_filter() is always called on the old instance in the
success path, decreasing filter_cnt of the still referenced class
and allowing it to be deleted, leading to a use-after-free.
Fix this by no longer copying the tcf_result struct from the old filter.
Fixes: e35a8ee5993b ("net: sched: fw use RCU") Reported-by: valis <sec@valis.email> Reported-by: Bing-Jhong Billy Jheng <billy@starlabs.sg> Signed-off-by: valis <sec@valis.email> Signed-off-by: Jamal Hadi Salim <jhs@mojatatu.com> Reviewed-by: Victor Nogueira <victor@mojatatu.com> Reviewed-by: Pedro Tammela <pctammela@mojatatu.com> Reviewed-by: M A Ramdhan <ramdhan@starlabs.sg> Link: https://lore.kernel.org/r/20230729123202.72406-3-jhs@mojatatu.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
When u32_change() is called on an existing filter, the whole
tcf_result struct is always copied into the new instance of the filter.
This causes a problem when updating a filter bound to a class,
as tcf_unbind_filter() is always called on the old instance in the
success path, decreasing filter_cnt of the still referenced class
and allowing it to be deleted, leading to a use-after-free.
Fix this by no longer copying the tcf_result struct from the old filter.
Fixes: de5df63228fc ("net: sched: cls_u32 changes to knode must appear atomic to readers") Reported-by: valis <sec@valis.email> Reported-by: M A Ramdhan <ramdhan@starlabs.sg> Signed-off-by: valis <sec@valis.email> Signed-off-by: Jamal Hadi Salim <jhs@mojatatu.com> Reviewed-by: Victor Nogueira <victor@mojatatu.com> Reviewed-by: Pedro Tammela <pctammela@mojatatu.com> Reviewed-by: M A Ramdhan <ramdhan@starlabs.sg> Link: https://lore.kernel.org/r/20230729123202.72406-2-jhs@mojatatu.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
The reason for the warning is twofold. One is due to the kthread
cpu_map_kthread_run() is stopped prematurely. Another one is
__cpu_map_ring_cleanup() doesn't handle skb mode and treats skbs in
ptr_ring as XDP frames.
Prematurely-stopped kthread will be fixed by the preceding patch and
ptr_ring will be empty when __cpu_map_ring_cleanup() is called. But
as the comments in __cpu_map_ring_cleanup() said, handling and freeing
skbs in ptr_ring as well to "catch any broken behaviour gracefully".
Fixes: 11941f8a8536 ("bpf: cpumap: Implement generic cpumap") Signed-off-by: Hou Tao <houtao1@huawei.com> Acked-by: Jesper Dangaard Brouer <hawk@kernel.org> Link: https://lore.kernel.org/r/20230729095107.1722450-3-houtao@huaweicloud.com Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
Fixes: 4cfd5779bd6e ("taprio: Add support for txtime-assist mode") Reported-by: syzkaller <syzkaller@googlegroups.com> Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com> Co-developed-by: Eric Dumazet <edumazet@google.com> Co-developed-by: Pedro Tammela <pctammela@mojatatu.com> Acked-by: Vinicius Costa Gomes <vinicius.gomes@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
sk_getsockopt() runs locklessly. This means sk->sk_priority
can be read while other threads are changing its value.
Other reads also happen without socket lock being held.
Add missing annotations where needed.
Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
In a prior commit I forgot that sk_getsockopt() reads
sk->sk_ll_usec without holding a lock.
Fixes: 0dbffbb5335a ("net: annotate data race around sk_ll_usec") Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
sk_getsockopt() runs locklessly, thus we need to annotate the read
of sk->sk_peek_off.
While we are at it, add corresponding annotations to sk_set_peek_off()
and unix_set_peek_off().
Fixes: b9bb53f3836f ("sock: convert sk_peek_offset functions to WRITE_ONCE") Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Willem de Bruijn <willemb@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
sk->sk_mark is often read while another thread could change the value.
Fixes: 4a19ec5800fc ("[NET]: Introducing socket mark socket option.") Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
In a prior commit, I forgot to change sk_getsockopt()
when reading sk->sk_rcvbuf locklessly.
Fixes: ebb3b78db7bf ("tcp: annotate sk->sk_rcvbuf lockless reads") Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
In a prior commit, I forgot to change sk_getsockopt()
when reading sk->sk_sndbuf locklessly.
Fixes: e292f05e0df7 ("tcp: annotate sk->sk_sndbuf lockless reads") Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
In a prior commit, I forgot to change sk_getsockopt()
when reading sk->sk_rcvlowat locklessly.
Fixes: eac66402d1c3 ("net: annotate sk->sk_rcvlowat lockless reads") Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
sk_getsockopt() runs locklessly. This means sk->sk_max_pacing_rate
can be read while other threads are changing its value.
Fixes: 62748f32d501 ("net: introduce SO_MAX_PACING_RATE") Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
sk_getsockopt() runs locklessly. This means sk->sk_txrehash
can be read while other threads are changing its value.
Other locations were handled in commit cb6cd2cec799
("tcp: Change SYN ACK retransmit behaviour to account for rehash")
Fixes: 26859240e4ee ("txhash: Add socket option to control TX hash rethink behavior") Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Akhmat Karakotov <hmukos@yandex-team.ru> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
sk_getsockopt() runs locklessly. This means sk->sk_reserved_mem
can be read while other threads are changing its value.
Add missing annotations where they are needed.
Fixes: 2bb2f5fb21b0 ("net: add new socket option SO_RESERVE_MEM") Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Wei Wang <weiwan@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
Fix this by making caller to provide the context whether it could be in
atomic context flow or not when getting stats from QED driver.
QED driver based on the context provided decide to schedule out or not
when acquiring the PTT BAR window.
We faced the BUG_ON() while getting vport stats, but according to the
code same issue could happen for fcoe and iscsi statistics as well, so
fixing them too.
Fixes: 6c75424612a7 ("qed: Add support for NCSI statistics.") Fixes: 1e128c81290a ("qed: Add support for hardware offloaded FCoE.") Fixes: 2f2b2614e893 ("qed: Provide iSCSI statistics to management") Cc: Sudarsana Kalluru <skalluru@marvell.com> Cc: David Miller <davem@davemloft.net> Cc: Manish Chopra <manishc@marvell.com> Signed-off-by: Konstantin Khorenko <khorenko@virtuozzo.com> Reviewed-by: Simon Horman <horms@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
As &hc->lock is acquired by both timer _hfcpci_softirq() and hardirq
hfcpci_int(), the timer should disable irq before lock acquisition
otherwise deadlock could happen if the timmer is preemtped by the hadr irq.
A match entry is uniquely identified with an "address" or "path" in the
form of: hashtable ID(12b):bucketid(8b):nodeid(12b).
When creating table match entries all of hash table id, bucket id and
node (match entry id) are needed to be either specified by the user or
reasonable in-kernel defaults are used. The in-kernel default for a table id is
0x800(omnipresent root table); for bucketid it is 0x0. Prior to this fix there
was none for a nodeid i.e. the code assumed that the user passed the correct
nodeid and if the user passes a nodeid of 0 (as Mingi Cho did) then that is what
was used. But nodeid of 0 is reserved for identifying the table. This is not
a problem until we dump. The dump code notices that the nodeid is zero and
assumes it is referencing a table and therefore references table struct
tc_u_hnode instead of what was created i.e match entry struct tc_u_knode.
Ming does an equivalent of:
tc filter add dev dummy0 parent 10: prio 1 handle 0x1000 \
protocol ip u32 match ip src 10.0.0.1/32 classid 10:1 action ok
Essentially specifying a table id 0, bucketid 1 and nodeid of zero
Tableid 0 is remapped to the default of 0x800.
Bucketid 1 is ignored and defaults to 0x00.
Nodeid was assumed to be what Ming passed - 0x000
dumping before fix shows:
~$ tc filter ls dev dummy0 parent 10:
filter protocol ip pref 1 u32 chain 0
filter protocol ip pref 1 u32 chain 0 fh 800: ht divisor 1
filter protocol ip pref 1 u32 chain 0 fh 800: ht divisor -30591
Note that the last line reports a table instead of a match entry
(you can tell this because it says "ht divisor...").
As a result of reporting the wrong data type (misinterpretting of struct
tc_u_knode as being struct tc_u_hnode) the divisor is reported with value
of -30591. Ming identified this as part of the heap address
(physmap_base is 0xffff8880 (-30591 - 1)).
The fix is to ensure that when table entry matches are added and no
nodeid is specified (i.e nodeid == 0) then we get the next available
nodeid from the table's pool.
After the fix, this is what the dump shows:
$ tc filter ls dev dummy0 parent 10:
filter protocol ip pref 1 u32 chain 0
filter protocol ip pref 1 u32 chain 0 fh 800: ht divisor 1
filter protocol ip pref 1 u32 chain 0 fh 800::800 order 2048 key ht 800 bkt 0 flowid 10:1 not_in_hw
match 0a000001/ffffffff at 12
action order 1: gact action pass
random type none pass val 0
index 1 ref 1 bind 1
Reported-by: Mingi Cho <mgcho.minic@gmail.com> Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Signed-off-by: Jamal Hadi Salim <jhs@mojatatu.com> Link: https://lore.kernel.org/r/20230726135151.416917-1-jhs@mojatatu.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
There are totally 9 ndo_bridge_setlink handlers in the current kernel,
which are 1) bnxt_bridge_setlink, 2) be_ndo_bridge_setlink 3)
i40e_ndo_bridge_setlink 4) ice_bridge_setlink 5)
ixgbe_ndo_bridge_setlink 6) mlx5e_bridge_setlink 7)
nfp_net_bridge_setlink 8) qeth_l2_bridge_setlink 9) br_setlink.
By investigating the code, we find that 1-7 parse and use nlattr
IFLA_BRIDGE_MODE but 3 and 4 forget to do the nla_len check. This can
lead to an out-of-attribute read and allow a malformed nlattr (e.g.,
length 0) to be viewed as a 2 byte integer.
To avoid such issues, also for other ndo_bridge_setlink handlers in the
future. This patch adds the nla_len check in rtnl_bridge_setlink and
does an early error return if length mismatches. To make it works, the
break is removed from the parsing for IFLA_BRIDGE_FLAGS to make sure
this nla_for_each_nested iterates every attribute.
Fixes: b1edc14a3fbf ("ice: Implement ice_bridge_getlink and ice_bridge_setlink") Fixes: 51616018dd1b ("i40e: Add support for getlink, setlink ndo ops") Suggested-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Lin Ma <linma@zju.edu.cn> Acked-by: Nikolay Aleksandrov <razor@blackwall.org> Reviewed-by: Hangbin Liu <liuhangbin@gmail.com> Link: https://lore.kernel.org/r/20230726075314.1059224-1-linma@zju.edu.cn Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
The nla_for_each_nested parsing in function bpf_sk_storage_diag_alloc
does not check the length of the nested attribute. This can lead to an
out-of-attribute read and allow a malformed nlattr (e.g., length 0) to
be viewed as a 4 byte integer.
This patch adds an additional check when the nlattr is getting counted.
This makes sure the latter nla_get_u32 can access the attributes with
the correct length.
Fixes: 1ed4d92458a9 ("bpf: INET_DIAG support in bpf_sk_storage") Suggested-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Lin Ma <linma@zju.edu.cn> Reviewed-by: Jakub Kicinski <kuba@kernel.org> Link: https://lore.kernel.org/r/20230725023330.422856-1-linma@zju.edu.cn Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
For IP tunnel encapsulation in ECMP (Equal-Cost Multipath) mode, as
the flow is duplicated to the peer eswitch, the related neighbour
information on the peer uplink representor is created as well.
In the cited commit, eswitch devcom unpair is moved to uplink unload
API, specifically the profile->cleanup_tx. If there is a encap rule
offloaded in ECMP mode, when one eswitch does unpair (because of
unloading the driver, for instance), and the peer rule from the peer
eswitch is going to be deleted, the use-after-free error is triggered
while accessing neigh info, as it is already cleaned up in uplink's
profile->disable, which is before its profile->cleanup_tx.
To fix this issue, move the neigh cleanup to profile's cleanup_tx
callback, and after mlx5e_cleanup_uplink_rep_tx is called. The neigh
init is moved to init_tx for symmeter.
[ 2453.376299] BUG: KASAN: slab-use-after-free in mlx5e_rep_neigh_entry_release+0x109/0x3a0 [mlx5_core]
[ 2453.379125] Read of size 4 at addr ffff888127af9008 by task modprobe/2496
Moving to switchdev mode with ntuple offload on causes the kernel to
crash since fs->arfs is freed during nic profile cleanup flow.
Ntuple offload is not supported in switchdev mode and it is already
unset by mlx5 fix feature ndo in switchdev mode. Verify fs->arfs is
valid before disabling it.
The memory pointed to by the priv->rx_res pointer is not freed in the error
path of mlx5e_init_rep_rx, which can lead to a memory leak. Fix by freeing
the memory in the error path, thereby making the error path identical to
mlx5e_cleanup_rep_rx().
Fixes: af8bbf730068 ("net/mlx5e: Convert mlx5e_flow_steering member of mlx5e_priv to pointer") Signed-off-by: Zhengchao Shao <shaozhengchao@huawei.com> Reviewed-by: Simon Horman <simon.horman@corigine.com> Reviewed-by: Tariq Toukan <tariqt@nvidia.com> Signed-off-by: Saeed Mahameed <saeedm@nvidia.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
when mlx5_cmd_exec failed in mlx5dr_cmd_create_reformat_ctx, the memory
pointed by 'in' is not released, which will cause memory leak. Move memory
release after mlx5_cmd_exec.
In function macsec_fs_tx_create_crypto_table_groups(), when the ft->g
memory is successfully allocated but the 'in' memory fails to be
allocated, the memory pointed to by ft->g is released once. And in function
macsec_fs_tx_create(), macsec_fs_tx_destroy() is called to release the
memory pointed to by ft->g again. This will cause double free problem.
When handling deduplicated compressed data, there can be multiple
decompressed extents pointing to the same compressed data in one shot.
In such cases, the bvecs which belong to the longest extent will be
selected as the primary bvecs for real decompressors to decode and the
other duplicated bvecs will be directly copied from the primary bvecs.
Previously, only relative offsets of the longest extent were checked to
decompress the primary bvecs. On rare occasions, it can be incorrect
if there are several extents with the same start relative offset.
As a result, some short bvecs could be selected for decompression and
then cause data corruption.
For example, as Shijie Sun reported off-list, considering the following
extents of a file:
117: 903345.. 915250 | 11905 : 385024.. 389120 | 4096
...
119: 919729.. 930323 | 10594 : 385024.. 389120 | 4096
...
124: 968881.. 980786 | 11905 : 385024.. 389120 | 4096
The start relative offset is the same: 2225, but extent 119 (919729..
930323) is shorter than the others.
Let's restrict the bvec length in addition to the start offset if bvecs
are not full.
Commit 9fb6c9b3fea1 ("s390/sthyi: add cache to store hypervisor info")
added cache handling for store hypervisor info. This also changed the
possible return code for sthyi_fill().
Instead of only returning a condition code like the sthyi instruction would
do, it can now also return a negative error value (-ENOMEM). handle_styhi()
was not changed accordingly. In case of an error, the negative error value
would incorrectly injected into the guest PSW.
Add proper error handling to prevent this, and update the comment which
describes the possible return values of sthyi_fill().
Fixes: 9fb6c9b3fea1 ("s390/sthyi: add cache to store hypervisor info") Reviewed-by: Christian Borntraeger <borntraeger@linux.ibm.com> Link: https://lore.kernel.org/r/20230727182939.2050744-1-hca@linux.ibm.com Signed-off-by: Heiko Carstens <hca@linux.ibm.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
Compiling big-endian targets with Clang produces the diagnostic:
fs/namei.c:2173:13: warning: use of bitwise '|' with boolean operands [-Wbitwise-instead-of-logical]
} while (!(has_zero(a, &adata, &constants) | has_zero(b, &bdata, &constants)));
~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
||
fs/namei.c:2173:13: note: cast one or both operands to int to silence this warning
It appears that when has_zero was introduced, two definitions were
produced with different signatures (in particular different return
types).
Looking at the usage in hash_name() in fs/namei.c, I suspect that
has_zero() is meant to be invoked twice per while loop iteration; using
logical-or would not update `bdata` when `a` did not have zeros. So I
think it's preferred to always return an unsigned long rather than a
bool than update the while loop in hash_name() to use a logical-or
rather than bitwise-or.
[ Also changed powerpc version to do the same - Linus ]
SCMI transport based on SMC can optionally use an additional IRQ to
signal message completion. The associated interrupt handler is currently
allocated using devres but on shutdown the core SCMI stack will call
.chan_free() well before any managed cleanup is invoked by devres.
As a consequence, the arrival of a late reply to an in-flight pending
transaction could still trigger the interrupt handler well after the
SCMI core has cleaned up the channels, with unpleasant results.
Inhibit further message processing on the IRQ path by explicitly freeing
the IRQ inside .chan_free() callback itself.
When building with Clang, and when KASAN and GCOV_PROFILE_ALL are both
enabled, the test fails to build [1]:
>> lib/test_bitmap.c:920:2: error: call to '__compiletime_assert_239' declared with 'error' attribute: BUILD_BUG_ON failed: !__builtin_constant_p(res)
BUILD_BUG_ON(!__builtin_constant_p(res));
^
include/linux/build_bug.h:50:2: note: expanded from macro 'BUILD_BUG_ON'
BUILD_BUG_ON_MSG(condition, "BUILD_BUG_ON failed: " #condition)
^
include/linux/build_bug.h:39:37: note: expanded from macro 'BUILD_BUG_ON_MSG'
#define BUILD_BUG_ON_MSG(cond, msg) compiletime_assert(!(cond), msg)
^
include/linux/compiler_types.h:352:2: note: expanded from macro 'compiletime_assert'
_compiletime_assert(condition, msg, __compiletime_assert_, __COUNTER__)
^
include/linux/compiler_types.h:340:2: note: expanded from macro '_compiletime_assert'
__compiletime_assert(condition, msg, prefix, suffix)
^
include/linux/compiler_types.h:333:4: note: expanded from macro '__compiletime_assert'
prefix ## suffix(); \
^
<scratch space>:185:1: note: expanded from here
__compiletime_assert_239
Originally it was attributed to s390, which now looks seemingly wrong. The
issue is not related to bitmap code itself, but it breaks build for a given
configuration.
Disabling the const_eval test under that config may potentially hide other
bugs. Instead, workaround it by disabling GCOV for the test_bitmap unless
the compiler will get fixed.
Commit 35727af2b15d ("irqchip/gicv3: Workaround for NVIDIA erratum
T241-FABRIC-4") moved the initialisation of the SoC version to
arm_smccc_version_init() but forgot to update the results structure
and it's usage.
Fix the use of the uninitialised results structure and update the
error strings.
Set VPU G2 clock to 300MHz like described in documentation.
This fixes pixels error occurring with large resolution ( >= 2560x1600)
HEVC test stream when using the postprocessor to produce NV12.
Fixes: 4ac7e4a81272 ("arm64: dts: imx8mq: Enable both G1 and G2 VPU's with vpu-blk-ctrl") Signed-off-by: Benjamin Gaignard <benjamin.gaignard@collabora.com> Signed-off-by: Shawn Guo <shawnguo@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
For SOMs with an onboard PHY, the RESET_N pull-up resistor is
currently deactivated in the pinmux configuration. When the pinmux
code selects the GPIO function for this pin, with a default direction
of input, this prevents the RESET_N pin from being taken to the proper
3.3V level (deasserted), and this results in the PHY being not
detected since it is held in reset.
Taken from RESET_N pin description in ADIN13000 datasheet:
This pin requires a 1K pull-up resistor to AVDD_3P3.
Activate the pull-up resistor to fix the issue.
Fixes: ade0176dd8a0 ("arm64: dts: imx8mn-var-som: Add Variscite VAR-SOM-MX8MN System on Module") Signed-off-by: Hugo Villeneuve <hvilleneuve@dimonoff.com> Reviewed-by: Fabio Estevam <festevam@gmail.com> Signed-off-by: Shawn Guo <shawnguo@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
The GW7904 does not connect the VDD_MIPI power rails thus MIPI is
disabled. However we must also disable disp_blk_ctrl as it uses the
pgc_mipi power domain and without it being disabled imx8m-blk-ctrl will
fail to probe:
imx8m-blk-ctrl 32e28000.blk-ctrl: error -ETIMEDOUT: failed to attach
power domain "mipi-dsi"
imx8m-blk-ctrl: probe of 32e28000.blk-ctrl failed with error -110
The GW7903 does not connect the VDD_MIPI power rails thus MIPI is
disabled. However we must also disable disp_blk_ctrl as it uses the
pgc_mipi power domain and without it being disabled imx8m-blk-ctrl will
fail to probe:
imx8m-blk-ctrl 32e28000.blk-ctrl: error -ETIMEDOUT: failed to attach power domain "mipi-dsi"
imx8m-blk-ctrl: probe of 32e28000.blk-ctrl failed with error -110
Both MMU-600 and MMU-700 have similar errata around TLB invalidation
while both stages of translation are active, which will need some
consideration once nesting support is implemented. For now, though,
it's very easy to make our implicit lack of nesting support explicit
for those cases, so they're less likely to be missed in future.
In certain cases we may want to refuse to allow nested translation even
when both stages are implemented, so let's add an explicit feature for
nesting support which we can control in its own right. For now this
merely serves as documentation, but it means a nice convenient check
will be ready and waiting for the future nesting code.
To work around MMU-700 erratum 2812531 we need to ensure that certain
sequences of commands cannot be issued without an intervening sync. In
practice this falls out of our current command-batching machinery
anyway - each batch only contains a single type of invalidation command,
and ends with a sync. The only exception is when a batch is sufficiently
large to need issuing across multiple command queue slots, wherein the
earlier slots will not contain a sync and thus may in theory interleave
with another batch being issued in parallel to create an affected
sequence across the slot boundary.
Since MMU-700 supports range invalidate commands and thus we will prefer
to use them (which also happens to avoid conditions for other errata),
I'm not entirely sure it's even possible for a single high-level
invalidate call to generate a batch of more than 63 commands, but for
the sake of robustness and documentation, wire up an option to enforce
that a sync is always inserted for every slot issued.
The other aspect is that the relative order of DVM commands cannot be
controlled, so DVM cannot be used. Again that is already the status quo,
but since we have at least defined ARM_SMMU_FEAT_BTM, we can explicitly
disable it for documentation purposes even if it's not wired up anywhere
yet.
MMU-600 versions prior to r1p0 fail to correctly generate a WFE wakeup
event when the command queue transitions fom full to non-full. We can
easily work around this by simply hiding the SEV capability such that we
fall back to polling for space in the queue - since MMU-600 implements
MSIs we wouldn't expect to need SEV for sync completion either, so this
should have little to no impact.
Last year, the code that manages GSI channel transactions switched
from using spinlock-protected linked lists to using indexes into the
ring buffer used for a channel. Recently, Google reported seeing
transaction reference count underflows occasionally during shutdown.
Doug Anderson found a way to reproduce the issue reliably, and
bisected the issue to the commit that eliminated the linked lists
and the lock. The root cause was ultimately determined to be
related to unused transactions being committed as part of the modem
shutdown cleanup activity. Unused transactions are not normally
expected (except in error cases).
The modem uses some ranges of IPA-resident memory, and whenever it
shuts down we zero those ranges. In ipa_filter_reset_table() a
transaction is allocated to zero modem filter table entries. If
hashing is not supported, hashed table memory should not be zeroed.
But currently nothing prevents that, and the result is an unused
transaction. Something similar occurs when we zero routing table
entries for the modem.
By preventing any attempt to clear hashed tables when hashing is not
supported, the reference count underflow is avoided in this case.
Note that there likely remains an issue with properly freeing unused
transactions (if they occur due to errors). This patch addresses
only the underflows that Google originally reported.
Cc: <stable@vger.kernel.org> # 6.1.x Fixes: d338ae28d8a8 ("net: ipa: kill all other transaction lists") Tested-by: Douglas Anderson <dianders@chromium.org> Signed-off-by: Alex Elder <elder@linaro.org> Link: https://lore.kernel.org/r/20230724224055.1688854-1-elder@linaro.org Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Alex Elder <elder@linaro.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Whenever a shutdown is invoked, free irqs only and keep mlx5_irq
synthetic wrapper intact in order to avoid use-after-free on
system shutdown.
for example:
==================================================================
BUG: KASAN: use-after-free in _find_first_bit+0x66/0x80
Read of size 8 at addr ffff88823fc0d318 by task kworker/u192:0/13608
The buggy address belongs to the object at ffff88823fc0d300
which belongs to the cache kmalloc-192 of size 192
The buggy address is located 24 bytes inside of
192-byte region [ffff88823fc0d300, ffff88823fc0d3c0)
Reported-by: Frederick Lawler <fred@cloudflare.com> Link: https://lore.kernel.org/netdev/be5b9271-7507-19c5-ded1-fa78f1980e69@cloudflare.com Signed-off-by: Shay Drory <shayd@nvidia.com> Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
[hardik: Refer to the irqn member of the mlx5_irq struct, instead of
the msi_map, since we don't have upstream v6.4 commit 235a25fe28de
("net/mlx5: Modify struct mlx5_irq to use struct msi_map")].
[hardik: Refer to the pf_pool member of the mlx5_irq_table struct,
instead of pcif_pool, since we don't have upstream v6.4 commit 8bebfd767909 ("net/mlx5: Improve naming of pci function vectors")]. Signed-off-by: Hardik Garg <hargar@linux.microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
A previous commit made all cqring waits marked as iowait, as a way to
improve performance for short schedules with pending IO. However, for
use cases that have a special reaper thread that does nothing but
wait on events on the ring, this causes a cosmetic issue where we
know have one core marked as being "busy" with 100% iowait.
While this isn't a grave issue, it is confusing to users. Rather than
always mark us as being in iowait, gate setting of current->in_iowait
to 1 by whether or not the waiting task has pending requests.
Due to the way the GDS and SRSO patches flowed into the stable tree, it
was a 50% chance that the merge of the which value GDS and SRSO should
be. Of course, I lost that bet, and chose the opposite of what Linus
chose in commit 64094e7e3118 ("Merge tag 'gds-for-linus-2023-08-01' of
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip")
Fix this up by switching the values to match what is now in Linus's tree
as that is the correct value to mirror.
It is possible that a guest can send a packet that contains a head + 18
slots and yet has a len <= XEN_NETBACK_TX_COPY_LEN. This causes nr_slots
to underflow in xenvif_get_requests() which then causes the subsequent
loop's termination condition to be wrong, causing a buffer overrun of
queue->tx_map_ops.
Rework the code to account for the extra frag_overflow slots.
This is CVE-2023-34319 / XSA-432.
Fixes: ad7f402ae4f4 ("xen/netback: Ensure protocol headers don't fall in the non-linear area") Signed-off-by: Ross Lagerwall <ross.lagerwall@citrix.com> Reviewed-by: Paul Durrant <paul@xen.org> Reviewed-by: Wei Liu <wei.liu@kernel.org> Signed-off-by: Juergen Gross <jgross@suse.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The SBPB bit in MSR_IA32_PRED_CMD is supported only after a microcode
patch has been applied so set X86_FEATURE_SBPB only then. Otherwise,
guests would attempt to set that bit and #GP on the MSR write.
While at it, make SMT detection more robust as some guests - depending
on how and what CPUID leafs their report - lead to cpu_smt_control
getting set to CPU_SMT_NOT_SUPPORTED but SRSO_NO should be set for any
guest incarnation where one simply cannot do SMT, for whatever reason.
Set X86_FEATURE_RETHUNK when enabling the SRSO mitigation so that
generated code (e.g., ftrace, static call, eBPF) generates "jmp
__x86_return_thunk" instead of RET.
Add the option to flush IBPB only on VMEXIT in order to protect from
malicious guests but one otherwise trusts the software that runs on the
hypervisor.
Add the option to mitigate using IBPB on a kernel entry. Pull in the
Retbleed alternative so that the IBPB call from there can be used. Also,
if Retbleed mitigation is done using IBPB, the same mitigation can and
must be used here.
Add support for the synthetic CPUID flag which "if this bit is 1,
it indicates that MSR 49h (PRED_CMD) bit 0 (IBPB) flushes all branch
type predictions from the CPU branch predictor."
This flag is there so that this capability in guests can be detected
easily (otherwise one would have to track microcode revisions which is
impossible for guests).
It is also needed only for Zen3 and -4. The other two (Zen1 and -2)
always flush branch type predictions by default.
Add a mitigation for the speculative return address stack overflow
vulnerability found on AMD processors.
The mitigation works by ensuring all RET instructions speculate to
a controlled location, similar to how speculation is controlled in the
retpoline sequence. To accomplish this, the __x86_return_thunk forces
the CPU to mispredict every function return using a 'safe return'
sequence.
To ensure the safety of this mitigation, the kernel must ensure that the
safe return sequence is itself free from attacker interference. In Zen3
and Zen4, this is accomplished by creating a BTB alias between the
untraining function srso_untrain_ret_alias() and the safe return
function srso_safe_ret_alias() which results in evicting a potentially
poisoned BTB entry and using that safe one for all function returns.
In older Zen1 and Zen2, this is accomplished using a reinterpretation
technique similar to Retbleed one: srso_untrain_ret() and
srso_safe_ret().
Instead of duplicating init_mm, allocate a fresh mm. The advantage is
that mm_alloc() has much simpler dependencies. Additionally it makes
more conceptual sense, init_mm has no (and must not have) user state
to duplicate.
Commit 3f4c8211d982 ("x86/mm: Use mm_alloc() in poking_init()") broke
the kernel for running as Xen PV guest.
It seems as if the new address space is never activated before being
used, resulting in Xen rejecting to accept the new CR3 value (the PGD
isn't pinned).
Fix that by adding the now missing call of paravirt_arch_dup_mmap() to
poking_init(). That call was previously done by dup_mm()->dup_mmap() and
it is a NOP for all cases but for Xen PV, where it is just doing the
pinning of the PGD.
Fixes: 3f4c8211d982 ("x86/mm: Use mm_alloc() in poking_init()") Signed-off-by: Juergen Gross <jgross@suse.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lkml.kernel.org/r/20230109150922.10578-1-jgross@suse.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Moving mem_encrypt_init() broke the AMD_MEM_ENCRYPT=n because the
declaration of that function was under #ifdef CONFIG_AMD_MEM_ENCRYPT and
the obvious placement for the inline stub was the #else path.
This is a leftover of commit 20f07a044a76 ("x86/sev: Move common memory
encryption code to mem_encrypt.c") which made mem_encrypt_init() depend on
X86_MEM_ENCRYPT without moving the prototype. That did not fail back then
because there was no stub inline as the core init code had a weak function.
Move both the declaration and the stub out of the CONFIG_AMD_MEM_ENCRYPT
section and guard it with CONFIG_X86_MEM_ENCRYPT.
Fixes: 439e17576eb4 ("init, x86: Move mem_encrypt_init() into arch_cpu_finalize_init()") Reported-by: kernel test robot <lkp@intel.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Closes: https://lore.kernel.org/oe-kbuild-all/202306170247.eQtCJPE8-lkp@intel.com/ Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Gather Data Sampling (GDS) is a transient execution attack using
gather instructions from the AVX2 and AVX512 extensions. This attack
allows malicious code to infer data that was previously stored in
vector registers. Systems that are not vulnerable to GDS will set the
GDS_NO bit of the IA32_ARCH_CAPABILITIES MSR. This is useful for VM
guests that may think they are on vulnerable systems that are, in
fact, not affected. Guests that are running on affected hosts where
the mitigation is enabled are protected as if they were running
on an unaffected system.
On all hosts that are not affected or that are mitigated, set the
GDS_NO bit.
Signed-off-by: Daniel Sneddon <daniel.sneddon@linux.intel.com> Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com> Acked-by: Josh Poimboeuf <jpoimboe@kernel.org> Signed-off-by: Daniel Sneddon <daniel.sneddon@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Gather Data Sampling (GDS) is mitigated in microcode. However, on
systems that haven't received the updated microcode, disabling AVX
can act as a mitigation. Add a Kconfig option that uses the microcode
mitigation if available and disables AVX otherwise. Setting this
option has no effect on systems not affected by GDS. This is the
equivalent of setting gather_data_sampling=force.
Signed-off-by: Daniel Sneddon <daniel.sneddon@linux.intel.com> Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com> Acked-by: Josh Poimboeuf <jpoimboe@kernel.org> Signed-off-by: Daniel Sneddon <daniel.sneddon@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The Gather Data Sampling (GDS) vulnerability allows malicious software
to infer stale data previously stored in vector registers. This may
include sensitive data such as cryptographic keys. GDS is mitigated in
microcode, and systems with up-to-date microcode are protected by
default. However, any affected system that is running with older
microcode will still be vulnerable to GDS attacks.
Since the gather instructions used by the attacker are part of the
AVX2 and AVX512 extensions, disabling these extensions prevents gather
instructions from being executed, thereby mitigating the system from
GDS. Disabling AVX2 is sufficient, but we don't have the granularity
to do this. The XCR0[2] disables AVX, with no option to just disable
AVX2.
Add a kernel parameter gather_data_sampling=force that will enable the
microcode mitigation if available, otherwise it will disable AVX on
affected systems.
This option will be ignored if cmdline mitigations=off.
This is a *big* hammer. It is known to break buggy userspace that
uses incomplete, buggy AVX enumeration. Unfortunately, such userspace
does exist in the wild:
Gather Data Sampling (GDS) is a hardware vulnerability which allows
unprivileged speculative access to data which was previously stored in
vector registers.
Intel processors that support AVX2 and AVX512 have gather instructions
that fetch non-contiguous data elements from memory. On vulnerable
hardware, when a gather instruction is transiently executed and
encounters a fault, stale data from architectural or internal vector
registers may get transiently stored to the destination vector
register allowing an attacker to infer the stale data using typical
side channel techniques like cache timing attacks.
This mitigation is different from many earlier ones for two reasons.
First, it is enabled by default and a bit must be set to *DISABLE* it.
This is the opposite of normal mitigation polarity. This means GDS can
be mitigated simply by updating microcode and leaving the new control
bit alone.
Second, GDS has a "lock" bit. This lock bit is there because the
mitigation affects the hardware security features KeyLocker and SGX.
It needs to be enabled and *STAY* enabled for these features to be
mitigated against GDS.
The mitigation is enabled in the microcode by default. Disable it by
setting gather_data_sampling=off or by disabling all mitigations with
mitigations=off. The mitigation status can be checked by reading:
Initializing the FPU during the early boot process is a pointless
exercise. Early boot is convoluted and fragile enough.
Nothing requires that the FPU is set up early. It has to be initialized
before fork_init() because the task_struct size depends on the FPU register
buffer size.
Move the initialization to arch_cpu_finalize_init() which is the perfect
place to do so.
No functional change.
This allows to remove quite some of the custom early command line parsing,
but that's subject to the next installment.
No point in doing this during really early boot. Move it to an early
initcall so that it is set up before possible user mode helpers are started
during device initialization.
X86 is reworking the boot process so that initializations which are not
required during early boot can be moved into the late boot process and out
of the fragile and restricted initial boot phase.
arch_cpu_finalize_init() is the obvious place to do such initializations,
but arch_cpu_finalize_init() is invoked too late in start_kernel() e.g. for
initializing the FPU completely. fork_init() requires that the FPU is
initialized as the size of task_struct on X86 depends on the size of the
required FPU register buffer.
Fortunately none of the init calls between calibrate_delay() and
arch_cpu_finalize_init() is relevant for the functionality of
arch_cpu_finalize_init().
Invoke it right after calibrate_delay() where everything which is relevant
for arch_cpu_finalize_init() has been set up already.
No functional change intended.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Reviewed-by: Rick Edgecombe <rick.p.edgecombe@intel.com> Link: https://lore.kernel.org/r/20230613224545.612182854@linutronix.de Signed-off-by: Daniel Sneddon <daniel.sneddon@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Everything is converted over to arch_cpu_finalize_init(). Remove the
check_bugs() leftovers including the empty stubs in asm-generic, alpha,
parisc, powerpc and xtensa.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Link: https://lore.kernel.org/r/20230613224545.553215951@linutronix.de Signed-off-by: Daniel Sneddon <daniel.sneddon@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
projects / kernel.git / log