August 2024 - Linux-stable-mirror

[PATCH] wifi: mt76: mt7996: fix NULL pointer dereference in mt7996_mcu_sta_bfer_eht

by Ma Ke

Fix the NULL pointer dereference in mt7996_mcu_sta_bfer_eht. Found by code review. Cc: stable(a)vger.kernel.org Fixes: ba01944adee9 ("wifi: mt76: mt7996: add EHT beamforming support") Signed-off-by: Ma Ke <make24(a)iscas.ac.cn> --- drivers/net/wireless/mediatek/mt76/mt7996/mcu.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/net/wireless/mediatek/mt76/mt7996/mcu.c b/drivers/net/wireless/mediatek/mt76/mt7996/mcu.c index 2e4fa9f48dfb..d0fe2505ddbf 100644 --- a/drivers/net/wireless/mediatek/mt76/mt7996/mcu.c +++ b/drivers/net/wireless/mediatek/mt76/mt7996/mcu.c @@ -1605,6 +1605,9 @@ mt7996_mcu_sta_bfer_eht(struct ieee80211_sta *sta, struct ieee80211_vif *vif, IEEE80211_EHT_MCS_NSS_RX) - 1; u8 snd_dim, sts; + if (!vc) + return; + bf->tx_mode = MT_PHY_TYPE_EHT_MU; mt7996_mcu_sta_sounding_rate(bf); -- 2.25.1

7 months, 1 week

1
0
0 0

[PATCH net v2 2/2] net: mctp-serial: Fix missing escapes on transmit

by Matt Johnston

0x7d and 0x7e bytes are meant to be escaped in the data portion of frames, but this didn't occur since next_chunk_len() had an off-by-one error. That also resulted in the final byte of a payload being written as a separate tty write op. The chunk prior to an escaped byte would be one byte short, and the next call would never test the txpos+1 case, which is where the escaped byte was located. That meant it never hit the escaping case in mctp_serial_tx_work(). Example Input: 01 00 08 c8 7e 80 02 Previous incorrect chunks from next_chunk_len(): 01 00 08 c8 7e 80 02 With this fix: 01 00 08 c8 7e 80 02 Cc: stable(a)vger.kernel.org Fixes: a0c2ccd9b5ad ("mctp: Add MCTP-over-serial transport binding") Signed-off-by: Matt Johnston <matt(a)codeconstruct.com.au> --- drivers/net/mctp/mctp-serial.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/net/mctp/mctp-serial.c b/drivers/net/mctp/mctp-serial.c index 7a40d07ff77b..f39bbe255497 100644 --- a/drivers/net/mctp/mctp-serial.c +++ b/drivers/net/mctp/mctp-serial.c @@ -91,8 +91,8 @@ static int next_chunk_len(struct mctp_serial *dev) * will be those non-escaped bytes, and does not include the escaped * byte. */ - for (i = 1; i + dev->txpos + 1 < dev->txlen; i++) { - if (needs_escape(dev->txbuf[dev->txpos + i + 1])) + for (i = 1; i + dev->txpos < dev->txlen; i++) { + if (needs_escape(dev->txbuf[dev->txpos + i])) break; }

7 months, 1 week

1
0
0 0

[PATCH net 2/3] i2c: designware: add device private data passing to lock functions

by Jiawen Wu

In order to add the hardware lock for Wangxun devices with minimal modification, pass struct dw_i2c_dev to the acquire and release lock functions. Cc: stable(a)vger.kernel.org Fixes: 2f8d1ed79345 ("i2c: designware: Add driver support for Wangxun 10Gb NIC") Signed-off-by: Jiawen Wu <jiawenwu(a)trustnetic.com> --- drivers/i2c/busses/i2c-designware-amdpsp.c | 4 ++-- drivers/i2c/busses/i2c-designware-baytrail.c | 14 ++++++++++++-- drivers/i2c/busses/i2c-designware-common.c | 4 ++-- drivers/i2c/busses/i2c-designware-core.h | 4 ++-- 4 files changed, 18 insertions(+), 8 deletions(-) diff --git a/drivers/i2c/busses/i2c-designware-amdpsp.c b/drivers/i2c/busses/i2c-designware-amdpsp.c index 63454b06e5da..ee7cc4b33f4b 100644 --- a/drivers/i2c/busses/i2c-designware-amdpsp.c +++ b/drivers/i2c/busses/i2c-designware-amdpsp.c @@ -167,7 +167,7 @@ static void psp_release_i2c_bus_deferred(struct work_struct *work) } static DECLARE_DELAYED_WORK(release_queue, psp_release_i2c_bus_deferred); -static int psp_acquire_i2c_bus(void) +static int psp_acquire_i2c_bus(struct dw_i2c_dev *dev) { int status; @@ -206,7 +206,7 @@ static int psp_acquire_i2c_bus(void) return 0; } -static void psp_release_i2c_bus(void) +static void psp_release_i2c_bus(struct dw_i2c_dev *dev) { mutex_lock(&psp_i2c_access_mutex); diff --git a/drivers/i2c/busses/i2c-designware-baytrail.c b/drivers/i2c/busses/i2c-designware-baytrail.c index 45774aa47c28..9dde796e0fcc 100644 --- a/drivers/i2c/busses/i2c-designware-baytrail.c +++ b/drivers/i2c/busses/i2c-designware-baytrail.c @@ -12,6 +12,16 @@ #include "i2c-designware-core.h" +static int iosf_mbi_block_punit_i2c_access_dev(struct dw_i2c_dev *dev) +{ + return iosf_mbi_block_punit_i2c_access(); +} + +static void iosf_mbi_unblock_punit_i2c_access_dev(struct dw_i2c_dev *dev) +{ + return iosf_mbi_unblock_punit_i2c_access(); +} + int i2c_dw_baytrail_probe_lock_support(struct dw_i2c_dev *dev) { acpi_status status; @@ -36,8 +46,8 @@ int i2c_dw_baytrail_probe_lock_support(struct dw_i2c_dev *dev) return -EPROBE_DEFER; dev_info(dev->dev, "I2C bus managed by PUNIT\n"); - dev->acquire_lock = iosf_mbi_block_punit_i2c_access; - dev->release_lock = iosf_mbi_unblock_punit_i2c_access; + dev->acquire_lock = iosf_mbi_block_punit_i2c_access_dev; + dev->release_lock = iosf_mbi_unblock_punit_i2c_access_dev; dev->shared_with_punit = true; return 0; diff --git a/drivers/i2c/busses/i2c-designware-common.c b/drivers/i2c/busses/i2c-designware-common.c index e8a688d04aee..743875090356 100644 --- a/drivers/i2c/busses/i2c-designware-common.c +++ b/drivers/i2c/busses/i2c-designware-common.c @@ -524,7 +524,7 @@ int i2c_dw_acquire_lock(struct dw_i2c_dev *dev) if (!dev->acquire_lock) return 0; - ret = dev->acquire_lock(); + ret = dev->acquire_lock(dev); if (!ret) return 0; @@ -536,7 +536,7 @@ int i2c_dw_acquire_lock(struct dw_i2c_dev *dev) void i2c_dw_release_lock(struct dw_i2c_dev *dev) { if (dev->release_lock) - dev->release_lock(); + dev->release_lock(dev); } /* diff --git a/drivers/i2c/busses/i2c-designware-core.h b/drivers/i2c/busses/i2c-designware-core.h index e9606c00b8d1..12b77f464fb5 100644 --- a/drivers/i2c/busses/i2c-designware-core.h +++ b/drivers/i2c/busses/i2c-designware-core.h @@ -291,8 +291,8 @@ struct dw_i2c_dev { u16 fp_lcnt; u16 hs_hcnt; u16 hs_lcnt; - int (*acquire_lock)(void); - void (*release_lock)(void); + int (*acquire_lock)(struct dw_i2c_dev *dev); + void (*release_lock)(struct dw_i2c_dev *dev); int semaphore_idx; bool shared_with_punit; void (*disable)(struct dw_i2c_dev *dev); -- 2.27.0

7 months, 1 week

4
3
0 0

[PATCH net-next 1/3] icmp: change the order of rate limits

by Eric Dumazet

ICMP messages are ratelimited : After the blamed commits, the two rate limiters are applied in this order: 1) host wide ratelimit (icmp_global_allow()) 2) Per destination ratelimit (inetpeer based) In order to avoid side-channels attacks, we need to apply the per destination check first. This patch makes the following change : 1) icmp_global_allow() checks if the host wide limit is reached. But credits are not yet consumed. This is deferred to 3) 2) The per destination limit is checked/updated. This might add a new node in inetpeer tree. 3) icmp_global_consume() consumes tokens if prior operations succeeded. This means that host wide ratelimit is still effective in keeping inetpeer tree small even under DDOS. As a bonus, I removed icmp_global.lock as the fast path can use a lock-free operation. Fixes: c0303efeab73 ("net: reduce cycles spend on ICMP replies that gets rate limited") Fixes: 4cdf507d5452 ("icmp: add a global rate limitation") Reported-by: Keyu Man <keyu.man(a)email.ucr.edu> Signed-off-by: Eric Dumazet <edumazet(a)google.com> Cc: Jesper Dangaard Brouer <hawk(a)kernel.org> Cc: stable(a)vger.kernel.org --- include/net/ip.h | 2 + net/ipv4/icmp.c | 103 ++++++++++++++++++++++++++--------------------- net/ipv6/icmp.c | 28 ++++++++----- 3 files changed, 76 insertions(+), 57 deletions(-) diff --git a/include/net/ip.h b/include/net/ip.h index c5606cadb1a552f3e282a5e1e721fd47b07432b2..82248813619e3f21e09d52976accbdc76c7668c2 100644 --- a/include/net/ip.h +++ b/include/net/ip.h @@ -795,6 +795,8 @@ static inline void ip_cmsg_recv(struct msghdr *msg, struct sk_buff *skb) } bool icmp_global_allow(void); +void icmp_global_consume(void); + extern int sysctl_icmp_msgs_per_sec; extern int sysctl_icmp_msgs_burst; diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c index b8f56d03fcbb62970a828e20dd9f05fcede2d552..0078e8fb2e86d0552ef85eb5bf5bef947b0f1c3d 100644 --- a/net/ipv4/icmp.c +++ b/net/ipv4/icmp.c @@ -224,57 +224,59 @@ int sysctl_icmp_msgs_per_sec __read_mostly = 1000; int sysctl_icmp_msgs_burst __read_mostly = 50; static struct { - spinlock_t lock; - u32 credit; + atomic_t credit; u32 stamp; -} icmp_global = { - .lock = __SPIN_LOCK_UNLOCKED(icmp_global.lock), -}; +} icmp_global; /** * icmp_global_allow - Are we allowed to send one more ICMP message ? * * Uses a token bucket to limit our ICMP messages to ~sysctl_icmp_msgs_per_sec. * Returns false if we reached the limit and can not send another packet. - * Note: called with BH disabled + * Works in tandem with icmp_global_consume(). */ bool icmp_global_allow(void) { - u32 credit, delta, incr = 0, now = (u32)jiffies; - bool rc = false; + u32 delta, now, oldstamp; + int incr, new, old; - /* Check if token bucket is empty and cannot be refilled - * without taking the spinlock. The READ_ONCE() are paired - * with the following WRITE_ONCE() in this same function. + /* Note: many cpus could find this condition true. + * Then later icmp_global_consume() could consume more credits, + * this is an acceptable race. */ - if (!READ_ONCE(icmp_global.credit)) { - delta = min_t(u32, now - READ_ONCE(icmp_global.stamp), HZ); - if (delta < HZ / 50) - return false; - } + if (atomic_read(&icmp_global.credit) > 0) + return true; - spin_lock(&icmp_global.lock); - delta = min_t(u32, now - icmp_global.stamp, HZ); - if (delta >= HZ / 50) { - incr = READ_ONCE(sysctl_icmp_msgs_per_sec) * delta / HZ; - if (incr) - WRITE_ONCE(icmp_global.stamp, now); - } - credit = min_t(u32, icmp_global.credit + incr, - READ_ONCE(sysctl_icmp_msgs_burst)); - if (credit) { - /* We want to use a credit of one in average, but need to randomize - * it for security reasons. - */ - credit = max_t(int, credit - get_random_u32_below(3), 0); - rc = true; + now = jiffies; + oldstamp = READ_ONCE(icmp_global.stamp); + delta = min_t(u32, now - oldstamp, HZ); + if (delta < HZ / 50) + return false; + + incr = READ_ONCE(sysctl_icmp_msgs_per_sec) * delta / HZ; + if (!incr) + return false; + + if (cmpxchg(&icmp_global.stamp, oldstamp, now) == oldstamp) { + old = atomic_read(&icmp_global.credit); + do { + new = min(old + incr, READ_ONCE(sysctl_icmp_msgs_burst)); + } while (!atomic_try_cmpxchg(&icmp_global.credit, &old, new)); } - WRITE_ONCE(icmp_global.credit, credit); - spin_unlock(&icmp_global.lock); - return rc; + return true; } EXPORT_SYMBOL(icmp_global_allow); +void icmp_global_consume(void) +{ + int credits = get_random_u32_below(3); + + /* Note: this might make icmp_global.credit negative. */ + if (credits) + atomic_sub(credits, &icmp_global.credit); +} +EXPORT_SYMBOL(icmp_global_consume); + static bool icmpv4_mask_allow(struct net *net, int type, int code) { if (type > NR_ICMP_TYPES) @@ -291,14 +293,16 @@ static bool icmpv4_mask_allow(struct net *net, int type, int code) return false; } -static bool icmpv4_global_allow(struct net *net, int type, int code) +static bool icmpv4_global_allow(struct net *net, int type, int code, + bool *apply_ratelimit) { if (icmpv4_mask_allow(net, type, code)) return true; - if (icmp_global_allow()) + if (icmp_global_allow()) { + *apply_ratelimit = true; return true; - + } __ICMP_INC_STATS(net, ICMP_MIB_RATELIMITGLOBAL); return false; } @@ -308,15 +312,16 @@ static bool icmpv4_global_allow(struct net *net, int type, int code) */ static bool icmpv4_xrlim_allow(struct net *net, struct rtable *rt, - struct flowi4 *fl4, int type, int code) + struct flowi4 *fl4, int type, int code, + bool apply_ratelimit) { struct dst_entry *dst = &rt->dst; struct inet_peer *peer; bool rc = true; int vif; - if (icmpv4_mask_allow(net, type, code)) - goto out; + if (!apply_ratelimit) + return true; /* No rate limit on loopback */ if (dst->dev && (dst->dev->flags&IFF_LOOPBACK)) @@ -331,6 +336,8 @@ static bool icmpv4_xrlim_allow(struct net *net, struct rtable *rt, out: if (!rc) __ICMP_INC_STATS(net, ICMP_MIB_RATELIMITHOST); + else + icmp_global_consume(); return rc; } @@ -402,6 +409,7 @@ static void icmp_reply(struct icmp_bxm *icmp_param, struct sk_buff *skb) struct ipcm_cookie ipc; struct rtable *rt = skb_rtable(skb); struct net *net = dev_net(rt->dst.dev); + bool apply_ratelimit = false; struct flowi4 fl4; struct sock *sk; struct inet_sock *inet; @@ -413,11 +421,11 @@ static void icmp_reply(struct icmp_bxm *icmp_param, struct sk_buff *skb) if (ip_options_echo(net, &icmp_param->replyopts.opt.opt, skb)) return; - /* Needed by both icmp_global_allow and icmp_xmit_lock */ + /* Needed by both icmpv4_global_allow and icmp_xmit_lock */ local_bh_disable(); - /* global icmp_msgs_per_sec */ - if (!icmpv4_global_allow(net, type, code)) + /* is global icmp_msgs_per_sec exhausted ? */ + if (!icmpv4_global_allow(net, type, code, &apply_ratelimit)) goto out_bh_enable; sk = icmp_xmit_lock(net); @@ -450,7 +458,7 @@ static void icmp_reply(struct icmp_bxm *icmp_param, struct sk_buff *skb) rt = ip_route_output_key(net, &fl4); if (IS_ERR(rt)) goto out_unlock; - if (icmpv4_xrlim_allow(net, rt, &fl4, type, code)) + if (icmpv4_xrlim_allow(net, rt, &fl4, type, code, apply_ratelimit)) icmp_push_reply(sk, icmp_param, &fl4, &ipc, &rt); ip_rt_put(rt); out_unlock: @@ -596,6 +604,7 @@ void __icmp_send(struct sk_buff *skb_in, int type, int code, __be32 info, int room; struct icmp_bxm icmp_param; struct rtable *rt = skb_rtable(skb_in); + bool apply_ratelimit = false; struct ipcm_cookie ipc; struct flowi4 fl4; __be32 saddr; @@ -677,7 +686,7 @@ void __icmp_send(struct sk_buff *skb_in, int type, int code, __be32 info, } } - /* Needed by both icmp_global_allow and icmp_xmit_lock */ + /* Needed by both icmpv4_global_allow and icmp_xmit_lock */ local_bh_disable(); /* Check global sysctl_icmp_msgs_per_sec ratelimit, unless @@ -685,7 +694,7 @@ void __icmp_send(struct sk_buff *skb_in, int type, int code, __be32 info, * loopback, then peer ratelimit still work (in icmpv4_xrlim_allow) */ if (!(skb_in->dev && (skb_in->dev->flags&IFF_LOOPBACK)) && - !icmpv4_global_allow(net, type, code)) + !icmpv4_global_allow(net, type, code, &apply_ratelimit)) goto out_bh_enable; sk = icmp_xmit_lock(net); @@ -744,7 +753,7 @@ void __icmp_send(struct sk_buff *skb_in, int type, int code, __be32 info, goto out_unlock; /* peer icmp_ratelimit */ - if (!icmpv4_xrlim_allow(net, rt, &fl4, type, code)) + if (!icmpv4_xrlim_allow(net, rt, &fl4, type, code, apply_ratelimit)) goto ende; /* RFC says return as much as we can without exceeding 576 bytes. */ diff --git a/net/ipv6/icmp.c b/net/ipv6/icmp.c index 7b31674644efc338ec458d92dfe495480825b0fd..46f70e4a835139ef7d8925c49440865355048193 100644 --- a/net/ipv6/icmp.c +++ b/net/ipv6/icmp.c @@ -175,14 +175,16 @@ static bool icmpv6_mask_allow(struct net *net, int type) return false; } -static bool icmpv6_global_allow(struct net *net, int type) +static bool icmpv6_global_allow(struct net *net, int type, + bool *apply_ratelimit) { if (icmpv6_mask_allow(net, type)) return true; - if (icmp_global_allow()) + if (icmp_global_allow()) { + *apply_ratelimit = true; return true; - + } __ICMP_INC_STATS(net, ICMP_MIB_RATELIMITGLOBAL); return false; } @@ -191,13 +193,13 @@ static bool icmpv6_global_allow(struct net *net, int type) * Check the ICMP output rate limit */ static bool icmpv6_xrlim_allow(struct sock *sk, u8 type, - struct flowi6 *fl6) + struct flowi6 *fl6, bool apply_ratelimit) { struct net *net = sock_net(sk); struct dst_entry *dst; bool res = false; - if (icmpv6_mask_allow(net, type)) + if (!apply_ratelimit) return true; /* @@ -228,6 +230,8 @@ static bool icmpv6_xrlim_allow(struct sock *sk, u8 type, if (!res) __ICMP6_INC_STATS(net, ip6_dst_idev(dst), ICMP6_MIB_RATELIMITHOST); + else + icmp_global_consume(); dst_release(dst); return res; } @@ -452,6 +456,7 @@ void icmp6_send(struct sk_buff *skb, u8 type, u8 code, __u32 info, struct net *net; struct ipv6_pinfo *np; const struct in6_addr *saddr = NULL; + bool apply_ratelimit = false; struct dst_entry *dst; struct icmp6hdr tmp_hdr; struct flowi6 fl6; @@ -533,11 +538,12 @@ void icmp6_send(struct sk_buff *skb, u8 type, u8 code, __u32 info, return; } - /* Needed by both icmp_global_allow and icmpv6_xmit_lock */ + /* Needed by both icmpv6_global_allow and icmpv6_xmit_lock */ local_bh_disable(); /* Check global sysctl_icmp_msgs_per_sec ratelimit */ - if (!(skb->dev->flags & IFF_LOOPBACK) && !icmpv6_global_allow(net, type)) + if (!(skb->dev->flags & IFF_LOOPBACK) && + !icmpv6_global_allow(net, type, &apply_ratelimit)) goto out_bh_enable; mip6_addr_swap(skb, parm); @@ -575,7 +581,7 @@ void icmp6_send(struct sk_buff *skb, u8 type, u8 code, __u32 info, np = inet6_sk(sk); - if (!icmpv6_xrlim_allow(sk, type, &fl6)) + if (!icmpv6_xrlim_allow(sk, type, &fl6, apply_ratelimit)) goto out; tmp_hdr.icmp6_type = type; @@ -717,6 +723,7 @@ static enum skb_drop_reason icmpv6_echo_reply(struct sk_buff *skb) struct ipv6_pinfo *np; const struct in6_addr *saddr = NULL; struct icmp6hdr *icmph = icmp6_hdr(skb); + bool apply_ratelimit = false; struct icmp6hdr tmp_hdr; struct flowi6 fl6; struct icmpv6_msg msg; @@ -781,8 +788,9 @@ static enum skb_drop_reason icmpv6_echo_reply(struct sk_buff *skb) goto out; /* Check the ratelimit */ - if ((!(skb->dev->flags & IFF_LOOPBACK) && !icmpv6_global_allow(net, ICMPV6_ECHO_REPLY)) || - !icmpv6_xrlim_allow(sk, ICMPV6_ECHO_REPLY, &fl6)) + if ((!(skb->dev->flags & IFF_LOOPBACK) && + !icmpv6_global_allow(net, ICMPV6_ECHO_REPLY, &apply_ratelimit)) || + !icmpv6_xrlim_allow(sk, ICMPV6_ECHO_REPLY, &fl6, apply_ratelimit)) goto out_dst_release; idev = __in6_dev_get(skb->dev); -- 2.46.0.295.g3b9ea8a38a-goog

7 months, 1 week

2
1
0 0

[PATCH] usb: dwc3: core: update LC timer as per USB Spec V3.2

by Faisal Hassan

This fix addresses STAR 9001285599, which only affects dwc3 core version 320a. The timer value for PM_LC_TIMER in 320a for the Link ECN changes is incorrect. If the PM TIMER ECN is enabled via GUCTL2[19], the link compliance test (TD7.21) may fail. If the ECN is not enabled (GUCTL2[19] = 0), the controller will use the old timer value (5us), which is still acceptable for the link compliance test. Therefore, clear GUCTL2[19] to pass the USB link compliance test: TD 7.21. Cc: stable(a)vger.kernel.org Signed-off-by: Faisal Hassan <quic_faisalh(a)quicinc.com> --- drivers/usb/dwc3/core.c | 15 +++++++++++++++ drivers/usb/dwc3/core.h | 2 ++ 2 files changed, 17 insertions(+) diff --git a/drivers/usb/dwc3/core.c b/drivers/usb/dwc3/core.c index 734de2a8bd21..d0bd3a0e1f9c 100644 --- a/drivers/usb/dwc3/core.c +++ b/drivers/usb/dwc3/core.c @@ -1378,6 +1378,21 @@ static int dwc3_core_init(struct dwc3 *dwc) dwc3_writel(dwc->regs, DWC3_GUCTL2, reg); } + /* + * STAR 9001285599: This issue affects dwc3 core version 3.20a + * only. If the PM TIMER ECM is enabled through GUCTL2[19], the + * link compliance test (TD7.21) may fail. If the ECN is not + * enabled (GUCTL2[19] = 0), the controller will use the old timer + * value (5us), which is still acceptable for the link compliance + * test. Therefore, do not enable PM TIMER ECM in 3.20a by + * setting GUCTL2[19] by default; instead, use GUCTL2[19] = 0. + */ + if (DWC3_VER_IS(DWC3, 320A)) { + reg = dwc3_readl(dwc->regs, DWC3_GUCTL2); + reg &= ~DWC3_GUCTL2_LC_TIMER; + dwc3_writel(dwc->regs, DWC3_GUCTL2, reg); + } + /* * When configured in HOST mode, after issuing U3/L2 exit controller * fails to send proper CRC checksum in CRC5 feild. Because of this diff --git a/drivers/usb/dwc3/core.h b/drivers/usb/dwc3/core.h index 1e561fd8b86e..c71240e8f7c7 100644 --- a/drivers/usb/dwc3/core.h +++ b/drivers/usb/dwc3/core.h @@ -421,6 +421,7 @@ /* Global User Control Register 2 */ #define DWC3_GUCTL2_RST_ACTBITLATER BIT(14) +#define DWC3_GUCTL2_LC_TIMER BIT(19) /* Global User Control Register 3 */ #define DWC3_GUCTL3_SPLITDISABLE BIT(14) @@ -1269,6 +1270,7 @@ struct dwc3 { #define DWC3_REVISION_290A 0x5533290a #define DWC3_REVISION_300A 0x5533300a #define DWC3_REVISION_310A 0x5533310a +#define DWC3_REVISION_320A 0x5533320a #define DWC3_REVISION_330A 0x5533330a #define DWC31_REVISION_ANY 0x0 -- 2.17.1

7 months, 1 week

2
1
0 0

[PATCH v5] usb: dwc3: Avoid waking up gadget during startxfer

by Prashanth K

When operating in High-Speed, it is observed that DSTS[USBLNKST] doesn't update link state immediately after receiving the wakeup interrupt. Since wakeup event handler calls the resume callbacks, there is a chance that function drivers can perform an ep queue, which in turn tries to perform remote wakeup from send_gadget_ep_cmd(STARTXFER). This happens because DSTS[[21:18] wasn't updated to U0 yet, it's observed that the latency of DSTS can be in order of milli-seconds. Hence avoid calling gadget_wakeup during startxfer to prevent unnecessarily issuing remote wakeup to host. Fixes: c36d8e947a56 ("usb: dwc3: gadget: put link to U0 before Start Transfer") Cc: <stable(a)vger.kernel.org> Suggested-by: Thinh Nguyen <Thinh.Nguyen(a)synopsys.com> Signed-off-by: Prashanth K <quic_prashk(a)quicinc.com> --- v5: Further rewording of the comment in function. v4: Rewording the comment in function definition. v3: Added notes on top the function definition. v2: Refactored the patch as suggested in v1 discussion. drivers/usb/dwc3/gadget.c | 41 ++++++++++++++++----------------------- 1 file changed, 17 insertions(+), 24 deletions(-) diff --git a/drivers/usb/dwc3/gadget.c b/drivers/usb/dwc3/gadget.c index 89fc690fdf34..291bc549935b 100644 --- a/drivers/usb/dwc3/gadget.c +++ b/drivers/usb/dwc3/gadget.c @@ -287,6 +287,23 @@ static int __dwc3_gadget_wakeup(struct dwc3 *dwc, bool async); * * Caller should handle locking. This function will issue @cmd with given * @params to @dep and wait for its completion. + * + * According to the programming guide, if the link state is in L1/L2/U3, + * then sending the Start Transfer command may not complete. The + * programming guide suggested to bring the link state back to ON/U0 by + * performing remote wakeup prior to sending the command. However, don't + * initiate remote wakeup when the user/function does not send wakeup + * request via wakeup ops. Send the command when it's allowed. + * + * Notes: + * For L1 link state, issuing a command requires the clearing of + * GUSB2PHYCFG.SUSPENDUSB2, which turns on the signal required to complete + * the given command (usually within 50us). This should happen within the + * command timeout set by driver. No additional step is needed. + * + * For L2 or U3 link state, the gadget is in USB suspend. Care should be + * taken when sending Start Transfer command to ensure that it's done after + * USB resume. */ int dwc3_send_gadget_ep_cmd(struct dwc3_ep *dep, unsigned int cmd, struct dwc3_gadget_ep_cmd_params *params) @@ -327,30 +344,6 @@ int dwc3_send_gadget_ep_cmd(struct dwc3_ep *dep, unsigned int cmd, dwc3_writel(dwc->regs, DWC3_GUSB2PHYCFG(0), reg); } - if (DWC3_DEPCMD_CMD(cmd) == DWC3_DEPCMD_STARTTRANSFER) { - int link_state; - - /* - * Initiate remote wakeup if the link state is in U3 when - * operating in SS/SSP or L1/L2 when operating in HS/FS. If the - * link state is in U1/U2, no remote wakeup is needed. The Start - * Transfer command will initiate the link recovery. - */ - link_state = dwc3_gadget_get_link_state(dwc); - switch (link_state) { - case DWC3_LINK_STATE_U2: - if (dwc->gadget->speed >= USB_SPEED_SUPER) - break; - - fallthrough; - case DWC3_LINK_STATE_U3: - ret = __dwc3_gadget_wakeup(dwc, false); - dev_WARN_ONCE(dwc->dev, ret, "wakeup failed --> %d\n", - ret); - break; - } - } - /* * For some commands such as Update Transfer command, DEPCMDPARn * registers are reserved. Since the driver often sends Update Transfer -- 2.25.1

7 months, 1 week

2
1
0 0

[PATCH 1/1] alloc_tag: fix allocation tag reporting when CONFIG_MODULES=n

by Suren Baghdasaryan

codetag_module_init() is used to initialize sections containing allocation tags. This function is used to initialize module sections as well as core kernel sections, in which case the module parameter is set to NULL. This function has to be called even when CONFIG_MODULES=n to initialize core kernel allocation tag sections. When CONFIG_MODULES=n, this function is a NOP, which is wrong. This leads to /proc/allocinfo reported as empty. Fix this by making it independent of CONFIG_MODULES. Fixes: 916cc5167cc6 ("lib: code tagging framework") Signed-off-by: Suren Baghdasaryan <surenb(a)google.com> Cc: stable(a)vger.kernel.org # v6.10 --- lib/codetag.c | 17 +++++++++++------ 1 file changed, 11 insertions(+), 6 deletions(-) diff --git a/lib/codetag.c b/lib/codetag.c index 5ace625f2328..afa8a2d4f317 100644 --- a/lib/codetag.c +++ b/lib/codetag.c @@ -125,7 +125,6 @@ static inline size_t range_size(const struct codetag_type *cttype, cttype->desc.tag_size; } -#ifdef CONFIG_MODULES static void *get_symbol(struct module *mod, const char *prefix, const char *name) { DECLARE_SEQ_BUF(sb, KSYM_NAME_LEN); @@ -155,6 +154,15 @@ static struct codetag_range get_section_range(struct module *mod, }; } +static const char *get_mod_name(__maybe_unused struct module *mod) +{ +#ifdef CONFIG_MODULES + if (mod) + return mod->name; +#endif + return "(built-in)"; +} + static int codetag_module_init(struct codetag_type *cttype, struct module *mod) { struct codetag_range range; @@ -164,8 +172,7 @@ static int codetag_module_init(struct codetag_type *cttype, struct module *mod) range = get_section_range(mod, cttype->desc.section); if (!range.start || !range.stop) { pr_warn("Failed to load code tags of type %s from the module %s\n", - cttype->desc.section, - mod ? mod->name : "(built-in)"); + cttype->desc.section, get_mod_name(mod)); return -EINVAL; } @@ -199,6 +206,7 @@ static int codetag_module_init(struct codetag_type *cttype, struct module *mod) return 0; } +#ifdef CONFIG_MODULES void codetag_load_module(struct module *mod) { struct codetag_type *cttype; @@ -248,9 +256,6 @@ bool codetag_unload_module(struct module *mod) return unload_ok; } - -#else /* CONFIG_MODULES */ -static int codetag_module_init(struct codetag_type *cttype, struct module *mod) { return 0; } #endif /* CONFIG_MODULES */ struct codetag_type * base-commit: 9287e4adbc6ab8fa04d25eb82e097fed877a4642 -- 2.46.0.295.g3b9ea8a38a-goog

7 months, 1 week

1
0
0 0

FAILED: patch "[PATCH] ata: libata-core: Fix null pointer dereference on error" failed to apply to 5.15-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 5.15-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-5.15.y git checkout FETCH_HEAD git cherry-pick -x 5d92c7c566dc76d96e0e19e481d926bbe6631c1e # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2024070108-cabana-swifter-f1c9@gregkh' --subject-prefix 'PATCH 5.15.y' HEAD^.. Possible dependencies: 5d92c7c566dc ("ata: libata-core: Fix null pointer dereference on error") thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From 5d92c7c566dc76d96e0e19e481d926bbe6631c1e Mon Sep 17 00:00:00 2001 From: Niklas Cassel <cassel(a)kernel.org> Date: Sat, 29 Jun 2024 14:42:11 +0200 Subject: [PATCH] ata: libata-core: Fix null pointer dereference on error If the ata_port_alloc() call in ata_host_alloc() fails, ata_host_release() will get called. However, the code in ata_host_release() tries to free ata_port struct members unconditionally, which can lead to the following: BUG: unable to handle page fault for address: 0000000000003990 PGD 0 P4D 0 Oops: Oops: 0000 [#1] PREEMPT SMP NOPTI CPU: 10 PID: 594 Comm: (udev-worker) Not tainted 6.10.0-rc5 #44 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.16.3-2.fc40 04/01/2014 RIP: 0010:ata_host_release.cold+0x2f/0x6e [libata] Code: e4 4d 63 f4 44 89 e2 48 c7 c6 90 ad 32 c0 48 c7 c7 d0 70 33 c0 49 83 c6 0e 41 RSP: 0018:ffffc90000ebb968 EFLAGS: 00010246 RAX: 0000000000000041 RBX: ffff88810fb52e78 RCX: 0000000000000000 RDX: 0000000000000000 RSI: ffff88813b3218c0 RDI: ffff88813b3218c0 RBP: ffff88810fb52e40 R08: 0000000000000000 R09: 6c65725f74736f68 R10: ffffc90000ebb738 R11: 73692033203a746e R12: 0000000000000004 R13: 0000000000000000 R14: 0000000000000011 R15: 0000000000000006 FS: 00007f6cc55b9980(0000) GS:ffff88813b300000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000000000003990 CR3: 00000001122a2000 CR4: 0000000000750ef0 PKRU: 55555554 Call Trace: <TASK> ? __die_body.cold+0x19/0x27 ? page_fault_oops+0x15a/0x2f0 ? exc_page_fault+0x7e/0x180 ? asm_exc_page_fault+0x26/0x30 ? ata_host_release.cold+0x2f/0x6e [libata] ? ata_host_release.cold+0x2f/0x6e [libata] release_nodes+0x35/0xb0 devres_release_group+0x113/0x140 ata_host_alloc+0xed/0x120 [libata] ata_host_alloc_pinfo+0x14/0xa0 [libata] ahci_init_one+0x6c9/0xd20 [ahci] Do not access ata_port struct members unconditionally. Fixes: 633273a3ed1c ("libata-pmp: hook PMP support and enable it") Cc: stable(a)vger.kernel.org Reviewed-by: Damien Le Moal <dlemoal(a)kernel.org> Reviewed-by: Hannes Reinecke <hare(a)suse.de> Reviewed-by: John Garry <john.g.garry(a)oracle.com> Link: https://lore.kernel.org/r/20240629124210.181537-7-cassel@kernel.org Signed-off-by: Niklas Cassel <cassel(a)kernel.org> diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c index efb5195da60c..bdccf4ea251a 100644 --- a/drivers/ata/libata-core.c +++ b/drivers/ata/libata-core.c @@ -5517,6 +5517,9 @@ static void ata_host_release(struct kref *kref) for (i = 0; i < host->n_ports; i++) { struct ata_port *ap = host->ports[i]; + if (!ap) + continue; + kfree(ap->pmp_link); kfree(ap->slave_link); kfree(ap->ncq_sense_buf);

7 months, 1 week

2
1
0 0

FAILED: patch "[PATCH] ata: libata-core: Fix null pointer dereference on error" failed to apply to 6.1-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 6.1-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-6.1.y git checkout FETCH_HEAD git cherry-pick -x 5d92c7c566dc76d96e0e19e481d926bbe6631c1e # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2024070107-underrate-unusable-ddb9@gregkh' --subject-prefix 'PATCH 6.1.y' HEAD^.. Possible dependencies: 5d92c7c566dc ("ata: libata-core: Fix null pointer dereference on error") thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From 5d92c7c566dc76d96e0e19e481d926bbe6631c1e Mon Sep 17 00:00:00 2001 From: Niklas Cassel <cassel(a)kernel.org> Date: Sat, 29 Jun 2024 14:42:11 +0200 Subject: [PATCH] ata: libata-core: Fix null pointer dereference on error If the ata_port_alloc() call in ata_host_alloc() fails, ata_host_release() will get called. However, the code in ata_host_release() tries to free ata_port struct members unconditionally, which can lead to the following: BUG: unable to handle page fault for address: 0000000000003990 PGD 0 P4D 0 Oops: Oops: 0000 [#1] PREEMPT SMP NOPTI CPU: 10 PID: 594 Comm: (udev-worker) Not tainted 6.10.0-rc5 #44 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.16.3-2.fc40 04/01/2014 RIP: 0010:ata_host_release.cold+0x2f/0x6e [libata] Code: e4 4d 63 f4 44 89 e2 48 c7 c6 90 ad 32 c0 48 c7 c7 d0 70 33 c0 49 83 c6 0e 41 RSP: 0018:ffffc90000ebb968 EFLAGS: 00010246 RAX: 0000000000000041 RBX: ffff88810fb52e78 RCX: 0000000000000000 RDX: 0000000000000000 RSI: ffff88813b3218c0 RDI: ffff88813b3218c0 RBP: ffff88810fb52e40 R08: 0000000000000000 R09: 6c65725f74736f68 R10: ffffc90000ebb738 R11: 73692033203a746e R12: 0000000000000004 R13: 0000000000000000 R14: 0000000000000011 R15: 0000000000000006 FS: 00007f6cc55b9980(0000) GS:ffff88813b300000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000000000003990 CR3: 00000001122a2000 CR4: 0000000000750ef0 PKRU: 55555554 Call Trace: <TASK> ? __die_body.cold+0x19/0x27 ? page_fault_oops+0x15a/0x2f0 ? exc_page_fault+0x7e/0x180 ? asm_exc_page_fault+0x26/0x30 ? ata_host_release.cold+0x2f/0x6e [libata] ? ata_host_release.cold+0x2f/0x6e [libata] release_nodes+0x35/0xb0 devres_release_group+0x113/0x140 ata_host_alloc+0xed/0x120 [libata] ata_host_alloc_pinfo+0x14/0xa0 [libata] ahci_init_one+0x6c9/0xd20 [ahci] Do not access ata_port struct members unconditionally. Fixes: 633273a3ed1c ("libata-pmp: hook PMP support and enable it") Cc: stable(a)vger.kernel.org Reviewed-by: Damien Le Moal <dlemoal(a)kernel.org> Reviewed-by: Hannes Reinecke <hare(a)suse.de> Reviewed-by: John Garry <john.g.garry(a)oracle.com> Link: https://lore.kernel.org/r/20240629124210.181537-7-cassel@kernel.org Signed-off-by: Niklas Cassel <cassel(a)kernel.org> diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c index efb5195da60c..bdccf4ea251a 100644 --- a/drivers/ata/libata-core.c +++ b/drivers/ata/libata-core.c @@ -5517,6 +5517,9 @@ static void ata_host_release(struct kref *kref) for (i = 0; i < host->n_ports; i++) { struct ata_port *ap = host->ports[i]; + if (!ap) + continue; + kfree(ap->pmp_link); kfree(ap->slave_link); kfree(ap->ncq_sense_buf);

7 months, 1 week

2
1
0 0

[PATCH v3] mm: Fix race between __split_huge_pmd_locked() and GUP-fast

by Ryan Roberts

__split_huge_pmd_locked() can be called for a present THP, devmap or (non-present) migration entry. It calls pmdp_invalidate() unconditionally on the pmdp and only determines if it is present or not based on the returned old pmd. This is a problem for the migration entry case because pmd_mkinvalid(), called by pmdp_invalidate() must only be called for a present pmd. On arm64 at least, pmd_mkinvalid() will mark the pmd such that any future call to pmd_present() will return true. And therefore any lockless pgtable walker could see the migration entry pmd in this state and start interpretting the fields as if it were present, leading to BadThings (TM). GUP-fast appears to be one such lockless pgtable walker. x86 does not suffer the above problem, but instead pmd_mkinvalid() will corrupt the offset field of the swap entry within the swap pte. See link below for discussion of that problem. Fix all of this by only calling pmdp_invalidate() for a present pmd. And for good measure let's add a warning to all implementations of pmdp_invalidate[_ad](). I've manually reviewed all other pmdp_invalidate[_ad]() call sites and believe all others to be conformant. This is a theoretical bug found during code review. I don't have any test case to trigger it in practice. Cc: stable(a)vger.kernel.org Link: https://lore.kernel.org/all/0dd7827a-6334-439a-8fd0-43c98e6af22b@arm.com/ Fixes: 84c3fc4e9c56 ("mm: thp: check pmd migration entry in common path") Signed-off-by: Ryan Roberts <ryan.roberts(a)arm.com> --- Right v3; this goes back to the original approach in v1 to fix core-mm rather than push the fix into arm64, since we discovered that x86 can't handle pmd_mkinvalid() being called for non-present pmds either. I'm pulling in more arch maintainers because this version adds some warnings in arch code to help spot incorrect usage. Although Catalin had already accepted v2 (fixing arm64) [2] into for-next/fixes, he's agreed to either remove or revert it. Changes since v1 [1] ==================== - Improve pmdp_mkinvalid() docs to make it clear it can only be called for present pmd (per JohnH, Zi Yan) - Added warnings to arch overrides of pmdp_invalidate[_ad]() (per Zi Yan) - Moved comment next to new location of pmpd_invalidate() (per Zi Yan) [1] https://lore.kernel.org/linux-mm/20240425170704.3379492-1-ryan.roberts@arm.… [2] https://lore.kernel.org/all/20240430133138.732088-1-ryan.roberts@arm.com/ Thanks, Ryan Documentation/mm/arch_pgtable_helpers.rst | 6 ++- arch/powerpc/mm/book3s64/pgtable.c | 1 + arch/s390/include/asm/pgtable.h | 4 +- arch/sparc/mm/tlb.c | 1 + arch/x86/mm/pgtable.c | 2 + mm/huge_memory.c | 49 ++++++++++++----------- mm/pgtable-generic.c | 2 + 7 files changed, 39 insertions(+), 26 deletions(-) diff --git a/Documentation/mm/arch_pgtable_helpers.rst b/Documentation/mm/arch_pgtable_helpers.rst index 2466d3363af7..ad50ca6f495e 100644 --- a/Documentation/mm/arch_pgtable_helpers.rst +++ b/Documentation/mm/arch_pgtable_helpers.rst @@ -140,7 +140,8 @@ PMD Page Table Helpers +---------------------------+--------------------------------------------------+ | pmd_swp_clear_soft_dirty | Clears a soft dirty swapped PMD | +---------------------------+--------------------------------------------------+ -| pmd_mkinvalid | Invalidates a mapped PMD [1] | +| pmd_mkinvalid | Invalidates a present PMD; do not call for | +| | non-present PMD [1] | +---------------------------+--------------------------------------------------+ | pmd_set_huge | Creates a PMD huge mapping | +---------------------------+--------------------------------------------------+ @@ -196,7 +197,8 @@ PUD Page Table Helpers +---------------------------+--------------------------------------------------+ | pud_mkdevmap | Creates a ZONE_DEVICE mapped PUD | +---------------------------+--------------------------------------------------+ -| pud_mkinvalid | Invalidates a mapped PUD [1] | +| pud_mkinvalid | Invalidates a present PUD; do not call for | +| | non-present PUD [1] | +---------------------------+--------------------------------------------------+ | pud_set_huge | Creates a PUD huge mapping | +---------------------------+--------------------------------------------------+ diff --git a/arch/powerpc/mm/book3s64/pgtable.c b/arch/powerpc/mm/book3s64/pgtable.c index 83823db3488b..2975ea0841ba 100644 --- a/arch/powerpc/mm/book3s64/pgtable.c +++ b/arch/powerpc/mm/book3s64/pgtable.c @@ -170,6 +170,7 @@ pmd_t pmdp_invalidate(struct vm_area_struct *vma, unsigned long address, { unsigned long old_pmd; + VM_WARN_ON_ONCE(!pmd_present(*pmdp)); old_pmd = pmd_hugepage_update(vma->vm_mm, address, pmdp, _PAGE_PRESENT, _PAGE_INVALID); flush_pmd_tlb_range(vma, address, address + HPAGE_PMD_SIZE); return __pmd(old_pmd); diff --git a/arch/s390/include/asm/pgtable.h b/arch/s390/include/asm/pgtable.h index 60950e7a25f5..480bea44559d 100644 --- a/arch/s390/include/asm/pgtable.h +++ b/arch/s390/include/asm/pgtable.h @@ -1768,8 +1768,10 @@ static inline pmd_t pmdp_huge_clear_flush(struct vm_area_struct *vma, static inline pmd_t pmdp_invalidate(struct vm_area_struct *vma, unsigned long addr, pmd_t *pmdp) { - pmd_t pmd = __pmd(pmd_val(*pmdp) | _SEGMENT_ENTRY_INVALID); + pmd_t pmd; + VM_WARN_ON_ONCE(!pmd_present(*pmdp)); + pmd = __pmd(pmd_val(*pmdp) | _SEGMENT_ENTRY_INVALID); return pmdp_xchg_direct(vma->vm_mm, addr, pmdp, pmd); } diff --git a/arch/sparc/mm/tlb.c b/arch/sparc/mm/tlb.c index b44d79d778c7..ef69127d7e5e 100644 --- a/arch/sparc/mm/tlb.c +++ b/arch/sparc/mm/tlb.c @@ -249,6 +249,7 @@ pmd_t pmdp_invalidate(struct vm_area_struct *vma, unsigned long address, { pmd_t old, entry; + VM_WARN_ON_ONCE(!pmd_present(*pmdp)); entry = __pmd(pmd_val(*pmdp) & ~_PAGE_VALID); old = pmdp_establish(vma, address, pmdp, entry); flush_tlb_range(vma, address, address + HPAGE_PMD_SIZE); diff --git a/arch/x86/mm/pgtable.c b/arch/x86/mm/pgtable.c index d007591b8059..103cbccf1d7d 100644 --- a/arch/x86/mm/pgtable.c +++ b/arch/x86/mm/pgtable.c @@ -631,6 +631,8 @@ int pmdp_clear_flush_young(struct vm_area_struct *vma, pmd_t pmdp_invalidate_ad(struct vm_area_struct *vma, unsigned long address, pmd_t *pmdp) { + VM_WARN_ON_ONCE(!pmd_present(*pmdp)); + /* * No flush is necessary. Once an invalid PTE is established, the PTE's * access and dirty bits cannot be updated. diff --git a/mm/huge_memory.c b/mm/huge_memory.c index 89f58c7603b2..dd1fc105f70b 100644 --- a/mm/huge_memory.c +++ b/mm/huge_memory.c @@ -2493,32 +2493,11 @@ static void __split_huge_pmd_locked(struct vm_area_struct *vma, pmd_t *pmd, return __split_huge_zero_page_pmd(vma, haddr, pmd); } - /* - * Up to this point the pmd is present and huge and userland has the - * whole access to the hugepage during the split (which happens in - * place). If we overwrite the pmd with the not-huge version pointing - * to the pte here (which of course we could if all CPUs were bug - * free), userland could trigger a small page size TLB miss on the - * small sized TLB while the hugepage TLB entry is still established in - * the huge TLB. Some CPU doesn't like that. - * See http://support.amd.com/TechDocs/41322_10h_Rev_Gd.pdf, Erratum - * 383 on page 105. Intel should be safe but is also warns that it's - * only safe if the permission and cache attributes of the two entries - * loaded in the two TLB is identical (which should be the case here). - * But it is generally safer to never allow small and huge TLB entries - * for the same virtual address to be loaded simultaneously. So instead - * of doing "pmd_populate(); flush_pmd_tlb_range();" we first mark the - * current pmd notpresent (atomically because here the pmd_trans_huge - * must remain set at all times on the pmd until the split is complete - * for this pmd), then we flush the SMP TLB and finally we write the - * non-huge version of the pmd entry with pmd_populate. - */ - old_pmd = pmdp_invalidate(vma, haddr, pmd); - - pmd_migration = is_pmd_migration_entry(old_pmd); + pmd_migration = is_pmd_migration_entry(*pmd); if (unlikely(pmd_migration)) { swp_entry_t entry; + old_pmd = *pmd; entry = pmd_to_swp_entry(old_pmd); page = pfn_swap_entry_to_page(entry); write = is_writable_migration_entry(entry); @@ -2529,6 +2508,30 @@ static void __split_huge_pmd_locked(struct vm_area_struct *vma, pmd_t *pmd, soft_dirty = pmd_swp_soft_dirty(old_pmd); uffd_wp = pmd_swp_uffd_wp(old_pmd); } else { + /* + * Up to this point the pmd is present and huge and userland has + * the whole access to the hugepage during the split (which + * happens in place). If we overwrite the pmd with the not-huge + * version pointing to the pte here (which of course we could if + * all CPUs were bug free), userland could trigger a small page + * size TLB miss on the small sized TLB while the hugepage TLB + * entry is still established in the huge TLB. Some CPU doesn't + * like that. See + * http://support.amd.com/TechDocs/41322_10h_Rev_Gd.pdf, Erratum + * 383 on page 105. Intel should be safe but is also warns that + * it's only safe if the permission and cache attributes of the + * two entries loaded in the two TLB is identical (which should + * be the case here). But it is generally safer to never allow + * small and huge TLB entries for the same virtual address to be + * loaded simultaneously. So instead of doing "pmd_populate(); + * flush_pmd_tlb_range();" we first mark the current pmd + * notpresent (atomically because here the pmd_trans_huge must + * remain set at all times on the pmd until the split is + * complete for this pmd), then we flush the SMP TLB and finally + * we write the non-huge version of the pmd entry with + * pmd_populate. + */ + old_pmd = pmdp_invalidate(vma, haddr, pmd); page = pmd_page(old_pmd); folio = page_folio(page); if (pmd_dirty(old_pmd)) { diff --git a/mm/pgtable-generic.c b/mm/pgtable-generic.c index 4fcd959dcc4d..a78a4adf711a 100644 --- a/mm/pgtable-generic.c +++ b/mm/pgtable-generic.c @@ -198,6 +198,7 @@ pgtable_t pgtable_trans_huge_withdraw(struct mm_struct *mm, pmd_t *pmdp) pmd_t pmdp_invalidate(struct vm_area_struct *vma, unsigned long address, pmd_t *pmdp) { + VM_WARN_ON_ONCE(!pmd_present(*pmdp)); pmd_t old = pmdp_establish(vma, address, pmdp, pmd_mkinvalid(*pmdp)); flush_pmd_tlb_range(vma, address, address + HPAGE_PMD_SIZE); return old; @@ -208,6 +209,7 @@ pmd_t pmdp_invalidate(struct vm_area_struct *vma, unsigned long address, pmd_t pmdp_invalidate_ad(struct vm_area_struct *vma, unsigned long address, pmd_t *pmdp) { + VM_WARN_ON_ONCE(!pmd_present(*pmdp)); return pmdp_invalidate(vma, address, pmdp); } #endif -- 2.25.1

7 months, 1 week

5
8
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror August 2024