April 2025 - Linux-stable-mirror

[PATCH] wifi: ath11k: fix rx completion meta data corruption

by Johan Hovold

Add the missing memory barrier to make sure that the REO dest ring descriptor is read after the head pointer to avoid using stale data on weakly ordered architectures like aarch64. This may fix the ring-buffer corruption worked around by commit f9fff67d2d7c ("wifi: ath11k: Fix SKB corruption in REO destination ring") by silently discarding data, and may possibly also address user reported errors like: ath11k_pci 0006:01:00.0: msdu_done bit in attention is not set Tested-on: WCN6855 hw2.1 WLAN.HSP.1.1-03125-QCAHSPSWPL_V1_V2_SILICONZ_LITE-3.6510.41 Fixes: d5c65159f289 ("ath11k: driver for Qualcomm IEEE 802.11ax devices") Cc: stable(a)vger.kernel.org # 5.6 Link: https://bugzilla.kernel.org/show_bug.cgi?id=218005 Signed-off-by: Johan Hovold <johan+linaro(a)kernel.org> --- As I reported here: https://lore.kernel.org/lkml/Z9G5zEOcTdGKm7Ei@hovoldconsulting.com/ the ath11k and ath12k appear to be missing a number of memory barriers that are required on weakly ordered architectures like aarch64 to avoid memory corruption issues. Here's a fix for one more such case which people already seem to be hitting. Note that I've seen one "msdu_done" bit not set warning also with this patch so whether it helps with that at all remains to be seen. I'm CCing Jens and Steev that see these warnings frequently and that may be able to help out with testing. Johan drivers/net/wireless/ath/ath11k/dp_rx.c | 25 ++++++++++++++++--------- 1 file changed, 16 insertions(+), 9 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/dp_rx.c b/drivers/net/wireless/ath/ath11k/dp_rx.c index 029ecf51c9ef..0a57b337e4c6 100644 --- a/drivers/net/wireless/ath/ath11k/dp_rx.c +++ b/drivers/net/wireless/ath/ath11k/dp_rx.c @@ -2646,7 +2646,7 @@ int ath11k_dp_process_rx(struct ath11k_base *ab, int ring_id, struct ath11k *ar; struct hal_reo_dest_ring *desc; enum hal_reo_dest_ring_push_reason push_reason; - u32 cookie; + u32 cookie, info0, rx_msdu_info0, rx_mpdu_info0; int i; for (i = 0; i < MAX_RADIOS; i++) @@ -2659,11 +2659,14 @@ int ath11k_dp_process_rx(struct ath11k_base *ab, int ring_id, try_again: ath11k_hal_srng_access_begin(ab, srng); + /* Make sure descriptor is read after the head pointer. */ + dma_rmb(); + while (likely(desc = (struct hal_reo_dest_ring *)ath11k_hal_srng_dst_get_next_entry(ab, srng))) { cookie = FIELD_GET(BUFFER_ADDR_INFO1_SW_COOKIE, - desc->buf_addr_info.info1); + READ_ONCE(desc->buf_addr_info.info1)); buf_id = FIELD_GET(DP_RXDMA_BUF_COOKIE_BUF_ID, cookie); mac_id = FIELD_GET(DP_RXDMA_BUF_COOKIE_PDEV_ID, cookie); @@ -2692,8 +2695,9 @@ int ath11k_dp_process_rx(struct ath11k_base *ab, int ring_id, num_buffs_reaped[mac_id]++; + info0 = READ_ONCE(desc->info0); push_reason = FIELD_GET(HAL_REO_DEST_RING_INFO0_PUSH_REASON, - desc->info0); + info0); if (unlikely(push_reason != HAL_REO_DEST_RING_PUSH_REASON_ROUTING_INSTRUCTION)) { dev_kfree_skb_any(msdu); @@ -2701,18 +2705,21 @@ int ath11k_dp_process_rx(struct ath11k_base *ab, int ring_id, continue; } - rxcb->is_first_msdu = !!(desc->rx_msdu_info.info0 & + rx_msdu_info0 = READ_ONCE(desc->rx_msdu_info.info0); + rx_mpdu_info0 = READ_ONCE(desc->rx_mpdu_info.info0); + + rxcb->is_first_msdu = !!(rx_msdu_info0 & RX_MSDU_DESC_INFO0_FIRST_MSDU_IN_MPDU); - rxcb->is_last_msdu = !!(desc->rx_msdu_info.info0 & + rxcb->is_last_msdu = !!(rx_msdu_info0 & RX_MSDU_DESC_INFO0_LAST_MSDU_IN_MPDU); - rxcb->is_continuation = !!(desc->rx_msdu_info.info0 & + rxcb->is_continuation = !!(rx_msdu_info0 & RX_MSDU_DESC_INFO0_MSDU_CONTINUATION); rxcb->peer_id = FIELD_GET(RX_MPDU_DESC_META_DATA_PEER_ID, - desc->rx_mpdu_info.meta_data); + READ_ONCE(desc->rx_mpdu_info.meta_data)); rxcb->seq_no = FIELD_GET(RX_MPDU_DESC_INFO0_SEQ_NUM, - desc->rx_mpdu_info.info0); + rx_mpdu_info0); rxcb->tid = FIELD_GET(HAL_REO_DEST_RING_INFO0_RX_QUEUE_NUM, - desc->info0); + info0); rxcb->mac_id = mac_id; __skb_queue_tail(&msdu_list[mac_id], msdu); -- 2.48.1

4 months

5
7
0 0

[PATCH] wifi: ath11k: fix ring-buffer corruption

by Johan Hovold

Users of the Lenovo ThinkPad X13s have reported that Wi-Fi sometimes breaks and the log fills up with errors like: ath11k_pci 0006:01:00.0: HTC Rx: insufficient length, got 1484, expected 1492 ath11k_pci 0006:01:00.0: HTC Rx: insufficient length, got 1460, expected 1484 which based on a quick look at the driver seemed to indicate some kind of ring-buffer corruption. Miaoqing Pan tracked it down to the host seeing the updated destination ring head pointer before the updated descriptor, and the error handling for that in turn leaves the ring buffer in an inconsistent state. Add the missing memory barrier to make sure that the descriptor is read after the head pointer to address the root cause of the corruption while fixing up the error handling in case there are ever any (ordering) bugs on the device side. Note that the READ_ONCE() are only needed to avoid compiler mischief in case the ring-buffer helpers are ever inlined. Tested-on: WCN6855 hw2.1 WLAN.HSP.1.1-03125-QCAHSPSWPL_V1_V2_SILICONZ_LITE-3.6510.41 Fixes: d5c65159f289 ("ath11k: driver for Qualcomm IEEE 802.11ax devices") Closes: https://bugzilla.kernel.org/show_bug.cgi?id=218623 Link: https://lore.kernel.org/20250310010217.3845141-3-quic_miaoqing@quicinc.com Cc: Miaoqing Pan <quic_miaoqing(a)quicinc.com> Cc: stable(a)vger.kernel.org # 5.6 Signed-off-by: Johan Hovold <johan+linaro(a)kernel.org> --- drivers/net/wireless/ath/ath11k/ce.c | 11 +++++------ drivers/net/wireless/ath/ath11k/hal.c | 4 ++-- 2 files changed, 7 insertions(+), 8 deletions(-) diff --git a/drivers/net/wireless/ath/ath11k/ce.c b/drivers/net/wireless/ath/ath11k/ce.c index e66e86bdec20..9d8efec46508 100644 --- a/drivers/net/wireless/ath/ath11k/ce.c +++ b/drivers/net/wireless/ath/ath11k/ce.c @@ -393,11 +393,10 @@ static int ath11k_ce_completed_recv_next(struct ath11k_ce_pipe *pipe, goto err; } + /* Make sure descriptor is read after the head pointer. */ + dma_rmb(); + *nbytes = ath11k_hal_ce_dst_status_get_length(desc); - if (*nbytes == 0) { - ret = -EIO; - goto err; - } *skb = pipe->dest_ring->skb[sw_index]; pipe->dest_ring->skb[sw_index] = NULL; @@ -430,8 +429,8 @@ static void ath11k_ce_recv_process_cb(struct ath11k_ce_pipe *pipe) dma_unmap_single(ab->dev, ATH11K_SKB_RXCB(skb)->paddr, max_nbytes, DMA_FROM_DEVICE); - if (unlikely(max_nbytes < nbytes)) { - ath11k_warn(ab, "rxed more than expected (nbytes %d, max %d)", + if (unlikely(max_nbytes < nbytes || nbytes == 0)) { + ath11k_warn(ab, "unexpected rx length (nbytes %d, max %d)", nbytes, max_nbytes); dev_kfree_skb_any(skb); continue; diff --git a/drivers/net/wireless/ath/ath11k/hal.c b/drivers/net/wireless/ath/ath11k/hal.c index 61f4b6dd5380..8cb1505a5a0c 100644 --- a/drivers/net/wireless/ath/ath11k/hal.c +++ b/drivers/net/wireless/ath/ath11k/hal.c @@ -599,7 +599,7 @@ u32 ath11k_hal_ce_dst_status_get_length(void *buf) struct hal_ce_srng_dst_status_desc *desc = buf; u32 len; - len = FIELD_GET(HAL_CE_DST_STATUS_DESC_FLAGS_LEN, desc->flags); + len = FIELD_GET(HAL_CE_DST_STATUS_DESC_FLAGS_LEN, READ_ONCE(desc->flags)); desc->flags &= ~HAL_CE_DST_STATUS_DESC_FLAGS_LEN; return len; @@ -829,7 +829,7 @@ void ath11k_hal_srng_access_begin(struct ath11k_base *ab, struct hal_srng *srng) srng->u.src_ring.cached_tp = *(volatile u32 *)srng->u.src_ring.tp_addr; } else { - srng->u.dst_ring.cached_hp = *srng->u.dst_ring.hp_addr; + srng->u.dst_ring.cached_hp = READ_ONCE(*srng->u.dst_ring.hp_addr); /* Try to prefetch the next descriptor in the ring */ if (srng->flags & HAL_SRNG_FLAGS_CACHED) -- 2.48.1

4 months

7
6
0 0

[PATCH stable 5.15/6.1/6.6] af_unix: Clear oob_skb in scan_inflight().

by Kuniyuki Iwashima

Embryo socket is not queued in gc_candidates, so we can't drop a reference held by its oob_skb. Let's say we create listener and embryo sockets, send the listener's fd to the embryo as OOB data, and close() them without recv()ing the OOB data. There is a self-reference cycle like listener -> embryo.oob_skb -> listener , so this must be cleaned up by GC. Otherwise, the listener's refcnt is not released and sockets are leaked: # unshare -n # cat /proc/net/protocols | grep UNIX-STREAM UNIX-STREAM 1024 0 -1 NI 0 yes kernel ... # python3 >>> from array import array >>> from socket import * >>> >>> s = socket(AF_UNIX, SOCK_STREAM) >>> s.bind('\0test\0') >>> s.listen() >>> >>> c = socket(AF_UNIX, SOCK_STREAM) >>> c.connect(s.getsockname()) >>> c.sendmsg([b'x'], [(SOL_SOCKET, SCM_RIGHTS, array('i', [s.fileno()]))], MSG_OOB) 1 >>> quit() # cat /proc/net/protocols | grep UNIX-STREAM UNIX-STREAM 1024 3 -1 NI 0 yes kernel ... ^^^ 3 sockets still in use after FDs are close()d Let's drop the embryo socket's oob_skb ref in scan_inflight(). This also fixes a racy access to oob_skb that commit 9841991a446c ("af_unix: Update unix_sk(sk)->oob_skb under sk_receive_queue lock.") fixed for the new Tarjan's algo-based GC. Fixes: 314001f0bf92 ("af_unix: Add OOB support") Reported-by: Lei Lu <llfamsec(a)gmail.com> Signed-off-by: Kuniyuki Iwashima <kuniyu(a)amazon.com> --- This has no upstream commit because I replaced the entire GC in 6.10 and the new GC does not have this bug, and this fix is only applicable to the old GC (<= 6.9), thus for 5.15/6.1/6.6. --- --- net/unix/garbage.c | 20 ++++++++++---------- 1 file changed, 10 insertions(+), 10 deletions(-) diff --git a/net/unix/garbage.c b/net/unix/garbage.c index 2a758531e102..b3fbdf129944 100644 --- a/net/unix/garbage.c +++ b/net/unix/garbage.c @@ -102,13 +102,14 @@ static void scan_inflight(struct sock *x, void (*func)(struct unix_sock *), /* Process the descriptors of this socket */ int nfd = UNIXCB(skb).fp->count; struct file **fp = UNIXCB(skb).fp->fp; + struct unix_sock *u; while (nfd--) { /* Get the socket the fd matches if it indeed does so */ struct sock *sk = unix_get_socket(*fp++); if (sk) { - struct unix_sock *u = unix_sk(sk); + u = unix_sk(sk); /* Ignore non-candidates, they could * have been added to the queues after @@ -122,6 +123,13 @@ static void scan_inflight(struct sock *x, void (*func)(struct unix_sock *), } } if (hit && hitlist != NULL) { +#if IS_ENABLED(CONFIG_AF_UNIX_OOB) + u = unix_sk(x); + if (u->oob_skb) { + WARN_ON_ONCE(skb_unref(u->oob_skb)); + u->oob_skb = NULL; + } +#endif __skb_unlink(skb, &x->sk_receive_queue); __skb_queue_tail(hitlist, skb); } @@ -299,17 +307,9 @@ void unix_gc(void) * which are creating the cycle(s). */ skb_queue_head_init(&hitlist); - list_for_each_entry(u, &gc_candidates, link) { + list_for_each_entry(u, &gc_candidates, link) scan_children(&u->sk, inc_inflight, &hitlist); -#if IS_ENABLED(CONFIG_AF_UNIX_OOB) - if (u->oob_skb) { - kfree_skb(u->oob_skb); - u->oob_skb = NULL; - } -#endif - } - /* not_cycle_list contains those sockets which do not make up a * cycle. Restore these to the inflight list. */ -- 2.39.5 (Apple Git-154)

4 months

5
8
0 0

[PATCH RESEND] drm/xe/display: Add check for alloc_ordered_workqueue()

by Haoxiang Li

Add check for the return value of alloc_ordered_workqueue() in xe_display_create() to catch potential exception. Fixes: 44e694958b95 ("drm/xe/display: Implement display support") Cc: stable(a)vger.kernel.org Signed-off-by: Haoxiang Li <haoxiang_li2024(a)163.com> --- drivers/gpu/drm/xe/display/xe_display.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/gpu/drm/xe/display/xe_display.c b/drivers/gpu/drm/xe/display/xe_display.c index 0b0aca7a25af..18062cfb265f 100644 --- a/drivers/gpu/drm/xe/display/xe_display.c +++ b/drivers/gpu/drm/xe/display/xe_display.c @@ -104,6 +104,8 @@ int xe_display_create(struct xe_device *xe) spin_lock_init(&xe->display.fb_tracking.lock); xe->display.hotplug.dp_wq = alloc_ordered_workqueue("xe-dp", 0); + if (!xe->display.hotplug.dp_wq) + return -ENOMEM; return drmm_add_action_or_reset(&xe->drm, display_destroy, NULL); } -- 2.25.1

4 months

2
1
0 0

[PATCH v2 RESEND] drm/i915/display: Add check for alloc_ordered_workqueue() and alloc_workqueue()

by Haoxiang Li

Add check for the return value of alloc_ordered_workqueue() and alloc_workqueue(). Furthermore, if some allocations fail, cleanup works are added to avoid potential memory leak problem. Fixes: 40053823baad ("drm/i915/display: move modeset probe/remove functions to intel_display_driver.c") Cc: stable(a)vger.kernel.org Signed-off-by: Haoxiang Li <haoxiang_li2024(a)163.com> --- Changes in v2: - Split the compound conditional statement into separate conditional statements to facilitate cleanup works. - Add cleanup works to destory work queues if allocations fail, and modify the later goto destination to do the full excercise. - modify the patch description. Thanks, Jani! --- .../drm/i915/display/intel_display_driver.c | 30 +++++++++++++++---- 1 file changed, 25 insertions(+), 5 deletions(-) diff --git a/drivers/gpu/drm/i915/display/intel_display_driver.c b/drivers/gpu/drm/i915/display/intel_display_driver.c index 31740a677dd8..ac94561715dc 100644 --- a/drivers/gpu/drm/i915/display/intel_display_driver.c +++ b/drivers/gpu/drm/i915/display/intel_display_driver.c @@ -241,31 +241,45 @@ int intel_display_driver_probe_noirq(struct intel_display *display) intel_dmc_init(display); display->wq.modeset = alloc_ordered_workqueue("i915_modeset", 0); + if (!display->wq.modeset) { + ret = -ENOMEM; + goto cleanup_vga_client_pw_domain_dmc; + } + display->wq.flip = alloc_workqueue("i915_flip", WQ_HIGHPRI | WQ_UNBOUND, WQ_UNBOUND_MAX_ACTIVE); + if (!display->wq.flip) { + ret = -ENOMEM; + goto cleanup_wq_modeset; + } + display->wq.cleanup = alloc_workqueue("i915_cleanup", WQ_HIGHPRI, 0); + if (!display->wq.cleanup) { + ret = -ENOMEM; + goto cleanup_wq_flip; + } intel_mode_config_init(display); ret = intel_cdclk_init(display); if (ret) - goto cleanup_vga_client_pw_domain_dmc; + goto cleanup_wq_cleanup; ret = intel_color_init(display); if (ret) - goto cleanup_vga_client_pw_domain_dmc; + goto cleanup_wq_cleanup; ret = intel_dbuf_init(i915); if (ret) - goto cleanup_vga_client_pw_domain_dmc; + goto cleanup_wq_cleanup; ret = intel_bw_init(i915); if (ret) - goto cleanup_vga_client_pw_domain_dmc; + goto cleanup_wq_cleanup; ret = intel_pmdemand_init(display); if (ret) - goto cleanup_vga_client_pw_domain_dmc; + goto cleanup_wq_cleanup; intel_init_quirks(display); @@ -273,6 +287,12 @@ int intel_display_driver_probe_noirq(struct intel_display *display) return 0; +cleanup_wq_cleanup: + destroy_workqueue(display->wq.cleanup); +cleanup_wq_flip: + destroy_workqueue(display->wq.flip); +cleanup_wq_modeset: + destroy_workqueue(display->wq.modeset); cleanup_vga_client_pw_domain_dmc: intel_dmc_fini(display); intel_power_domains_driver_remove(display); -- 2.25.1

4 months

2
1
0 0

[PATCH v2 0/5] Fix onboard USB hub instability on RK3399 Puma SoM

by Lukasz Czechowski

The RK3399 Puma SoM contains the internal Cypress CYUSB3304 USB hub, that shows instability due to improper reset pin configuration. Currently reset pin is modeled as a vcc5v0_host regulator, that might result in too short reset pulse duration. Starting with the v6.6, the Onboard USB hub driver (later renamed to Onboard USB dev) contains support for Cypress HX3 hub family. It can be now used to correctly model the RK3399 Puma SoM hardware. The first commits in this series fix the onboard USB dev driver to support all HX3 hub variants, including the CYUSB3304 found in the RK3399 Puma SoM. This allows to introduce fix for internal USB hub instability on RK3399 Puma, by replacing the vcc5v0_host regulator with cy3304_reset, used inside the hub node. Please be aware that the patch that fixes USB hub instability in arch/arm64/boot/dts/rockchip/rk3399-puma.dtsi can me merged only after updating the Onboard USB dev driver, otherwise the hub will not work. Two last commits in the series disable unrouted USB controllers and PHYs on RK3399 Puma SOM and Haikou carrier board, with no intended functional changes. This series depends on the patch: Link: https://lore.kernel.org/linux-usb/20250418-dt-binding-usb-device-compatible… ("dt-bindings: usb: usb-device: relax compatible pattern to a contains") Signed-off-by: Lukasz Czechowski <lukasz.czechowski(a)thaumatec.com> --- Changes in v2: - Removed additional entries from onboard_dev_match table and updated dt-bindings list, as suggested by Krzysztof and Conor. Fallback compatible entry in SoM's dtsi file is used instead. - Added vdd-supply and vdd2-supply entries to onboard hub nodes to satisfy bindings checks. - Changed the default cy3304-reset pin configuration to pcfg_output_high. - Added dependency to: change-id: 20250415-dt-binding-usb-device-compatibles-188f7b0a81b4 - Link to v1: https://lore.kernel.org/r/20250326-onboard_usb_dev-v1-0-a4b0a5d1b32c@thauma… --- Lukasz Czechowski (3): usb: misc: onboard_usb_dev: fix support for Cypress HX3 hubs dt-bindings: usb: cypress,hx3: Add support for all variants arm64: dts: rockchip: fix internal USB hub instability on RK3399 Puma Quentin Schulz (2): arm64: dts: rockchip: disable unrouted USB controllers and PHY on RK3399 Puma arm64: dts: rockchip: disable unrouted USB controllers and PHY on RK3399 Puma with Haikou .../devicetree/bindings/usb/cypress,hx3.yaml | 19 +++++++-- .../arm64/boot/dts/rockchip/rk3399-puma-haikou.dts | 8 ---- arch/arm64/boot/dts/rockchip/rk3399-puma.dtsi | 48 +++++++++++----------- drivers/usb/misc/onboard_usb_dev.c | 10 ++++- 4 files changed, 48 insertions(+), 37 deletions(-) --- base-commit: 834a4a689699090a406d1662b03affa8b155d025 change-id: 20250326-onboard_usb_dev-a7c063a8a515 prerequisite-change-id: 20250415-dt-binding-usb-device-compatibles-188f7b0a81b4:v2 prerequisite-patch-id: f5b90f95302ac9065fbbe5244cc7845c2a772ab6 Best regards, -- Lukasz Czechowski <lukasz.czechowski(a)thaumatec.com>

4 months

5
9
0 0

[PATCH 0/2] PCI: of: Fix OF device node refcount leakages

by Zijun Hu

This patch series is to fix OF device node refcount leakage for - of_irq_parse_and_map_pci() - of_pci_prop_intr_map() Signed-off-by: Zijun Hu <quic_zijuhu(a)quicinc.com> --- Zijun Hu (2): PCI: of: Fix OF device node refcount leakage in API of_irq_parse_and_map_pci() PCI: of: Fix OF device node refcount leakages in of_pci_prop_intr_map() drivers/pci/of.c | 2 ++ drivers/pci/of_property.c | 20 +++++++++++--------- 2 files changed, 13 insertions(+), 9 deletions(-) --- base-commit: 7d06015d936c861160803e020f68f413b5c3cd9d change-id: 20250407-fix_of_pci-20b45dcc26b5 Best regards, -- Zijun Hu <quic_zijuhu(a)quicinc.com>

4 months

1
2
0 0

[PATCH v2 RESEND] phy: Fix error handling in tegra_xusb_port_init

by Ma Ke

If device_add() fails, do not use device_unregister() for error handling. device_unregister() consists two functions: device_del() and put_device(). device_unregister() should only be called after device_add() succeeded because device_del() undoes what device_add() does if successful. Change device_unregister() to put_device() call before returning from the function. As comment of device_add() says, 'if device_add() succeeds, you should call device_del() when you want to get rid of it. If device_add() has not succeeded, use only put_device() to drop the reference count'. Found by code review. Cc: stable(a)vger.kernel.org Fixes: 53d2a715c240 ("phy: Add Tegra XUSB pad controller support") Signed-off-by: Ma Ke <make24(a)iscas.ac.cn> --- Changes in v2: - modified the bug description as suggestions. --- drivers/phy/tegra/xusb.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/phy/tegra/xusb.c b/drivers/phy/tegra/xusb.c index 79d4814d758d..c89df95aa6ca 100644 --- a/drivers/phy/tegra/xusb.c +++ b/drivers/phy/tegra/xusb.c @@ -548,16 +548,16 @@ static int tegra_xusb_port_init(struct tegra_xusb_port *port, err = dev_set_name(&port->dev, "%s-%u", name, index); if (err < 0) - goto unregister; + goto put_device; err = device_add(&port->dev); if (err < 0) - goto unregister; + goto put_device; return 0; -unregister: - device_unregister(&port->dev); +put_device: + put_device(&port->dev); return err; } -- 2.25.1

4 months

3
3
0 0

[PATCH v3] x86/sev: Do not touch VMSA pages during kdump of SNP guest memory

by Ashish Kalra

From: Ashish Kalra <ashish.kalra(a)amd.com> When kdump is running makedumpfile to generate vmcore and dumping SNP guest memory it touches the VMSA page of the vCPU executing kdump which then results in unrecoverable #NPF/RMP faults as the VMSA page is marked busy/in-use when the vCPU is running and subsequently causes guest softlockup/hang. Additionally other APs may be halted in guest mode and their VMSA pages are marked busy and touching these VMSA pages during guest memory dump will also cause #NPF. Issue AP_DESTROY GHCB calls on other APs to ensure they are kicked out of guest mode and then clear the VMSA bit on their VMSA pages. If the vCPU running kdump is an AP, mark it's VMSA page as offline to ensure that makedumpfile excludes that page while dumping guest memory. Cc: stable(a)vger.kernel.org Fixes: 3074152e56c9 ("x86/sev: Convert shared memory back to private on kexec") Signed-off-by: Ashish Kalra <ashish.kalra(a)amd.com> --- arch/x86/coco/sev/core.c | 241 +++++++++++++++++++++++++-------------- 1 file changed, 155 insertions(+), 86 deletions(-) diff --git a/arch/x86/coco/sev/core.c b/arch/x86/coco/sev/core.c index dcfaa698d6cf..f4eb5b645239 100644 --- a/arch/x86/coco/sev/core.c +++ b/arch/x86/coco/sev/core.c @@ -877,6 +877,99 @@ void snp_accept_memory(phys_addr_t start, phys_addr_t end) set_pages_state(vaddr, npages, SNP_PAGE_STATE_PRIVATE); } +static int vmgexit_ap_control(u64 event, struct sev_es_save_area *vmsa, u32 apic_id) +{ + struct ghcb_state state; + unsigned long flags; + struct ghcb *ghcb; + int ret = 0; + + local_irq_save(flags); + + ghcb = __sev_get_ghcb(&state); + + vc_ghcb_invalidate(ghcb); + if (event == SVM_VMGEXIT_AP_CREATE) + ghcb_set_rax(ghcb, vmsa->sev_features); + ghcb_set_sw_exit_code(ghcb, SVM_VMGEXIT_AP_CREATION); + ghcb_set_sw_exit_info_1(ghcb, + ((u64)apic_id << 32) | + ((u64)snp_vmpl << 16) | + event); + ghcb_set_sw_exit_info_2(ghcb, __pa(vmsa)); + + sev_es_wr_ghcb_msr(__pa(ghcb)); + VMGEXIT(); + + if (!ghcb_sw_exit_info_1_is_valid(ghcb) || + lower_32_bits(ghcb->save.sw_exit_info_1)) { + pr_err("SNP AP %s error\n", (event == SVM_VMGEXIT_AP_CREATE ? "CREATE" : "DESTROY")); + ret = -EINVAL; + } + + __sev_put_ghcb(&state); + + local_irq_restore(flags); + + return ret; +} + +static int snp_set_vmsa(void *va, void *caa, int apic_id, bool make_vmsa) +{ + int ret; + + if (snp_vmpl) { + struct svsm_call call = {}; + unsigned long flags; + + local_irq_save(flags); + + call.caa = this_cpu_read(svsm_caa); + call.rcx = __pa(va); + + if (make_vmsa) { + /* Protocol 0, Call ID 2 */ + call.rax = SVSM_CORE_CALL(SVSM_CORE_CREATE_VCPU); + call.rdx = __pa(caa); + call.r8 = apic_id; + } else { + /* Protocol 0, Call ID 3 */ + call.rax = SVSM_CORE_CALL(SVSM_CORE_DELETE_VCPU); + } + + ret = svsm_perform_call_protocol(&call); + + local_irq_restore(flags); + } else { + /* + * If the kernel runs at VMPL0, it can change the VMSA + * bit for a page using the RMPADJUST instruction. + * However, for the instruction to succeed it must + * target the permissions of a lesser privileged (higher + * numbered) VMPL level, so use VMPL1. + */ + u64 attrs = 1; + + if (make_vmsa) + attrs |= RMPADJUST_VMSA_PAGE_BIT; + + ret = rmpadjust((unsigned long)va, RMP_PG_SIZE_4K, attrs); + } + + return ret; +} + +static void snp_cleanup_vmsa(struct sev_es_save_area *vmsa, int apic_id) +{ + int err; + + err = snp_set_vmsa(vmsa, NULL, apic_id, false); + if (err) + pr_err("clear VMSA page failed (%u), leaking page\n", err); + else + free_page((unsigned long)vmsa); +} + static void set_pte_enc(pte_t *kpte, int level, void *va) { struct pte_enc_desc d = { @@ -973,6 +1066,65 @@ void snp_kexec_begin(void) pr_warn("Failed to stop shared<->private conversions\n"); } +/* + * Shutdown all APs except the one handling kexec/kdump and clearing + * the VMSA tag on AP's VMSA pages as they are not being used as + * VMSA page anymore. + */ +static void shutdown_all_aps(void) +{ + struct sev_es_save_area *vmsa; + int apic_id, this_cpu, cpu; + + this_cpu = get_cpu(); + + /* + * APs are already in HLT loop when enc_kexec_finish() callback + * is invoked. + */ + for_each_present_cpu(cpu) { + vmsa = per_cpu(sev_vmsa, cpu); + + /* + * BSP does not have guest allocated VMSA and there is no need + * to clear the VMSA tag for this page. + */ + if (!vmsa) + continue; + + /* + * Cannot clear the VMSA tag for the currently running vCPU. + */ + if (this_cpu == cpu) { + unsigned long pa; + struct page *p; + + pa = __pa(vmsa); + /* + * Mark the VMSA page of the running vCPU as offline + * so that is excluded and not touched by makedumpfile + * while generating vmcore during kdump. + */ + p = pfn_to_online_page(pa >> PAGE_SHIFT); + if (p) + __SetPageOffline(p); + continue; + } + + apic_id = cpuid_to_apicid[cpu]; + + /* + * Issue AP destroy to ensure AP gets kicked out of guest mode + * to allow using RMPADJUST to remove the VMSA tag on it's + * VMSA page. + */ + vmgexit_ap_control(SVM_VMGEXIT_AP_DESTROY, vmsa, apic_id); + snp_cleanup_vmsa(vmsa, apic_id); + } + + put_cpu(); +} + void snp_kexec_finish(void) { struct sev_es_runtime_data *data; @@ -987,6 +1139,8 @@ void snp_kexec_finish(void) if (!IS_ENABLED(CONFIG_KEXEC_CORE)) return; + shutdown_all_aps(); + unshare_all_memory(); /* @@ -1008,51 +1162,6 @@ void snp_kexec_finish(void) } } -static int snp_set_vmsa(void *va, void *caa, int apic_id, bool make_vmsa) -{ - int ret; - - if (snp_vmpl) { - struct svsm_call call = {}; - unsigned long flags; - - local_irq_save(flags); - - call.caa = this_cpu_read(svsm_caa); - call.rcx = __pa(va); - - if (make_vmsa) { - /* Protocol 0, Call ID 2 */ - call.rax = SVSM_CORE_CALL(SVSM_CORE_CREATE_VCPU); - call.rdx = __pa(caa); - call.r8 = apic_id; - } else { - /* Protocol 0, Call ID 3 */ - call.rax = SVSM_CORE_CALL(SVSM_CORE_DELETE_VCPU); - } - - ret = svsm_perform_call_protocol(&call); - - local_irq_restore(flags); - } else { - /* - * If the kernel runs at VMPL0, it can change the VMSA - * bit for a page using the RMPADJUST instruction. - * However, for the instruction to succeed it must - * target the permissions of a lesser privileged (higher - * numbered) VMPL level, so use VMPL1. - */ - u64 attrs = 1; - - if (make_vmsa) - attrs |= RMPADJUST_VMSA_PAGE_BIT; - - ret = rmpadjust((unsigned long)va, RMP_PG_SIZE_4K, attrs); - } - - return ret; -} - #define __ATTR_BASE (SVM_SELECTOR_P_MASK | SVM_SELECTOR_S_MASK) #define INIT_CS_ATTRIBS (__ATTR_BASE | SVM_SELECTOR_READ_MASK | SVM_SELECTOR_CODE_MASK) #define INIT_DS_ATTRIBS (__ATTR_BASE | SVM_SELECTOR_WRITE_MASK) @@ -1084,24 +1193,10 @@ static void *snp_alloc_vmsa_page(int cpu) return page_address(p + 1); } -static void snp_cleanup_vmsa(struct sev_es_save_area *vmsa, int apic_id) -{ - int err; - - err = snp_set_vmsa(vmsa, NULL, apic_id, false); - if (err) - pr_err("clear VMSA page failed (%u), leaking page\n", err); - else - free_page((unsigned long)vmsa); -} - static int wakeup_cpu_via_vmgexit(u32 apic_id, unsigned long start_ip) { struct sev_es_save_area *cur_vmsa, *vmsa; - struct ghcb_state state; struct svsm_ca *caa; - unsigned long flags; - struct ghcb *ghcb; u8 sipi_vector; int cpu, ret; u64 cr4; @@ -1215,33 +1310,7 @@ static int wakeup_cpu_via_vmgexit(u32 apic_id, unsigned long start_ip) } /* Issue VMGEXIT AP Creation NAE event */ - local_irq_save(flags); - - ghcb = __sev_get_ghcb(&state); - - vc_ghcb_invalidate(ghcb); - ghcb_set_rax(ghcb, vmsa->sev_features); - ghcb_set_sw_exit_code(ghcb, SVM_VMGEXIT_AP_CREATION); - ghcb_set_sw_exit_info_1(ghcb, - ((u64)apic_id << 32) | - ((u64)snp_vmpl << 16) | - SVM_VMGEXIT_AP_CREATE); - ghcb_set_sw_exit_info_2(ghcb, __pa(vmsa)); - - sev_es_wr_ghcb_msr(__pa(ghcb)); - VMGEXIT(); - - if (!ghcb_sw_exit_info_1_is_valid(ghcb) || - lower_32_bits(ghcb->save.sw_exit_info_1)) { - pr_err("SNP AP Creation error\n"); - ret = -EINVAL; - } - - __sev_put_ghcb(&state); - - local_irq_restore(flags); - - /* Perform cleanup if there was an error */ + ret = vmgexit_ap_control(SVM_VMGEXIT_AP_CREATE, vmsa, apic_id); if (ret) { snp_cleanup_vmsa(vmsa, apic_id); vmsa = NULL; -- 2.34.1

4 months

6
8
0 0

[PATCH] dt-bindings: remoteproc: qcom,sm8150-pas: Add missing SC8180X compatible

by Krzysztof Kozlowski

Commit 4b4ab93ddc5f ("dt-bindings: remoteproc: Consolidate SC8180X and SM8150 PAS files") moved SC8180X bindings from separate file into this one, but it forgot to add actual compatibles in top-level properties section making the entire binding un-selectable (no-op) for SC8180X PAS. Fixes: 4b4ab93ddc5f ("dt-bindings: remoteproc: Consolidate SC8180X and SM8150 PAS files") Cc: <stable(a)vger.kernel.org> Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski(a)linaro.org> --- .../devicetree/bindings/remoteproc/qcom,sm8150-pas.yaml | 3 +++ 1 file changed, 3 insertions(+) diff --git a/Documentation/devicetree/bindings/remoteproc/qcom,sm8150-pas.yaml b/Documentation/devicetree/bindings/remoteproc/qcom,sm8150-pas.yaml index 56ff6386534d..5dcc2a32c080 100644 --- a/Documentation/devicetree/bindings/remoteproc/qcom,sm8150-pas.yaml +++ b/Documentation/devicetree/bindings/remoteproc/qcom,sm8150-pas.yaml @@ -16,6 +16,9 @@ description: properties: compatible: enum: + - qcom,sc8180x-adsp-pas + - qcom,sc8180x-cdsp-pas + - qcom,sc8180x-slpi-pas - qcom,sm8150-adsp-pas - qcom,sm8150-cdsp-pas - qcom,sm8150-mpss-pas -- 2.45.2

4 months

3
2
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror April 2025