March 2025 - Linux-stable-mirror

[REGRESSION][BISECTED] Commit 60e3318e3e900 in stable/linux-6.1.y breaks cifs client failover to another server in DFS namespace

by Andrew Paniakin

Commit 60e3318e3e900 ("cifs: use fs_context for automounts") was released in v6.1.54 and broke the failover when one of the servers inside DFS becomes unavailable. We reproduced the problem on the EC2 instances of different types. Reverting aforementioned commint on top of the latest stable verison v6.1.94 helps to resolve the problem. Earliest working version is v6.2-rc1. There were two big merges of CIFS fixes: [1] and [2]. We would like to ask for the help to investigate this problem and if some of those patches need to be backported. Also, is it safe to just revert problematic commit until proper fixes/backports will be available? We will help to do testing and confirm if fix works, but let me also list the steps we used to reproduce the problem if it will help to identify the problem: 1. Create Active Directory domain eg. 'corp.fsxtest.local' in AWS Directory Service with: - three AWS FSX file systems filesystem1..filesystem3 - three Windows servers; They have DFS installed as per https://learn.microsoft.com/en-us/windows-server/storage/dfs-namespaces/dfs…: - dfs-srv1: EC2AMAZ-2EGTM59 - dfs-srv2: EC2AMAZ-1N36PRD - dfs-srv3: EC2AMAZ-0PAUH2U 2. Create DFS namespace eg. 'dfs-namespace' in Windows server 2008 mode and three folders targets in it: - referral-a mapped to filesystem1.corp.local - referral-b mapped to filesystem2.corp.local - referral-c mapped to filesystem3.corp.local - local folders dfs-srv1..dfs-srv3 in C:\DFSRoots\dfs-namespace of every Windows server. This helps to quickly define underlying server when DFS is mounted. 3. Enabled cifs debug logs: ``` echo 'module cifs +p' > /sys/kernel/debug/dynamic_debug/control echo 'file fs/cifs/* +p' > /sys/kernel/debug/dynamic_debug/control echo 7 > /proc/fs/cifs/cifsFYI ``` 4. Mount DFS namespace on Amazon Linux 2023 instance running any vanilla kernel v6.1.54+: ``` dmesg -c &>/dev/null cd /mnt mount -t cifs -o cred=/mnt/creds,echo_interval=5 \ //corp.fsxtest.local/dfs-namespace \ ./dfs-namespace ``` 5. List DFS root, it's also required to avoid recursive mounts that happen during regular 'ls' run: ``` sh -c 'ls dfs-namespace' dfs-srv2 referral-a referral-b ``` The DFS server is EC2AMAZ-1N36PRD, it's also listed in mount: ``` [root@ip-172-31-2-82 mnt]# mount | grep dfs //corp.fsxtest.local/dfs-namespace on /mnt/dfs-namespace type cifs (rw,relatime,vers=3.1.1,cache=strict,username=Admin,domain=corp.fsxtest.local,uid=0,noforceuid,gid=0,noforcegid,addr=172.31.11.26,file_mode=0755,dir_mode=0755,soft,nounix,mapposix,rsize=4194304,wsize=4194304,bsize=1048576,echo_interval=5,actimeo=1,closetimeo=1) //EC2AMAZ-1N36PRD.corp.fsxtest.local/dfs-namespace/referral-a on /mnt/dfs-namespace/referral-a type cifs (rw,relatime,vers=3.1.1,cache=strict,username=Admin,domain=corp.fsxtest.local,uid=0,noforceuid,gid=0,noforcegid,addr=172.31.12.80,file_mode=0755,dir_mode=0755,soft,nounix,mapposix,rsize=4194304,wsize=4194304,bsize=1048576,echo_interval=5,actimeo=1,closetimeo=1) ``` List files in first folder: ``` sh -c 'ls dfs-namespace/referral-a' filea.txt.txt ``` 6. Shutdown DFS server-2. List DFS root again, server changed from dfs-srv2 to dfs-srv1 EC2AMAZ-2EGTM59: ``` sh -c 'ls dfs-namespace' dfs-srv1 referral-a referral-b ``` 7. Try to list files in another folder, this causes ls to fail with error: ``` sh -c 'ls dfs-namespace/referral-b' ls: cannot access 'dfs-namespace/referral-b': No route to host``` Sometimes it's also 'Operation now in progress' error. mount shows the same output: ``` //corp.fsxtest.local/dfs-namespace on /mnt/dfs-namespace type cifs (rw,relatime,vers=3.1.1,cache=strict,username=Admin,domain=corp.fsxtest.local,uid=0,noforceuid,gid=0,noforcegid,addr=172.31.11.26,file_mode=0755,dir_mode=0755,soft,nounix,mapposix,rsize=4194304,wsize=4194304,bsize=1048576,echo_interval=5,actimeo=1,closetimeo=1) //EC2AMAZ-1N36PRD.corp.fsxtest.local/dfs-namespace/referral-a on /mnt/dfs-namespace/referral-a type cifs (rw,relatime,vers=3.1.1,cache=strict,username=Admin,domain=corp.fsxtest.local,uid=0,noforceuid,gid=0,noforcegid,addr=172.31.12.80,file_mode=0755,dir_mode=0755,soft,nounix,mapposix,rsize=4194304,wsize=4194304,bsize=1048576,echo_interval=5,actimeo=1,closetimeo=1) ``` I also attached kernel debug logs from this test. [1] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… [2] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… Reported-by: Andrei Paniakin <apanyaki(a)amazon.com> Bisected-by: Simba Bonga <simbarb(a)amazon.com> --- #regzbot introduced: v6.1.54..v6.2-rc1

1 month

3
12
0 0

[PATCH] soc: qcom: pmic_glink_altmode: fix spurious DP hotplug events

by Johan Hovold

The PMIC GLINK driver is currently generating DisplayPort hotplug notifications whenever something is connected to (or disconnected from) a port regardless of the type of notification sent by the firmware. These notifications are forwarded to user space by the DRM subsystem as connector "change" uevents: KERNEL[1556.223776] change /devices/platform/soc(a)0/ae00000.display-subsystem/ae01000.display-controller/drm/card0 (drm) ACTION=change DEVPATH=/devices/platform/soc(a)0/ae00000.display-subsystem/ae01000.display-controller/drm/card0 SUBSYSTEM=drm HOTPLUG=1 CONNECTOR=36 DEVNAME=/dev/dri/card0 DEVTYPE=drm_minor SEQNUM=4176 MAJOR=226 MINOR=0 On the Lenovo ThinkPad X13s and T14s, the PMIC GLINK firmware sends two identical notifications with orientation information when connecting a charger, each generating a bogus DRM hotplug event. On the X13s, two such notification are also sent every 90 seconds while a charger remains connected, which again are forwarded to user space: port = 1, svid = ff00, mode = 255, hpd_state = 0 payload = 01 00 00 00 00 00 00 ff 00 00 00 00 00 00 00 00 Note that the firmware only sends on of these when connecting an ethernet adapter. Fix the spurious hotplug events by only forwarding hotplug notifications for the Type-C DisplayPort service id. This also reduces the number of uevents from four to two when an actual DisplayPort altmode device is connected: port = 0, svid = ff01, mode = 2, hpd_state = 0 payload = 00 01 02 00 f2 0c 01 ff 03 00 00 00 00 00 00 00 port = 0, svid = ff01, mode = 2, hpd_state = 1 payload = 00 01 02 00 f2 0c 01 ff 43 00 00 00 00 00 00 00 Fixes: 080b4e24852b ("soc: qcom: pmic_glink: Introduce altmode support") Cc: stable(a)vger.kernel.org # 6.3 Cc: Bjorn Andersson <andersson(a)kernel.org> Reported-by: Clayton Craft <clayton(a)craftyguy.net> Signed-off-by: Johan Hovold <johan+linaro(a)kernel.org> --- Clayton reported seeing display flickering with recent RC kernels, which may possibly be related to these spurious events being generated with even greater frequency. That still remains to be fully understood, but the spurious events, that on the X13s are generated every 90 seconds, should be fixed either way. Johan drivers/soc/qcom/pmic_glink_altmode.c | 30 +++++++++++++++++---------- 1 file changed, 19 insertions(+), 11 deletions(-) diff --git a/drivers/soc/qcom/pmic_glink_altmode.c b/drivers/soc/qcom/pmic_glink_altmode.c index bd06ce161804..7f11acd33323 100644 --- a/drivers/soc/qcom/pmic_glink_altmode.c +++ b/drivers/soc/qcom/pmic_glink_altmode.c @@ -218,21 +218,29 @@ static void pmic_glink_altmode_worker(struct work_struct *work) { struct pmic_glink_altmode_port *alt_port = work_to_altmode_port(work); struct pmic_glink_altmode *altmode = alt_port->altmode; + enum drm_connector_status conn_status; typec_switch_set(alt_port->typec_switch, alt_port->orientation); - if (alt_port->svid == USB_TYPEC_DP_SID && alt_port->mode == 0xff) - pmic_glink_altmode_safe(altmode, alt_port); - else if (alt_port->svid == USB_TYPEC_DP_SID) - pmic_glink_altmode_enable_dp(altmode, alt_port, alt_port->mode, - alt_port->hpd_state, alt_port->hpd_irq); - else - pmic_glink_altmode_enable_usb(altmode, alt_port); + if (alt_port->svid == USB_TYPEC_DP_SID) { + if (alt_port->mode == 0xff) { + pmic_glink_altmode_safe(altmode, alt_port); + } else { + pmic_glink_altmode_enable_dp(altmode, alt_port, + alt_port->mode, + alt_port->hpd_state, + alt_port->hpd_irq); + } - drm_aux_hpd_bridge_notify(&alt_port->bridge->dev, - alt_port->hpd_state ? - connector_status_connected : - connector_status_disconnected); + if (alt_port->hpd_state) + conn_status = connector_status_connected; + else + conn_status = connector_status_disconnected; + + drm_aux_hpd_bridge_notify(&alt_port->bridge->dev, conn_status); + } else { + pmic_glink_altmode_enable_usb(altmode, alt_port); + } pmic_glink_altmode_request(altmode, ALTMODE_PAN_ACK, alt_port->index); } -- 2.48.1

1 month

5
9
0 0

[PATCHv3] thunderbolt: do not double dequeue a request

by Sergey Senozhatsky

Some of our devices crash in tb_cfg_request_dequeue(): general protection fault, probably for non-canonical address 0xdead000000000122 CPU: 6 PID: 91007 Comm: kworker/6:2 Tainted: G U W 6.6.65 RIP: 0010:tb_cfg_request_dequeue+0x2d/0xa0 Call Trace: <TASK> ? tb_cfg_request_dequeue+0x2d/0xa0 tb_cfg_request_work+0x33/0x80 worker_thread+0x386/0x8f0 kthread+0xed/0x110 ret_from_fork+0x38/0x50 ret_from_fork_asm+0x1b/0x30 The circumstances are unclear, however, the theory is that tb_cfg_request_work() can be scheduled twice for a request: first time via frame.callback from ring_work() and second time from tb_cfg_request(). Both times kworkers will execute tb_cfg_request_dequeue(), which results in double list_del() from the ctl->request_queue (the list poison deference hints at it: 0xdead000000000122). Do not dequeue requests that don't have TB_CFG_REQUEST_ACTIVE bit set. Signed-off-by: Sergey Senozhatsky <senozhatsky(a)chromium.org> Cc: stable(a)vger.kernel.org --- v3: tweaked commit message drivers/thunderbolt/ctl.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/thunderbolt/ctl.c b/drivers/thunderbolt/ctl.c index cd15e84c47f4..1db2e951b53f 100644 --- a/drivers/thunderbolt/ctl.c +++ b/drivers/thunderbolt/ctl.c @@ -151,6 +151,11 @@ static void tb_cfg_request_dequeue(struct tb_cfg_request *req) struct tb_ctl *ctl = req->ctl; mutex_lock(&ctl->request_queue_lock); + if (!test_bit(TB_CFG_REQUEST_ACTIVE, &req->flags)) { + mutex_unlock(&ctl->request_queue_lock); + return; + } + list_del(&req->list); clear_bit(TB_CFG_REQUEST_ACTIVE, &req->flags); if (test_bit(TB_CFG_REQUEST_CANCELED, &req->flags)) -- 2.49.0.395.g12beb8f557-goog

1 month

2
2
0 0

[PATCH] clk: s2mps11: initialise clk_hw_onecell_data::num before accessing ::hws[] in probe()

by André Draszik

With UBSAN enabled, we're getting the following trace: UBSAN: array-index-out-of-bounds in .../drivers/clk/clk-s2mps11.c:186:3 index 0 is out of range for type 'struct clk_hw *[] __counted_by(num)' (aka 'struct clk_hw *[]') This is because commit f316cdff8d67 ("clk: Annotate struct clk_hw_onecell_data with __counted_by") annotated the hws member of that struct with __counted_by, which informs the bounds sanitizer about the number of elements in hws, so that it can warn when hws is accessed out of bounds. As noted in that change, the __counted_by member must be initialised with the number of elements before the first array access happens, otherwise there will be a warning from each access prior to the initialisation because the number of elements is zero. This occurs in s2mps11_clk_probe() due to ::num being assigned after ::hws access. Move the assignment to satisfy the requirement of assign-before-access. Cc: stable(a)vger.kernel.org Fixes: f316cdff8d67 ("clk: Annotate struct clk_hw_onecell_data with __counted_by") Signed-off-by: André Draszik <andre.draszik(a)linaro.org> --- drivers/clk/clk-s2mps11.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/clk/clk-s2mps11.c b/drivers/clk/clk-s2mps11.c index 014db6386624071e173b5b940466301d2596400a..8ddf3a9a53dfd5bb52a05a3e02788a357ea77ad3 100644 --- a/drivers/clk/clk-s2mps11.c +++ b/drivers/clk/clk-s2mps11.c @@ -137,6 +137,8 @@ static int s2mps11_clk_probe(struct platform_device *pdev) if (!clk_data) return -ENOMEM; + clk_data->num = S2MPS11_CLKS_NUM; + switch (hwid) { case S2MPS11X: s2mps11_reg = S2MPS11_REG_RTC_CTRL; @@ -186,7 +188,6 @@ static int s2mps11_clk_probe(struct platform_device *pdev) clk_data->hws[i] = &s2mps11_clks[i].hw; } - clk_data->num = S2MPS11_CLKS_NUM; of_clk_add_hw_provider(s2mps11_clks->clk_np, of_clk_hw_onecell_get, clk_data); --- base-commit: 9388ec571cb1adba59d1cded2300eeb11827679c change-id: 20250326-s2mps11-ubsan-c90978e7bc04 Best regards, -- André Draszik <andre.draszik(a)linaro.org>

1 month

3
2
0 0

[PATCH RESEND] drm/tegra: fix a possible null pointer dereference

by Qiu-ji Chen

In tegra_crtc_reset(), new memory is allocated with kzalloc(), but no check is performed. Before calling __drm_atomic_helper_crtc_reset, state should be checked to prevent possible null pointer dereference. Fixes: b7e0b04ae450 ("drm/tegra: Convert to using __drm_atomic_helper_crtc_reset() for reset.") Cc: stable(a)vger.kernel.org Signed-off-by: Qiu-ji Chen <chenqiuji666(a)gmail.com> --- drivers/gpu/drm/tegra/dc.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/tegra/dc.c b/drivers/gpu/drm/tegra/dc.c index be61c9d1a4f0..1ed30853bd9e 100644 --- a/drivers/gpu/drm/tegra/dc.c +++ b/drivers/gpu/drm/tegra/dc.c @@ -1388,7 +1388,10 @@ static void tegra_crtc_reset(struct drm_crtc *crtc) if (crtc->state) tegra_crtc_atomic_destroy_state(crtc, crtc->state); - __drm_atomic_helper_crtc_reset(crtc, &state->base); + if (state) + __drm_atomic_helper_crtc_reset(crtc, &state->base); + else + __drm_atomic_helper_crtc_reset(crtc, NULL); } static struct drm_crtc_state * -- 2.34.1

1 month

2
1
0 0

[PATCH AUTOSEL 5.10 1/6] pm: cpupower: bench: Prevent NULL dereference on malloc failure

by Sasha Levin

From: Zhongqiu Han <quic_zhonhan(a)quicinc.com> [ Upstream commit 208baa3ec9043a664d9acfb8174b332e6b17fb69 ] If malloc returns NULL due to low memory, 'config' pointer can be NULL. Add a check to prevent NULL dereference. Link: https://lore.kernel.org/r/20250219122715.3892223-1-quic_zhonhan@quicinc.com Signed-off-by: Zhongqiu Han <quic_zhonhan(a)quicinc.com> Signed-off-by: Shuah Khan <skhan(a)linuxfoundation.org> Signed-off-by: Sasha Levin <sashal(a)kernel.org> --- tools/power/cpupower/bench/parse.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/tools/power/cpupower/bench/parse.c b/tools/power/cpupower/bench/parse.c index e63dc11fa3a53..48e25be6e1635 100644 --- a/tools/power/cpupower/bench/parse.c +++ b/tools/power/cpupower/bench/parse.c @@ -120,6 +120,10 @@ FILE *prepare_output(const char *dirname) struct config *prepare_default_config() { struct config *config = malloc(sizeof(struct config)); + if (!config) { + perror("malloc"); + return NULL; + } dprintf("loading defaults\n"); -- 2.39.5

1 month, 1 week

5
24
0 0

[PATCH 6.1 6.6 6.12 6.13] scripts/sorttable: fix ELF64 mcount_loc address parsing when compiling on 32-bit

by Sahil Gupta

The ftrace __mcount_loc buildtime sort does not work properly when the host is 32-bit and the target is 64-bit. sorttable parses the start and stop addresses by calling strtoul on the buffer holding the hexadecimal string. Since the target is 64-bit but unsigned long on 32-bit machines is 32 bits, strtoul, and by extension the start and stop addresses, can max out to 2^32 - 1. This patch adds a new macro, parse_addr, that corresponds to a strtoul or strtoull call based on whether you are operating on a 32-bit ELF or a 64-bit ELF. This way, the correct width is guaranteed whether or not the host is 32-bit. This should cleanly apply on all of the 6.x stable kernels. Manually verified that the __mcount_loc section is sorted by parsing the ELF and verified tests corresponding to CONFIG_FTRACE_SORT_STARTUP_TEST for kernels built on a 32-bit and a 64-bit host. Signed-off-by: Sahil Gupta <s.gupta(a)arista.com> --- scripts/sorttable.h | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/scripts/sorttable.h b/scripts/sorttable.h index 7bd0184380d3..9ed7acca9f30 100644 --- a/scripts/sorttable.h +++ b/scripts/sorttable.h @@ -40,6 +40,7 @@ #undef uint_t #undef _r #undef _w +#undef parse_addr #ifdef SORTTABLE_64 # define extable_ent_size 16 @@ -65,6 +66,7 @@ # define uint_t uint64_t # define _r r8 # define _w w8 +# define parse_addr(buf) strtoull(buf, NULL, 16) #else # define extable_ent_size 8 # define compare_extable compare_extable_32 @@ -89,6 +91,7 @@ # define uint_t uint32_t # define _r r # define _w w +# define parse_addr(buf) strtoul(buf, NULL, 16) #endif #if defined(SORTTABLE_64) && defined(UNWINDER_ORC_ENABLED) @@ -246,13 +249,13 @@ static void get_mcount_loc(uint_t *_start, uint_t *_stop) len = strlen(start_buff); start_buff[len - 1] = '\0'; } - *_start = strtoul(start_buff, NULL, 16); + *_start = parse_addr(start_buff); while (fgets(stop_buff, sizeof(stop_buff), file_stop) != NULL) { len = strlen(stop_buff); stop_buff[len - 1] = '\0'; } - *_stop = strtoul(stop_buff, NULL, 16); + *_stop = parse_addr(stop_buff); pclose(file_start); pclose(file_stop); -- 2.47.0

1 month, 1 week

4
8
0 0

FAILED: patch "[PATCH] KVM: x86: Load DR6 with guest value only before entering" 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 c2fee09fc167c74a64adb08656cb993ea475197e # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2025021815-protract-greasily-cdea@gregkh' --subject-prefix 'PATCH 5.15.y' HEAD^.. Possible dependencies: thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From c2fee09fc167c74a64adb08656cb993ea475197e Mon Sep 17 00:00:00 2001 From: Sean Christopherson <seanjc(a)google.com> Date: Fri, 24 Jan 2025 17:18:33 -0800 Subject: [PATCH] KVM: x86: Load DR6 with guest value only before entering .vcpu_run() loop Move the conditional loading of hardware DR6 with the guest's DR6 value out of the core .vcpu_run() loop to fix a bug where KVM can load hardware with a stale vcpu->arch.dr6. When the guest accesses a DR and host userspace isn't debugging the guest, KVM disables DR interception and loads the guest's values into hardware on VM-Enter and saves them on VM-Exit. This allows the guest to access DRs at will, e.g. so that a sequence of DR accesses to configure a breakpoint only generates one VM-Exit. For DR0-DR3, the logic/behavior is identical between VMX and SVM, and also identical between KVM_DEBUGREG_BP_ENABLED (userspace debugging the guest) and KVM_DEBUGREG_WONT_EXIT (guest using DRs), and so KVM handles loading DR0-DR3 in common code, _outside_ of the core kvm_x86_ops.vcpu_run() loop. But for DR6, the guest's value doesn't need to be loaded into hardware for KVM_DEBUGREG_BP_ENABLED, and SVM provides a dedicated VMCB field whereas VMX requires software to manually load the guest value, and so loading the guest's value into DR6 is handled by {svm,vmx}_vcpu_run(), i.e. is done _inside_ the core run loop. Unfortunately, saving the guest values on VM-Exit is initiated by common x86, again outside of the core run loop. If the guest modifies DR6 (in hardware, when DR interception is disabled), and then the next VM-Exit is a fastpath VM-Exit, KVM will reload hardware DR6 with vcpu->arch.dr6 and clobber the guest's actual value. The bug shows up primarily with nested VMX because KVM handles the VMX preemption timer in the fastpath, and the window between hardware DR6 being modified (in guest context) and DR6 being read by guest software is orders of magnitude larger in a nested setup. E.g. in non-nested, the VMX preemption timer would need to fire precisely between #DB injection and the #DB handler's read of DR6, whereas with a KVM-on-KVM setup, the window where hardware DR6 is "dirty" extends all the way from L1 writing DR6 to VMRESUME (in L1). L1's view: ========== <L1 disables DR interception> CPU 0/KVM-7289 [023] d.... 2925.640961: kvm_entry: vcpu 0 A: L1 Writes DR6 CPU 0/KVM-7289 [023] d.... 2925.640963: <hack>: Set DRs, DR6 = 0xffff0ff1 B: CPU 0/KVM-7289 [023] d.... 2925.640967: kvm_exit: vcpu 0 reason EXTERNAL_INTERRUPT intr_info 0x800000ec D: L1 reads DR6, arch.dr6 = 0 CPU 0/KVM-7289 [023] d.... 2925.640969: <hack>: Sync DRs, DR6 = 0xffff0ff0 CPU 0/KVM-7289 [023] d.... 2925.640976: kvm_entry: vcpu 0 L2 reads DR6, L1 disables DR interception CPU 0/KVM-7289 [023] d.... 2925.640980: kvm_exit: vcpu 0 reason DR_ACCESS info1 0x0000000000000216 CPU 0/KVM-7289 [023] d.... 2925.640983: kvm_entry: vcpu 0 CPU 0/KVM-7289 [023] d.... 2925.640983: <hack>: Set DRs, DR6 = 0xffff0ff0 L2 detects failure CPU 0/KVM-7289 [023] d.... 2925.640987: kvm_exit: vcpu 0 reason HLT L1 reads DR6 (confirms failure) CPU 0/KVM-7289 [023] d.... 2925.640990: <hack>: Sync DRs, DR6 = 0xffff0ff0 L0's view: ========== L2 reads DR6, arch.dr6 = 0 CPU 23/KVM-5046 [001] d.... 3410.005610: kvm_exit: vcpu 23 reason DR_ACCESS info1 0x0000000000000216 CPU 23/KVM-5046 [001] ..... 3410.005610: kvm_nested_vmexit: vcpu 23 reason DR_ACCESS info1 0x0000000000000216 L2 => L1 nested VM-Exit CPU 23/KVM-5046 [001] ..... 3410.005610: kvm_nested_vmexit_inject: reason: DR_ACCESS ext_inf1: 0x0000000000000216 CPU 23/KVM-5046 [001] d.... 3410.005610: kvm_entry: vcpu 23 CPU 23/KVM-5046 [001] d.... 3410.005611: kvm_exit: vcpu 23 reason VMREAD CPU 23/KVM-5046 [001] d.... 3410.005611: kvm_entry: vcpu 23 CPU 23/KVM-5046 [001] d.... 3410.005612: kvm_exit: vcpu 23 reason VMREAD CPU 23/KVM-5046 [001] d.... 3410.005612: kvm_entry: vcpu 23 L1 writes DR7, L0 disables DR interception CPU 23/KVM-5046 [001] d.... 3410.005612: kvm_exit: vcpu 23 reason DR_ACCESS info1 0x0000000000000007 CPU 23/KVM-5046 [001] d.... 3410.005613: kvm_entry: vcpu 23 L0 writes DR6 = 0 (arch.dr6) CPU 23/KVM-5046 [001] d.... 3410.005613: <hack>: Set DRs, DR6 = 0xffff0ff0 A: <L1 writes DR6 = 1, no interception, arch.dr6 is still '0'> B: CPU 23/KVM-5046 [001] d.... 3410.005614: kvm_exit: vcpu 23 reason PREEMPTION_TIMER CPU 23/KVM-5046 [001] d.... 3410.005614: kvm_entry: vcpu 23 C: L0 writes DR6 = 0 (arch.dr6) CPU 23/KVM-5046 [001] d.... 3410.005614: <hack>: Set DRs, DR6 = 0xffff0ff0 L1 => L2 nested VM-Enter CPU 23/KVM-5046 [001] d.... 3410.005616: kvm_exit: vcpu 23 reason VMRESUME L0 reads DR6, arch.dr6 = 0 Reported-by: John Stultz <jstultz(a)google.com> Closes: https://lkml.kernel.org/r/CANDhNCq5_F3HfFYABqFGCA1bPd_%2BxgNj-iDQhH4tDk%2Bw… Fixes: 375e28ffc0cf ("KVM: X86: Set host DR6 only on VMX and for KVM_DEBUGREG_WONT_EXIT") Fixes: d67668e9dd76 ("KVM: x86, SVM: isolate vcpu->arch.dr6 from vmcb->save.dr6") Cc: stable(a)vger.kernel.org Cc: Jim Mattson <jmattson(a)google.com> Tested-by: John Stultz <jstultz(a)google.com> Link: https://lore.kernel.org/r/20250125011833.3644371-1-seanjc@google.com Signed-off-by: Sean Christopherson <seanjc(a)google.com> diff --git a/arch/x86/include/asm/kvm-x86-ops.h b/arch/x86/include/asm/kvm-x86-ops.h index c35550581da0..823c0434bbad 100644 --- a/arch/x86/include/asm/kvm-x86-ops.h +++ b/arch/x86/include/asm/kvm-x86-ops.h @@ -48,6 +48,7 @@ KVM_X86_OP(set_idt) KVM_X86_OP(get_gdt) KVM_X86_OP(set_gdt) KVM_X86_OP(sync_dirty_debug_regs) +KVM_X86_OP(set_dr6) KVM_X86_OP(set_dr7) KVM_X86_OP(cache_reg) KVM_X86_OP(get_rflags) diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h index b15cde0a9b5c..0b7af5902ff7 100644 --- a/arch/x86/include/asm/kvm_host.h +++ b/arch/x86/include/asm/kvm_host.h @@ -1696,6 +1696,7 @@ struct kvm_x86_ops { void (*get_gdt)(struct kvm_vcpu *vcpu, struct desc_ptr *dt); void (*set_gdt)(struct kvm_vcpu *vcpu, struct desc_ptr *dt); void (*sync_dirty_debug_regs)(struct kvm_vcpu *vcpu); + void (*set_dr6)(struct kvm_vcpu *vcpu, unsigned long value); void (*set_dr7)(struct kvm_vcpu *vcpu, unsigned long value); void (*cache_reg)(struct kvm_vcpu *vcpu, enum kvm_reg reg); unsigned long (*get_rflags)(struct kvm_vcpu *vcpu); diff --git a/arch/x86/kvm/svm/svm.c b/arch/x86/kvm/svm/svm.c index 7640a84e554a..a713c803a3a3 100644 --- a/arch/x86/kvm/svm/svm.c +++ b/arch/x86/kvm/svm/svm.c @@ -1991,11 +1991,11 @@ static void new_asid(struct vcpu_svm *svm, struct svm_cpu_data *sd) svm->asid = sd->next_asid++; } -static void svm_set_dr6(struct vcpu_svm *svm, unsigned long value) +static void svm_set_dr6(struct kvm_vcpu *vcpu, unsigned long value) { - struct vmcb *vmcb = svm->vmcb; + struct vmcb *vmcb = to_svm(vcpu)->vmcb; - if (svm->vcpu.arch.guest_state_protected) + if (vcpu->arch.guest_state_protected) return; if (unlikely(value != vmcb->save.dr6)) { @@ -4247,10 +4247,8 @@ static __no_kcsan fastpath_t svm_vcpu_run(struct kvm_vcpu *vcpu, * Run with all-zero DR6 unless needed, so that we can get the exact cause * of a #DB. */ - if (unlikely(vcpu->arch.switch_db_regs & KVM_DEBUGREG_WONT_EXIT)) - svm_set_dr6(svm, vcpu->arch.dr6); - else - svm_set_dr6(svm, DR6_ACTIVE_LOW); + if (likely(!(vcpu->arch.switch_db_regs & KVM_DEBUGREG_WONT_EXIT))) + svm_set_dr6(vcpu, DR6_ACTIVE_LOW); clgi(); kvm_load_guest_xsave_state(vcpu); @@ -5043,6 +5041,7 @@ static struct kvm_x86_ops svm_x86_ops __initdata = { .set_idt = svm_set_idt, .get_gdt = svm_get_gdt, .set_gdt = svm_set_gdt, + .set_dr6 = svm_set_dr6, .set_dr7 = svm_set_dr7, .sync_dirty_debug_regs = svm_sync_dirty_debug_regs, .cache_reg = svm_cache_reg, diff --git a/arch/x86/kvm/vmx/main.c b/arch/x86/kvm/vmx/main.c index 2427f918e763..43ee9ed11291 100644 --- a/arch/x86/kvm/vmx/main.c +++ b/arch/x86/kvm/vmx/main.c @@ -61,6 +61,7 @@ struct kvm_x86_ops vt_x86_ops __initdata = { .set_idt = vmx_set_idt, .get_gdt = vmx_get_gdt, .set_gdt = vmx_set_gdt, + .set_dr6 = vmx_set_dr6, .set_dr7 = vmx_set_dr7, .sync_dirty_debug_regs = vmx_sync_dirty_debug_regs, .cache_reg = vmx_cache_reg, diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c index f72835e85b6d..6c56d5235f0f 100644 --- a/arch/x86/kvm/vmx/vmx.c +++ b/arch/x86/kvm/vmx/vmx.c @@ -5648,6 +5648,12 @@ void vmx_sync_dirty_debug_regs(struct kvm_vcpu *vcpu) set_debugreg(DR6_RESERVED, 6); } +void vmx_set_dr6(struct kvm_vcpu *vcpu, unsigned long val) +{ + lockdep_assert_irqs_disabled(); + set_debugreg(vcpu->arch.dr6, 6); +} + void vmx_set_dr7(struct kvm_vcpu *vcpu, unsigned long val) { vmcs_writel(GUEST_DR7, val); @@ -7417,10 +7423,6 @@ fastpath_t vmx_vcpu_run(struct kvm_vcpu *vcpu, bool force_immediate_exit) vmx->loaded_vmcs->host_state.cr4 = cr4; } - /* When KVM_DEBUGREG_WONT_EXIT, dr6 is accessible in guest. */ - if (unlikely(vcpu->arch.switch_db_regs & KVM_DEBUGREG_WONT_EXIT)) - set_debugreg(vcpu->arch.dr6, 6); - /* When single-stepping over STI and MOV SS, we must clear the * corresponding interruptibility bits in the guest state. Otherwise * vmentry fails as it then expects bit 14 (BS) in pending debug diff --git a/arch/x86/kvm/vmx/x86_ops.h b/arch/x86/kvm/vmx/x86_ops.h index ce3295a67c04..430773a5ef8e 100644 --- a/arch/x86/kvm/vmx/x86_ops.h +++ b/arch/x86/kvm/vmx/x86_ops.h @@ -73,6 +73,7 @@ void vmx_get_idt(struct kvm_vcpu *vcpu, struct desc_ptr *dt); void vmx_set_idt(struct kvm_vcpu *vcpu, struct desc_ptr *dt); void vmx_get_gdt(struct kvm_vcpu *vcpu, struct desc_ptr *dt); void vmx_set_gdt(struct kvm_vcpu *vcpu, struct desc_ptr *dt); +void vmx_set_dr6(struct kvm_vcpu *vcpu, unsigned long val); void vmx_set_dr7(struct kvm_vcpu *vcpu, unsigned long val); void vmx_sync_dirty_debug_regs(struct kvm_vcpu *vcpu); void vmx_cache_reg(struct kvm_vcpu *vcpu, enum kvm_reg reg); diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index 8e77e61d4fbd..02159c967d29 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -10961,6 +10961,9 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu) set_debugreg(vcpu->arch.eff_db[1], 1); set_debugreg(vcpu->arch.eff_db[2], 2); set_debugreg(vcpu->arch.eff_db[3], 3); + /* When KVM_DEBUGREG_WONT_EXIT, dr6 is accessible in guest. */ + if (unlikely(vcpu->arch.switch_db_regs & KVM_DEBUGREG_WONT_EXIT)) + kvm_x86_call(set_dr6)(vcpu, vcpu->arch.dr6); } else if (unlikely(hw_breakpoint_active())) { set_debugreg(0, 7); }

1 month, 1 week

3
2
0 0

FAILED: patch "[PATCH] KVM: x86: Load DR6 with guest value only before entering" failed to apply to 6.6-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 6.6-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-6.6.y git checkout FETCH_HEAD git cherry-pick -x c2fee09fc167c74a64adb08656cb993ea475197e # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2025021809-census-mammal-224b@gregkh' --subject-prefix 'PATCH 6.6.y' HEAD^.. Possible dependencies: thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From c2fee09fc167c74a64adb08656cb993ea475197e Mon Sep 17 00:00:00 2001 From: Sean Christopherson <seanjc(a)google.com> Date: Fri, 24 Jan 2025 17:18:33 -0800 Subject: [PATCH] KVM: x86: Load DR6 with guest value only before entering .vcpu_run() loop Move the conditional loading of hardware DR6 with the guest's DR6 value out of the core .vcpu_run() loop to fix a bug where KVM can load hardware with a stale vcpu->arch.dr6. When the guest accesses a DR and host userspace isn't debugging the guest, KVM disables DR interception and loads the guest's values into hardware on VM-Enter and saves them on VM-Exit. This allows the guest to access DRs at will, e.g. so that a sequence of DR accesses to configure a breakpoint only generates one VM-Exit. For DR0-DR3, the logic/behavior is identical between VMX and SVM, and also identical between KVM_DEBUGREG_BP_ENABLED (userspace debugging the guest) and KVM_DEBUGREG_WONT_EXIT (guest using DRs), and so KVM handles loading DR0-DR3 in common code, _outside_ of the core kvm_x86_ops.vcpu_run() loop. But for DR6, the guest's value doesn't need to be loaded into hardware for KVM_DEBUGREG_BP_ENABLED, and SVM provides a dedicated VMCB field whereas VMX requires software to manually load the guest value, and so loading the guest's value into DR6 is handled by {svm,vmx}_vcpu_run(), i.e. is done _inside_ the core run loop. Unfortunately, saving the guest values on VM-Exit is initiated by common x86, again outside of the core run loop. If the guest modifies DR6 (in hardware, when DR interception is disabled), and then the next VM-Exit is a fastpath VM-Exit, KVM will reload hardware DR6 with vcpu->arch.dr6 and clobber the guest's actual value. The bug shows up primarily with nested VMX because KVM handles the VMX preemption timer in the fastpath, and the window between hardware DR6 being modified (in guest context) and DR6 being read by guest software is orders of magnitude larger in a nested setup. E.g. in non-nested, the VMX preemption timer would need to fire precisely between #DB injection and the #DB handler's read of DR6, whereas with a KVM-on-KVM setup, the window where hardware DR6 is "dirty" extends all the way from L1 writing DR6 to VMRESUME (in L1). L1's view: ========== <L1 disables DR interception> CPU 0/KVM-7289 [023] d.... 2925.640961: kvm_entry: vcpu 0 A: L1 Writes DR6 CPU 0/KVM-7289 [023] d.... 2925.640963: <hack>: Set DRs, DR6 = 0xffff0ff1 B: CPU 0/KVM-7289 [023] d.... 2925.640967: kvm_exit: vcpu 0 reason EXTERNAL_INTERRUPT intr_info 0x800000ec D: L1 reads DR6, arch.dr6 = 0 CPU 0/KVM-7289 [023] d.... 2925.640969: <hack>: Sync DRs, DR6 = 0xffff0ff0 CPU 0/KVM-7289 [023] d.... 2925.640976: kvm_entry: vcpu 0 L2 reads DR6, L1 disables DR interception CPU 0/KVM-7289 [023] d.... 2925.640980: kvm_exit: vcpu 0 reason DR_ACCESS info1 0x0000000000000216 CPU 0/KVM-7289 [023] d.... 2925.640983: kvm_entry: vcpu 0 CPU 0/KVM-7289 [023] d.... 2925.640983: <hack>: Set DRs, DR6 = 0xffff0ff0 L2 detects failure CPU 0/KVM-7289 [023] d.... 2925.640987: kvm_exit: vcpu 0 reason HLT L1 reads DR6 (confirms failure) CPU 0/KVM-7289 [023] d.... 2925.640990: <hack>: Sync DRs, DR6 = 0xffff0ff0 L0's view: ========== L2 reads DR6, arch.dr6 = 0 CPU 23/KVM-5046 [001] d.... 3410.005610: kvm_exit: vcpu 23 reason DR_ACCESS info1 0x0000000000000216 CPU 23/KVM-5046 [001] ..... 3410.005610: kvm_nested_vmexit: vcpu 23 reason DR_ACCESS info1 0x0000000000000216 L2 => L1 nested VM-Exit CPU 23/KVM-5046 [001] ..... 3410.005610: kvm_nested_vmexit_inject: reason: DR_ACCESS ext_inf1: 0x0000000000000216 CPU 23/KVM-5046 [001] d.... 3410.005610: kvm_entry: vcpu 23 CPU 23/KVM-5046 [001] d.... 3410.005611: kvm_exit: vcpu 23 reason VMREAD CPU 23/KVM-5046 [001] d.... 3410.005611: kvm_entry: vcpu 23 CPU 23/KVM-5046 [001] d.... 3410.005612: kvm_exit: vcpu 23 reason VMREAD CPU 23/KVM-5046 [001] d.... 3410.005612: kvm_entry: vcpu 23 L1 writes DR7, L0 disables DR interception CPU 23/KVM-5046 [001] d.... 3410.005612: kvm_exit: vcpu 23 reason DR_ACCESS info1 0x0000000000000007 CPU 23/KVM-5046 [001] d.... 3410.005613: kvm_entry: vcpu 23 L0 writes DR6 = 0 (arch.dr6) CPU 23/KVM-5046 [001] d.... 3410.005613: <hack>: Set DRs, DR6 = 0xffff0ff0 A: <L1 writes DR6 = 1, no interception, arch.dr6 is still '0'> B: CPU 23/KVM-5046 [001] d.... 3410.005614: kvm_exit: vcpu 23 reason PREEMPTION_TIMER CPU 23/KVM-5046 [001] d.... 3410.005614: kvm_entry: vcpu 23 C: L0 writes DR6 = 0 (arch.dr6) CPU 23/KVM-5046 [001] d.... 3410.005614: <hack>: Set DRs, DR6 = 0xffff0ff0 L1 => L2 nested VM-Enter CPU 23/KVM-5046 [001] d.... 3410.005616: kvm_exit: vcpu 23 reason VMRESUME L0 reads DR6, arch.dr6 = 0 Reported-by: John Stultz <jstultz(a)google.com> Closes: https://lkml.kernel.org/r/CANDhNCq5_F3HfFYABqFGCA1bPd_%2BxgNj-iDQhH4tDk%2Bw… Fixes: 375e28ffc0cf ("KVM: X86: Set host DR6 only on VMX and for KVM_DEBUGREG_WONT_EXIT") Fixes: d67668e9dd76 ("KVM: x86, SVM: isolate vcpu->arch.dr6 from vmcb->save.dr6") Cc: stable(a)vger.kernel.org Cc: Jim Mattson <jmattson(a)google.com> Tested-by: John Stultz <jstultz(a)google.com> Link: https://lore.kernel.org/r/20250125011833.3644371-1-seanjc@google.com Signed-off-by: Sean Christopherson <seanjc(a)google.com> diff --git a/arch/x86/include/asm/kvm-x86-ops.h b/arch/x86/include/asm/kvm-x86-ops.h index c35550581da0..823c0434bbad 100644 --- a/arch/x86/include/asm/kvm-x86-ops.h +++ b/arch/x86/include/asm/kvm-x86-ops.h @@ -48,6 +48,7 @@ KVM_X86_OP(set_idt) KVM_X86_OP(get_gdt) KVM_X86_OP(set_gdt) KVM_X86_OP(sync_dirty_debug_regs) +KVM_X86_OP(set_dr6) KVM_X86_OP(set_dr7) KVM_X86_OP(cache_reg) KVM_X86_OP(get_rflags) diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h index b15cde0a9b5c..0b7af5902ff7 100644 --- a/arch/x86/include/asm/kvm_host.h +++ b/arch/x86/include/asm/kvm_host.h @@ -1696,6 +1696,7 @@ struct kvm_x86_ops { void (*get_gdt)(struct kvm_vcpu *vcpu, struct desc_ptr *dt); void (*set_gdt)(struct kvm_vcpu *vcpu, struct desc_ptr *dt); void (*sync_dirty_debug_regs)(struct kvm_vcpu *vcpu); + void (*set_dr6)(struct kvm_vcpu *vcpu, unsigned long value); void (*set_dr7)(struct kvm_vcpu *vcpu, unsigned long value); void (*cache_reg)(struct kvm_vcpu *vcpu, enum kvm_reg reg); unsigned long (*get_rflags)(struct kvm_vcpu *vcpu); diff --git a/arch/x86/kvm/svm/svm.c b/arch/x86/kvm/svm/svm.c index 7640a84e554a..a713c803a3a3 100644 --- a/arch/x86/kvm/svm/svm.c +++ b/arch/x86/kvm/svm/svm.c @@ -1991,11 +1991,11 @@ static void new_asid(struct vcpu_svm *svm, struct svm_cpu_data *sd) svm->asid = sd->next_asid++; } -static void svm_set_dr6(struct vcpu_svm *svm, unsigned long value) +static void svm_set_dr6(struct kvm_vcpu *vcpu, unsigned long value) { - struct vmcb *vmcb = svm->vmcb; + struct vmcb *vmcb = to_svm(vcpu)->vmcb; - if (svm->vcpu.arch.guest_state_protected) + if (vcpu->arch.guest_state_protected) return; if (unlikely(value != vmcb->save.dr6)) { @@ -4247,10 +4247,8 @@ static __no_kcsan fastpath_t svm_vcpu_run(struct kvm_vcpu *vcpu, * Run with all-zero DR6 unless needed, so that we can get the exact cause * of a #DB. */ - if (unlikely(vcpu->arch.switch_db_regs & KVM_DEBUGREG_WONT_EXIT)) - svm_set_dr6(svm, vcpu->arch.dr6); - else - svm_set_dr6(svm, DR6_ACTIVE_LOW); + if (likely(!(vcpu->arch.switch_db_regs & KVM_DEBUGREG_WONT_EXIT))) + svm_set_dr6(vcpu, DR6_ACTIVE_LOW); clgi(); kvm_load_guest_xsave_state(vcpu); @@ -5043,6 +5041,7 @@ static struct kvm_x86_ops svm_x86_ops __initdata = { .set_idt = svm_set_idt, .get_gdt = svm_get_gdt, .set_gdt = svm_set_gdt, + .set_dr6 = svm_set_dr6, .set_dr7 = svm_set_dr7, .sync_dirty_debug_regs = svm_sync_dirty_debug_regs, .cache_reg = svm_cache_reg, diff --git a/arch/x86/kvm/vmx/main.c b/arch/x86/kvm/vmx/main.c index 2427f918e763..43ee9ed11291 100644 --- a/arch/x86/kvm/vmx/main.c +++ b/arch/x86/kvm/vmx/main.c @@ -61,6 +61,7 @@ struct kvm_x86_ops vt_x86_ops __initdata = { .set_idt = vmx_set_idt, .get_gdt = vmx_get_gdt, .set_gdt = vmx_set_gdt, + .set_dr6 = vmx_set_dr6, .set_dr7 = vmx_set_dr7, .sync_dirty_debug_regs = vmx_sync_dirty_debug_regs, .cache_reg = vmx_cache_reg, diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c index f72835e85b6d..6c56d5235f0f 100644 --- a/arch/x86/kvm/vmx/vmx.c +++ b/arch/x86/kvm/vmx/vmx.c @@ -5648,6 +5648,12 @@ void vmx_sync_dirty_debug_regs(struct kvm_vcpu *vcpu) set_debugreg(DR6_RESERVED, 6); } +void vmx_set_dr6(struct kvm_vcpu *vcpu, unsigned long val) +{ + lockdep_assert_irqs_disabled(); + set_debugreg(vcpu->arch.dr6, 6); +} + void vmx_set_dr7(struct kvm_vcpu *vcpu, unsigned long val) { vmcs_writel(GUEST_DR7, val); @@ -7417,10 +7423,6 @@ fastpath_t vmx_vcpu_run(struct kvm_vcpu *vcpu, bool force_immediate_exit) vmx->loaded_vmcs->host_state.cr4 = cr4; } - /* When KVM_DEBUGREG_WONT_EXIT, dr6 is accessible in guest. */ - if (unlikely(vcpu->arch.switch_db_regs & KVM_DEBUGREG_WONT_EXIT)) - set_debugreg(vcpu->arch.dr6, 6); - /* When single-stepping over STI and MOV SS, we must clear the * corresponding interruptibility bits in the guest state. Otherwise * vmentry fails as it then expects bit 14 (BS) in pending debug diff --git a/arch/x86/kvm/vmx/x86_ops.h b/arch/x86/kvm/vmx/x86_ops.h index ce3295a67c04..430773a5ef8e 100644 --- a/arch/x86/kvm/vmx/x86_ops.h +++ b/arch/x86/kvm/vmx/x86_ops.h @@ -73,6 +73,7 @@ void vmx_get_idt(struct kvm_vcpu *vcpu, struct desc_ptr *dt); void vmx_set_idt(struct kvm_vcpu *vcpu, struct desc_ptr *dt); void vmx_get_gdt(struct kvm_vcpu *vcpu, struct desc_ptr *dt); void vmx_set_gdt(struct kvm_vcpu *vcpu, struct desc_ptr *dt); +void vmx_set_dr6(struct kvm_vcpu *vcpu, unsigned long val); void vmx_set_dr7(struct kvm_vcpu *vcpu, unsigned long val); void vmx_sync_dirty_debug_regs(struct kvm_vcpu *vcpu); void vmx_cache_reg(struct kvm_vcpu *vcpu, enum kvm_reg reg); diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index 8e77e61d4fbd..02159c967d29 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -10961,6 +10961,9 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu) set_debugreg(vcpu->arch.eff_db[1], 1); set_debugreg(vcpu->arch.eff_db[2], 2); set_debugreg(vcpu->arch.eff_db[3], 3); + /* When KVM_DEBUGREG_WONT_EXIT, dr6 is accessible in guest. */ + if (unlikely(vcpu->arch.switch_db_regs & KVM_DEBUGREG_WONT_EXIT)) + kvm_x86_call(set_dr6)(vcpu, vcpu->arch.dr6); } else if (unlikely(hw_breakpoint_active())) { set_debugreg(0, 7); }

1 month, 1 week

3
2
0 0

FAILED: patch "[PATCH] KVM: x86: Load DR6 with guest value only before entering" 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 c2fee09fc167c74a64adb08656cb993ea475197e # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2025021812-candied-hamper-7f08@gregkh' --subject-prefix 'PATCH 6.1.y' HEAD^.. Possible dependencies: thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From c2fee09fc167c74a64adb08656cb993ea475197e Mon Sep 17 00:00:00 2001 From: Sean Christopherson <seanjc(a)google.com> Date: Fri, 24 Jan 2025 17:18:33 -0800 Subject: [PATCH] KVM: x86: Load DR6 with guest value only before entering .vcpu_run() loop Move the conditional loading of hardware DR6 with the guest's DR6 value out of the core .vcpu_run() loop to fix a bug where KVM can load hardware with a stale vcpu->arch.dr6. When the guest accesses a DR and host userspace isn't debugging the guest, KVM disables DR interception and loads the guest's values into hardware on VM-Enter and saves them on VM-Exit. This allows the guest to access DRs at will, e.g. so that a sequence of DR accesses to configure a breakpoint only generates one VM-Exit. For DR0-DR3, the logic/behavior is identical between VMX and SVM, and also identical between KVM_DEBUGREG_BP_ENABLED (userspace debugging the guest) and KVM_DEBUGREG_WONT_EXIT (guest using DRs), and so KVM handles loading DR0-DR3 in common code, _outside_ of the core kvm_x86_ops.vcpu_run() loop. But for DR6, the guest's value doesn't need to be loaded into hardware for KVM_DEBUGREG_BP_ENABLED, and SVM provides a dedicated VMCB field whereas VMX requires software to manually load the guest value, and so loading the guest's value into DR6 is handled by {svm,vmx}_vcpu_run(), i.e. is done _inside_ the core run loop. Unfortunately, saving the guest values on VM-Exit is initiated by common x86, again outside of the core run loop. If the guest modifies DR6 (in hardware, when DR interception is disabled), and then the next VM-Exit is a fastpath VM-Exit, KVM will reload hardware DR6 with vcpu->arch.dr6 and clobber the guest's actual value. The bug shows up primarily with nested VMX because KVM handles the VMX preemption timer in the fastpath, and the window between hardware DR6 being modified (in guest context) and DR6 being read by guest software is orders of magnitude larger in a nested setup. E.g. in non-nested, the VMX preemption timer would need to fire precisely between #DB injection and the #DB handler's read of DR6, whereas with a KVM-on-KVM setup, the window where hardware DR6 is "dirty" extends all the way from L1 writing DR6 to VMRESUME (in L1). L1's view: ========== <L1 disables DR interception> CPU 0/KVM-7289 [023] d.... 2925.640961: kvm_entry: vcpu 0 A: L1 Writes DR6 CPU 0/KVM-7289 [023] d.... 2925.640963: <hack>: Set DRs, DR6 = 0xffff0ff1 B: CPU 0/KVM-7289 [023] d.... 2925.640967: kvm_exit: vcpu 0 reason EXTERNAL_INTERRUPT intr_info 0x800000ec D: L1 reads DR6, arch.dr6 = 0 CPU 0/KVM-7289 [023] d.... 2925.640969: <hack>: Sync DRs, DR6 = 0xffff0ff0 CPU 0/KVM-7289 [023] d.... 2925.640976: kvm_entry: vcpu 0 L2 reads DR6, L1 disables DR interception CPU 0/KVM-7289 [023] d.... 2925.640980: kvm_exit: vcpu 0 reason DR_ACCESS info1 0x0000000000000216 CPU 0/KVM-7289 [023] d.... 2925.640983: kvm_entry: vcpu 0 CPU 0/KVM-7289 [023] d.... 2925.640983: <hack>: Set DRs, DR6 = 0xffff0ff0 L2 detects failure CPU 0/KVM-7289 [023] d.... 2925.640987: kvm_exit: vcpu 0 reason HLT L1 reads DR6 (confirms failure) CPU 0/KVM-7289 [023] d.... 2925.640990: <hack>: Sync DRs, DR6 = 0xffff0ff0 L0's view: ========== L2 reads DR6, arch.dr6 = 0 CPU 23/KVM-5046 [001] d.... 3410.005610: kvm_exit: vcpu 23 reason DR_ACCESS info1 0x0000000000000216 CPU 23/KVM-5046 [001] ..... 3410.005610: kvm_nested_vmexit: vcpu 23 reason DR_ACCESS info1 0x0000000000000216 L2 => L1 nested VM-Exit CPU 23/KVM-5046 [001] ..... 3410.005610: kvm_nested_vmexit_inject: reason: DR_ACCESS ext_inf1: 0x0000000000000216 CPU 23/KVM-5046 [001] d.... 3410.005610: kvm_entry: vcpu 23 CPU 23/KVM-5046 [001] d.... 3410.005611: kvm_exit: vcpu 23 reason VMREAD CPU 23/KVM-5046 [001] d.... 3410.005611: kvm_entry: vcpu 23 CPU 23/KVM-5046 [001] d.... 3410.005612: kvm_exit: vcpu 23 reason VMREAD CPU 23/KVM-5046 [001] d.... 3410.005612: kvm_entry: vcpu 23 L1 writes DR7, L0 disables DR interception CPU 23/KVM-5046 [001] d.... 3410.005612: kvm_exit: vcpu 23 reason DR_ACCESS info1 0x0000000000000007 CPU 23/KVM-5046 [001] d.... 3410.005613: kvm_entry: vcpu 23 L0 writes DR6 = 0 (arch.dr6) CPU 23/KVM-5046 [001] d.... 3410.005613: <hack>: Set DRs, DR6 = 0xffff0ff0 A: <L1 writes DR6 = 1, no interception, arch.dr6 is still '0'> B: CPU 23/KVM-5046 [001] d.... 3410.005614: kvm_exit: vcpu 23 reason PREEMPTION_TIMER CPU 23/KVM-5046 [001] d.... 3410.005614: kvm_entry: vcpu 23 C: L0 writes DR6 = 0 (arch.dr6) CPU 23/KVM-5046 [001] d.... 3410.005614: <hack>: Set DRs, DR6 = 0xffff0ff0 L1 => L2 nested VM-Enter CPU 23/KVM-5046 [001] d.... 3410.005616: kvm_exit: vcpu 23 reason VMRESUME L0 reads DR6, arch.dr6 = 0 Reported-by: John Stultz <jstultz(a)google.com> Closes: https://lkml.kernel.org/r/CANDhNCq5_F3HfFYABqFGCA1bPd_%2BxgNj-iDQhH4tDk%2Bw… Fixes: 375e28ffc0cf ("KVM: X86: Set host DR6 only on VMX and for KVM_DEBUGREG_WONT_EXIT") Fixes: d67668e9dd76 ("KVM: x86, SVM: isolate vcpu->arch.dr6 from vmcb->save.dr6") Cc: stable(a)vger.kernel.org Cc: Jim Mattson <jmattson(a)google.com> Tested-by: John Stultz <jstultz(a)google.com> Link: https://lore.kernel.org/r/20250125011833.3644371-1-seanjc@google.com Signed-off-by: Sean Christopherson <seanjc(a)google.com> diff --git a/arch/x86/include/asm/kvm-x86-ops.h b/arch/x86/include/asm/kvm-x86-ops.h index c35550581da0..823c0434bbad 100644 --- a/arch/x86/include/asm/kvm-x86-ops.h +++ b/arch/x86/include/asm/kvm-x86-ops.h @@ -48,6 +48,7 @@ KVM_X86_OP(set_idt) KVM_X86_OP(get_gdt) KVM_X86_OP(set_gdt) KVM_X86_OP(sync_dirty_debug_regs) +KVM_X86_OP(set_dr6) KVM_X86_OP(set_dr7) KVM_X86_OP(cache_reg) KVM_X86_OP(get_rflags) diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h index b15cde0a9b5c..0b7af5902ff7 100644 --- a/arch/x86/include/asm/kvm_host.h +++ b/arch/x86/include/asm/kvm_host.h @@ -1696,6 +1696,7 @@ struct kvm_x86_ops { void (*get_gdt)(struct kvm_vcpu *vcpu, struct desc_ptr *dt); void (*set_gdt)(struct kvm_vcpu *vcpu, struct desc_ptr *dt); void (*sync_dirty_debug_regs)(struct kvm_vcpu *vcpu); + void (*set_dr6)(struct kvm_vcpu *vcpu, unsigned long value); void (*set_dr7)(struct kvm_vcpu *vcpu, unsigned long value); void (*cache_reg)(struct kvm_vcpu *vcpu, enum kvm_reg reg); unsigned long (*get_rflags)(struct kvm_vcpu *vcpu); diff --git a/arch/x86/kvm/svm/svm.c b/arch/x86/kvm/svm/svm.c index 7640a84e554a..a713c803a3a3 100644 --- a/arch/x86/kvm/svm/svm.c +++ b/arch/x86/kvm/svm/svm.c @@ -1991,11 +1991,11 @@ static void new_asid(struct vcpu_svm *svm, struct svm_cpu_data *sd) svm->asid = sd->next_asid++; } -static void svm_set_dr6(struct vcpu_svm *svm, unsigned long value) +static void svm_set_dr6(struct kvm_vcpu *vcpu, unsigned long value) { - struct vmcb *vmcb = svm->vmcb; + struct vmcb *vmcb = to_svm(vcpu)->vmcb; - if (svm->vcpu.arch.guest_state_protected) + if (vcpu->arch.guest_state_protected) return; if (unlikely(value != vmcb->save.dr6)) { @@ -4247,10 +4247,8 @@ static __no_kcsan fastpath_t svm_vcpu_run(struct kvm_vcpu *vcpu, * Run with all-zero DR6 unless needed, so that we can get the exact cause * of a #DB. */ - if (unlikely(vcpu->arch.switch_db_regs & KVM_DEBUGREG_WONT_EXIT)) - svm_set_dr6(svm, vcpu->arch.dr6); - else - svm_set_dr6(svm, DR6_ACTIVE_LOW); + if (likely(!(vcpu->arch.switch_db_regs & KVM_DEBUGREG_WONT_EXIT))) + svm_set_dr6(vcpu, DR6_ACTIVE_LOW); clgi(); kvm_load_guest_xsave_state(vcpu); @@ -5043,6 +5041,7 @@ static struct kvm_x86_ops svm_x86_ops __initdata = { .set_idt = svm_set_idt, .get_gdt = svm_get_gdt, .set_gdt = svm_set_gdt, + .set_dr6 = svm_set_dr6, .set_dr7 = svm_set_dr7, .sync_dirty_debug_regs = svm_sync_dirty_debug_regs, .cache_reg = svm_cache_reg, diff --git a/arch/x86/kvm/vmx/main.c b/arch/x86/kvm/vmx/main.c index 2427f918e763..43ee9ed11291 100644 --- a/arch/x86/kvm/vmx/main.c +++ b/arch/x86/kvm/vmx/main.c @@ -61,6 +61,7 @@ struct kvm_x86_ops vt_x86_ops __initdata = { .set_idt = vmx_set_idt, .get_gdt = vmx_get_gdt, .set_gdt = vmx_set_gdt, + .set_dr6 = vmx_set_dr6, .set_dr7 = vmx_set_dr7, .sync_dirty_debug_regs = vmx_sync_dirty_debug_regs, .cache_reg = vmx_cache_reg, diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c index f72835e85b6d..6c56d5235f0f 100644 --- a/arch/x86/kvm/vmx/vmx.c +++ b/arch/x86/kvm/vmx/vmx.c @@ -5648,6 +5648,12 @@ void vmx_sync_dirty_debug_regs(struct kvm_vcpu *vcpu) set_debugreg(DR6_RESERVED, 6); } +void vmx_set_dr6(struct kvm_vcpu *vcpu, unsigned long val) +{ + lockdep_assert_irqs_disabled(); + set_debugreg(vcpu->arch.dr6, 6); +} + void vmx_set_dr7(struct kvm_vcpu *vcpu, unsigned long val) { vmcs_writel(GUEST_DR7, val); @@ -7417,10 +7423,6 @@ fastpath_t vmx_vcpu_run(struct kvm_vcpu *vcpu, bool force_immediate_exit) vmx->loaded_vmcs->host_state.cr4 = cr4; } - /* When KVM_DEBUGREG_WONT_EXIT, dr6 is accessible in guest. */ - if (unlikely(vcpu->arch.switch_db_regs & KVM_DEBUGREG_WONT_EXIT)) - set_debugreg(vcpu->arch.dr6, 6); - /* When single-stepping over STI and MOV SS, we must clear the * corresponding interruptibility bits in the guest state. Otherwise * vmentry fails as it then expects bit 14 (BS) in pending debug diff --git a/arch/x86/kvm/vmx/x86_ops.h b/arch/x86/kvm/vmx/x86_ops.h index ce3295a67c04..430773a5ef8e 100644 --- a/arch/x86/kvm/vmx/x86_ops.h +++ b/arch/x86/kvm/vmx/x86_ops.h @@ -73,6 +73,7 @@ void vmx_get_idt(struct kvm_vcpu *vcpu, struct desc_ptr *dt); void vmx_set_idt(struct kvm_vcpu *vcpu, struct desc_ptr *dt); void vmx_get_gdt(struct kvm_vcpu *vcpu, struct desc_ptr *dt); void vmx_set_gdt(struct kvm_vcpu *vcpu, struct desc_ptr *dt); +void vmx_set_dr6(struct kvm_vcpu *vcpu, unsigned long val); void vmx_set_dr7(struct kvm_vcpu *vcpu, unsigned long val); void vmx_sync_dirty_debug_regs(struct kvm_vcpu *vcpu); void vmx_cache_reg(struct kvm_vcpu *vcpu, enum kvm_reg reg); diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index 8e77e61d4fbd..02159c967d29 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -10961,6 +10961,9 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu) set_debugreg(vcpu->arch.eff_db[1], 1); set_debugreg(vcpu->arch.eff_db[2], 2); set_debugreg(vcpu->arch.eff_db[3], 3); + /* When KVM_DEBUGREG_WONT_EXIT, dr6 is accessible in guest. */ + if (unlikely(vcpu->arch.switch_db_regs & KVM_DEBUGREG_WONT_EXIT)) + kvm_x86_call(set_dr6)(vcpu, vcpu->arch.dr6); } else if (unlikely(hw_breakpoint_active())) { set_debugreg(0, 7); }

1 month, 1 week

3
2
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror March 2025