Skip to contentSkip to conversation

gpio: pxa: handle corner case of unprobed device

PATCH In review Sasha Levin 2 participants 29 messages Mar 30, 2019 – Apr 3, 2019

Subsystems (17)
lore
Share & tools
lore

Sasha Levin

author maintainer
#0 lore raw

PATCH 1/37 gpio: pxa: handle corner case of unprobed device

From: Robert Jarzmik <[email protected]>

[ Upstream commit 9ce3ebe973bf4073426f35f282c6b955ed802765 ]

In the corner case where the gpio driver probe fails, for whatever reason, the suspend and resume handlers will still be called as they have to be registered as syscore operations. This applies as well when no probe was called while the driver has been built in the kernel.

Nicolas tracked this in :
https://bugzilla.kernel.org/show_bug.cgi?id=200905

Therefore, add a failsafe in these function, and test if a proper probe succeeded and the driver is functional.

---
 drivers/gpio/gpio-pxa.c | 6 ++++++
 1 file changed, 6 insertions(+)
drivers/gpio/gpio-pxa.c +6 -0
index 2943dfc4c470..822ad220f0af 100644
@@ -776,6 +776,9 @@ static int pxa_gpio_suspend(void)
 	struct pxa_gpio_bank *c;
 	int gpio;
 
+	if (!pchip)
+		return 0;
+
 	for_each_gpio_bank(gpio, c, pchip) {
 		c->saved_gplr = readl_relaxed(c->regbase + GPLR_OFFSET);
 		c->saved_gpdr = readl_relaxed(c->regbase + GPDR_OFFSET);
@@ -794,6 +797,9 @@ static void pxa_gpio_resume(void)
 	struct pxa_gpio_bank *c;
 	int gpio;
 
+	if (!pchip)
+		return;
+
 	for_each_gpio_bank(gpio, c, pchip) {
 		/* restore level with set/clear */
 		writel_relaxed(c->saved_gplr, c->regbase + GPSR_OFFSET);
Signature · 1 line
2.19.1
author maintainer
#1 lore raw

PATCH 2/37 rsi: improve kernel thread handling to fix kernel panic

From: Siva Rebbagondla <[email protected]>

[ Upstream commit 4c62764d0fc21a34ffc44eec1210038c3a2e4473 ]

While running regressions, observed below kernel panic when sdio disconnect called. This is because of, kthread_stop() is taking care of wait_for_completion() by default. When wait_for_completion triggered in kthread_stop and as it was done already, giving kernel panic. Hence, removing redundant wait_for_completion() from rsi_kill_thread().

... skipping ...

BUG: unable to handle kernel NULL pointer dereference at           (null)
IP: [<ffffffff810a63df>] exit_creds+0x1f/0x50
PGD 0
Oops: 0002 [#1] SMP
CPU: 0 PID: 6502 Comm: rmmod Tainted: G  OE   4.15.9-Generic #154-Ubuntu
Hardware name: Dell Inc. Edge Gateway 3003/ , BIOS 01.00.00 04/17/2017
Stack:
ffff88007392e600 ffff880075847dc0 ffffffff8108160a 0000000000000000
ffff88007392e600 ffff880075847de8 ffffffff810a484b ffff880076127000
ffff88003cd3a800 ffff880074f12a00 ffff880075847e28 ffffffffc09bed15
Call Trace:
[<ffffffff8108160a>] __put_task_struct+0x5a/0x140
[<ffffffff810a484b>] kthread_stop+0x10b/0x110
[<ffffffffc09bed15>] rsi_disconnect+0x2f5/0x300 [ven_rsi_sdio]
[<ffffffff81578bcb>] ? __pm_runtime_resume+0x5b/0x80
[<ffffffff816f0918>] sdio_bus_remove+0x38/0x100
[<ffffffff8156cc64>] __device_release_driver+0xa4/0x150
[<ffffffff8156d7a5>] driver_detach+0xb5/0xc0
[<ffffffff8156c6c5>] bus_remove_driver+0x55/0xd0
[<ffffffff8156dfbc>] driver_unregister+0x2c/0x50
[<ffffffff816f0b8a>] sdio_unregister_driver+0x1a/0x20
[<ffffffffc09bf0f5>] rsi_module_exit+0x15/0x30 [ven_rsi_sdio]
[<ffffffff8110cad8>] SyS_delete_module+0x1b8/0x210
[<ffffffff81851dc8>] entry_SYSCALL_64_fastpath+0x1c/0xbb
---
 drivers/net/wireless/rsi/rsi_common.h | 1 -
 1 file changed, 1 deletion(-)
drivers/net/wireless/rsi/rsi_common.h +0 -1
index e579d694d13c..21986ba56a3c 100644
@@ -74,7 +74,6 @@ static inline int rsi_kill_thread(struct rsi_thread *handle)
 	atomic_inc(&handle->thread_done);
 	rsi_set_event(&handle->event);
 
-	wait_for_completion(&handle->completion);
 	return kthread_stop(handle->task);
 }
 
Signature · 1 line
2.19.1
author maintainer
#2 lore raw

PATCH 3/37 9p: do not trust pdu content for stat item size

In reply to Sasha Levin

From: Gertjan Halkes <[email protected]>

[ Upstream commit 2803cf4379ed252894f046cb8812a48db35294e3 ]

v9fs_dir_readdir() could deadloop if a struct was sent with a size set to -2

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=88021

---
 fs/9p/vfs_dir.c   | 8 +++-----
 net/9p/protocol.c | 3 ++-
 2 files changed, 5 insertions(+), 6 deletions(-)
fs/9p/vfs_dir.c +3 -5
index 48db9a9f13f9..cb6c4031af55 100644
@@ -105,7 +105,6 @@ static int v9fs_dir_readdir(struct file *file, struct dir_context *ctx)
 	int err = 0;
 	struct p9_fid *fid;
 	int buflen;
-	int reclen = 0;
 	struct p9_rdir *rdir;
 	struct kvec kvec;
 
@@ -138,11 +137,10 @@ static int v9fs_dir_readdir(struct file *file, struct dir_context *ctx)
 		while (rdir->head < rdir->tail) {
 			err = p9stat_read(fid->clnt, rdir->buf + rdir->head,
 					  rdir->tail - rdir->head, &st);
-			if (err) {
+			if (err <= 0) {
 				p9_debug(P9_DEBUG_VFS, "returned %d\n", err);
 				return -EIO;
 			}
-			reclen = st.size+2;
 
 			over = !dir_emit(ctx, st.name, strlen(st.name),
 					 v9fs_qid2ino(&st.qid), dt_type(&st));
@@ -150,8 +148,8 @@ static int v9fs_dir_readdir(struct file *file, struct dir_context *ctx)
 			if (over)
 				return 0;
 
-			rdir->head += reclen;
-			ctx->pos += reclen;
+			rdir->head += err;
+			ctx->pos += err;
 		}
 	}
 }
net/9p/protocol.c +2 -1
index 9743837aebc6..766d1ef4640a 100644
@@ -570,9 +570,10 @@ int p9stat_read(struct p9_client *clnt, char *buf, int len, struct p9_wstat *st)
 	if (ret) {
 		p9_debug(P9_DEBUG_9P, "<<< p9stat_read failed: %d\n", ret);
 		trace_9p_protocol_dump(clnt, &fake_pdu);
+		return ret;
 	}
 
-	return ret;
+	return fake_pdu.offset;
 }
 EXPORT_SYMBOL(p9stat_read);
 
Signature · 1 line
2.19.1
author maintainer
#3 lore raw

PATCH 4/37 9p locks: add mount option for lock retry interval

In reply to Sasha Levin

From: Dinu-Razvan Chis-Serban <[email protected]>

[ Upstream commit 5e172f75e51e3de1b4274146d9b990f803cb5c2a ]

The default P9_LOCK_TIMEOUT can be too long for some users exporting a local file system to a guest VM (30s), make this configurable at mount time.

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=195727

---
 fs/9p/v9fs.c     | 21 +++++++++++++++++++++
 fs/9p/v9fs.h     |  1 +
 fs/9p/vfs_file.c |  6 +++++-
 3 files changed, 27 insertions(+), 1 deletion(-)
fs/9p/v9fs.c +21 -0
index 8fb89ddc6cc7..c52f10efdc9c 100644
@@ -61,6 +61,8 @@ enum {
 	Opt_cache_loose, Opt_fscache, Opt_mmap,
 	/* Access options */
 	Opt_access, Opt_posixacl,
+	/* Lock timeout option */
+	Opt_locktimeout,
 	/* Error token */
 	Opt_err
 };
@@ -80,6 +82,7 @@ static const match_table_t tokens = {
 	{Opt_cachetag, "cachetag=%s"},
 	{Opt_access, "access=%s"},
 	{Opt_posixacl, "posixacl"},
+	{Opt_locktimeout, "locktimeout=%u"},
 	{Opt_err, NULL}
 };
 
@@ -187,6 +190,7 @@ static int v9fs_parse_options(struct v9fs_session_info *v9ses, char *opts)
 #ifdef CONFIG_9P_FSCACHE
 	v9ses->cachetag = NULL;
 #endif
+	v9ses->session_lock_timeout = P9_LOCK_TIMEOUT;
 
 	if (!opts)
 		return 0;
@@ -360,6 +364,23 @@ static int v9fs_parse_options(struct v9fs_session_info *v9ses, char *opts)
 #endif
 			break;
 
+		case Opt_locktimeout:
+			r = match_int(&args[0], &option);
+			if (r < 0) {
+				p9_debug(P9_DEBUG_ERROR,
+					 "integer field, but no integer?\n");
+				ret = r;
+				continue;
+			}
+			if (option < 1) {
+				p9_debug(P9_DEBUG_ERROR,
+					 "locktimeout must be a greater than zero integer.\n");
+				ret = -EINVAL;
+				continue;
+			}
+			v9ses->session_lock_timeout = (long)option * HZ;
+			break;
+
 		default:
 			continue;
 		}
fs/9p/v9fs.h +1 -0
index 982e017acadb..129e5243a6bf 100644
@@ -116,6 +116,7 @@ struct v9fs_session_info {
 	struct p9_client *clnt;	/* 9p client */
 	struct list_head slist; /* list of sessions registered with v9fs */
 	struct rw_semaphore rename_sem;
+	long session_lock_timeout; /* retry interval for blocking locks */
 };
 
 /* cache_validity flags */
fs/9p/vfs_file.c +5 -1
index af8cac975a74..89e69904976a 100644
@@ -154,6 +154,7 @@ static int v9fs_file_do_lock(struct file *filp, int cmd, struct file_lock *fl)
 	uint8_t status = P9_LOCK_ERROR;
 	int res = 0;
 	unsigned char fl_type;
+	struct v9fs_session_info *v9ses;
 
 	fid = filp->private_data;
 	BUG_ON(fid == NULL);
@@ -189,6 +190,8 @@ static int v9fs_file_do_lock(struct file *filp, int cmd, struct file_lock *fl)
 	if (IS_SETLKW(cmd))
 		flock.flags = P9_LOCK_FLAGS_BLOCK;
 
+	v9ses = v9fs_inode2v9ses(file_inode(filp));
+
 	/*
 	 * if its a blocked request and we get P9_LOCK_BLOCKED as the status
 	 * for lock request, keep on trying
@@ -202,7 +205,8 @@ static int v9fs_file_do_lock(struct file *filp, int cmd, struct file_lock *fl)
 			break;
 		if (status == P9_LOCK_BLOCKED && !IS_SETLKW(cmd))
 			break;
-		if (schedule_timeout_interruptible(P9_LOCK_TIMEOUT) != 0)
+		if (schedule_timeout_interruptible(v9ses->session_lock_timeout)
+				!= 0)
 			break;
 		/*
 		 * p9_client_lock_dotl overwrites flock.client_id with the
Signature · 1 line
2.19.1
author maintainer
#4 lore raw

PATCH 5/37 f2fs: fix to do sanity check with current segment number

In reply to Sasha Levin

From: Chao Yu <[email protected]>

[ Upstream commit 042be0f849e5fc24116d0afecfaf926eed5cac63 ]

https://bugzilla.kernel.org/show_bug.cgi?id=200219

Reproduction way:
- mount image
- run poc code
- umount image

F2FS-fs (loop1): Bitmap was wrongly set, blk:15364

------------[ cut here ]------------
kernel BUG at /home/yuchao/git/devf2fs/segment.c:2061!
invalid opcode: 0000 [#1] PREEMPT SMP
CPU: 2 PID: 17686 Comm: umount Tainted: G        W  O      4.18.0-rc2+ #39
Hardware name: innotek GmbH VirtualBox/VirtualBox, BIOS VirtualBox 12/01/2006
EIP: update_sit_entry+0x459/0x4e0 [f2fs]
Code: e8 1c b5 fd ff 0f 0b 0f 0b 8b 45 e4 c7 44 24 08 9c 7a 6c f8 c7 44 24 04 bc 4a 6c f8 89 44 24 0c 8b 06 89 04 24 e8 f7 b4 fd ff <0f> 0b 8b 45 e4 0f b6 d2 89 54 24 10 c7 44 24 08 60 7a 6c f8 c7 44
EAX: 00000032 EBX: 000000f8 ECX: 00000002 EDX: 00000001
ESI: d7177000 EDI: f520fe68 EBP: d6477c6c ESP: d6477c34
DS: 007b ES: 007b FS: 00d8 GS: 00e0 SS: 0068 EFLAGS: 00010282
CR0: 80050033 CR2: b7fbe000 CR3: 2a99b3c0 CR4: 000406f0
Call Trace:
 f2fs_allocate_data_block+0x124/0x580 [f2fs]
 do_write_page+0x78/0x150 [f2fs]
 f2fs_do_write_node_page+0x25/0xa0 [f2fs]
 __write_node_page+0x2bf/0x550 [f2fs]
 f2fs_sync_node_pages+0x60e/0x6d0 [f2fs]
 ? sync_inode_metadata+0x2f/0x40
 ? f2fs_write_checkpoint+0x28f/0x7d0 [f2fs]
 ? up_write+0x1e/0x80
 f2fs_write_checkpoint+0x2a9/0x7d0 [f2fs]
 ? mark_held_locks+0x5d/0x80
 ? _raw_spin_unlock_irq+0x27/0x50
 kill_f2fs_super+0x68/0x90 [f2fs]
 deactivate_locked_super+0x3d/0x70
 deactivate_super+0x40/0x60
 cleanup_mnt+0x39/0x70
 __cleanup_mnt+0x10/0x20
 task_work_run+0x81/0xa0
 exit_to_usermode_loop+0x59/0xa7
 do_fast_syscall_32+0x1f5/0x22c
 entry_SYSENTER_32+0x53/0x86
EIP: 0xb7f95c51
Code: c1 1e f7 ff ff 89 e5 8b 55 08 85 d2 8b 81 64 cd ff ff 74 02 89 02 5d c3 8b 0c 24 c3 8b 1c 24 c3 90 51 52 55 89 e5 0f 34 cd 80 <5d> 5a 59 c3 90 90 90 90 8d 76 00 58 b8 77 00 00 00 cd 80 90 8d 76
EAX: 00000000 EBX: 0871ab90 ECX: bfb2cd00 EDX: 00000000
ESI: 00000000 EDI: 0871ab90 EBP: 0871ab90 ESP: bfb2cd7c
DS: 007b ES: 007b FS: 0000 GS: 0033 SS: 007b EFLAGS: 00000246
Modules linked in: f2fs(O) crc32_generic bnep rfcomm bluetooth ecdh_generic snd_intel8x0 snd_ac97_codec ac97_bus snd_pcm snd_seq_midi snd_seq_midi_event snd_rawmidi snd_seq pcbc joydev aesni_intel snd_seq_device aes_i586 snd_timer crypto_simd snd cryptd soundcore mac_hid serio_raw video i2c_piix4 parport_pc ppdev lp parport hid_generic psmouse usbhid hid e1000 [last unloaded: f2fs]
---[ end trace d423f83982cfcdc5 ]---

The reason is, different log headers using the same segment, once one log's next block address is used by another log, it will cause panic as above.

Main area: 24 segs, 24 secs 24 zones

  - COLD  data: 0, 0, 0
  - WARM  data: 1, 1, 1
  - HOT   data: 20, 20, 20
  - Dir   dnode: 22, 22, 22

- File dnode: 22, 22, 22
- Indir nodes: 21, 21, 21

So this patch adds sanity check to detect such condition to avoid this issue.

---
 fs/f2fs/super.c | 34 +++++++++++++++++++++++++++++++++-
 1 file changed, 33 insertions(+), 1 deletion(-)
fs/f2fs/super.c +33 -1
index fc5c41257e68..4c169ba50c0f 100644
@@ -1959,7 +1959,7 @@ int sanity_check_ckpt(struct f2fs_sb_info *sbi)
 	unsigned int segment_count_main;
 	unsigned int cp_pack_start_sum, cp_payload;
 	block_t user_block_count;
-	int i;
+	int i, j;
 
 	total = le32_to_cpu(raw_super->segment_count);
 	fsmeta = le32_to_cpu(raw_super->segment_count_ckpt);
@@ -2000,11 +2000,43 @@ int sanity_check_ckpt(struct f2fs_sb_info *sbi)
 		if (le32_to_cpu(ckpt->cur_node_segno[i]) >= main_segs ||
 			le16_to_cpu(ckpt->cur_node_blkoff[i]) >= blocks_per_seg)
 			return 1;
+		for (j = i + 1; j < NR_CURSEG_NODE_TYPE; j++) {
+			if (le32_to_cpu(ckpt->cur_node_segno[i]) ==
+				le32_to_cpu(ckpt->cur_node_segno[j])) {
+				f2fs_msg(sbi->sb, KERN_ERR,
+					"Node segment (%u, %u) has the same "
+					"segno: %u", i, j,
+					le32_to_cpu(ckpt->cur_node_segno[i]));
+				return 1;
+			}
+		}
 	}
 	for (i = 0; i < NR_CURSEG_DATA_TYPE; i++) {
 		if (le32_to_cpu(ckpt->cur_data_segno[i]) >= main_segs ||
 			le16_to_cpu(ckpt->cur_data_blkoff[i]) >= blocks_per_seg)
 			return 1;
+		for (j = i + 1; j < NR_CURSEG_DATA_TYPE; j++) {
+			if (le32_to_cpu(ckpt->cur_data_segno[i]) ==
+				le32_to_cpu(ckpt->cur_data_segno[j])) {
+				f2fs_msg(sbi->sb, KERN_ERR,
+					"Data segment (%u, %u) has the same "
+					"segno: %u", i, j,
+					le32_to_cpu(ckpt->cur_data_segno[i]));
+				return 1;
+			}
+		}
+	}
+	for (i = 0; i < NR_CURSEG_NODE_TYPE; i++) {
+		for (j = i; j < NR_CURSEG_DATA_TYPE; j++) {
+			if (le32_to_cpu(ckpt->cur_node_segno[i]) ==
+				le32_to_cpu(ckpt->cur_data_segno[j])) {
+				f2fs_msg(sbi->sb, KERN_ERR,
+					"Data segment (%u) and Data segment (%u)"
+					" has the same segno: %u", i, j,
+					le32_to_cpu(ckpt->cur_node_segno[i]));
+				return 1;
+			}
+		}
 	}
 
 	sit_bitmap_size = le32_to_cpu(ckpt->sit_ver_bitmap_bytesize);
Signature · 1 line
2.19.1
author maintainer
#5 lore raw

PATCH 6/37 netfilter: xt_cgroup: shrink size of v2 path

In reply to Sasha Levin

From: Pablo Neira Ayuso <[email protected]>

[ Upstream commit 0d704967f4a49cc2212350b3e4a8231f8b4283ed ]

cgroup v2 path field is PATH_MAX which is too large, this is placing too much pressure on memory allocation for people with many rules doing cgroup v1 classid matching, side effects of this are bug reports like:

https://bugzilla.kernel.org/show_bug.cgi?id=200639

This patch registers a new revision that shrinks the cgroup path to 512 bytes, which is the same approach we follow in similar extensions that have a path field.

---
 include/uapi/linux/netfilter/xt_cgroup.h | 16 ++++++
 net/netfilter/xt_cgroup.c                | 72 ++++++++++++++++++++++++
 2 files changed, 88 insertions(+)
include/uapi/linux/netfilter/xt_cgroup.h +16 -0
index e96dfa1b34f7..b74e370d6133 100644
@@ -22,4 +22,20 @@ struct xt_cgroup_info_v1 {
 	void		*priv __attribute__((aligned(8)));
 };
 
+#define XT_CGROUP_PATH_MAX	512
+
+struct xt_cgroup_info_v2 {
+	__u8		has_path;
+	__u8		has_classid;
+	__u8		invert_path;
+	__u8		invert_classid;
+	union {
+		char	path[XT_CGROUP_PATH_MAX];
+		__u32	classid;
+	};
+
+	/* kernel internal data */
+	void		*priv __attribute__((aligned(8)));
+};
+
 #endif /* _UAPI_XT_CGROUP_H */
net/netfilter/xt_cgroup.c +72 -0

96 lines not on this page · raw message

Signature · 1 line
2.19.1
author maintainer
#6 lore raw

PATCH 7/37 serial: uartps: console_setup() can't be placed to init section

In reply to Sasha Levin

From: Michal Simek <[email protected]>

[ Upstream commit 4bb1ce2350a598502b23088b169e16b43d4bc639 ]

When console device is rebinded, console_setup() is called again. But marking it as __init means that function will be clear after boot is complete. If console device is binded again console_setup() is not found and error "Unable to handle kernel paging request at virtual address" is reported.

---
 drivers/tty/serial/xilinx_uartps.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/tty/serial/xilinx_uartps.c +1 -1
index f438a2158006..b0da63737aa1 100644
@@ -1270,7 +1270,7 @@ static void cdns_uart_console_write(struct console *co, const char *s,
  *
  * Return: 0 on success, negative errno otherwise.
  */
-static int __init cdns_uart_console_setup(struct console *co, char *options)
+static int cdns_uart_console_setup(struct console *co, char *options)
 {
 	struct uart_port *port = &cdns_uart_port[co->index];
 	int baud = 9600;
Signature · 1 line
2.19.1
author maintainer
#7 lore raw

PATCH 8/37 powerpc/pseries: Remove prrn_work workqueue

In reply to Sasha Levin

From: Nathan Fontenot <[email protected]>

[ Upstream commit cd24e457fd8b2d087d9236700c8d2957054598bf ]

When a PRRN event is received we are already running in a worker thread. Instead of spawning off another worker thread on the prrn_work workqueue to handle the PRRN event we can just call the PRRN handler routine directly.

With this update we can also pass the scope variable for the PRRN event directly to the handler instead of it being a global variable.

This patch fixes the following oops mnessage we are seeing in PRRN testing:

  Oops: Bad kernel stack pointer, sig: 6 [#1]
  SMP NR_CPUS=2048 NUMA pSeries
  Modules linked in: nfsv3 nfs_acl rpcsec_gss_krb5 auth_rpcgss nfsv4 dns_resolver nfs lockd grace sunrpc fscache binfmt_misc reiserfs vfat fat rpadlpar_io(X) rpaphp(X) tcp_diag udp_diag inet_diag unix_diag af_packet_diag netlink_diag af_packet xfs libcrc32c dm_service_time ibmveth(X) ses enclosure scsi_transport_sas rtc_generic btrfs xor raid6_pq sd_mod ibmvscsi(X) scsi_transport_srp ipr(X) libata sg dm_multipath dm_mod scsi_dh_rdac scsi_dh_emc scsi_dh_alua scsi_mod autofs4
  Supported: Yes, External                                                     54
  CPU: 7 PID: 18967 Comm: kworker/u96:0 Tainted: G                 X 4.4.126-94.22-default #1
  Workqueue: pseries hotplug workque pseries_hp_work_fn
  task: c000000775367790 ti: c00000001ebd4000 task.ti: c00000070d140000
  NIP: 0000000000000000 LR: 000000001fb3d050 CTR: 0000000000000000
  REGS: c00000001ebd7d40 TRAP: 0700   Tainted: G                 X  (4.4.126-94.22-default)
  MSR: 8000000102081000 <41,VEC,ME5  CR: 28000002  XER: 20040018   4
  CFAR: 000000001fb3d084 40 419   1                                3
  GPR00: 000000000000000040000000000010007 000000001ffff400 000000041fffe200
  GPR04: 000000000000008050000000000000000 000000001fb15fa8 0000000500000500
  GPR08: 000000000001f40040000000000000001 0000000000000000 000005:5200040002
  GPR12: 00000000000000005c000000007a05400 c0000000000e89f8 000000001ed9f668
  GPR16: 000000001fbeff944000000001fbeff94 000000001fb545e4 0000006000000060
  GPR20: ffffffffffffffff4ffffffffffffffff 0000000000000000 0000000000000000
  GPR24: 00000000000000005400000001fb3c000 0000000000000000 000000001fb1b040
  GPR28: 000000001fb240004000000001fb440d8 0000000000000008 0000000000000000
  NIP [0000000000000000] 5         (null)
  LR [000000001fb3d050] 031fb3d050
  Call Trace:            4
  Instruction dump:      4                                       5:47 12    2
  XXXXXXXX XXXXXXXX XXXXX4XX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX
  XXXXXXXX XXXXXXXX XXXXX5XX XXXXXXXX 60000000 60000000 60000000 60000000
  ---[ end trace aa5627b04a7d9d6b ]---                                       3NMI watchdog: BUG: soft lockup - CPU#27 stuck for 23s! [kworker/27:0:13903]
  Modules linked in: nfsv3 nfs_acl rpcsec_gss_krb5 auth_rpcgss nfsv4 dns_resolver nfs lockd grace sunrpc fscache binfmt_misc reiserfs vfat fat rpadlpar_io(X) rpaphp(X) tcp_diag udp_diag inet_diag unix_diag af_packet_diag netlink_diag af_packet xfs libcrc32c dm_service_time ibmveth(X) ses enclosure scsi_transport_sas rtc_generic btrfs xor raid6_pq sd_mod ibmvscsi(X) scsi_transport_srp ipr(X) libata sg dm_multipath dm_mod scsi_dh_rdac scsi_dh_emc scsi_dh_alua scsi_mod autofs4
  Supported: Yes, External
  CPU: 27 PID: 13903 Comm: kworker/27:0 Tainted: G      D          X 4.4.126-94.22-default #1
  Workqueue: events prrn_work_fn
  task: c000000747cfa390 ti: c00000074712c000 task.ti: c00000074712c000
  NIP: c0000000008002a8 LR: c000000000090770 CTR: 000000000032e088
  REGS: c00000074712f7b0 TRAP: 0901   Tainted: G      D          X  (4.4.126-94.22-default)
  MSR: 8000000100009033 <SF,EE,ME,IR,DR,RI,LE>  CR: 22482044  XER: 20040000
  CFAR: c0000000008002c4 SOFTE: 1
  GPR00: c000000000090770 c00000074712fa30 c000000000f09800 c000000000fa1928 6:02
  GPR04: c000000775f5e000 fffffffffffffffe 0000000000000001 c000000000f42db8
  GPR08: 0000000000000001 0000000080000007 0000000000000000 0000000000000000
  GPR12: 8006210083180000 c000000007a14400
  NIP [c0000000008002a8] _raw_spin_lock+0x68/0xd0
  LR [c000000000090770] mobility_rtas_call+0x50/0x100
  Call Trace:            59                                        5
  [c00000074712fa60] [c000000000090770] mobility_rtas_call+0x50/0x100
  [c00000074712faf0] [c000000000090b08] pseries_devicetree_update+0xf8/0x530
  [c00000074712fc20] [c000000000031ba4] prrn_work_fn+0x34/0x50
  [c00000074712fc40] [c0000000000e0390] process_one_work+0x1a0/0x4e0
  [c00000074712fcd0] [c0000000000e0870] worker_thread+0x1a0/0x6105:57       2
  [c00000074712fd80] [c0000000000e8b18] kthread+0x128/0x150
  [c00000074712fe30] [c0000000000096f8] ret_from_kernel_thread+0x5c/0x64
  Instruction dump:
  2c090000 40c20010 7d40192d 40c2fff0 7c2004ac 2fa90000 40de0018 5:540030   3
  e8010010 ebe1fff8 7c0803a6 4e800020 <7c210b78> e92d0000 89290009 792affe3
---
 arch/powerpc/kernel/rtasd.c | 17 +++--------------
 1 file changed, 3 insertions(+), 14 deletions(-)
arch/powerpc/kernel/rtasd.c +3 -14
index 0f0b1b2f3b60..7caeae73348d 100644
@@ -274,27 +274,16 @@ void pSeries_log_error(char *buf, unsigned int err_type, int fatal)
 }
 
 #ifdef CONFIG_PPC_PSERIES
-static s32 prrn_update_scope;
-
-static void prrn_work_fn(struct work_struct *work)
+static void handle_prrn_event(s32 scope)
 {
 	/*
 	 * For PRRN, we must pass the negative of the scope value in
 	 * the RTAS event.
 	 */
-	pseries_devicetree_update(-prrn_update_scope);
+	pseries_devicetree_update(-scope);
 	numa_update_cpu_topology(false);
 }
 
-static DECLARE_WORK(prrn_work, prrn_work_fn);
-
-static void prrn_schedule_update(u32 scope)
-{
-	flush_work(&prrn_work);
-	prrn_update_scope = scope;
-	schedule_work(&prrn_work);
-}
-
 static void handle_rtas_event(const struct rtas_error_log *log)
 {
 	if (rtas_error_type(log) != RTAS_TYPE_PRRN || !prrn_is_enabled())
@@ -303,7 +292,7 @@ static void handle_rtas_event(const struct rtas_error_log *log)
 	/* For PRRN Events the extended log length is used to denote
 	 * the scope for calling rtas update-nodes.
 	 */
-	prrn_schedule_update(rtas_error_extended_log_length(log));
+	handle_prrn_event(rtas_error_extended_log_length(log));
 }
 
 #else
Signature · 1 line
2.19.1
author maintainer
#8 lore raw

PATCH 9/37 media: au0828: cannot kfree dev before usb disconnect

In reply to Sasha Levin

From: Brad Love <[email protected]>

[ Upstream commit 4add7104919f9e94e0db03e234caeadbfcc02ea9 ]

If au0828_analog_register fails, the dev is kfree'd and then flow jumps to done, which can call au0828_usb_disconnect. Since all USB error codes are negative, au0828_usb_disconnect will be called. The problem is au0828_usb_disconnect uses dev, if dev is NULL then there is immediate oops encountered.

[    7.454307] au0828: au0828_usb_probe() au0282_dev_register failed to register on V4L2
[    7.454323] BUG: unable to handle kernel NULL pointer dereference at 0000000000000050
[    7.454421] PGD 0 P4D 0
[    7.454457] Oops: 0002 [#1] SMP PTI
[    7.454500] CPU: 1 PID: 262 Comm: systemd-udevd Tainted: P           O      4.18.3 #1
[    7.454584] Hardware name: Google Panther/Panther, BIOS MattDevo 04/27/2015
[    7.454670] RIP: 0010:_raw_spin_lock_irqsave+0x2c/0x50
[    7.454725] Code: 44 00 00 55 48 89 e5 41 54 53 48 89 fb 9c 58 0f 1f 44 00 00 49 89 c4 fa 66 0f 1f 44 00 00 e8 db 23 1b ff 31 c0 ba 01 00 00 00 <f0> 0f b1 13 85 c0 75 08 4c 89 e0 5b 41 5c 5d c3 89 c6 48 89 df e8
[    7.455004] RSP: 0018:ffff9130f53ef988 EFLAGS: 00010046
[    7.455063] RAX: 0000000000000000 RBX: 0000000000000050 RCX: 0000000000000000
[    7.455139] RDX: 0000000000000001 RSI: 0000000000000003 RDI: 0000000000000050
[    7.455216] RBP: ffff9130f53ef998 R08: 0000000000000018 R09: 0000000000000090
[    7.455292] R10: ffffed4cc53cb000 R11: ffffed4cc53cb108 R12: 0000000000000082
[    7.455369] R13: ffff9130cf2c6188 R14: 0000000000000000 R15: 0000000000000018
[    7.455447] FS:  00007f2ff8514cc0(0000) GS:ffff9130fcb00000(0000) knlGS:0000000000000000
[    7.455535] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[    7.455597] CR2: 0000000000000050 CR3: 00000001753f0002 CR4: 00000000000606a0
[    7.455675] Call Trace:
[    7.455713]  __wake_up_common_lock+0x65/0xc0
[    7.455764]  __wake_up+0x13/0x20
[    7.455808]  ir_lirc_unregister+0x57/0xe0 [rc_core]
[    7.455865]  rc_unregister_device+0xa0/0xc0 [rc_core]
[    7.455935]  au0828_rc_unregister+0x25/0x40 [au0828]
[    7.455999]  au0828_usb_disconnect+0x33/0x80 [au0828]
[    7.456064]  au0828_usb_probe.cold.16+0x8d/0x2aa [au0828]
[    7.456130]  usb_probe_interface+0xf1/0x300
[    7.456184]  driver_probe_device+0x2e3/0x460
[    7.456235]  __driver_attach+0xe4/0x110
[    7.456282]  ? driver_probe_device+0x460/0x460
[    7.456335]  bus_for_each_dev+0x74/0xb0
[    7.456385]  ? kmem_cache_alloc_trace+0x15d/0x1d0
[    7.456441]  driver_attach+0x1e/0x20
[    7.456485]  bus_add_driver+0x159/0x230
[    7.456532]  driver_register+0x70/0xc0
[    7.456578]  usb_register_driver+0x7f/0x140
[    7.456626]  ? 0xffffffffc0474000
[    7.456674]  au0828_init+0xbc/0x1000 [au0828]
[    7.456725]  do_one_initcall+0x4a/0x1c9
[    7.456771]  ? _cond_resched+0x19/0x30
[    7.456817]  ? kmem_cache_alloc_trace+0x15d/0x1d0
[    7.456873]  do_init_module+0x60/0x210
[    7.456918]  load_module+0x221b/0x2710
[    7.456966]  ? vfs_read+0xf5/0x120
[    7.457010]  __do_sys_finit_module+0xbd/0x120
[    7.457061]  ? __do_sys_finit_module+0xbd/0x120
[    7.457115]  __x64_sys_finit_module+0x1a/0x20
[    7.457166]  do_syscall_64+0x5b/0x110
[    7.457210]  entry_SYSCALL_64_after_hwframe+0x49/0xbe
---
 drivers/media/usb/au0828/au0828-core.c | 1 -
 1 file changed, 1 deletion(-)
drivers/media/usb/au0828/au0828-core.c +0 -1
index cd363a2100d4..257ae0d8cfe2 100644
@@ -629,7 +629,6 @@ static int au0828_usb_probe(struct usb_interface *interface,
 		pr_err("%s() au0282_dev_register failed to register on V4L2\n",
 			__func__);
 		mutex_unlock(&dev->lock);
-		kfree(dev);
 		goto done;
 	}
 
Signature · 1 line
2.19.1
author maintainer
#9 lore raw

PATCH 10/37 HID: i2c-hid: override HID descriptors for certain devices

In reply to Sasha Levin

From: Julian Sax <[email protected]>

[ Upstream commit 9ee3e06610fdb8a601cde59c92089fb6c1deb4aa ]

A particular touchpad (SIPODEV SP1064) refuses to supply the HID descriptors. This patch provides the framework for overriding these descriptors based on DMI data. It also includes the descriptors for said touchpad, which were extracted by listening to the traffic of the windows filter driver, as well as the DMI data for the laptops known to use this device.

Relevant Bug: https://bugzilla.redhat.com/show_bug.cgi?id=1526312

---
 drivers/hid/i2c-hid/Makefile                  |   3 +
 .../hid/i2c-hid/{i2c-hid.c => i2c-hid-core.c} |  56 ++-
 drivers/hid/i2c-hid/i2c-hid-dmi-quirks.c      | 376 ++++++++++++++++++
 drivers/hid/i2c-hid/i2c-hid.h                 |  20 +
 4 files changed, 437 insertions(+), 18 deletions(-)
 rename drivers/hid/i2c-hid/{i2c-hid.c => i2c-hid-core.c} (96%)
 create mode 100644 drivers/hid/i2c-hid/i2c-hid-dmi-quirks.c
 create mode 100644 drivers/hid/i2c-hid/i2c-hid.h
drivers/hid/i2c-hid/Makefile +3 -0
index 832d8f9aaba2..099e1ce2f234 100644
@@ -3,3 +3,6 @@
 #
 
 obj-$(CONFIG_I2C_HID)				+= i2c-hid.o
+
+i2c-hid-objs					=  i2c-hid-core.o
+i2c-hid-$(CONFIG_DMI)				+= i2c-hid-dmi-quirks.o
drivers/hid/i2c-hid/i2c-hid-core.c renamed from drivers/hid/i2c-hid/i2c-hid.c +38 -18

88 lines not on this page · raw message

drivers/hid/i2c-hid/i2c-hid-dmi-quirks.c new file +376 -0

376 lines not on this page · raw message

drivers/hid/i2c-hid/i2c-hid.h new file +20 -0
new file mode 100644
index 000000000000..a8c19aef5824
@@ -0,0 +1,20 @@
+/* SPDX-License-Identifier: GPL-2.0+ */
+
+#ifndef I2C_HID_H
+#define I2C_HID_H
+
+
+#ifdef CONFIG_DMI
+struct i2c_hid_desc *i2c_hid_get_dmi_i2c_hid_desc_override(uint8_t *i2c_name);
+char *i2c_hid_get_dmi_hid_report_desc_override(uint8_t *i2c_name,
+					       unsigned int *size);
+#else
+static inline struct i2c_hid_desc
+		   *i2c_hid_get_dmi_i2c_hid_desc_override(uint8_t *i2c_name)
+{ return NULL; }
+static inline char *i2c_hid_get_dmi_hid_report_desc_override(uint8_t *i2c_name,
+							     unsigned int *size)
+{ return NULL; }
+#endif
+
+#endif
Signature · 1 line
2.19.1
author maintainer
#10 lore raw

PATCH 11/37 ARM: samsung: Limit SAMSUNG_PM_CHECK config option to non-Exynos platforms

In reply to Sasha Levin

From: Bartlomiej Zolnierkiewicz <[email protected]>

[ Upstream commit 6862fdf2201ab67cd962dbf0643d37db909f4860 ]

"S3C2410 PM Suspend Memory CRC" feature (controlled by SAMSUNG_PM_CHECK config option) is incompatible with highmem (uses phys_to_virt() instead of proper mapping) which is used by the majority of Exynos boards. The issue manifests itself in OOPS on affected boards, i.e. on Odroid-U3 I got the following one:

Unable to handle kernel paging request at virtual address f0000000
pgd = 1c0f9bb4
[f0000000] *pgd=00000000
Internal error: Oops: 5 [#1] PREEMPT SMP ARM
[<c0458034>] (crc32_le) from [<c0121f8c>] (s3c_pm_makecheck+0x34/0x54)
[<c0121f8c>] (s3c_pm_makecheck) from [<c0121efc>] (s3c_pm_run_res+0x74/0x8c)
[<c0121efc>] (s3c_pm_run_res) from [<c0121ecc>] (s3c_pm_run_res+0x44/0x8c)
[<c0121ecc>] (s3c_pm_run_res) from [<c01210b8>] (exynos_suspend_enter+0x64/0x148)
[<c01210b8>] (exynos_suspend_enter) from [<c018893c>] (suspend_devices_and_enter+0x9ec/0xe74)
[<c018893c>] (suspend_devices_and_enter) from [<c0189534>] (pm_suspend+0x770/0xc04)
[<c0189534>] (pm_suspend) from [<c0186ce8>] (state_store+0x6c/0xcc)
[<c0186ce8>] (state_store) from [<c09db434>] (kobj_attr_store+0x14/0x20)
[<c09db434>] (kobj_attr_store) from [<c02fa63c>] (sysfs_kf_write+0x4c/0x50)
[<c02fa63c>] (sysfs_kf_write) from [<c02f97a4>] (kernfs_fop_write+0xfc/0x1e4)
[<c02f97a4>] (kernfs_fop_write) from [<c027b198>] (__vfs_write+0x2c/0x140)
[<c027b198>] (__vfs_write) from [<c027b418>] (vfs_write+0xa4/0x160)
[<c027b418>] (vfs_write) from [<c027b5d8>] (ksys_write+0x40/0x8c)
[<c027b5d8>] (ksys_write) from [<c0101000>] (ret_fast_syscall+0x0/0x28)

Add PLAT_S3C24XX, ARCH_S3C64XX and ARCH_S5PV210 dependencies to SAMSUNG_PM_CHECK config option to hide it on Exynos platforms.

---
 arch/arm/plat-samsung/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/arm/plat-samsung/Kconfig +1 -1
index e8229b9fee4a..3265b8f86069 100644
@@ -258,7 +258,7 @@ config S3C_PM_DEBUG_LED_SMDK
 
 config SAMSUNG_PM_CHECK
 	bool "S3C2410 PM Suspend Memory CRC"
-	depends on PM
+	depends on PM && (PLAT_S3C24XX || ARCH_S3C64XX || ARCH_S5PV210)
 	select CRC32
 	help
 	  Enable the PM code's memory area checksum over sleep. This option
Signature · 1 line
2.19.1
author maintainer
#11 lore raw

PATCH 12/37 usbip: fix vhci_hcd controller counting

In reply to Sasha Levin

From: Maciej Żenczykowski <[email protected]>

[ Upstream commit e0a2e73e501c77037c8756137e87b12c7c3c9793 ]

Without this usbip fails on a machine with devices that lexicographically come after vhci_hcd.

ie.

  $ ls -l /sys/devices/platform
  ...
  drwxr-xr-x. 4 root root    0 Sep 19 16:21 serial8250
  -rw-r--r--. 1 root root 4096 Sep 19 23:50 uevent
  drwxr-xr-x. 6 root root    0 Sep 20 13:15 vhci_hcd.0
  drwxr-xr-x. 4 root root    0 Sep 19 16:22 w83627hf.656

Because it detects 'w83627hf.656' as another vhci_hcd controller, and then fails to be able to talk to it.

Note: this doesn't actually fix usbip's support for multiple controllers... that's still broken for other reasons ("vhci_hcd.0" is hardcoded in a string macro), but is enough to actually make it work on the above machine.

See also:
https://bugzilla.redhat.com/show_bug.cgi?id=1631148

---
 tools/usb/usbip/libsrc/vhci_driver.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
tools/usb/usbip/libsrc/vhci_driver.c +1 -1
index ed8c9d360c0f..4225d462701d 100644
@@ -150,7 +150,7 @@ static int get_nports(struct udev_device *hc_device)
 
 static int vhci_hcd_filter(const struct dirent *dirent)
 {
-	return strcmp(dirent->d_name, "vhci_hcd") >= 0;
+	return !strncmp(dirent->d_name, "vhci_hcd.", 9);
 }
 
 static int get_ncontrollers(void)
Signature · 1 line
2.19.1
author maintainer
#12 lore raw

PATCH 13/37 ACPI / SBS: Fix GPE storm on recent MacBookPro's

In reply to Sasha Levin

From: Ronald Tschalär <[email protected]>

[ Upstream commit ca1721c5bee77105829cbd7baab8ee0eab85b06d ]

On Apple machines, plugging-in or unplugging the power triggers a GPE for the EC. Since these machines expose an SBS device, this GPE ends up triggering the acpi_sbs_callback(). This in turn tries to get the status of the SBS charger. However, on MBP13,* and MBP14,* machines, performing the smbus-read operation to get the charger's status triggers the EC's GPE again. The result is an endless re-triggering and handling of that GPE, consuming significant CPU resources (> 50% in irq).

In the end this is quite similar to commit 3031cddea633 (ACPI / SBS:
Don't assume the existence of an SBS charger), except that on the above machines a status of all 1's is returned. And like there, we just want ignore the charger here.

---
 drivers/acpi/sbs.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)
drivers/acpi/sbs.c +6 -2
index a2428e9462dd..3c092f07d7e3 100644
@@ -441,9 +441,13 @@ static int acpi_ac_get_present(struct acpi_sbs *sbs)
 
 	/*
 	 * The spec requires that bit 4 always be 1. If it's not set, assume
-	 * that the implementation doesn't support an SBS charger
+	 * that the implementation doesn't support an SBS charger.
+	 *
+	 * And on some MacBooks a status of 0xffff is always returned, no
+	 * matter whether the charger is plugged in or not, which is also
+	 * wrong, so ignore the SBS charger for those too.
 	 */
-	if (!((status >> 4) & 0x1))
+	if (!((status >> 4) & 0x1) || status == 0xffff)
 		return -ENODEV;
 
 	sbs->charger_present = (status >> 15) & 0x1;
Signature · 1 line
2.19.1
author maintainer
#13 lore raw

PATCH 14/37 KVM: nVMX: restore host state in nested_vmx_vmexit for VMFail

In reply to Sasha Levin

From: Sean Christopherson <[email protected]>

[ Upstream commit bd18bffca35397214ae68d85cf7203aca25c3c1d ]

A VMEnter that VMFails (as opposed to VMExits) does not touch host state beyond registers that are explicitly noted in the VMFail path, e.g. EFLAGS. Host state does not need to be loaded because VMFail is only signaled for consistency checks that occur before the CPU starts to load guest state, i.e. there is no need to restore any state as nothing has been modified. But in the case where a VMFail is detected by hardware and not by KVM (due to deferring consistency checks to hardware), KVM has already loaded some amount of guest state. Luckily, "loaded" only means loaded to KVM's software model, i.e. vmcs01 has not been modified. So, unwind our software model to the pre-VMEntry host state.

Not restoring host state in this VMFail path leads to a variety of failures because we end up with stale data in vcpu->arch, e.g. CR0, CR4, EFER, etc... will all be out of sync relative to vmcs01. Any significant delta in the stale data is all but guaranteed to crash L1, e.g. emulation of SMEP, SMAP, UMIP, WP, etc... will be wrong.

An alternative to this "soft" reload would be to load host state from vmcs12 as if we triggered a VMExit (as opposed to VMFail), but that is wildly inconsistent with respect to the VMX architecture, e.g. an L1 VMM with separate VMExit and VMFail paths would explode.

Note that this approach does not mean KVM is 100% accurate with respect to VMX hardware behavior, even at an architectural level (the exact order of consistency checks is microarchitecture specific). But 100% emulation accuracy isn't the goal (with this patch), rather the goal is to be consistent in the information delivered to L1, e.g. a VMExit should not fall-through VMENTER, and a VMFail should not jump to HOST_RIP.

This technically reverts commit "5af4157388ad (KVM: nVMX: Fix mmu context after VMLAUNCH/VMRESUME failure)", but retains the core aspects of that patch, just in an open coded form due to the need to pull state from vmcs01 instead of vmcs12. Restoring host state resolves a variety of issues introduced by commit "4f350c6dbcb9 (kvm: nVMX: Handle deferred early VMLAUNCH/VMRESUME failure properly)", which remedied the incorrect behavior of treating VMFail like VMExit but in doing so neglected to restore arch state that had been modified prior to attempting nested VMEnter.

A sample failure that occurs due to stale vcpu.arch state is a fault of some form while emulating an LGDT (due to emulated UMIP) from L1 after a failed VMEntry to L3, in this case when running the KVM unit test test_tpr_threshold_values in L1. L0 also hits a WARN in this case due to a stale arch.cr4.UMIP.

L1:

  BUG: unable to handle kernel paging request at ffffc90000663b9e
  PGD 276512067 P4D 276512067 PUD 276513067 PMD 274efa067 PTE 8000000271de2163
  Oops: 0009 [#1] SMP
  CPU: 5 PID: 12495 Comm: qemu-system-x86 Tainted: G        W         4.18.0-rc2+ #2
  Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 0.0.0 02/06/2015
  RIP: 0010:native_load_gdt+0x0/0x10

...

  Call Trace:
   load_fixmap_gdt+0x22/0x30
   __vmx_load_host_state+0x10e/0x1c0 [kvm_intel]
   vmx_switch_vmcs+0x2d/0x50 [kvm_intel]
   nested_vmx_vmexit+0x222/0x9c0 [kvm_intel]
   vmx_handle_exit+0x246/0x15a0 [kvm_intel]
   kvm_arch_vcpu_ioctl_run+0x850/0x1830 [kvm]
   kvm_vcpu_ioctl+0x3a1/0x5c0 [kvm]
   do_vfs_ioctl+0x9f/0x600
   ksys_ioctl+0x66/0x70
   __x64_sys_ioctl+0x16/0x20
   do_syscall_64+0x4f/0x100
   entry_SYSCALL_64_after_hwframe+0x44/0xa9

L0:

  WARNING: CPU: 2 PID: 3529 at arch/x86/kvm/vmx.c:6618 handle_desc+0x28/0x30 [kvm_intel]
  ...
  CPU: 2 PID: 3529 Comm: qemu-system-x86 Not tainted 4.17.2-coffee+ #76
  Hardware name: Intel Corporation Kabylake Client platform/KBL S
  RIP: 0010:handle_desc+0x28/0x30 [kvm_intel]

...

  Call Trace:
   kvm_arch_vcpu_ioctl_run+0x863/0x1840 [kvm]
   kvm_vcpu_ioctl+0x3a1/0x5c0 [kvm]
   do_vfs_ioctl+0x9f/0x5e0
   ksys_ioctl+0x66/0x70
   __x64_sys_ioctl+0x16/0x20
   do_syscall_64+0x49/0xf0
   entry_SYSCALL_64_after_hwframe+0x44/0xa9
Fixes 5af4157388ad (KVM: nVMX: Fix mmu context after VMLAUNCH/VMRESUME failure)
Fixes 4f350c6dbcb9 (kvm: nVMX: Handle deferred early VMLAUNCH/VMRESUME failure properly)
Cc Jim Mattson <[email protected]>
Cc Krish Sadhukhan <[email protected]>
Cc Paolo Bonzini <[email protected]>
Cc Radim KrÄmář <[email protected]>
Cc Wanpeng Li <[email protected]>
---
 arch/x86/kvm/vmx.c | 173 +++++++++++++++++++++++++++++++++++++++------
 1 file changed, 153 insertions(+), 20 deletions(-)
arch/x86/kvm/vmx.c +153 -20
index 229d5e39f5c0..28faf781e6fd 100644
@@ -11860,24 +11860,6 @@ static void prepare_vmcs12(struct kvm_vcpu *vcpu, struct vmcs12 *vmcs12,
 	kvm_clear_interrupt_queue(vcpu);
 }
 
-static void load_vmcs12_mmu_host_state(struct kvm_vcpu *vcpu,
-			struct vmcs12 *vmcs12)
-{
-	u32 entry_failure_code;
-
-	nested_ept_uninit_mmu_context(vcpu);
-
-	/*
-	 * Only PDPTE load can fail as the value of cr3 was checked on entry and
-	 * couldn't have changed.
-	 */
-	if (nested_vmx_load_cr3(vcpu, vmcs12->host_cr3, false, &entry_failure_code))
-		nested_vmx_abort(vcpu, VMX_ABORT_LOAD_HOST_PDPTE_FAIL);
-
-	if (!enable_ept)
-		vcpu->arch.walk_mmu->inject_page_fault = kvm_inject_page_fault;
-}
-
 /*
  * A part of what we need to when the nested L2 guest exits and we want to
  * run its L1 parent, is to reset L1's guest state to the host state specified
@@ -11891,6 +11873,7 @@ static void load_vmcs12_host_state(struct kvm_vcpu *vcpu,
 				   struct vmcs12 *vmcs12)
 {
 	struct kvm_segment seg;
+	u32 entry_failure_code;
 
 	if (vmcs12->vm_exit_controls & VM_EXIT_LOAD_IA32_EFER)
 		vcpu->arch.efer = vmcs12->host_ia32_efer;
@@ -11917,7 +11900,17 @@ static void load_vmcs12_host_state(struct kvm_vcpu *vcpu,
 	vcpu->arch.cr4_guest_owned_bits = ~vmcs_readl(CR4_GUEST_HOST_MASK);
 	vmx_set_cr4(vcpu, vmcs12->host_cr4);
 
-	load_vmcs12_mmu_host_state(vcpu, vmcs12);
+	nested_ept_uninit_mmu_context(vcpu);
+
+	/*
+	 * Only PDPTE load can fail as the value of cr3 was checked on entry and
+	 * couldn't have changed.
+	 */
+	if (nested_vmx_load_cr3(vcpu, vmcs12->host_cr3, false, &entry_failure_code))
+		nested_vmx_abort(vcpu, VMX_ABORT_LOAD_HOST_PDPTE_FAIL);
+
+	if (!enable_ept)
+		vcpu->arch.walk_mmu->inject_page_fault = kvm_inject_page_fault;
 
 	if (enable_vpid) {
 		/*
@@ -12008,6 +12001,140 @@ static void load_vmcs12_host_state(struct kvm_vcpu *vcpu,
 		nested_vmx_abort(vcpu, VMX_ABORT_LOAD_HOST_MSR_FAIL);
 }
 
+static inline u64 nested_vmx_get_vmcs01_guest_efer(struct vcpu_vmx *vmx)
+{
+	struct shared_msr_entry *efer_msr;
+	unsigned int i;
+
+	if (vm_entry_controls_get(vmx) & VM_ENTRY_LOAD_IA32_EFER)
+		return vmcs_read64(GUEST_IA32_EFER);
+
+	if (cpu_has_load_ia32_efer)
+		return host_efer;
+
+	for (i = 0; i < vmx->msr_autoload.guest.nr; ++i) {
+		if (vmx->msr_autoload.guest.val[i].index == MSR_EFER)
+			return vmx->msr_autoload.guest.val[i].value;
+	}
+
+	efer_msr = find_msr_entry(vmx, MSR_EFER);
+	if (efer_msr)
+		return efer_msr->data;
+
+	return host_efer;
+}
+
+static void nested_vmx_restore_host_state(struct kvm_vcpu *vcpu)
+{
+	struct vmcs12 *vmcs12 = get_vmcs12(vcpu);
+	struct vcpu_vmx *vmx = to_vmx(vcpu);
+	struct vmx_msr_entry g, h;
+	struct msr_data msr;
+	gpa_t gpa;
+	u32 i, j;
+
+	vcpu->arch.pat = vmcs_read64(GUEST_IA32_PAT);
+
+	if (vmcs12->vm_entry_controls & VM_ENTRY_LOAD_DEBUG_CONTROLS) {
+		/*
+		 * L1's host DR7 is lost if KVM_GUESTDBG_USE_HW_BP is set
+		 * as vmcs01.GUEST_DR7 contains a userspace defined value
+		 * and vcpu->arch.dr7 is not squirreled away before the
+		 * nested VMENTER (not worth adding a variable in nested_vmx).
+		 */
+		if (vcpu->guest_debug & KVM_GUESTDBG_USE_HW_BP)
+			kvm_set_dr(vcpu, 7, DR7_FIXED_1);
+		else
+			WARN_ON(kvm_set_dr(vcpu, 7, vmcs_readl(GUEST_DR7)));
+	}
+
+	/*
+	 * Note that calling vmx_set_{efer,cr0,cr4} is important as they
+	 * handle a variety of side effects to KVM's software model.
+	 */
+	vmx_set_efer(vcpu, nested_vmx_get_vmcs01_guest_efer(vmx));
+
+	vcpu->arch.cr0_guest_owned_bits = X86_CR0_TS;
+	vmx_set_cr0(vcpu, vmcs_readl(CR0_READ_SHADOW));
+
+	vcpu->arch.cr4_guest_owned_bits = ~vmcs_readl(CR4_GUEST_HOST_MASK);
+	vmx_set_cr4(vcpu, vmcs_readl(CR4_READ_SHADOW));
+
+	nested_ept_uninit_mmu_context(vcpu);
+	vcpu->arch.cr3 = vmcs_readl(GUEST_CR3);
+	__set_bit(VCPU_EXREG_CR3, (ulong *)&vcpu->arch.regs_avail);
+
+	/*
+	 * Use ept_save_pdptrs(vcpu) to load the MMU's cached PDPTRs
+	 * from vmcs01 (if necessary).  The PDPTRs are not loaded on
+	 * VMFail, like everything else we just need to ensure our
+	 * software model is up-to-date.
+	 */
+	ept_save_pdptrs(vcpu);
+
+	kvm_mmu_reset_context(vcpu);
+
+	if (cpu_has_vmx_msr_bitmap())
+		vmx_update_msr_bitmap(vcpu);
+
+	/*
+	 * This nasty bit of open coding is a compromise between blindly
+	 * loading L1's MSRs using the exit load lists (incorrect emulation
+	 * of VMFail), leaving the nested VM's MSRs in the software model
+	 * (incorrect behavior) and snapshotting the modified MSRs (too
+	 * expensive since the lists are unbound by hardware).  For each
+	 * MSR that was (prematurely) loaded from the nested VMEntry load
+	 * list, reload it from the exit load list if it exists and differs
+	 * from the guest value.  The intent is to stuff host state as
+	 * silently as possible, not to fully process the exit load list.
+	 */
+	msr.host_initiated = false;
+	for (i = 0; i < vmcs12->vm_entry_msr_load_count; i++) {
+		gpa = vmcs12->vm_entry_msr_load_addr + (i * sizeof(g));
+		if (kvm_vcpu_read_guest(vcpu, gpa, &g, sizeof(g))) {
+			pr_debug_ratelimited(
+				"%s read MSR index failed (%u, 0x%08llx)\n",
+				__func__, i, gpa);
+			goto vmabort;
+		}
+
+		for (j = 0; j < vmcs12->vm_exit_msr_load_count; j++) {
+			gpa = vmcs12->vm_exit_msr_load_addr + (j * sizeof(h));
+			if (kvm_vcpu_read_guest(vcpu, gpa, &h, sizeof(h))) {
+				pr_debug_ratelimited(
+					"%s read MSR failed (%u, 0x%08llx)\n",
+					__func__, j, gpa);
+				goto vmabort;
+			}
+			if (h.index != g.index)
+				continue;
+			if (h.value == g.value)
+				break;
+
+			if (nested_vmx_load_msr_check(vcpu, &h)) {
+				pr_debug_ratelimited(
+					"%s check failed (%u, 0x%x, 0x%x)\n",
+					__func__, j, h.index, h.reserved);
+				goto vmabort;
+			}
+
+			msr.index = h.index;
+			msr.data = h.value;
+			if (kvm_set_msr(vcpu, &msr)) {
+				pr_debug_ratelimited(
+					"%s WRMSR failed (%u, 0x%x, 0x%llx)\n",
+					__func__, j, h.index, h.value);
+				goto vmabort;
+			}
+		}
+	}
+
+	return;
+
+vmabort:
+	nested_vmx_abort(vcpu, VMX_ABORT_LOAD_HOST_MSR_FAIL);
+}
+
 /*
  * Emulate an exit from nested guest (L2) to L1, i.e., prepare to run L1
  * and modify vmcs12 to make it see what it would expect to see there if
@@ -12140,7 +12267,13 @@ static void nested_vmx_vmexit(struct kvm_vcpu *vcpu, u32 exit_reason,
 	 */
 	nested_vmx_failValid(vcpu, VMXERR_ENTRY_INVALID_CONTROL_FIELD);
 
-	load_vmcs12_mmu_host_state(vcpu, vmcs12);
+	/*
+	 * Restore L1's host state to KVM's software model.  We're here
+	 * because a consistency check was caught by hardware, which
+	 * means some amount of guest state has been propagated to KVM's
+	 * model and needs to be unwound to the host's state.
+	 */
+	nested_vmx_restore_host_state(vcpu);
 
 	/*
 	 * The emulated instruction was already skipped in
Signature · 1 line
2.19.1
author maintainer
#14 lore raw

PATCH 15/37 compiler.h: update definition of unreachable()

In reply to Sasha Levin

From: "[email protected]" <[email protected]>

[ Upstream commit fe0640eb30b7da261ae84d252ed9ed3c7e68dfd8 ]

Fixes the objtool warning seen with Clang:
arch/x86/mm/fault.o: warning: objtool: no_context()+0x220: unreachable instruction

Fixes commit 815f0ddb346c ("include/linux/compiler*.h: make compiler-*.h mutually exclusive")

Josh noted that the fallback definition was meant to work around a pre-gcc-4.6 bug. GCC still needs to work around
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82365, so compiler-gcc.h defines its own version of unreachable(). Clang and ICC can use this shared definition.

---
 include/linux/compiler.h | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
include/linux/compiler.h +4 -1
index a704d032713b..67c3934fb9ed 100644
@@ -119,7 +119,10 @@ void ftrace_likely_update(struct ftrace_likely_data *f, int val,
 # define ASM_UNREACHABLE
 #endif
 #ifndef unreachable
-# define unreachable() do { annotate_reachable(); do { } while (1); } while (0)
+# define unreachable() do {		\
+	annotate_unreachable();		\
+	__builtin_unreachable();	\
+} while (0)
 #endif
 
 /*
Signature · 1 line
2.19.1
author maintainer
#15 lore raw

PATCH 16/37 cifs: fallback to older infolevels on findfirst queryinfo retry

In reply to Sasha Levin

From: Steve French <[email protected]>

[ Upstream commit 3b7960caceafdfc2cdfe2850487f8d091eb41144 ]

In cases where queryinfo fails, we have cases in cifs (vers=1.0) where with backupuid mounts we retry the query info with findfirst. This doesn't work to some NetApp servers which don't support WindowsXP (and later) infolevel 261 (SMB_FIND_FILE_ID_FULL_DIR_INFO) so in this case use other info levels (in this case it will usually be level 257, SMB_FIND_FILE_DIRECTORY_INFO).

(Also fixes some indentation)

See kernel bugzilla 201435

---
 fs/cifs/inode.c | 67 +++++++++++++++++++++++++++----------------------
 1 file changed, 37 insertions(+), 30 deletions(-)
fs/cifs/inode.c +37 -30
index a90a637ae79a..6fd4a6a75234 100644
@@ -779,43 +779,50 @@ cifs_get_inode_info(struct inode **inode, const char *full_path,
 	} else if ((rc == -EACCES) && backup_cred(cifs_sb) &&
 		   (strcmp(server->vals->version_string, SMB1_VERSION_STRING)
 		      == 0)) {
-			/*
-			 * For SMB2 and later the backup intent flag is already
-			 * sent if needed on open and there is no path based
-			 * FindFirst operation to use to retry with
-			 */
+		/*
+		 * For SMB2 and later the backup intent flag is already
+		 * sent if needed on open and there is no path based
+		 * FindFirst operation to use to retry with
+		 */
 
-			srchinf = kzalloc(sizeof(struct cifs_search_info),
-						GFP_KERNEL);
-			if (srchinf == NULL) {
-				rc = -ENOMEM;
-				goto cgii_exit;
-			}
+		srchinf = kzalloc(sizeof(struct cifs_search_info),
+					GFP_KERNEL);
+		if (srchinf == NULL) {
+			rc = -ENOMEM;
+			goto cgii_exit;
+		}
 
-			srchinf->endOfSearch = false;
+		srchinf->endOfSearch = false;
+		if (tcon->unix_ext)
+			srchinf->info_level = SMB_FIND_FILE_UNIX;
+		else if ((tcon->ses->capabilities &
+			 tcon->ses->server->vals->cap_nt_find) == 0)
+			srchinf->info_level = SMB_FIND_FILE_INFO_STANDARD;
+		else if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SERVER_INUM)
 			srchinf->info_level = SMB_FIND_FILE_ID_FULL_DIR_INFO;
+		else /* no srvino useful for fallback to some netapp */
+			srchinf->info_level = SMB_FIND_FILE_DIRECTORY_INFO;
 
-			srchflgs = CIFS_SEARCH_CLOSE_ALWAYS |
-					CIFS_SEARCH_CLOSE_AT_END |
-					CIFS_SEARCH_BACKUP_SEARCH;
+		srchflgs = CIFS_SEARCH_CLOSE_ALWAYS |
+				CIFS_SEARCH_CLOSE_AT_END |
+				CIFS_SEARCH_BACKUP_SEARCH;
 
-			rc = CIFSFindFirst(xid, tcon, full_path,
-				cifs_sb, NULL, srchflgs, srchinf, false);
-			if (!rc) {
-				data =
-				(FILE_ALL_INFO *)srchinf->srch_entries_start;
+		rc = CIFSFindFirst(xid, tcon, full_path,
+			cifs_sb, NULL, srchflgs, srchinf, false);
+		if (!rc) {
+			data = (FILE_ALL_INFO *)srchinf->srch_entries_start;
 
-				cifs_dir_info_to_fattr(&fattr,
-				(FILE_DIRECTORY_INFO *)data, cifs_sb);
-				fattr.cf_uniqueid = le64_to_cpu(
-				((SEARCH_ID_FULL_DIR_INFO *)data)->UniqueId);
-				validinum = true;
+			cifs_dir_info_to_fattr(&fattr,
+			(FILE_DIRECTORY_INFO *)data, cifs_sb);
+			fattr.cf_uniqueid = le64_to_cpu(
+			((SEARCH_ID_FULL_DIR_INFO *)data)->UniqueId);
+			validinum = true;
 
-				cifs_buf_release(srchinf->ntwrk_buf_start);
-			}
-			kfree(srchinf);
-			if (rc)
-				goto cgii_exit;
+			cifs_buf_release(srchinf->ntwrk_buf_start);
+		}
+		kfree(srchinf);
+		if (rc)
+			goto cgii_exit;
 	} else
 		goto cgii_exit;
 
Signature · 1 line
2.19.1
author maintainer
#16 lore raw

PATCH 17/37 kernel: hung_task.c: disable on suspend

In reply to Sasha Levin

From: Vitaly Kuznetsov <[email protected]>

[ Upstream commit a1c6ca3c6de763459a6e93b644ec6518c890ba1c ]

It is possible to observe hung_task complaints when system goes to suspend-to-idle state:

 # echo freeze > /sys/power/state

PM: Syncing filesystems ... done.
Freezing user space processes ... (elapsed 0.001 seconds) done. OOM killer disabled.
Freezing remaining freezable tasks ... (elapsed 0.002 seconds) done. sd 0:0:0:0: [sda] Synchronizing SCSI cache

 INFO: task bash:1569 blocked for more than 120 seconds.
       Not tainted 4.19.0-rc3_+ #687
 "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
 bash            D    0  1569    604 0x00000000
 Call Trace:
  ? __schedule+0x1fe/0x7e0
  schedule+0x28/0x80
  suspend_devices_and_enter+0x4ac/0x750
  pm_suspend+0x2c0/0x310

Register a PM notifier to disable the detector on suspend and re-enable back on wakeup.

---
 kernel/hung_task.c | 30 +++++++++++++++++++++++++++++-
 1 file changed, 29 insertions(+), 1 deletion(-)
kernel/hung_task.c +29 -1
index f9aaf4994062..2e4869fa66c9 100644
@@ -15,6 +15,7 @@
 #include <linux/lockdep.h>
 #include <linux/export.h>
 #include <linux/sysctl.h>
+#include <linux/suspend.h>
 #include <linux/utsname.h>
 #include <linux/sched/signal.h>
 #include <linux/sched/debug.h>
@@ -232,6 +233,28 @@ void reset_hung_task_detector(void)
 }
 EXPORT_SYMBOL_GPL(reset_hung_task_detector);
 
+static bool hung_detector_suspended;
+
+static int hungtask_pm_notify(struct notifier_block *self,
+			      unsigned long action, void *hcpu)
+{
+	switch (action) {
+	case PM_SUSPEND_PREPARE:
+	case PM_HIBERNATION_PREPARE:
+	case PM_RESTORE_PREPARE:
+		hung_detector_suspended = true;
+		break;
+	case PM_POST_SUSPEND:
+	case PM_POST_HIBERNATION:
+	case PM_POST_RESTORE:
+		hung_detector_suspended = false;
+		break;
+	default:
+		break;
+	}
+	return NOTIFY_OK;
+}
+
 /*
  * kthread which checks for tasks stuck in D state
  */
@@ -246,7 +269,8 @@ static int watchdog(void *dummy)
 		long t = hung_timeout_jiffies(hung_last_checked, timeout);
 
 		if (t <= 0) {
-			if (!atomic_xchg(&reset_hung_task, 0))
+			if (!atomic_xchg(&reset_hung_task, 0) &&
+			    !hung_detector_suspended)
 				check_hung_uninterruptible_tasks(timeout);
 			hung_last_checked = jiffies;
 			continue;
@@ -260,6 +284,10 @@ static int watchdog(void *dummy)
 static int __init hung_task_init(void)
 {
 	atomic_notifier_chain_register(&panic_notifier_list, &panic_block);
+
+	/* Disable hung task detector on suspend */
+	pm_notifier(hungtask_pm_notify, 0);
+
 	watchdog_task = kthread_run(watchdog, NULL, "khungtaskd");
 
 	return 0;
Signature · 1 line
2.19.1
author maintainer
#17 lore raw

PATCH 18/37 nvme-pci: fix conflicting p2p resource adds

In reply to Sasha Levin

From: Keith Busch <[email protected]>

[ Upstream commit 9fe5c59ff6a1e5e26a39b75489a1420e7eaaf0b1 ]

The nvme pci driver had been adding its CMB resource to the P2P DMA subsystem everytime on on a controller reset. This results in the following warning:

    ------------[ cut here ]------------
    nvme 0000:00:03.0: Conflicting mapping in same section
    WARNING: CPU: 7 PID: 81 at kernel/memremap.c:155 devm_memremap_pages+0xa6/0x380
    ...
    Call Trace:
     pci_p2pdma_add_resource+0x153/0x370
     nvme_reset_work+0x28c/0x17b1 [nvme]
     ? add_timer+0x107/0x1e0
     ? dequeue_entity+0x81/0x660
     ? dequeue_entity+0x3b0/0x660
     ? pick_next_task_fair+0xaf/0x610
     ? __switch_to+0xbc/0x410
     process_one_work+0x1cf/0x350
     worker_thread+0x215/0x3d0
     ? process_one_work+0x350/0x350
     kthread+0x107/0x120
     ? kthread_park+0x80/0x80
     ret_from_fork+0x1f/0x30
    ---[ end trace f7ea76ac6ee72727 ]---
    nvme nvme0: failed to register the CMB

This patch fixes this by registering the CMB with P2P only once.

Signed-off-by Keith Busch <keith.busch at intel.com>
Reviewed-by Logan Gunthorpe <logang at deltatee.com>
Signed-off-by Christoph Hellwig <hch at lst.de>
Signed-off-by Jens Axboe <axboe at kernel.dk>
Signed-off-by Sasha Levin <sashal at kernel.org>
---
 drivers/nvme/host/pci.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
drivers/nvme/host/pci.c +4 -1
index 06355ca832db..ae03a5b70324 100644
@@ -1546,6 +1546,9 @@ static void __iomem *nvme_map_cmb(struct nvme_dev *dev)
 	void __iomem *cmb;
 	int bar;
 
+	if (dev->cmb_size)
+		return;
+
 	dev->cmbsz = readl(dev->bar + NVME_REG_CMBSZ);
 	if (!(NVME_CMB_SZ(dev->cmbsz)))
 		return NULL;
@@ -2034,7 +2037,6 @@ static void nvme_pci_disable(struct nvme_dev *dev)
 {
 	struct pci_dev *pdev = to_pci_dev(dev->dev);
 
-	nvme_release_cmb(dev);
 	pci_free_irq_vectors(pdev);
 
 	if (pci_is_enabled(pdev)) {
@@ -2434,6 +2436,7 @@ static void nvme_remove(struct pci_dev *pdev)
 	nvme_stop_ctrl(&dev->ctrl);
 	nvme_remove_namespaces(&dev->ctrl);
 	nvme_dev_disable(dev, true);
+	nvme_release_cmb(dev);
 	nvme_free_host_mem(dev);
 	nvme_dev_remove_admin(dev);
 	nvme_free_queues(dev, 0);
Signature · 1 line
2.19.1
#18 lore raw

Re: [PATCH AUTOSEL 4.14 18/37] nvme-pci: fix conflicting p2p resource adds

Quoted 64 lines · Sasha Levin

From: Keith Busch <keith.busch at intel.com>

[ Upstream commit 9fe5c59ff6a1e5e26a39b75489a1420e7eaaf0b1 ]

The nvme pci driver had been adding its CMB resource to the P2P DMA subsystem everytime on on a controller reset. This results in the following warning:

     ------------[ cut here ]------------
     nvme 0000:00:03.0: Conflicting mapping in same section
     WARNING: CPU: 7 PID: 81 at kernel/memremap.c:155 devm_memremap_pages+0xa6/0x380
     ...
     Call Trace:
      pci_p2pdma_add_resource+0x153/0x370
      nvme_reset_work+0x28c/0x17b1 [nvme]
      ? add_timer+0x107/0x1e0
      ? dequeue_entity+0x81/0x660
      ? dequeue_entity+0x3b0/0x660
      ? pick_next_task_fair+0xaf/0x610
      ? __switch_to+0xbc/0x410
      process_one_work+0x1cf/0x350
      worker_thread+0x215/0x3d0
      ? process_one_work+0x350/0x350
      kthread+0x107/0x120
      ? kthread_park+0x80/0x80
      ret_from_fork+0x1f/0x30
     ---[ end trace f7ea76ac6ee72727 ]---
     nvme nvme0: failed to register the CMB

This patch fixes this by registering the CMB with P2P only once.

Signed-off-by Keith Busch <keith.busch at intel.com>
Reviewed-by Logan Gunthorpe <logang at deltatee.com>
Signed-off-by Christoph Hellwig <hch at lst.de>
Signed-off-by Jens Axboe <axboe at kernel.dk>
Signed-off-by Sasha Levin <sashal at kernel.org>
---
  drivers/nvme/host/pci.c | 5 ++++-
  1 file changed, 4 insertions(+), 1 deletion(-)
drivers/nvme/host/pci.c +3 -1
index 06355ca832db..ae03a5b70324 100644
@@ -1546,6 +1546,9 @@ static void __iomem *nvme_map_cmb(struct nvme_dev *dev)
  	void __iomem *cmb;
  	int bar;
 
+	if (dev->cmb_size)
+		return;
+
  	dev->cmbsz = readl(dev->bar + NVME_REG_CMBSZ);
  	if (!(NVME_CMB_SZ(dev->cmbsz)))
  		return NULL;
@@ -2034,7 +2037,6 @@ static void nvme_pci_disable(struct nvme_dev *dev)
  {
  	struct pci_dev *pdev = to_pci_dev(dev->dev);
 
-	nvme_release_cmb(dev);
  	pci_free_irq_vectors(pdev);
 
  	if (pci_is_enabled(pdev)) {

drivers/nvme/host/pci.c static void nvme_remove(struct pci_dev *pdev) · line 2442

Quoted 7 lines
drivers/nvme/host/pci.c +1 -0
@@ -2434,6 +2436,7 @@ static void nvme_remove(struct pci_dev *pdev)
  	nvme_stop_ctrl(&dev->ctrl);
  	nvme_remove_namespaces(&dev->ctrl);
  	nvme_dev_disable(dev, true);
+	nvme_release_cmb(dev);
  	nvme_free_host_mem(dev);
  	nvme_dev_remove_admin(dev);
  	nvme_free_queues(dev, 0);

There is no P2P support in the 4.14 kernel so I'm not sure this change is necessary. I would also worry that it might break things if the BARs need to be remapped after a controller reset.

author maintainer
#19 lore raw

Re: [PATCH AUTOSEL 4.14 18/37] nvme-pci: fix conflicting p2p resource adds

Quoted 66 lines · Kenneth
Quoted 64 lines · Sasha Levin

From: Keith Busch <keith.busch at intel.com>

[ Upstream commit 9fe5c59ff6a1e5e26a39b75489a1420e7eaaf0b1 ]

The nvme pci driver had been adding its CMB resource to the P2P DMA subsystem everytime on on a controller reset. This results in the following warning:

    ------------[ cut here ]------------
    nvme 0000:00:03.0: Conflicting mapping in same section
    WARNING: CPU: 7 PID: 81 at kernel/memremap.c:155 devm_memremap_pages+0xa6/0x380
    ...
    Call Trace:
     pci_p2pdma_add_resource+0x153/0x370
     nvme_reset_work+0x28c/0x17b1 [nvme]
     ? add_timer+0x107/0x1e0
     ? dequeue_entity+0x81/0x660
     ? dequeue_entity+0x3b0/0x660
     ? pick_next_task_fair+0xaf/0x610
     ? __switch_to+0xbc/0x410
     process_one_work+0x1cf/0x350
     worker_thread+0x215/0x3d0
     ? process_one_work+0x350/0x350
     kthread+0x107/0x120
     ? kthread_park+0x80/0x80
     ret_from_fork+0x1f/0x30
    ---[ end trace f7ea76ac6ee72727 ]---
    nvme nvme0: failed to register the CMB

This patch fixes this by registering the CMB with P2P only once.

Signed-off-by Keith Busch <keith.busch at intel.com>
Reviewed-by Logan Gunthorpe <logang at deltatee.com>
Signed-off-by Christoph Hellwig <hch at lst.de>
Signed-off-by Jens Axboe <axboe at kernel.dk>
Signed-off-by Sasha Levin <sashal at kernel.org>
---
 drivers/nvme/host/pci.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
drivers/nvme/host/pci.c +4 -1
index 06355ca832db..ae03a5b70324 100644
@@ -1546,6 +1546,9 @@ static void __iomem *nvme_map_cmb(struct nvme_dev *dev)
 	void __iomem *cmb;
 	int bar;
+	if (dev->cmb_size)
+		return;
+
 	dev->cmbsz = readl(dev->bar + NVME_REG_CMBSZ);
 	if (!(NVME_CMB_SZ(dev->cmbsz)))
 		return NULL;
@@ -2034,7 +2037,6 @@ static void nvme_pci_disable(struct nvme_dev *dev)
 {
 	struct pci_dev *pdev = to_pci_dev(dev->dev);
-	nvme_release_cmb(dev);
 	pci_free_irq_vectors(pdev);
 	if (pci_is_enabled(pdev)) {
@@ -2434,6 +2436,7 @@ static void nvme_remove(struct pci_dev *pdev)
 	nvme_stop_ctrl(&dev->ctrl);
 	nvme_remove_namespaces(&dev->ctrl);
 	nvme_dev_disable(dev, true);
+	nvme_release_cmb(dev);
Quoted 8 lines

drivers/nvme/host/pci.c static void nvme_remove(struct pci_dev *pdev) · line 2442

Quoted 4 lines
drivers/nvme/host/pci.c +0 -0
 	nvme_free_host_mem(dev);
 	nvme_dev_remove_admin(dev);
 	nvme_free_queues(dev, 0);

There is no P2P support in the 4.14 kernel so I'm not sure this change is necessary. I would also worry that it might break things if the BARs need to be remapped after a controller reset.

I've removed it, thanks!

--
Thanks,
Sasha

author maintainer
#20 lore raw

PATCH 19/37 blk-mq: protect debugfs_create_files() from failures

In reply to Sasha Levin

From: Greg Kroah-Hartman <[email protected]>

[ Upstream commit 36991ca68db9dd43bac7f3519f080ee3939263ef ]

If debugfs were to return a non-NULL error for a debugfs call, using that pointer later in debugfs_create_files() would crash.

Fix that by properly checking the pointer before referencing it.

---
 block/blk-mq-debugfs.c | 3 +++
 1 file changed, 3 insertions(+)
block/blk-mq-debugfs.c +3 -0
index d95439154556..da0453277f7f 100644
@@ -791,6 +791,9 @@ static const struct blk_mq_debugfs_attr blk_mq_debugfs_ctx_attrs[] = {
 static bool debugfs_create_files(struct dentry *parent, void *data,
 				 const struct blk_mq_debugfs_attr *attr)
 {
+	if (IS_ERR_OR_NULL(parent))
+		return false;
+
 	d_inode(parent)->i_private = data;
 
 	for (; attr->name; attr++) {
Signature · 1 line
2.19.1
author maintainer
#21 lore raw

PATCH 20/37 crypto: sha256/arm - fix crash bug in Thumb2 build

In reply to Sasha Levin

From: Ard Biesheuvel <[email protected]>

[ Upstream commit 69216a545cf81b2b32d01948f7039315abaf75a0 ]

The SHA256 code we adopted from the OpenSSL project uses a rather peculiar way to take the address of the round constant table: it takes the address of the sha256_block_data_order() routine, and substracts a constant known quantity to arrive at the base of the table, which is emitted by the same assembler code right before the routine's entry point.

However, recent versions of binutils have helpfully changed the behavior of references emitted via an ADR instruction when running in Thumb2 mode: it now takes the Thumb execution mode bit into account, which is bit 0 af the address. This means the produced table address also has bit 0 set, and so we end up with an address value pointing 1 byte past the start of the table, which results in crashes such as

  Unable to handle kernel paging request at virtual address bf825000
  pgd = 42f44b11
  [bf825000] *pgd=80000040206003, *pmd=5f1bd003, *pte=00000000
  Internal error: Oops: 207 [#1] PREEMPT SMP THUMB2
  Modules linked in: sha256_arm(+) sha1_arm_ce sha1_arm ...
  CPU: 7 PID: 396 Comm: cryptomgr_test Not tainted 5.0.0-rc6+ #144
  Hardware name: QEMU KVM Virtual Machine, BIOS 0.0.0 02/06/2015
  PC is at sha256_block_data_order+0xaaa/0xb30 [sha256_arm]
  LR is at __this_module+0x17fd/0xffffe800 [sha256_arm]
  pc : [<bf820bca>]    lr : [<bf824ffd>]    psr: 800b0033
  sp : ebc8bbe8  ip : faaabe1c  fp : 2fdd3433
  r10: 4c5f1692  r9 : e43037df  r8 : b04b0a5a
  r7 : c369d722  r6 : 39c3693e  r5 : 7a013189  r4 : 1580d26b
  r3 : 8762a9b0  r2 : eea9c2cd  r1 : 3e9ab536  r0 : 1dea4ae7
  Flags: Nzcv  IRQs on  FIQs on  Mode SVC_32  ISA Thumb  Segment user
  Control: 70c5383d  Table: 6b8467c0  DAC: dbadc0de
  Process cryptomgr_test (pid: 396, stack limit = 0x69e1fe23)
  Stack: (0xebc8bbe8 to 0xebc8c000)
  ...
  unwind: Unknown symbol address bf820bca
  unwind: Index not found bf820bca
  Code: 441a ea80 40f9 440a (f85e) 3b04
  ---[ end trace e560cce92700ef8a ]---

Given that this affects older kernels as well, in case they are built with a recent toolchain, apply a minimal backportable fix, which is to emit another non-code label at the start of the routine, and reference that instead. (This is similar to the current upstream state of this file in OpenSSL)

---
 arch/arm/crypto/sha256-armv4.pl       | 3 ++-
 arch/arm/crypto/sha256-core.S_shipped | 3 ++-
 2 files changed, 4 insertions(+), 2 deletions(-)
arch/arm/crypto/sha256-armv4.pl +2 -1
index fac0533ea633..f64e8413ab9a 100644
@@ -205,10 +205,11 @@ K256:
 .global	sha256_block_data_order
 .type	sha256_block_data_order,%function
 sha256_block_data_order:
+.Lsha256_block_data_order:
 #if __ARM_ARCH__<7
 	sub	r3,pc,#8		@ sha256_block_data_order
 #else
-	adr	r3,sha256_block_data_order
+	adr	r3,.Lsha256_block_data_order
 #endif
 #if __ARM_MAX_ARCH__>=7 && !defined(__KERNEL__)
 	ldr	r12,.LOPENSSL_armcap
arch/arm/crypto/sha256-core.S_shipped +2 -1
index 555a1a8eec90..72c248081d27 100644
@@ -86,10 +86,11 @@ K256:
 .global	sha256_block_data_order
 .type	sha256_block_data_order,%function
 sha256_block_data_order:
+.Lsha256_block_data_order:
 #if __ARM_ARCH__<7
 	sub	r3,pc,#8		@ sha256_block_data_order
 #else
-	adr	r3,sha256_block_data_order
+	adr	r3,.Lsha256_block_data_order
 #endif
 #if __ARM_MAX_ARCH__>=7 && !defined(__KERNEL__)
 	ldr	r12,.LOPENSSL_armcap
Signature · 1 line
2.19.1
author maintainer
#22 lore raw

PATCH 21/37 crypto: sha512/arm - fix crash bug in Thumb2 build

In reply to Sasha Levin

From: Ard Biesheuvel <[email protected]>

[ Upstream commit c64316502008064c158fa40cc250665e461b0f2a ]

The SHA512 code we adopted from the OpenSSL project uses a rather peculiar way to take the address of the round constant table: it takes the address of the sha256_block_data_order() routine, and substracts a constant known quantity to arrive at the base of the table, which is emitted by the same assembler code right before the routine's entry point.

However, recent versions of binutils have helpfully changed the behavior of references emitted via an ADR instruction when running in Thumb2 mode: it now takes the Thumb execution mode bit into account, which is bit 0 af the address. This means the produced table address also has bit 0 set, and so we end up with an address value pointing 1 byte past the start of the table, which results in crashes such as

  Unable to handle kernel paging request at virtual address bf825000
  pgd = 42f44b11
  [bf825000] *pgd=80000040206003, *pmd=5f1bd003, *pte=00000000
  Internal error: Oops: 207 [#1] PREEMPT SMP THUMB2
  Modules linked in: sha256_arm(+) sha1_arm_ce sha1_arm ...
  CPU: 7 PID: 396 Comm: cryptomgr_test Not tainted 5.0.0-rc6+ #144
  Hardware name: QEMU KVM Virtual Machine, BIOS 0.0.0 02/06/2015
  PC is at sha256_block_data_order+0xaaa/0xb30 [sha256_arm]
  LR is at __this_module+0x17fd/0xffffe800 [sha256_arm]
  pc : [<bf820bca>]    lr : [<bf824ffd>]    psr: 800b0033
  sp : ebc8bbe8  ip : faaabe1c  fp : 2fdd3433
  r10: 4c5f1692  r9 : e43037df  r8 : b04b0a5a
  r7 : c369d722  r6 : 39c3693e  r5 : 7a013189  r4 : 1580d26b
  r3 : 8762a9b0  r2 : eea9c2cd  r1 : 3e9ab536  r0 : 1dea4ae7
  Flags: Nzcv  IRQs on  FIQs on  Mode SVC_32  ISA Thumb  Segment user
  Control: 70c5383d  Table: 6b8467c0  DAC: dbadc0de
  Process cryptomgr_test (pid: 396, stack limit = 0x69e1fe23)
  Stack: (0xebc8bbe8 to 0xebc8c000)
  ...
  unwind: Unknown symbol address bf820bca
  unwind: Index not found bf820bca
  Code: 441a ea80 40f9 440a (f85e) 3b04
  ---[ end trace e560cce92700ef8a ]---

Given that this affects older kernels as well, in case they are built with a recent toolchain, apply a minimal backportable fix, which is to emit another non-code label at the start of the routine, and reference that instead. (This is similar to the current upstream state of this file in OpenSSL)

---
 arch/arm/crypto/sha512-armv4.pl       | 3 ++-
 arch/arm/crypto/sha512-core.S_shipped | 3 ++-
 2 files changed, 4 insertions(+), 2 deletions(-)
arch/arm/crypto/sha512-armv4.pl +2 -1
index a2b11a844357..5fe336420bcf 100644
@@ -267,10 +267,11 @@ WORD64(0x5fcb6fab,0x3ad6faec, 0x6c44198c,0x4a475817)
 .global	sha512_block_data_order
 .type	sha512_block_data_order,%function
 sha512_block_data_order:
+.Lsha512_block_data_order:
 #if __ARM_ARCH__<7
 	sub	r3,pc,#8		@ sha512_block_data_order
 #else
-	adr	r3,sha512_block_data_order
+	adr	r3,.Lsha512_block_data_order
 #endif
 #if __ARM_MAX_ARCH__>=7 && !defined(__KERNEL__)
 	ldr	r12,.LOPENSSL_armcap
arch/arm/crypto/sha512-core.S_shipped +2 -1
index 3694c4d4ca2b..de9bd7f55242 100644
@@ -134,10 +134,11 @@ WORD64(0x5fcb6fab,0x3ad6faec, 0x6c44198c,0x4a475817)
 .global	sha512_block_data_order
 .type	sha512_block_data_order,%function
 sha512_block_data_order:
+.Lsha512_block_data_order:
 #if __ARM_ARCH__<7
 	sub	r3,pc,#8		@ sha512_block_data_order
 #else
-	adr	r3,sha512_block_data_order
+	adr	r3,.Lsha512_block_data_order
 #endif
 #if __ARM_MAX_ARCH__>=7 && !defined(__KERNEL__)
 	ldr	r12,.LOPENSSL_armcap
Signature · 1 line
2.19.1
author maintainer
#23 lore raw

PATCH 22/37 iommu/dmar: Fix buffer overflow during PCI bus notification

In reply to Sasha Levin

From: Julia Cartwright <[email protected]>

[ Upstream commit cffaaf0c816238c45cd2d06913476c83eb50f682 ]

Commit 57384592c433 ("iommu/vt-d: Store bus information in RMRR PCI device path") changed the type of the path data, however, the change in path type was not reflected in size calculations. Update to use the correct type and prevent a buffer overflow.

This bug manifests in systems with deep PCI hierarchies, and can lead to an overflow of the static allocated buffer (dmar_pci_notify_info_buf), or can lead to overflow of slab-allocated data.

   BUG: KASAN: global-out-of-bounds in dmar_alloc_pci_notify_info+0x1d5/0x2e0
   Write of size 1 at addr ffffffff90445d80 by task swapper/0/1
   CPU: 0 PID: 1 Comm: swapper/0 Tainted: G        W       4.14.87-rt49-02406-gd0a0e96 #1
   Call Trace:
    ? dump_stack+0x46/0x59
    ? print_address_description+0x1df/0x290
    ? dmar_alloc_pci_notify_info+0x1d5/0x2e0
    ? kasan_report+0x256/0x340
    ? dmar_alloc_pci_notify_info+0x1d5/0x2e0
    ? e820__memblock_setup+0xb0/0xb0
    ? dmar_dev_scope_init+0x424/0x48f
    ? __down_write_common+0x1ec/0x230
    ? dmar_dev_scope_init+0x48f/0x48f
    ? dmar_free_unused_resources+0x109/0x109
    ? cpumask_next+0x16/0x20
    ? __kmem_cache_create+0x392/0x430
    ? kmem_cache_create+0x135/0x2f0
    ? e820__memblock_setup+0xb0/0xb0
    ? intel_iommu_init+0x170/0x1848
    ? _raw_spin_unlock_irqrestore+0x32/0x60
    ? migrate_enable+0x27a/0x5b0
    ? sched_setattr+0x20/0x20
    ? migrate_disable+0x1fc/0x380
    ? task_rq_lock+0x170/0x170
    ? try_to_run_init_process+0x40/0x40
    ? locks_remove_file+0x85/0x2f0
    ? dev_prepare_static_identity_mapping+0x78/0x78
    ? rt_spin_unlock+0x39/0x50
    ? lockref_put_or_lock+0x2a/0x40
    ? dput+0x128/0x2f0
    ? __rcu_read_unlock+0x66/0x80
    ? __fput+0x250/0x300
    ? __rcu_read_lock+0x1b/0x30
    ? mntput_no_expire+0x38/0x290
    ? e820__memblock_setup+0xb0/0xb0
    ? pci_iommu_init+0x25/0x63
    ? pci_iommu_init+0x25/0x63
    ? do_one_initcall+0x7e/0x1c0
    ? initcall_blacklisted+0x120/0x120
    ? kernel_init_freeable+0x27b/0x307
    ? rest_init+0xd0/0xd0
    ? kernel_init+0xf/0x120
    ? rest_init+0xd0/0xd0
    ? ret_from_fork+0x1f/0x40
   The buggy address belongs to the variable:
    dmar_pci_notify_info_buf+0x40/0x60
Fixes 57384592c433 ("iommu/vt-d: Store bus information in RMRR PCI device path")
---
 drivers/iommu/dmar.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/iommu/dmar.c +1 -1
index c0d1c4db5794..38d0128b8135 100644
@@ -144,7 +144,7 @@ dmar_alloc_pci_notify_info(struct pci_dev *dev, unsigned long event)
 		for (tmp = dev; tmp; tmp = tmp->bus->self)
 			level++;
 
-	size = sizeof(*info) + level * sizeof(struct acpi_dmar_pci_path);
+	size = sizeof(*info) + level * sizeof(info->path[0]);
 	if (size <= sizeof(dmar_pci_notify_info_buf)) {
 		info = (struct dmar_pci_notify_info *)dmar_pci_notify_info_buf;
 	} else {
Signature · 1 line
2.19.1
author maintainer
#24 lore raw

PATCH 23/37 kvm: properly check debugfs dentry before using it

In reply to Sasha Levin

From: Greg Kroah-Hartman <[email protected]>

[ Upstream commit 8ed0579c12b2fe56a1fac2f712f58fc26c1dc49b ]

debugfs can now report an error code if something went wrong instead of just NULL. So if the return value is to be used as a "real" dentry, it needs to be checked if it is an error before dereferencing it.

This is now happening because of ff9fb72bc077 ("debugfs: return error values, not NULL"). syzbot has found a way to trigger multiple debugfs files attempting to be created, which fails, and then the error code gets passed to dentry_path_raw() which obviously does not like it.

---
 virt/kvm/kvm_main.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
virt/kvm/kvm_main.c +1 -1
index 66cc315efa6d..cf6a86631267 100644
@@ -3911,7 +3911,7 @@ static void kvm_uevent_notify_change(unsigned int type, struct kvm *kvm)
 	}
 	add_uevent_var(env, "PID=%d", kvm->userspace_pid);
 
-	if (kvm->debugfs_dentry) {
+	if (!IS_ERR_OR_NULL(kvm->debugfs_dentry)) {
 		char *tmp, *p = kmalloc(PATH_MAX, GFP_KERNEL);
 
 		if (p) {
Signature · 1 line
2.19.1
author maintainer
#25 lore raw

PATCH 24/37 soc/tegra: pmc: Drop locking from tegra_powergate_is_powered()

In reply to Sasha Levin

From: Dmitry Osipenko <[email protected]>

[ Upstream commit b6e1fd17a38bd1d97c11d69fd3207b3ef9bfa4b3 ]

This fixes splats like the one below if CONFIG_DEBUG_ATOMIC_SLEEP=y and machine (Tegra30) booted with SMP=n or all secondary CPU's are put offline. Locking isn't needed because it protects atomic operation.

BUG: sleeping function called from invalid context at kernel/locking/mutex.c:254
in_atomic(): 1, irqs_disabled(): 128, pid: 0, name: swapper/0
CPU: 0 PID: 0 Comm: swapper/0 Tainted: G         C        4.18.0-next-20180821-00180-gc3ebb6544e44-dirty #823
Hardware name: NVIDIA Tegra SoC (Flattened Device Tree)
[<c01134f4>] (unwind_backtrace) from [<c010db2c>] (show_stack+0x20/0x24)
[<c010db2c>] (show_stack) from [<c0bd0f3c>] (dump_stack+0x94/0xa8)
[<c0bd0f3c>] (dump_stack) from [<c0151df8>] (___might_sleep+0x13c/0x174)
[<c0151df8>] (___might_sleep) from [<c0151ea0>] (__might_sleep+0x70/0xa8)
[<c0151ea0>] (__might_sleep) from [<c0bec2b8>] (mutex_lock+0x2c/0x70)
[<c0bec2b8>] (mutex_lock) from [<c0589844>] (tegra_powergate_is_powered+0x44/0xa8)
[<c0589844>] (tegra_powergate_is_powered) from [<c0581a60>] (tegra30_cpu_rail_off_ready+0x30/0x74)
[<c0581a60>] (tegra30_cpu_rail_off_ready) from [<c0122244>] (tegra30_idle_lp2+0xa0/0x108)
[<c0122244>] (tegra30_idle_lp2) from [<c0853438>] (cpuidle_enter_state+0x140/0x540)
[<c0853438>] (cpuidle_enter_state) from [<c08538a4>] (cpuidle_enter+0x40/0x4c)
[<c08538a4>] (cpuidle_enter) from [<c01595e0>] (call_cpuidle+0x30/0x48)
[<c01595e0>] (call_cpuidle) from [<c01599f8>] (do_idle+0x238/0x28c)
[<c01599f8>] (do_idle) from [<c0159d28>] (cpu_startup_entry+0x28/0x2c)
[<c0159d28>] (cpu_startup_entry) from [<c0be76c8>] (rest_init+0xd8/0xdc)
[<c0be76c8>] (rest_init) from [<c1200f50>] (start_kernel+0x41c/0x430)
---
 drivers/soc/tegra/pmc.c | 8 +-------
 1 file changed, 1 insertion(+), 7 deletions(-)
drivers/soc/tegra/pmc.c +1 -7
index 7e9ef3431bea..2422ed56895a 100644
@@ -521,16 +521,10 @@ EXPORT_SYMBOL(tegra_powergate_power_off);
  */
 int tegra_powergate_is_powered(unsigned int id)
 {
-	int status;
-
 	if (!tegra_powergate_is_valid(id))
 		return -EINVAL;
 
-	mutex_lock(&pmc->powergates_lock);
-	status = tegra_powergate_state(id);
-	mutex_unlock(&pmc->powergates_lock);
-
-	return status;
+	return tegra_powergate_state(id);
 }
 
 /**
Signature · 1 line
2.19.1
author maintainer
#26 lore raw

PATCH 25/37 ext4: prohibit fstrim in norecovery mode

In reply to Sasha Levin

From: "Darrick J. Wong" <[email protected]>

[ Upstream commit 18915b5873f07e5030e6fb108a050fa7c71c59fb ]

The ext4 fstrim implementation uses the block bitmaps to find free space that can be discarded. If we haven't replayed the journal, the bitmaps will be stale and we absolutely *cannot* use stale metadata to zap the underlying storage.

---
 fs/ext4/ioctl.c | 7 +++++++
 1 file changed, 7 insertions(+)
fs/ext4/ioctl.c +7 -0
index 7917cc89ab21..3dbf4e414706 100644
@@ -940,6 +940,13 @@ long ext4_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 		if (!blk_queue_discard(q))
 			return -EOPNOTSUPP;
 
+		/*
+		 * We haven't replayed the journal, so we cannot use our
+		 * block-bitmap-guided storage zapping commands.
+		 */
+		if (test_opt(sb, NOLOAD) && ext4_has_feature_journal(sb))
+			return -EROFS;
+
 		if (copy_from_user(&range, (struct fstrim_range __user *)arg,
 		    sizeof(range)))
 			return -EFAULT;
Signature · 1 line
2.19.1
author maintainer
#27 lore raw

PATCH 26/37 lkdtm: Print real addresses

In reply to Sasha Levin

From: Christophe Leroy <[email protected]>

[ Upstream commit 4c411157a42f122051ae3469bee0b5cabe89e139 ]

Today, when doing a lkdtm test before the readiness of the random generator, (ptrval) is printed instead of the address at which it perform the fault:

[ 1597.337030] lkdtm: Performing direct entry EXEC_USERSPACE
[ 1597.337142] lkdtm: attempting ok execution at (ptrval)
[ 1597.337398] lkdtm: attempting bad execution at (ptrval)
[ 1597.337460] kernel tried to execute user page (77858000) -exploit attempt? (uid: 0)
[ 1597.344769] Unable to handle kernel paging request for instruction fetch
[ 1597.351392] Faulting instruction address: 0x77858000
[ 1597.356312] Oops: Kernel access of bad area, sig: 11 [#1]

If the lkdtm test is done later on, it prints an hashed address.

In both cases this is pointless. The purpose of the test is to ensure the kernel generates an Oops at the expected address, so real addresses needs to be printed. This patch fixes that.

---
 drivers/misc/lkdtm_perms.c | 18 +++++++++---------
 1 file changed, 9 insertions(+), 9 deletions(-)
drivers/misc/lkdtm_perms.c +9 -9
index 53b85c9d16b8..fa54add6375a 100644
@@ -47,7 +47,7 @@ static noinline void execute_location(void *dst, bool write)
 {
 	void (*func)(void) = dst;
 
-	pr_info("attempting ok execution at %p\n", do_nothing);
+	pr_info("attempting ok execution at %px\n", do_nothing);
 	do_nothing();
 
 	if (write == CODE_WRITE) {
@@ -55,7 +55,7 @@ static noinline void execute_location(void *dst, bool write)
 		flush_icache_range((unsigned long)dst,
 				   (unsigned long)dst + EXEC_SIZE);
 	}
-	pr_info("attempting bad execution at %p\n", func);
+	pr_info("attempting bad execution at %px\n", func);
 	func();
 }
 
@@ -66,14 +66,14 @@ static void execute_user_location(void *dst)
 	/* Intentionally crossing kernel/user memory boundary. */
 	void (*func)(void) = dst;
 
-	pr_info("attempting ok execution at %p\n", do_nothing);
+	pr_info("attempting ok execution at %px\n", do_nothing);
 	do_nothing();
 
 	copied = access_process_vm(current, (unsigned long)dst, do_nothing,
 				   EXEC_SIZE, FOLL_WRITE);
 	if (copied < EXEC_SIZE)
 		return;
-	pr_info("attempting bad execution at %p\n", func);
+	pr_info("attempting bad execution at %px\n", func);
 	func();
 }
 
@@ -82,7 +82,7 @@ void lkdtm_WRITE_RO(void)
 	/* Explicitly cast away "const" for the test. */
 	unsigned long *ptr = (unsigned long *)&rodata;
 
-	pr_info("attempting bad rodata write at %p\n", ptr);
+	pr_info("attempting bad rodata write at %px\n", ptr);
 	*ptr ^= 0xabcd1234;
 }
 
@@ -100,7 +100,7 @@ void lkdtm_WRITE_RO_AFTER_INIT(void)
 		return;
 	}
 
-	pr_info("attempting bad ro_after_init write at %p\n", ptr);
+	pr_info("attempting bad ro_after_init write at %px\n", ptr);
 	*ptr ^= 0xabcd1234;
 }
 
@@ -112,7 +112,7 @@ void lkdtm_WRITE_KERN(void)
 	size = (unsigned long)do_overwritten - (unsigned long)do_nothing;
 	ptr = (unsigned char *)do_overwritten;
 
-	pr_info("attempting bad %zu byte write at %p\n", size, ptr);
+	pr_info("attempting bad %zu byte write at %px\n", size, ptr);
 	memcpy(ptr, (unsigned char *)do_nothing, size);
 	flush_icache_range((unsigned long)ptr, (unsigned long)(ptr + size));
 
@@ -185,11 +185,11 @@ void lkdtm_ACCESS_USERSPACE(void)
 
 	ptr = (unsigned long *)user_addr;
 
-	pr_info("attempting bad read at %p\n", ptr);
+	pr_info("attempting bad read at %px\n", ptr);
 	tmp = *ptr;
 	tmp += 0xc0dec0de;
 
-	pr_info("attempting bad write at %p\n", ptr);
+	pr_info("attempting bad write at %px\n", ptr);
 	*ptr = tmp;
 
 	vm_munmap(user_addr, PAGE_SIZE);
Signature · 1 line
2.19.1
author maintainer
#28 lore raw

PATCH 27/37 lkdtm: Add tests for NULL pointer dereference

In reply to Sasha Levin

From: Christophe Leroy <[email protected]>

[ Upstream commit 59a12205d3c32aee4c13ca36889fdf7cfed31126 ]

Introduce lkdtm tests for NULL pointer dereference: check access or exec at NULL address, since these errors tend to be reported differently from the general fault error text. For example from x86:

pr_alert("BUG: unable to handle kernel %s at %px\n",
address < PAGE_SIZE ? "NULL pointer dereference" : "paging request", (void *)address);

---
 drivers/misc/lkdtm.h       |  2 ++
 drivers/misc/lkdtm_core.c  |  2 ++
 drivers/misc/lkdtm_perms.c | 18 ++++++++++++++++++
 3 files changed, 22 insertions(+)
drivers/misc/lkdtm.h +2 -0
index 687a0dbbe199..614612325332 100644
@@ -45,7 +45,9 @@ void lkdtm_EXEC_KMALLOC(void);
 void lkdtm_EXEC_VMALLOC(void);
 void lkdtm_EXEC_RODATA(void);
 void lkdtm_EXEC_USERSPACE(void);
+void lkdtm_EXEC_NULL(void);
 void lkdtm_ACCESS_USERSPACE(void);
+void lkdtm_ACCESS_NULL(void);
 
 /* lkdtm_refcount.c */
 void lkdtm_REFCOUNT_INC_OVERFLOW(void);
drivers/misc/lkdtm_core.c +2 -0
index 981b3ef71e47..199271708aed 100644
@@ -220,7 +220,9 @@ struct crashtype crashtypes[] = {
 	CRASHTYPE(EXEC_VMALLOC),
 	CRASHTYPE(EXEC_RODATA),
 	CRASHTYPE(EXEC_USERSPACE),
+	CRASHTYPE(EXEC_NULL),
 	CRASHTYPE(ACCESS_USERSPACE),
+	CRASHTYPE(ACCESS_NULL),
 	CRASHTYPE(WRITE_RO),
 	CRASHTYPE(WRITE_RO_AFTER_INIT),
 	CRASHTYPE(WRITE_KERN),
drivers/misc/lkdtm_perms.c +18 -0
index fa54add6375a..62f76d506f04 100644
@@ -164,6 +164,11 @@ void lkdtm_EXEC_USERSPACE(void)
 	vm_munmap(user_addr, PAGE_SIZE);
 }
 
+void lkdtm_EXEC_NULL(void)
+{
+	execute_location(NULL, CODE_AS_IS);
+}
+
 void lkdtm_ACCESS_USERSPACE(void)
 {
 	unsigned long user_addr, tmp = 0;
@@ -195,6 +200,19 @@ void lkdtm_ACCESS_USERSPACE(void)
 	vm_munmap(user_addr, PAGE_SIZE);
 }
 
+void lkdtm_ACCESS_NULL(void)
+{
+	unsigned long tmp;
+	unsigned long *ptr = (unsigned long *)NULL;
+
+	pr_info("attempting bad read at %px\n", ptr);
+	tmp = *ptr;
+	tmp += 0xc0dec0de;
+
+	pr_info("attempting bad write at %px\n", ptr);
+	*ptr = tmp;
+}
+
 void __init lkdtm_perms_init(void)
 {
 	/* Make sure we can write to __ro_after_init values during __init */
Signature · 1 line
2.19.1