Skip to contentSkip to conversation

3.16.55-rc1 review

PATCH In review Ben Hutchings 4 participants 261 messages Feb 28, 2018 – Mar 3, 2018 Page 2 of 2

Subsystems (5)
Stable branch KVM PCI MIPS +1 more subsystems
lore
Share & tools
lore
#172 lore raw

PATCH 168/254 sh_eth: fix TSU resource handling

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Sergei Shtylyov <[email protected]>

commit dfe8266b8dd10e12a731c985b725fcf7f0e537f0 upstream.

When switching the driver to the managed device API, I managed to break the case of a dual Ether devices sharing a single TSU: the 2nd Ether port wouldn't probe. Iwamatsu-san has tried to fix this but his patch was buggy and he then dropped the ball...

The solution is to limit calling devm_request_mem_region() to the first of the two ports sharing the same TSU, so devm_ioremap_resource() can't be used anymore for the TSU resource...

Fixes d5e07e69218f ("sh_eth: use managed device API")
---
 drivers/net/ethernet/renesas/sh_eth.c | 25 ++++++++++++++++++++++---
 1 file changed, 22 insertions(+), 3 deletions(-)
drivers/net/ethernet/renesas/sh_eth.c +22 -3
@@ -2873,10 +2873,29 @@ static int sh_eth_drv_probe(struct platf
 	/* ioremap the TSU registers */
 	if (mdp->cd->tsu) {
 		struct resource *rtsu;
+
 		rtsu = platform_get_resource(pdev, IORESOURCE_MEM, 1);
-		mdp->tsu_addr = devm_ioremap_resource(&pdev->dev, rtsu);
-		if (IS_ERR(mdp->tsu_addr)) {
-			ret = PTR_ERR(mdp->tsu_addr);
+		if (!rtsu) {
+			dev_err(&pdev->dev, "no TSU resource\n");
+			ret = -ENODEV;
+			goto out_release;
+		}
+		/* We can only request the  TSU region  for the first port
+		 * of the two  sharing this TSU for the probe to succeed...
+		 */
+		if (devno % 2 == 0 &&
+		    !devm_request_mem_region(&pdev->dev, rtsu->start,
+					     resource_size(rtsu),
+					     dev_name(&pdev->dev))) {
+			dev_err(&pdev->dev, "can't request TSU resource.\n");
+			ret = -EBUSY;
+			goto out_release;
+		}
+		mdp->tsu_addr = devm_ioremap(&pdev->dev, rtsu->start,
+					     resource_size(rtsu));
+		if (!mdp->tsu_addr) {
+			dev_err(&pdev->dev, "TSU region ioremap() failed.\n");
+			ret = -ENOMEM;
 			goto out_release;
 		}
 		mdp->port = devno % 2;
#173 lore raw

PATCH 169/254 kernel/acct.c: fix the acct->needcheck check in check_free_space()

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Oleg Nesterov <[email protected]>

commit 4d9570158b6260f449e317a5f9ed030c2504a615 upstream.

As Tsukada explains, the time_is_before_jiffies(acct->needcheck) check is very wrong, we need time_is_after_jiffies() to make sys_acct() work.

Ignoring the overflows, the code should "goto out" if needcheck > jiffies, while currently it checks "needcheck < jiffies" and thus in the likely case check_free_space() does nothing until jiffies overflow.

In particular this means that sys_acct() is simply broken, acct_on() sets acct->needcheck = jiffies and expects that check_free_space() should set acct->active = 1 after the free-space check, but this won't happen if jiffies increments in between.

This was broken by commit 32dc73086015 ("get rid of timer in kern/acct.c") in 2011, then another (correct) commit 795a2f22a8ea ("acct() should honour the limits from the very beginning") made the problem more visible.

Fixes 32dc73086015 ("get rid of timer in kern/acct.c")
Cc Al Viro <[email protected]>
[bwh: Backported to 3.16: adjust context]
---
 kernel/acct.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
kernel/acct.c +1 -1
@@ -107,7 +107,7 @@ static int check_free_space(struct bsd_a
 
 	spin_lock(&acct_lock);
 	res = acct->active;
-	if (!file || time_is_before_jiffies(acct->needcheck))
+	if (!file || time_is_after_jiffies(acct->needcheck))
 		goto out;
 	spin_unlock(&acct_lock);
 
#174 lore raw

PATCH 170/254 mm/mprotect: add a cond_resched() inside change_pmd_range()

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Anshuman Khandual <[email protected]>

commit 4991c09c7c812dba13ea9be79a68b4565bb1fa4e upstream.

While testing on a large CPU system, detected the following RCU stall many times over the span of the workload. This problem is solved by adding a cond_resched() in the change_pmd_range() function.

  INFO: rcu_sched detected stalls on CPUs/tasks:
   154-....: (670 ticks this GP) idle=022/140000000000000/0 softirq=2825/2825 fqs=612
   (detected by 955, t=6002 jiffies, g=4486, c=4485, q=90864)
  Sending NMI from CPU 955 to CPUs 154:
  NMI backtrace for cpu 154
  CPU: 154 PID: 147071 Comm: workload Not tainted 4.15.0-rc3+ #3
  NIP:  c0000000000b3f64 LR: c0000000000b33d4 CTR: 000000000000aa18
  REGS: 00000000a4b0fb44 TRAP: 0501   Not tainted  (4.15.0-rc3+)
  MSR:  8000000000009033 <SF,EE,ME,IR,DR,RI,LE>  CR: 22422082  XER: 00000000
  CFAR: 00000000006cf8f0 SOFTE: 1
  GPR00: 0010000000000000 c00003ef9b1cb8c0 c0000000010cc600 0000000000000000
  GPR04: 8e0000018c32b200 40017b3858fd6e00 8e0000018c32b208 40017b3858fd6e00
  GPR08: 8e0000018c32b210 40017b3858fd6e00 8e0000018c32b218 40017b3858fd6e00
  GPR12: ffffffffffffffff c00000000fb25100
  NIP [c0000000000b3f64] plpar_hcall9+0x44/0x7c
  LR [c0000000000b33d4] pSeries_lpar_flush_hash_range+0x384/0x420
  Call Trace:
    flush_hash_range+0x48/0x100
    __flush_tlb_pending+0x44/0xd0
    hpte_need_flush+0x408/0x470
    change_protection_range+0xaac/0xf10
    change_prot_numa+0x30/0xb0
    task_numa_work+0x2d0/0x3e0
    task_work_run+0x130/0x190
    do_notify_resume+0x118/0x120
    ret_from_except_lite+0x70/0x74
  Instruction dump:
  60000000 f8810028 7ca42b78 7cc53378 7ce63b78 7d074378 7d284b78 7d495378
  e9410060 e9610068 e9810070 44000022 <7d806378> e9810028 f88c0000 f8ac0008
---
 mm/mprotect.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
mm/mprotect.c +4 -2
@@ -152,7 +152,7 @@ static inline unsigned long change_pmd_r
 
 		next = pmd_addr_end(addr, end);
 		if (!pmd_trans_huge(*pmd) && pmd_none_or_clear_bad(pmd))
-			continue;
+			goto next;
 
 		/* invoke the mmu notifier if the pmd is populated */
 		if (!mni_start) {
@@ -174,7 +174,7 @@ static inline unsigned long change_pmd_r
 					}
 
 					/* huge pmd was handled */
-					continue;
+					goto next;
 				}
 			}
 			/* fall through, the trans huge pmd just split */
@@ -182,6 +182,8 @@ static inline unsigned long change_pmd_r
 		this_pages = change_pte_range(vma, pmd, addr, next, newprot,
 				 dirty_accountable, prot_numa);
 		pages += this_pages;
+next:
+		cond_resched();
 	} while (pmd++, addr = next, addr != end);
 
 	if (mni_start)
#175 lore raw

PATCH 171/254 crypto: algapi - fix NULL dereference in crypto_remove_spawns()

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Eric Biggers <[email protected]>

commit 9a00674213a3f00394f4e3221b88f2d21fc05789 upstream.

syzkaller triggered a NULL pointer dereference in crypto_remove_spawns() via a program that repeatedly and concurrently requests AEADs "authenc(cmac(des3_ede-asm),pcbc-aes-aesni)" and hashes "cmac(des3_ede)" through AF_ALG, where the hashes are requested as "untested" (CRYPTO_ALG_TESTED is set in ->salg_mask but clear in ->salg_feat; this causes the template to be instantiated for every request).

Although AF_ALG users really shouldn't be able to request an "untested" algorithm, the NULL pointer dereference is actually caused by a longstanding race condition where crypto_remove_spawns() can encounter an instance which has had spawn(s) "grabbed" but hasn't yet been registered, resulting in ->cra_users still being NULL.

We probably should properly initialize ->cra_users earlier, but that would require updating many templates individually. For now just fix the bug in a simple way that can easily be backported: make crypto_remove_spawns() treat a NULL ->cra_users list as empty.

---
 crypto/algapi.c | 12 ++++++++++++
 1 file changed, 12 insertions(+)
crypto/algapi.c +12 -0
@@ -147,6 +147,18 @@ void crypto_remove_spawns(struct crypto_
 
 			spawn->alg = NULL;
 			spawns = &inst->alg.cra_users;
+
+			/*
+			 * We may encounter an unregistered instance here, since
+			 * an instance's spawns are set up prior to the instance
+			 * being registered.  An unregistered instance will have
+			 * NULL ->cra_users.next, since ->cra_users isn't
+			 * properly initialized until registration.  But an
+			 * unregistered instance cannot have any users, so treat
+			 * it the same as ->cra_users being empty.
+			 */
+			if (spawns->next == NULL)
+				break;
 		}
 	} while ((spawns = crypto_more_spawns(alg, &stack, &top,
 					      &secondary_spawns)));
#176 lore raw

PATCH 172/254 xfrm: Use __skb_queue_tail in xfrm_trans_queue

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Herbert Xu <[email protected]>

commit d16b46e4fd8bc6063624605f25b8c0835bb1fbe3 upstream.

We do not need locking in xfrm_trans_queue because it is designed to use per-CPU buffers. However, the original code incorrectly used skb_queue_tail which takes the lock. This patch switches it to __skb_queue_tail instead.

Reported-and-tested-by Artem Savkov <[email protected]>
Fixes acf568ee859f ("xfrm: Reinject transport-mode packets...")
---
 net/xfrm/xfrm_input.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/xfrm/xfrm_input.c +1 -1
@@ -414,7 +414,7 @@ int xfrm_trans_queue(struct sk_buff *skb
 		return -ENOBUFS;
 
 	XFRM_TRANS_SKB_CB(skb)->finish = finish;
-	skb_queue_tail(&trans->queue, skb);
+	__skb_queue_tail(&trans->queue, skb);
 	tasklet_schedule(&trans->tasklet);
 	return 0;
 }
#177 lore raw

PATCH 173/254 ARM: dts: kirkwood: fix pin-muxing of MPP7 on OpenBlocks A7

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Thomas Petazzoni <[email protected]>

commit 56aeb07c914a616ab84357d34f8414a69b140cdf upstream.

MPP7 is currently muxed as "gpio", but this function doesn't exist for MPP7, only "gpo" is available. This causes the following error:

kirkwood-pinctrl f1010000.pin-controller: unsupported function gpio on pin mpp7 pinctrl core: failed to register map default (6): invalid type given kirkwood-pinctrl f1010000.pin-controller: error claiming hogs: -22 kirkwood-pinctrl f1010000.pin-controller: could not claim hogs: -22 kirkwood-pinctrl f1010000.pin-controller: unable to register pinctrl driver
kirkwood-pinctrl: probe of f1010000.pin-controller failed with error -22

So the pinctrl driver is not probed, all device drivers (including the UART driver) do a -EPROBE_DEFER, and therefore the system doesn't really boot (well, it boots, but with no UART, and no devices that require pin-muxing).

Back when the Device Tree file for this board was introduced, the definition was already wrong. The pinctrl driver also always described as "gpo" this function for MPP7. However, between Linux 4.10 and 4.11, a hog pin failing to be muxed was turned from a simple warning to a hard error that caused the entire pinctrl driver probe to bail out. This is probably the result of commit 6118714275f0a ("pinctrl:
core: Fix pinctrl_register_and_init() with pinctrl_enable()").

This commit fixes the Device Tree to use the proper "gpo" function for MPP7, which fixes the boot of OpenBlocks A7, which was broken since Linux 4.11.

Fixes f24b56cbcd9d ("ARM: kirkwood: add support for OpenBlocks A7 platform")
---
 arch/arm/boot/dts/kirkwood-openblocks_a7.dts | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)
arch/arm/boot/dts/kirkwood-openblocks_a7.dts +8 -2
@@ -53,7 +53,8 @@
 		};
 
 		pinctrl: pin-controller@10000 {
-			pinctrl-0 = <&pmx_dip_switches &pmx_gpio_header>;
+			pinctrl-0 = <&pmx_dip_switches &pmx_gpio_header
+				     &pmx_gpio_header_gpo>;
 			pinctrl-names = "default";
 
 			pmx_uart0: pmx-uart0 {
@@ -85,11 +86,16 @@
 			 * ground.
 			 */
 			pmx_gpio_header: pmx-gpio-header {
-				marvell,pins = "mpp17", "mpp7", "mpp29", "mpp28",
+				marvell,pins = "mpp17", "mpp29", "mpp28",
 					       "mpp35", "mpp34", "mpp40";
 				marvell,function = "gpio";
 			};
 
+			pmx_gpio_header_gpo: pxm-gpio-header-gpo {
+				marvell,pins = "mpp7";
+				marvell,function = "gpo";
+			};
+
 			pmx_gpio_init: pmx-init {
 				marvell,pins = "mpp38";
 				marvell,function = "gpio";
#178 lore raw

PATCH 174/254 x86/alternatives: Add missing '\n' at end of ALTERNATIVE inline asm

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: David Woodhouse <[email protected]>

commit b9e705ef7cfaf22db0daab91ad3cd33b0fa32eb9 upstream.

Where an ALTERNATIVE is used in the middle of an inline asm block, this would otherwise lead to the following instruction being appended directly to the trailing ".popsection", and a failed compile.

Fixes 9cebed423c84 ("x86, alternative: Use .pushsection/.popsection")
Cc Rik van Riel <[email protected]>
Cc Tim Chen <[email protected]>
Cc Peter Zijlstra <[email protected]>
Cc Paul Turner <[email protected]>
Cc Jiri Kosina <[email protected]>
Cc Andy Lutomirski <[email protected]>
Cc Dave Hansen <[email protected]>
Cc Kees Cook <[email protected]>
Cc Linus Torvalds <[email protected]>
Cc Greg Kroah-Hartman <[email protected]>
---
 arch/x86/include/asm/alternative.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
arch/x86/include/asm/alternative.h +2 -2
@@ -124,7 +124,7 @@ static inline int alternatives_text_rese
 	".popsection\n"							\
 	".pushsection .altinstr_replacement, \"ax\"\n"			\
 	ALTINSTR_REPLACEMENT(newinstr, feature, 1)			\
-	".popsection"
+	".popsection\n"
 
 #define ALTERNATIVE_2(oldinstr, newinstr1, feature1, newinstr2, feature2)\
 	OLDINSTR_2(oldinstr, 1, 2)					\
@@ -135,7 +135,7 @@ static inline int alternatives_text_rese
 	".pushsection .altinstr_replacement, \"ax\"\n"			\
 	ALTINSTR_REPLACEMENT(newinstr1, feature1, 1)			\
 	ALTINSTR_REPLACEMENT(newinstr2, feature2, 2)			\
-	".popsection"
+	".popsection\n"
 
 /*
  * This must be included *after* the definition of ALTERNATIVE due to
#179 lore raw

PATCH 175/254 ALSA: aloop: Release cable upon open error path

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Takashi Iwai <[email protected]>

commit 9685347aa0a5c2869058ca6ab79fd8e93084a67f upstream.

The aloop runtime object and its assignment in the cable are left even when opening a substream fails. This doesn't mean any memory leak, but it still keeps the invalid pointer that may be referred by the another side of the cable spontaneously, which is a potential Oops cause.

Clean up the cable assignment and the empty cable upon the error path properly.

Fixes 597603d615d2 ("ALSA: introduce the snd-aloop module for the PCM loopback")
---
 sound/drivers/aloop.c | 38 +++++++++++++++++++++++++-------------
 1 file changed, 25 insertions(+), 13 deletions(-)
sound/drivers/aloop.c +25 -13
@@ -659,12 +659,31 @@ static int rule_channels(struct snd_pcm_
 	return snd_interval_refine(hw_param_interval(params, rule->var), &t);
 }
 
+static void free_cable(struct snd_pcm_substream *substream)
+{
+	struct loopback *loopback = substream->private_data;
+	int dev = get_cable_index(substream);
+	struct loopback_cable *cable;
+
+	cable = loopback->cables[substream->number][dev];
+	if (!cable)
+		return;
+	if (cable->streams[!substream->stream]) {
+		/* other stream is still alive */
+		cable->streams[substream->stream] = NULL;
+	} else {
+		/* free the cable */
+		loopback->cables[substream->number][dev] = NULL;
+		kfree(cable);
+	}
+}
+
 static int loopback_open(struct snd_pcm_substream *substream)
 {
 	struct snd_pcm_runtime *runtime = substream->runtime;
 	struct loopback *loopback = substream->private_data;
 	struct loopback_pcm *dpcm;
-	struct loopback_cable *cable;
+	struct loopback_cable *cable = NULL;
 	int err = 0;
 	int dev = get_cable_index(substream);
 
@@ -683,7 +702,6 @@ static int loopback_open(struct snd_pcm_
 	if (!cable) {
 		cable = kzalloc(sizeof(*cable), GFP_KERNEL);
 		if (!cable) {
-			kfree(dpcm);
 			err = -ENOMEM;
 			goto unlock;
 		}
@@ -725,6 +743,10 @@ static int loopback_open(struct snd_pcm_
 	else
 		runtime->hw = cable->hw;
  unlock:
+	if (err < 0) {
+		free_cable(substream);
+		kfree(dpcm);
+	}
 	mutex_unlock(&loopback->cable_lock);
 	return err;
 }
@@ -733,20 +755,10 @@ static int loopback_close(struct snd_pcm
 {
 	struct loopback *loopback = substream->private_data;
 	struct loopback_pcm *dpcm = substream->runtime->private_data;
-	struct loopback_cable *cable;
-	int dev = get_cable_index(substream);
 
 	loopback_timer_stop(dpcm);
 	mutex_lock(&loopback->cable_lock);
-	cable = loopback->cables[substream->number][dev];
-	if (cable->streams[!substream->stream]) {
-		/* other stream is still alive */
-		cable->streams[substream->stream] = NULL;
-	} else {
-		/* free the cable */
-		loopback->cables[substream->number][dev] = NULL;
-		kfree(cable);
-	}
+	free_cable(substream);
 	mutex_unlock(&loopback->cable_lock);
 	return 0;
 }
#180 lore raw

PATCH 176/254 ALSA: aloop: Fix inconsistent format due to incomplete rule

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Takashi Iwai <[email protected]>

commit b088b53e20c7d09b5ab84c5688e609f478e5c417 upstream.

The extra hw constraint rule for the formats the aloop driver introduced has a slight flaw, where it doesn't return a positive value when the mask got changed. It came from the fact that it's basically a copy&paste from snd_hw_constraint_mask64(). The original code is supposed to be a single-shot and it modifies the mask bits only once and never after, while what we need for aloop is the dynamic hw rule that limits the mask bits.

This difference results in the inconsistent state, as the hw_refine doesn't apply the dependencies fully. The worse and surprisingly result is that it causes a crash in OSS emulation when multiple full-duplex reads/writes are performed concurrently (I leave why it triggers Oops to readers as a homework).

For fixing this, replace a few open-codes with the standard snd_mask_*() macros.

Reported-by [email protected]
Fixes b1c73fc8e697 ("ALSA: snd-aloop: Fix hw_params restrictions and checking")
---
 sound/drivers/aloop.c | 13 ++++++-------
 1 file changed, 6 insertions(+), 7 deletions(-)
sound/drivers/aloop.c +6 -7
@@ -39,6 +39,7 @@
 #include <sound/core.h>
 #include <sound/control.h>
 #include <sound/pcm.h>
+#include <sound/pcm_params.h>
 #include <sound/info.h>
 #include <sound/initval.h>
 
@@ -623,14 +624,12 @@ static int rule_format(struct snd_pcm_hw
 {
 
 	struct snd_pcm_hardware *hw = rule->private;
-	struct snd_mask *maskp = hw_param_mask(params, rule->var);
+	struct snd_mask m;
 
-	maskp->bits[0] &= (u_int32_t)hw->formats;
-	maskp->bits[1] &= (u_int32_t)(hw->formats >> 32);
-	memset(maskp->bits + 2, 0, (SNDRV_MASK_MAX-64) / 8); /* clear rest */
-	if (! maskp->bits[0] && ! maskp->bits[1])
-		return -EINVAL;
-	return 0;
+	snd_mask_none(&m);
+	m.bits[0] = (u_int32_t)hw->formats;
+	m.bits[1] = (u_int32_t)(hw->formats >> 32);
+	return snd_mask_refine(hw_param_mask(params, rule->var), &m);
 }
 
 static int rule_rate(struct snd_pcm_hw_params *params,
#181 lore raw

PATCH 177/254 ALSA: aloop: Fix racy hw constraints adjustment

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Takashi Iwai <[email protected]>

commit 898dfe4687f460ba337a01c11549f87269a13fa2 upstream.

The aloop driver tries to update the hw constraints of the connected target on the cable of the opened PCM substream. This is done by adding the extra hw constraints rules referring to the substream runtime->hw fields, while the other substream may update the runtime hw of another side on the fly.

This is, however, racy and may result in the inconsistent values when both PCM streams perform the prepare concurrently. One of the reason is that it overwrites the other's runtime->hw field; which is not only racy but also broken when it's called before the open of another side finishes. And, since the reference to runtime->hw isn't protected, the concurrent write may give the partial value update and become inconsistent.

This patch is an attempt to fix and clean up:
- The prepare doesn't change the runtime->hw of other side any longer, but only update the cable->hw that is referred commonly.
- The extra rules refer to the loopback_pcm object instead of the runtime->hw. The actual hw is deduced from cable->hw.
- The extra rules take the cable_lock to protect against the race.

Fixes b1c73fc8e697 ("ALSA: snd-aloop: Fix hw_params restrictions and checking")
---
 sound/drivers/aloop.c | 51 +++++++++++++++++++++------------------------------
 1 file changed, 21 insertions(+), 30 deletions(-)
sound/drivers/aloop.c +21 -30
@@ -307,19 +307,6 @@ static int loopback_trigger(struct snd_p
 	return 0;
 }
 
-static void params_change_substream(struct loopback_pcm *dpcm,
-				    struct snd_pcm_runtime *runtime)
-{
-	struct snd_pcm_runtime *dst_runtime;
-
-	if (dpcm == NULL || dpcm->substream == NULL)
-		return;
-	dst_runtime = dpcm->substream->runtime;
-	if (dst_runtime == NULL)
-		return;
-	dst_runtime->hw = dpcm->cable->hw;
-}
-
 static void params_change(struct snd_pcm_substream *substream)
 {
 	struct snd_pcm_runtime *runtime = substream->runtime;
@@ -331,10 +318,6 @@ static void params_change(struct snd_pcm
 	cable->hw.rate_max = runtime->rate;
 	cable->hw.channels_min = runtime->channels;
 	cable->hw.channels_max = runtime->channels;
-	params_change_substream(cable->streams[SNDRV_PCM_STREAM_PLAYBACK],
-				runtime);
-	params_change_substream(cable->streams[SNDRV_PCM_STREAM_CAPTURE],
-				runtime);
 }
 
 static int loopback_prepare(struct snd_pcm_substream *substream)
@@ -622,24 +605,29 @@ static unsigned int get_cable_index(stru
 static int rule_format(struct snd_pcm_hw_params *params,
 		       struct snd_pcm_hw_rule *rule)
 {
-
-	struct snd_pcm_hardware *hw = rule->private;
+	struct loopback_pcm *dpcm = rule->private;
+	struct loopback_cable *cable = dpcm->cable;
 	struct snd_mask m;
 
 	snd_mask_none(&m);
-	m.bits[0] = (u_int32_t)hw->formats;
-	m.bits[1] = (u_int32_t)(hw->formats >> 32);
+	mutex_lock(&dpcm->loopback->cable_lock);
+	m.bits[0] = (u_int32_t)cable->hw.formats;
+	m.bits[1] = (u_int32_t)(cable->hw.formats >> 32);
+	mutex_unlock(&dpcm->loopback->cable_lock);
 	return snd_mask_refine(hw_param_mask(params, rule->var), &m);
 }
 
 static int rule_rate(struct snd_pcm_hw_params *params,
 		     struct snd_pcm_hw_rule *rule)
 {
-	struct snd_pcm_hardware *hw = rule->private;
+	struct loopback_pcm *dpcm = rule->private;
+	struct loopback_cable *cable = dpcm->cable;
 	struct snd_interval t;
 
-        t.min = hw->rate_min;
-        t.max = hw->rate_max;
+	mutex_lock(&dpcm->loopback->cable_lock);
+	t.min = cable->hw.rate_min;
+	t.max = cable->hw.rate_max;
+	mutex_unlock(&dpcm->loopback->cable_lock);
         t.openmin = t.openmax = 0;
         t.integer = 0;
 	return snd_interval_refine(hw_param_interval(params, rule->var), &t);
@@ -648,11 +636,14 @@ static int rule_rate(struct snd_pcm_hw_p
 static int rule_channels(struct snd_pcm_hw_params *params,
 			 struct snd_pcm_hw_rule *rule)
 {
-	struct snd_pcm_hardware *hw = rule->private;
+	struct loopback_pcm *dpcm = rule->private;
+	struct loopback_cable *cable = dpcm->cable;
 	struct snd_interval t;
 
-        t.min = hw->channels_min;
-        t.max = hw->channels_max;
+	mutex_lock(&dpcm->loopback->cable_lock);
+	t.min = cable->hw.channels_min;
+	t.max = cable->hw.channels_max;
+	mutex_unlock(&dpcm->loopback->cable_lock);
         t.openmin = t.openmax = 0;
         t.integer = 0;
 	return snd_interval_refine(hw_param_interval(params, rule->var), &t);
@@ -718,19 +709,19 @@ static int loopback_open(struct snd_pcm_
 	/* are cached -> they do not reflect the actual state */
 	err = snd_pcm_hw_rule_add(runtime, 0,
 				  SNDRV_PCM_HW_PARAM_FORMAT,
-				  rule_format, &runtime->hw,
+				  rule_format, dpcm,
 				  SNDRV_PCM_HW_PARAM_FORMAT, -1);
 	if (err < 0)
 		goto unlock;
 	err = snd_pcm_hw_rule_add(runtime, 0,
 				  SNDRV_PCM_HW_PARAM_RATE,
-				  rule_rate, &runtime->hw,
+				  rule_rate, dpcm,
 				  SNDRV_PCM_HW_PARAM_RATE, -1);
 	if (err < 0)
 		goto unlock;
 	err = snd_pcm_hw_rule_add(runtime, 0,
 				  SNDRV_PCM_HW_PARAM_CHANNELS,
-				  rule_channels, &runtime->hw,
+				  rule_channels, dpcm,
 				  SNDRV_PCM_HW_PARAM_CHANNELS, -1);
 	if (err < 0)
 		goto unlock;
#182 lore raw

PATCH 178/254 sh_eth: fix SH7757 GEther initialization

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Sergei Shtylyov <[email protected]>

commit 5133550296d43236439494aa955bfb765a89f615 upstream.

Renesas SH7757 has 2 Fast and 2 Gigabit Ether controllers, while the 'sh_eth' driver can only reset and initialize TSU of the first controller pair. Shimoda-san tried to solve that adding the 'needs_init' member to the 'struct sh_eth_plat_data', however the platform code still never sets this flag. I think that we can infer this information from the 'devno' variable (set to 'platform_device::id') and reset/init the Ether controller pair only for an even 'devno'; therefore 'sh_eth_plat_data::needs_init' can be removed...

Fixes 150647fb2c31 ("net: sh_eth: change the condition of initialization")
---
 drivers/net/ethernet/renesas/sh_eth.c | 4 ++--
 include/linux/sh_eth.h                | 1 -
 2 files changed, 2 insertions(+), 3 deletions(-)
drivers/net/ethernet/renesas/sh_eth.c +2 -2
@@ -2902,8 +2902,8 @@ static int sh_eth_drv_probe(struct platf
 		ndev->features = NETIF_F_HW_VLAN_CTAG_FILTER;
 	}
 
-	/* initialize first or needed device */
-	if (!devno || pd->needs_init) {
+	/* Need to init only the first port of the two sharing a TSU */
+	if (devno % 2 == 0) {
 		if (mdp->cd->chip_reset)
 			mdp->cd->chip_reset(ndev);
 
include/linux/sh_eth.h +0 -1
@@ -16,7 +16,6 @@ struct sh_eth_plat_data {
 	unsigned char mac_addr[ETH_ALEN];
 	unsigned no_ether_link:1;
 	unsigned ether_link_active_low:1;
-	unsigned needs_init:1;
 };
 
 #endif
#183 lore raw

PATCH 179/254 x86/microcode/intel: Extend BDW late-loading with a revision check

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Jia Zhang <[email protected]>

commit b94b7373317164402ff7728d10f7023127a02b60 upstream.

Instead of blacklisting all model 79 CPUs when attempting a late microcode loading, limit that only to CPUs with microcode revisions < 0x0b000021 because only on those late loading may cause a system hang.

For such processors either:

a) a BIOS update which might contain a newer microcode revision

or

b) the early microcode loading method

should be considered.

Processors with revisions 0x0b000021 or higher will not experience such hangs.

For more details, see erratum BDF90 in document #334165 (Intel Xeon Processor E7-8800/4800 v4 Product Family Specification Update) from September 2017.

[ bp: Heavily massage commit message and pr_* statements. ]

Fixes 723f2828a98c ("x86/microcode/intel: Disable late loading on model 79")
Cc x86-ml <[email protected]>
[bwh: Backported to 3.16: adjust context]
---
 arch/x86/kernel/cpu/microcode/intel.c | 13 +++++++++++--
 1 file changed, 11 insertions(+), 2 deletions(-)
arch/x86/kernel/cpu/microcode/intel.c +11 -2
@@ -271,8 +271,17 @@ static bool is_blacklisted(unsigned int
 {
 	struct cpuinfo_x86 *c = &cpu_data(cpu);
 
-	if (c->x86 == 6 && c->x86_model == 0x4F) {
-		pr_err_once("late loading on model 79 is disabled.\n");
+	/*
+	 * Late loading on model 79 with microcode revision less than 0x0b000021
+	 * may result in a system hang. This behavior is documented in item
+	 * BDF90, #334165 (Intel Xeon Processor E7-8800/4800 v4 Product Family).
+	 */
+	if (c->x86 == 6 &&
+	    c->x86_model == 0x4F &&
+	    c->x86_mask == 0x01 &&
+	    c->microcode < 0x0b000021) {
+		pr_err_once("Erratum BDF90: late loading with revision < 0x0b000021 (0x%x) disabled.\n", c->microcode);
+		pr_err_once("Please consider either early loading through initrd/built-in or a potential BIOS update.\n");
 		return true;
 	}
 
#184 lore raw

PATCH 180/254 xfrm: Return error on unknown encap_type in init_state

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Herbert Xu <[email protected]>

commit bcfd09f7837f5240c30fd2f52ee7293516641faa upstream.

Currently esp will happily create an xfrm state with an unknown encap type for IPv4, without setting the necessary state parameters. This patch fixes it by returning -EINVAL.

There is a similar problem in IPv6 where if the mode is unknown we will skip initialisation while returning zero. However, this is harmless as the mode has already been checked further up the stack. This patch removes this anomaly by aligning the IPv6 behaviour with IPv4 and treating unknown modes (which cannot actually happen) as transport mode.

Fixes 38320c70d282 ("[IPSEC]: Use crypto_aead and authenc in ESP")
---
 net/ipv4/esp4.c | 1 +
 net/ipv6/esp6.c | 3 +--
 2 files changed, 2 insertions(+), 2 deletions(-)
net/ipv4/esp4.c +1 -0
@@ -657,6 +657,7 @@ static int esp_init_state(struct xfrm_st
 
 		switch (encap->encap_type) {
 		default:
+			err = -EINVAL;
 			goto error;
 		case UDP_ENCAP_ESPINUDP:
 			x->props.header_len += sizeof(struct udphdr);
net/ipv6/esp6.c +1 -2
@@ -600,13 +600,12 @@ static int esp6_init_state(struct xfrm_s
 			x->props.header_len += IPV4_BEET_PHMAXLEN +
 				               (sizeof(struct ipv6hdr) - sizeof(struct iphdr));
 		break;
+	default:
 	case XFRM_MODE_TRANSPORT:
 		break;
 	case XFRM_MODE_TUNNEL:
 		x->props.header_len += sizeof(struct ipv6hdr);
 		break;
-	default:
-		goto error;
 	}
 
 	align = ALIGN(crypto_aead_blocksize(aead), 4);
#185 lore raw

PATCH 181/254 ALSA: pcm: Abort properly at pending signal in OSS read/write loops

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Takashi Iwai <[email protected]>

commit 29159a4ed7044c52e3e2cf1a9fb55cec4745c60b upstream.

The loops for read and write in PCM OSS emulation have no proper check of pending signals, and they keep processing even after user tries to break. This results in a very long delay, often seen as RCU stall when a huge unprocessed bytes remain queued. The bug could be easily triggered by syzkaller.

As a simple workaround, this patch adds the proper check of pending signals and aborts the loop appropriately.

---
 sound/core/oss/pcm_oss.c | 8 ++++++++
 1 file changed, 8 insertions(+)
sound/core/oss/pcm_oss.c +8 -0
@@ -1417,6 +1417,10 @@ static ssize_t snd_pcm_oss_write1(struct
 			    tmp != runtime->oss.period_bytes)
 				break;
 		}
+		if (signal_pending(current)) {
+			tmp = -ERESTARTSYS;
+			goto err;
+		}
 	}
 	mutex_unlock(&runtime->oss.params_lock);
 	return xfer;
@@ -1502,6 +1506,10 @@ static ssize_t snd_pcm_oss_read1(struct
 			bytes -= tmp;
 			xfer += tmp;
 		}
+		if (signal_pending(current)) {
+			tmp = -ERESTARTSYS;
+			goto err;
+		}
 	}
 	mutex_unlock(&runtime->oss.params_lock);
 	return xfer;
#186 lore raw

PATCH 182/254 ALSA: pcm: Allow aborting mutex lock at OSS read/write loops

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Takashi Iwai <[email protected]>

commit 900498a34a3ac9c611e9b425094c8106bdd7dc1c upstream.

PCM OSS read/write loops keep taking the mutex lock for the whole read/write, and this might take very long when the exceptionally high amount of data is given. Also, since it invokes with mutex_lock(), the concurrent read/write becomes unbreakable.

This patch tries to address these issues by replacing mutex_lock() with mutex_lock_interruptible(), and also splits / re-takes the lock at each read/write period chunk, so that it can switch the context more finely if requested.

---
 sound/core/oss/pcm_oss.c | 36 +++++++++++++++++++++---------------
 1 file changed, 21 insertions(+), 15 deletions(-)
sound/core/oss/pcm_oss.c +21 -15
@@ -1370,8 +1370,11 @@ static ssize_t snd_pcm_oss_write1(struct
 
 	if ((tmp = snd_pcm_oss_make_ready(substream)) < 0)
 		return tmp;
-	mutex_lock(&runtime->oss.params_lock);
 	while (bytes > 0) {
+		if (mutex_lock_interruptible(&runtime->oss.params_lock)) {
+			tmp = -ERESTARTSYS;
+			break;
+		}
 		if (bytes < runtime->oss.period_bytes || runtime->oss.buffer_used > 0) {
 			tmp = bytes;
 			if (tmp + runtime->oss.buffer_used > runtime->oss.period_bytes)
@@ -1415,18 +1418,18 @@ static ssize_t snd_pcm_oss_write1(struct
 			xfer += tmp;
 			if ((substream->f_flags & O_NONBLOCK) != 0 &&
 			    tmp != runtime->oss.period_bytes)
-				break;
+				tmp = -EAGAIN;
 		}
+ err:
+		mutex_unlock(&runtime->oss.params_lock);
+		if (tmp < 0)
+			break;
 		if (signal_pending(current)) {
 			tmp = -ERESTARTSYS;
-			goto err;
+			break;
 		}
+		tmp = 0;
 	}
-	mutex_unlock(&runtime->oss.params_lock);
-	return xfer;
-
- err:
-	mutex_unlock(&runtime->oss.params_lock);
 	return xfer > 0 ? (snd_pcm_sframes_t)xfer : tmp;
 }
 
@@ -1474,8 +1477,11 @@ static ssize_t snd_pcm_oss_read1(struct
 
 	if ((tmp = snd_pcm_oss_make_ready(substream)) < 0)
 		return tmp;
-	mutex_lock(&runtime->oss.params_lock);
 	while (bytes > 0) {
+		if (mutex_lock_interruptible(&runtime->oss.params_lock)) {
+			tmp = -ERESTARTSYS;
+			break;
+		}
 		if (bytes < runtime->oss.period_bytes || runtime->oss.buffer_used > 0) {
 			if (runtime->oss.buffer_used == 0) {
 				tmp = snd_pcm_oss_read2(substream, runtime->oss.buffer, runtime->oss.period_bytes, 1);
@@ -1506,16 +1512,16 @@ static ssize_t snd_pcm_oss_read1(struct
 			bytes -= tmp;
 			xfer += tmp;
 		}
+ err:
+		mutex_unlock(&runtime->oss.params_lock);
+		if (tmp < 0)
+			break;
 		if (signal_pending(current)) {
 			tmp = -ERESTARTSYS;
-			goto err;
+			break;
 		}
+		tmp = 0;
 	}
-	mutex_unlock(&runtime->oss.params_lock);
-	return xfer;
-
- err:
-	mutex_unlock(&runtime->oss.params_lock);
 	return xfer > 0 ? (snd_pcm_sframes_t)xfer : tmp;
 }
 
#187 lore raw

PATCH 183/254 mdio-sun4i: Fix a memory leak

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Christophe JAILLET <[email protected]>

commit 56c0290202ab94a2f2780c449395d4ae8495fab4 upstream.

If the probing of the regulator is deferred, the memory allocated by 'mdiobus_alloc_size()' will be leaking.
It should be freed before the next call to 'sun4i_mdio_probe()' which will reallocate it.

Fixes 4bdcb1dd9feb ("net: Add MDIO bus driver for the Allwinner EMAC")
---
 drivers/net/phy/mdio-sun4i.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/net/phy/mdio-sun4i.c +4 -2
@@ -128,8 +128,10 @@ static int sun4i_mdio_probe(struct platf
 
 	data->regulator = devm_regulator_get(&pdev->dev, "phy");
 	if (IS_ERR(data->regulator)) {
-		if (PTR_ERR(data->regulator) == -EPROBE_DEFER)
-			return -EPROBE_DEFER;
+		if (PTR_ERR(data->regulator) == -EPROBE_DEFER) {
+			ret = -EPROBE_DEFER;
+			goto err_out_free_mdiobus;
+		}
 
 		dev_info(&pdev->dev, "no regulator found\n");
 		data->regulator = NULL;
#188 lore raw

PATCH 184/254 sh_eth: fix TXALCR1 offsets

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Sergei Shtylyov <[email protected]>

commit 50f3d740d376f664f6accc7e86c9afd8f1c7e1e4 upstream.

The TXALCR1 offsets are incorrect in the register offset tables, most probably due to copy&paste error. Luckily, the driver never uses this register. :-)

Fixes 4a55530f38e4 ("net: sh_eth: modify the definitions of register")
---
 drivers/net/ethernet/renesas/sh_eth.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/net/ethernet/renesas/sh_eth.c +2 -2
@@ -142,7 +142,7 @@ static const u16 sh_eth_offset_gigabit[S
 	[FWNLCR0]	= 0x0090,
 	[FWALCR0]	= 0x0094,
 	[TXNLCR1]	= 0x00a0,
-	[TXALCR1]	= 0x00a0,
+	[TXALCR1]	= 0x00a4,
 	[RXNLCR1]	= 0x00a8,
 	[RXALCR1]	= 0x00ac,
 	[FWNLCR1]	= 0x00b0,
@@ -384,7 +384,7 @@ static const u16 sh_eth_offset_fast_sh3_
 	[FWNLCR0]	= 0x0090,
 	[FWALCR0]	= 0x0094,
 	[TXNLCR1]	= 0x00a0,
-	[TXALCR1]	= 0x00a0,
+	[TXALCR1]	= 0x00a4,
 	[RXNLCR1]	= 0x00a8,
 	[RXALCR1]	= 0x00ac,
 	[FWNLCR1]	= 0x00b0,
#189 lore raw

PATCH 185/254 Input: twl4030-vibra - fix ERROR: Bad of_node_put() warning

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Marek Belisko <[email protected]>

commit e661d0a04462dd98667f8947141bd8defab5b34a upstream.

Fix following:

[    8.862274] ERROR: Bad of_node_put() on /ocp/i2c@48070000/twl@48/audio
[    8.869293] CPU: 0 PID: 1003 Comm: modprobe Not tainted 4.2.0-rc2-letux+ #1175
[    8.876922] Hardware name: Generic OMAP36xx (Flattened Device Tree)
[    8.883514] [<c00159e0>] (unwind_backtrace) from [<c0012488>] (show_stack+0x10/0x14)
[    8.891693] [<c0012488>] (show_stack) from [<c05cb810>] (dump_stack+0x78/0x94)
[    8.899322] [<c05cb810>] (dump_stack) from [<c02cfd5c>] (kobject_release+0x68/0x7c)
[    8.907409] [<c02cfd5c>] (kobject_release) from [<bf0040c4>] (twl4030_vibra_probe+0x74/0x188 [twl4030_vibra])
[    8.917877] [<bf0040c4>] (twl4030_vibra_probe [twl4030_vibra]) from [<c03816ac>] (platform_drv_probe+0x48/0x90)
[    8.928497] [<c03816ac>] (platform_drv_probe) from [<c037feb4>] (really_probe+0xd4/0x238)
[    8.937103] [<c037feb4>] (really_probe) from [<c0380160>] (driver_probe_device+0x30/0x48)
[    8.945678] [<c0380160>] (driver_probe_device) from [<c03801e0>] (__driver_attach+0x68/0x8c)
[    8.954589] [<c03801e0>] (__driver_attach) from [<c037ea60>] (bus_for_each_dev+0x50/0x84)
[    8.963226] [<c037ea60>] (bus_for_each_dev) from [<c037f828>] (bus_add_driver+0xcc/0x1e4)
[    8.971832] [<c037f828>] (bus_add_driver) from [<c0380b60>] (driver_register+0x9c/0xe0)
[    8.980255] [<c0380b60>] (driver_register) from [<c00097e0>] (do_one_initcall+0x100/0x1b8)
[    8.988983] [<c00097e0>] (do_one_initcall) from [<c00b8008>] (do_init_module+0x58/0x1c0)
[    8.997497] [<c00b8008>] (do_init_module) from [<c00b8cac>] (SyS_init_module+0x54/0x64)
[    9.005950] [<c00b8cac>] (SyS_init_module) from [<c000ed20>] (ret_fast_syscall+0x0/0x54)
[    9.015838] input: twl4030:vibrator as /devices/platform/68000000.ocp/48070000.i2c/i2c-0/0-0048/48070000.i2c:twl@48:audio/input/input2

node passed to of_find_node_by_name is put inside that function and new node is returned if found. Free returned node not already freed node.

---
 drivers/input/misc/twl4030-vibra.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/input/misc/twl4030-vibra.c +2 -1
@@ -185,7 +185,8 @@ static bool twl4030_vibra_check_coexist(
 	if (pdata && pdata->coexist)
 		return true;
 
-	if (of_find_node_by_name(node, "codec")) {
+	node = of_find_node_by_name(node, "codec");
+	if (node) {
 		of_node_put(node);
 		return true;
 	}
#190 lore raw

PATCH 186/254 Input: twl4030-vibra - fix sibling-node lookup

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Johan Hovold <[email protected]>

commit 5b189201993ab03001a398de731045bfea90c689 upstream.

A helper purported to look up a child node based on its name was using the wrong of-helper and ended up prematurely freeing the parent of-node while searching the whole device tree depth-first starting at the parent node.

Fixes 64b9e4d803b1 ("input: twl4030-vibra: Support for DT booted kernel")
Fixes e661d0a04462 ("Input: twl4030-vibra - fix ERROR: Bad of_node_put() warning")
---
 drivers/input/misc/twl4030-vibra.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/input/misc/twl4030-vibra.c +4 -2
@@ -180,12 +180,14 @@ static SIMPLE_DEV_PM_OPS(twl4030_vibra_p
 			 twl4030_vibra_suspend, twl4030_vibra_resume);
 
 static bool twl4030_vibra_check_coexist(struct twl4030_vibra_data *pdata,
-			      struct device_node *node)
+			      struct device_node *parent)
 {
+	struct device_node *node;
+
 	if (pdata && pdata->coexist)
 		return true;
 
-	node = of_find_node_by_name(node, "codec");
+	node = of_get_child_by_name(parent, "codec");
 	if (node) {
 		of_node_put(node);
 		return true;
#191 lore raw

PATCH 187/254 Input: twl6040-vibra - fix DT node memory management

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: "H. Nikolaus Schaller" <[email protected]>

commit c52c545ead97fcc2f4f8ea38f1ae3c23211e09a8 upstream.

commit e7ec014a47e4 ("Input: twl6040-vibra - update for device tree support")

made the separate vibra DT node to a subnode of the twl6040.

It now calls of_find_node_by_name() to locate the "vibra" subnode. This function has a side effect to call of_node_put on() for the twl6040 parent node passed in as a parameter. This causes trouble later on.

Solution: we must call of_node_get() before of_find_node_by_name()

---
 drivers/input/misc/twl6040-vibra.c | 1 +
 1 file changed, 1 insertion(+)
drivers/input/misc/twl6040-vibra.c +1 -0
@@ -264,6 +264,7 @@ static int twl6040_vibra_probe(struct pl
 	int vddvibr_uV = 0;
 	int error;
 
+	of_node_get(twl6040_core_dev->of_node);
 	twl6040_core_node = of_find_node_by_name(twl6040_core_dev->of_node,
 						 "vibra");
 	if (!twl6040_core_node) {
#192 lore raw

PATCH 188/254 Input: twl6040-vibra - fix child-node lookup

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Johan Hovold <[email protected]>

commit dcaf12a8b0bbdbfcfa2be8dff2c4948d9844b4ad upstream.

Fix child-node lookup during probe, which ended up searching the whole device tree depth-first starting at parent rather than just matching on its children.

Later sanity checks on node properties (which would likely be missing) should prevent this from causing much trouble however, especially as the original premature free of the parent node has already been fixed separately (but that "fix" was apparently never backported to stable).

Fixes e7ec014a47e4 ("Input: twl6040-vibra - update for device tree support")
Fixes c52c545ead97 ("Input: twl6040-vibra - fix DT node memory management")
---
 drivers/input/misc/twl6040-vibra.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
drivers/input/misc/twl6040-vibra.c +1 -2
@@ -264,8 +264,7 @@ static int twl6040_vibra_probe(struct pl
 	int vddvibr_uV = 0;
 	int error;
 
-	of_node_get(twl6040_core_dev->of_node);
-	twl6040_core_node = of_find_node_by_name(twl6040_core_dev->of_node,
+	twl6040_core_node = of_get_child_by_name(twl6040_core_dev->of_node,
 						 "vibra");
 	if (!twl6040_core_node) {
 		dev_err(&pdev->dev, "parent of node is missing?\n");
#193 lore raw

PATCH 189/254 Input: 88pm860x-ts - fix child-node lookup

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Johan Hovold <[email protected]>

commit 906bf7daa0618d0ef39f4872ca42218c29a3631f upstream.

Fix child node-lookup during probe, which ended up searching the whole device tree depth-first starting at parent rather than just matching on its children.

To make things worse, the parent node was prematurely freed, while the child node was leaked.

Fixes 2e57d56747e6 ("mfd: 88pm860x: Device tree support")
---
 drivers/input/touchscreen/88pm860x-ts.c | 16 ++++++++++++----
 1 file changed, 12 insertions(+), 4 deletions(-)
drivers/input/touchscreen/88pm860x-ts.c +12 -4
@@ -126,7 +126,7 @@ static int pm860x_touch_dt_init(struct p
 	int data, n, ret;
 	if (!np)
 		return -ENODEV;
-	np = of_find_node_by_name(np, "touch");
+	np = of_get_child_by_name(np, "touch");
 	if (!np) {
 		dev_err(&pdev->dev, "Can't find touch node\n");
 		return -EINVAL;
@@ -144,13 +144,13 @@ static int pm860x_touch_dt_init(struct p
 	if (data) {
 		ret = pm860x_reg_write(i2c, PM8607_GPADC_MISC1, data);
 		if (ret < 0)
-			return -EINVAL;
+			goto err_put_node;
 	}
 	/* set tsi prebias time */
 	if (!of_property_read_u32(np, "marvell,88pm860x-tsi-prebias", &data)) {
 		ret = pm860x_reg_write(i2c, PM8607_TSI_PREBIAS, data);
 		if (ret < 0)
-			return -EINVAL;
+			goto err_put_node;
 	}
 	/* set prebias & prechg time of pen detect */
 	data = 0;
@@ -161,10 +161,18 @@ static int pm860x_touch_dt_init(struct p
 	if (data) {
 		ret = pm860x_reg_write(i2c, PM8607_PD_PREBIAS, data);
 		if (ret < 0)
-			return -EINVAL;
+			goto err_put_node;
 	}
 	of_property_read_u32(np, "marvell,88pm860x-resistor-X", res_x);
+
+	of_node_put(np);
+
 	return 0;
+
+err_put_node:
+	of_node_put(np);
+
+	return -EINVAL;
 }
 #else
 #define pm860x_touch_dt_init(x, y, z)	(-1)
#194 lore raw

PATCH 190/254 USB: fix usbmon BUG trigger

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Pete Zaitcev <[email protected]>

commit 46eb14a6e1585d99c1b9f58d0e7389082a5f466b upstream.

Automated tests triggered this by opening usbmon and accessing the mmap while simultaneously resizing the buffers. This bug was with us since 2006, because typically applications only size the buffers once and thus avoid racing. Reported by Kirill A. Shutemov.

---
 drivers/usb/mon/mon_bin.c | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)
drivers/usb/mon/mon_bin.c +7 -1
@@ -1000,7 +1000,9 @@ static long mon_bin_ioctl(struct file *f
 		break;
 
 	case MON_IOCQ_RING_SIZE:
+		mutex_lock(&rp->fetch_lock);
 		ret = rp->b_size;
+		mutex_unlock(&rp->fetch_lock);
 		break;
 
 	case MON_IOCT_RING_SIZE:
@@ -1227,12 +1229,16 @@ static int mon_bin_vma_fault(struct vm_a
 	unsigned long offset, chunk_idx;
 	struct page *pageptr;
 
+	mutex_lock(&rp->fetch_lock);
 	offset = vmf->pgoff << PAGE_SHIFT;
-	if (offset >= rp->b_size)
+	if (offset >= rp->b_size) {
+		mutex_unlock(&rp->fetch_lock);
 		return VM_FAULT_SIGBUS;
+	}
 	chunk_idx = offset / CHUNK_SIZE;
 	pageptr = rp->b_vec[chunk_idx].pg;
 	get_page(pageptr);
+	mutex_unlock(&rp->fetch_lock);
 	vmf->page = pageptr;
 	return 0;
 }
#195 lore raw

PATCH 191/254 usb: udc: core: add device_del() call to error pathway

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Alan Stern <[email protected]>

commit c93e64e91248becd0edb8f01723dff9da890e2ab upstream.

This patch fixes a bug in the error pathway of usb_add_gadget_udc_release() in udc-core.c. If the udc registration fails, the gadget registration is not fully undone; there's a put_device(&gadget->dev) call but no device_del().

[bwh: Backported to 3.16: adjust filename]
---
 drivers/usb/gadget/udc-core.c | 1 +
 1 file changed, 1 insertion(+)
drivers/usb/gadget/udc-core.c +1 -0
@@ -252,6 +252,7 @@ err4:
 
 err3:
 	put_device(&udc->dev);
+	device_del(&gadget->dev);
 
 err2:
 	put_device(&gadget->dev);
#196 lore raw

PATCH 192/254 USB: Gadget core: fix inconsistency in the interface tousb_add_gadget_udc_release()

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Alan Stern <[email protected]>

commit afd7fd81f22bf90474216515dbd6088f9bd70343 upstream.

The usb_add_gadget_udc_release() routine in the USB gadget core will sometimes but not always call the gadget's release function when an error occurs. More specifically, if the struct usb_udc allocation fails then the release function is not called, and for other errors it is.

As a result, users of this routine cannot know whether they need to deallocate the memory containing the gadget structure following an error. This leads to unavoidable memory leaks or double frees.

This patch fixes the problem by splitting the existing device_register() call into device_initialize() and device_add(), and doing the udc allocation in between. That way, even if the allocation fails it is still possible to call device_del(), and so the release function will be always called following an error.

[bwh: Backported to 3.16: adjust filename]
---
 drivers/usb/gadget/udc-core.c | 15 +++++++++------
 1 file changed, 9 insertions(+), 6 deletions(-)
drivers/usb/gadget/udc-core.c +9 -6
@@ -192,6 +192,7 @@ static void usb_udc_nop_release(struct d
  * @release: a gadget release function.
  *
  * Returns zero on success, negative errno otherwise.
+ * Calls the gadget release function in the latter case.
  */
 int usb_add_gadget_udc_release(struct device *parent, struct usb_gadget *gadget,
 		void (*release)(struct device *dev))
@@ -199,10 +200,6 @@ int usb_add_gadget_udc_release(struct de
 	struct usb_udc		*udc;
 	int			ret = -ENOMEM;
 
-	udc = kzalloc(sizeof(*udc), GFP_KERNEL);
-	if (!udc)
-		goto err1;
-
 	dev_set_name(&gadget->dev, "gadget");
 	INIT_WORK(&gadget->work, usb_gadget_state_work);
 	gadget->dev.parent = parent;
@@ -218,7 +215,13 @@ int usb_add_gadget_udc_release(struct de
 	else
 		gadget->dev.release = usb_udc_nop_release;
 
-	ret = device_register(&gadget->dev);
+	device_initialize(&gadget->dev);
+
+	udc = kzalloc(sizeof(*udc), GFP_KERNEL);
+	if (!udc)
+		goto err1;
+
+	ret = device_add(&gadget->dev);
 	if (ret)
 		goto err2;
 
@@ -255,10 +258,10 @@ err3:
 	device_del(&gadget->dev);
 
 err2:
-	put_device(&gadget->dev);
 	kfree(udc);
 
 err1:
+	put_device(&gadget->dev);
 	return ret;
 }
 EXPORT_SYMBOL_GPL(usb_add_gadget_udc_release);
#197 lore raw

PATCH 193/254 USB: UDC core: fix double-free in usb_add_gadget_udc_release

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Alan Stern <[email protected]>

commit 7ae2c3c280db183ca9ada2675c34ec2f7378abfa upstream.

The error-handling pathways in usb_add_gadget_udc_release() are messed up. Aside from the uninformative statement labels, they can deallocate the udc structure after calling put_device(), which is a double-free. This was observed by KASAN in automatic testing.

This patch cleans up the routine. It preserves the requirement that when any failure occurs, we call put_device(&gadget->dev).

[bwh: Backported to 3.16:
- The err5 path is not presnet
- Adjust filename]

---

drivers/usb/gadget/udc-core.c +11 -13
@@ -219,11 +219,7 @@ int usb_add_gadget_udc_release(struct de
 
 	udc = kzalloc(sizeof(*udc), GFP_KERNEL);
 	if (!udc)
-		goto err1;
-
-	ret = device_add(&gadget->dev);
-	if (ret)
-		goto err2;
+		goto err_put_gadget;
 
 	device_initialize(&udc->dev);
 	udc->dev.release = usb_udc_release;
@@ -232,7 +228,11 @@ int usb_add_gadget_udc_release(struct de
 	udc->dev.parent = parent;
 	ret = dev_set_name(&udc->dev, "%s", kobject_name(&parent->kobj));
 	if (ret)
-		goto err3;
+		goto err_put_udc;
+
+	ret = device_add(&gadget->dev);
+	if (ret)
+		goto err_put_udc;
 
 	udc->gadget = gadget;
 
@@ -241,7 +241,7 @@ int usb_add_gadget_udc_release(struct de
 
 	ret = device_add(&udc->dev);
 	if (ret)
-		goto err4;
+		goto err_unlist_udc;
 
 	usb_gadget_set_state(gadget, USB_STATE_NOTATTACHED);
 
@@ -249,18 +249,16 @@ int usb_add_gadget_udc_release(struct de
 
 	return 0;
 
-err4:
+ err_unlist_udc:
 	list_del(&udc->list);
 	mutex_unlock(&udc_lock);
 
-err3:
-	put_device(&udc->dev);
 	device_del(&gadget->dev);
 
-err2:
-	kfree(udc);
+ err_put_udc:
+	put_device(&udc->dev);
 
-err1:
+ err_put_gadget:
 	put_device(&gadget->dev);
 	return ret;
 }
#198 lore raw

PATCH 194/254 uas: ignore UAS for Norelsys NS1068(X) chips

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Icenowy Zheng <[email protected]>

commit 928afc85270753657b5543e052cc270c279a3fe9 upstream.

The UAS mode of Norelsys NS1068(X) is reported to fail to work on several platforms with the following error message:

xhci-hcd xhci-hcd.0.auto: ERROR Transfer event for unknown stream ring slot 1 ep 8 xhci-hcd xhci-hcd.0.auto: @00000000bf04a400 00000000 00000000 1b000000 01098001

And when trying to mount a partition on the disk the disk will disconnect from the USB controller, then after re-connecting the device will be offlined and not working at all.

Falling back to USB mass storage can solve this problem, so ignore UAS function of this chip.

---
 drivers/usb/storage/unusual_uas.h | 7 +++++++
 1 file changed, 7 insertions(+)
drivers/usb/storage/unusual_uas.h +7 -0
@@ -145,6 +145,13 @@ UNUSUAL_DEV(0x2109, 0x0711, 0x0000, 0x99
 		USB_SC_DEVICE, USB_PR_DEVICE, NULL,
 		US_FL_NO_ATA_1X),
 
+/* Reported-by: Icenowy Zheng <[email protected]> */
+UNUSUAL_DEV(0x2537, 0x1068, 0x0000, 0x9999,
+		"Norelsys",
+		"NS1068X",
+		USB_SC_DEVICE, USB_PR_DEVICE, NULL,
+		US_FL_IGNORE_UAS),
+
 /* Reported-by: Takeo Nakayama <[email protected]> */
 UNUSUAL_DEV(0x357d, 0x7788, 0x0000, 0x9999,
 		"JMicron",
#199 lore raw

PATCH 195/254 net: ipv4: emulate READ_ONCE() on ->hdrincl bit-field in raw_sendmsg()

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Nicolai Stange <[email protected]>

commit 20b50d79974ea3192e8c3ab7faf4e536e5f14d8f upstream.

Commit 8f659a03a0ba ("net: ipv4: fix for a race condition in raw_sendmsg") fixed the issue of possibly inconsistent ->hdrincl handling due to concurrent updates by reading this bit-field member into a local variable and using the thus stabilized value in subsequent tests.

However, aforementioned commit also adds the (correct) comment that

/* hdrincl should be READ_ONCE(inet->hdrincl)
* but READ_ONCE() doesn't work with bit fields
*/

because as it stands, the compiler is free to shortcut or even eliminate the local variable at its will.

Note that I have not seen anything like this happening in reality and thus, the concern is a theoretical one.

However, in order to be on the safe side, emulate a READ_ONCE() on the bit-field by doing it on the local 'hdrincl' variable itself:

	int hdrincl = inet->hdrincl;
	hdrincl = READ_ONCE(hdrincl);

This breaks the chain in the sense that the compiler is not allowed to replace subsequent reads from hdrincl with reloads from inet->hdrincl.

Fixes 8f659a03a0ba ("net: ipv4: fix for a race condition in raw_sendmsg")
[bwh: Backported to 3.16: use ACCESS_ONCE()]
---
 net/ipv4/raw.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
net/ipv4/raw.c +3 -1
@@ -497,9 +497,11 @@ static int raw_sendmsg(struct kiocb *ioc
 		goto out;
 
 	/* hdrincl should be READ_ONCE(inet->hdrincl)
-	 * but READ_ONCE() doesn't work with bit fields
+	 * but READ_ONCE() doesn't work with bit fields.
+	 * Doing this indirectly yields the same result.
 	 */
 	hdrincl = inet->hdrincl;
+	hdrincl = ACCESS_ONCE(hdrincl);
 	/*
 	 *	Check the flags.
 	 */
#200 lore raw

PATCH 196/254 SolutionEngine771x: fix Ether platform data

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Sergei Shtylyov <[email protected]>

commit 195e2addbce09e5afbc766efc1e6567c9ce840d3 upstream.

The 'sh_eth' driver's probe() method would fail on the SolutionEngine7710 board and crash on SolutionEngine7712 board as the platform code is hopelessly behind the driver's platform data -- it passes the PHY address instead of 'struct sh_eth_plat_data *'; pass the latter to the driver in order to fix the bug...

Fixes 71557a37adb5 ("[netdrvr] sh_eth: Add SH7619 support")
---
 arch/sh/boards/mach-se/770x/setup.c | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)
arch/sh/boards/mach-se/770x/setup.c +8 -2
@@ -8,6 +8,7 @@
  */
 #include <linux/init.h>
 #include <linux/platform_device.h>
+#include <linux/sh_eth.h>
 #include <mach-se/mach/se.h>
 #include <mach-se/mach/mrshpc.h>
 #include <asm/machvec.h>
@@ -114,6 +115,11 @@ static struct platform_device heartbeat_
 #if defined(CONFIG_CPU_SUBTYPE_SH7710) ||\
 	defined(CONFIG_CPU_SUBTYPE_SH7712)
 /* SH771X Ethernet driver */
+static struct sh_eth_plat_data sh_eth_plat = {
+	.phy = PHY_ID,
+	.phy_interface = PHY_INTERFACE_MODE_MII,
+};
+
 static struct resource sh_eth0_resources[] = {
 	[0] = {
 		.start = SH_ETH0_BASE,
@@ -131,7 +137,7 @@ static struct platform_device sh_eth0_de
 	.name = "sh771x-ether",
 	.id = 0,
 	.dev = {
-		.platform_data = PHY_ID,
+		.platform_data = &sh_eth_plat,
 	},
 	.num_resources = ARRAY_SIZE(sh_eth0_resources),
 	.resource = sh_eth0_resources,
@@ -154,7 +160,7 @@ static struct platform_device sh_eth1_de
 	.name = "sh771x-ether",
 	.id = 1,
 	.dev = {
-		.platform_data = PHY_ID,
+		.platform_data = &sh_eth_plat,
 	},
 	.num_resources = ARRAY_SIZE(sh_eth1_resources),
 	.resource = sh_eth1_resources,
#201 lore raw

PATCH 197/254 SolutionEngine771x: add Ether TSU resource

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Sergei Shtylyov <[email protected]>

commit f9a531d6731d74f1e24298d9641c2dc1fef2631b upstream.

After the Ether platform data is fixed, the driver probe() method would still fail since the 'struct sh_eth_cpu_data' corresponding to SH771x indicates the presence of TSU but the memory resource for it is absent. Add the missing TSU resource to both Ether devices and fix the harmless off-by-one error in the main memory resources, while at it...

Fixes 4986b996882d ("net: sh_eth: remove the SH_TSU_ADDR")
---
 arch/sh/boards/mach-se/770x/setup.c | 14 ++++++++++++--
 arch/sh/include/mach-se/mach/se.h   |  1 +
 2 files changed, 13 insertions(+), 2 deletions(-)
arch/sh/boards/mach-se/770x/setup.c +12 -2
@@ -123,10 +123,15 @@ static struct sh_eth_plat_data sh_eth_pl
 static struct resource sh_eth0_resources[] = {
 	[0] = {
 		.start = SH_ETH0_BASE,
-		.end = SH_ETH0_BASE + 0x1B8,
+		.end = SH_ETH0_BASE + 0x1B8 - 1,
 		.flags = IORESOURCE_MEM,
 	},
 	[1] = {
+		.start = SH_TSU_BASE,
+		.end = SH_TSU_BASE + 0x200 - 1,
+		.flags = IORESOURCE_MEM,
+	},
+	[2] = {
 		.start = SH_ETH0_IRQ,
 		.end = SH_ETH0_IRQ,
 		.flags = IORESOURCE_IRQ,
@@ -146,10 +151,15 @@ static struct platform_device sh_eth0_de
 static struct resource sh_eth1_resources[] = {
 	[0] = {
 		.start = SH_ETH1_BASE,
-		.end = SH_ETH1_BASE + 0x1B8,
+		.end = SH_ETH1_BASE + 0x1B8 - 1,
 		.flags = IORESOURCE_MEM,
 	},
 	[1] = {
+		.start = SH_TSU_BASE,
+		.end = SH_TSU_BASE + 0x200 - 1,
+		.flags = IORESOURCE_MEM,
+	},
+	[2] = {
 		.start = SH_ETH1_IRQ,
 		.end = SH_ETH1_IRQ,
 		.flags = IORESOURCE_IRQ,
arch/sh/include/mach-se/mach/se.h +1 -0
@@ -99,6 +99,7 @@
 /* Base address */
 #define SH_ETH0_BASE 0xA7000000
 #define SH_ETH1_BASE 0xA7000400
+#define SH_TSU_BASE  0xA7000800
 /* PHY ID */
 #if defined(CONFIG_CPU_SUBTYPE_SH7710)
 # define PHY_ID 0x00
#202 lore raw

PATCH 198/254 ALSA: hda - Apply the existing quirk to iMac 14,1

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Takashi Iwai <[email protected]>

commit 031f335cda879450095873003abb03ae8ed3b74a upstream.

iMac 14,1 requires the same quirk as iMac 12,2, using GPIO 2 and 3 for headphone and speaker output amps. Add the codec SSID quirk entry (106b:0600) accordingly.

---
 sound/pci/hda/patch_cirrus.c | 1 +
 1 file changed, 1 insertion(+)
sound/pci/hda/patch_cirrus.c +1 -0
@@ -396,6 +396,7 @@ static const struct snd_pci_quirk cs420x
 	/*SND_PCI_QUIRK(0x8086, 0x7270, "IMac 27 Inch", CS420X_IMAC27),*/
 
 	/* codec SSID */
+	SND_PCI_QUIRK(0x106b, 0x0600, "iMac 14,1", CS420X_IMAC27_122),
 	SND_PCI_QUIRK(0x106b, 0x1c00, "MacBookPro 8,1", CS420X_MBP81),
 	SND_PCI_QUIRK(0x106b, 0x2000, "iMac 12,2", CS420X_IMAC27_122),
 	SND_PCI_QUIRK(0x106b, 0x2800, "MacBookPro 10,1", CS420X_MBP101),
#203 lore raw

PATCH 199/254 powerpc: Don't preempt_disable() in show_cpuinfo()

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Benjamin Herrenschmidt <[email protected]>

commit 349524bc0da698ec77f2057cf4a4948eb6349265 upstream.

This causes warnings from cpufreq mutex code. This is also rather unnecessary and ineffective. If we really want to prevent concurrent unplug, we could take the unplug read lock but I don't see this being critical.

Fixes cd77b5ce208c ("powerpc/powernv/cpufreq: Fix the frequency read by /proc/cpuinfo")
---
 arch/powerpc/kernel/setup-common.c | 11 -----------
 1 file changed, 11 deletions(-)
arch/powerpc/kernel/setup-common.c +0 -11
@@ -216,14 +216,6 @@ static int show_cpuinfo(struct seq_file
 	unsigned short maj;
 	unsigned short min;
 
-	/* We only show online cpus: disable preempt (overzealous, I
-	 * knew) to prevent cpu going down. */
-	preempt_disable();
-	if (!cpu_online(cpu_id)) {
-		preempt_enable();
-		return 0;
-	}
-
 #ifdef CONFIG_SMP
 	pvr = per_cpu(cpu_pvr, cpu_id);
 #else
@@ -328,9 +320,6 @@ static int show_cpuinfo(struct seq_file
 #ifdef CONFIG_SMP
 	seq_printf(m, "\n");
 #endif
-
-	preempt_enable();
-
 	/* If this is the last cpu, print the summary */
 	if (cpumask_next(cpu_id, cpu_online_mask) >= nr_cpu_ids)
 		show_cpuinfo_summary(m);
#204 lore raw

PATCH 200/254 8021q: fix a memory leak for VLAN 0 device

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Cong Wang <[email protected]>

commit 78bbb15f2239bc8e663aa20bbe1987c91a0b75f6 upstream.

A vlan device with vid 0 is allow to creat by not able to be fully cleaned up by unregister_vlan_dev() which checks for vlan_id!=0.

Also, VLAN 0 is probably not a valid number and it is kinda "reserved" for HW accelerating devices, but it is probably too late to reject it from creation even if makes sense. Instead, just remove the check in unregister_vlan_dev().

Fixes ad1afb003939 ("vlan_dev: VLAN 0 should be treated as "no vlan tag" (802.1p packet)")
Cc Vlad Yasevich <[email protected]>
Cc Ben Hutchings <[email protected]>
---
 net/8021q/vlan.c | 7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)
net/8021q/vlan.c +1 -6
@@ -111,12 +111,7 @@ void unregister_vlan_dev(struct net_devi
 		vlan_gvrp_uninit_applicant(real_dev);
 	}
 
-	/* Take it out of our own structures, but be sure to interlock with
-	 * HW accelerating devices or SW vlan input packet processing if
-	 * VLAN is not 0 (leave it there for 802.1p).
-	 */
-	if (vlan_id)
-		vlan_vid_del(real_dev, vlan->vlan_proto, vlan_id);
+	vlan_vid_del(real_dev, vlan->vlan_proto, vlan_id);
 
 	/* Get rid of the vlan's reference to real_dev */
 	dev_put(real_dev);
#205 lore raw

PATCH 201/254 ALSA: pcm: Remove yet superfluous WARN_ON()

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Takashi Iwai <[email protected]>

commit 23b19b7b50fe1867da8d431eea9cd3e4b6328c2c upstream.

muldiv32() contains a snd_BUG_ON() (which is morphed as WARN_ON() with debug option) for checking the case of 0 / 0. This would be helpful if this happens only as a logical error; however, since the hw refine is performed with any data set provided by user, the inconsistent values that can trigger such a condition might be passed easily. Actually, syzbot caught this by passing some zero'ed old hw_params ioctl.

So, having snd_BUG_ON() there is simply superfluous and rather harmful to give unnecessary confusions. Let's get rid of it.

---
 sound/core/pcm_lib.c | 1 -
 1 file changed, 1 deletion(-)
sound/core/pcm_lib.c +0 -1
@@ -644,7 +644,6 @@ static inline unsigned int muldiv32(unsi
 {
 	u_int64_t n = (u_int64_t) a * b;
 	if (c == 0) {
-		snd_BUG_ON(!n);
 		*r = 0;
 		return UINT_MAX;
 	}
#206 lore raw

PATCH 202/254 KVM: x86: Add memory barrier on vmcs field lookup

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Andrew Honig <[email protected]>

commit 75f139aaf896d6fdeec2e468ddfa4b2fe469bf40 upstream.

This adds a memory barrier when performing a lookup into the vmcs_field_to_offset_table. This is related to CVE-2017-5753.

[bwh: Backported to 3.16: adjust context]

---

arch/x86/kvm/vmx.c +11 -1
@@ -698,8 +698,18 @@ static const int max_vmcs_field = ARRAY_
 
 static inline short vmcs_field_to_offset(unsigned long field)
 {
-	if (field >= max_vmcs_field || vmcs_field_to_offset_table[field] == 0)
+	if (field >= max_vmcs_field)
 		return -1;
+
+	/*
+	 * FIXME: Mitigation for CVE-2017-5753.  To be replaced with a
+	 * generic mechanism.
+	 */
+	asm("lfence");
+
+	if (vmcs_field_to_offset_table[field] == 0)
+		return -1;
+
 	return vmcs_field_to_offset_table[field];
 }
 
#207 lore raw

PATCH 203/254 usb: misc: usb3503: make sure reset is low for at least 100us

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Stefan Agner <[email protected]>

commit b8626f1dc29d3eee444bfaa92146ec7b291ef41c upstream.

When using a GPIO which is high by default, and initialize the driver in USB Hub mode, initialization fails with:

  [  111.757794] usb3503 0-0008: SP_ILOCK failed (-5)

The reason seems to be that the chip is not properly reset. Probe does initialize reset low, however some lines later the code already set it back high, which is not long enouth.

Make sure reset is asserted for at least 100us by inserting a delay after initializing the reset pin during probe.

---
 drivers/usb/misc/usb3503.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/usb/misc/usb3503.c +2 -0
@@ -292,6 +292,8 @@ static int usb3503_probe(struct usb3503
 	if (gpio_is_valid(hub->gpio_reset)) {
 		err = devm_gpio_request_one(dev, hub->gpio_reset,
 				GPIOF_OUT_INIT_LOW, "usb3503 reset");
+		/* Datasheet defines a hardware reset to be at least 100us */
+		usleep_range(100, 10000);
 		if (err) {
 			dev_err(dev,
 				"unable to request GPIO %d as reset pin (%d)\n",
#208 lore raw

PATCH 204/254 futex: Prevent overflow by strengthen input validation

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Li Jinyue <[email protected]>

commit fbe0e839d1e22d88810f3ee3e2f1479be4c0aa4a upstream.

UBSAN reports signed integer overflow in kernel/futex.c:

 UBSAN: Undefined behaviour in kernel/futex.c:2041:18
 signed integer overflow:
 0 - -2147483648 cannot be represented in type 'int'

Add a sanity check to catch negative values of nr_wake and nr_requeue.

---
 kernel/futex.c | 3 +++
 1 file changed, 3 insertions(+)
kernel/futex.c +3 -0
@@ -1531,6 +1531,9 @@ static int futex_requeue(u32 __user *uad
 	struct futex_hash_bucket *hb1, *hb2;
 	struct futex_q *this, *next;
 
+	if (nr_wake < 0 || nr_requeue < 0)
+		return -EINVAL;
+
 	if (requeue_pi) {
 		/*
 		 * Requeue PI only works on two distinct uaddrs. This
#209 lore raw

PATCH 205/254 nl80211: take RCU read lock when calling ieee80211_bss_get_ie()

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Dominik Brodowski <[email protected]>

commit 7a94b8c2eee7083ddccd0515830f8c81a8e44b1a upstream.

As ieee80211_bss_get_ie() derefences an RCU to return ssid_ie, both the call to this function and any operation on this variable need protection by the RCU read lock.

Fixes 44905265bc15 ("nl80211: don't expose wdev->ssid for most interfaces")
---
 net/wireless/nl80211.c | 11 +++++++----
 1 file changed, 7 insertions(+), 4 deletions(-)
net/wireless/nl80211.c +7 -4
@@ -2354,12 +2354,13 @@ static int nl80211_send_iface(struct sk_
 		const u8 *ssid_ie;
 		if (!wdev->current_bss)
 			break;
+		rcu_read_lock();
 		ssid_ie = ieee80211_bss_get_ie(&wdev->current_bss->pub,
 					       WLAN_EID_SSID);
-		if (!ssid_ie)
-			break;
-		if (nla_put(msg, NL80211_ATTR_SSID, ssid_ie[1], ssid_ie + 2))
-			goto nla_put_failure_locked;
+		if (ssid_ie &&
+		    nla_put(msg, NL80211_ATTR_SSID, ssid_ie[1], ssid_ie + 2))
+			goto nla_put_failure_rcu_locked;
+		rcu_read_unlock();
 		break;
 		}
 	default:
@@ -2370,6 +2371,8 @@ static int nl80211_send_iface(struct sk_
 
 	return genlmsg_end(msg, hdr);
 
+ nla_put_failure_rcu_locked:
+	rcu_read_unlock();
  nla_put_failure_locked:
 	wdev_unlock(wdev);
  nla_put_failure:
#210 lore raw

PATCH 206/254 mac80211_hwsim: validate number of different channels

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Johannes Berg <[email protected]>

commit 51a1aaa631c90223888d8beac4d649dc11d2ca55 upstream.

When creating a new radio on the fly, hwsim allows this to be done with an arbitrary number of channels, but cfg80211 only supports a limited number of simultaneous channels, leading to a warning.

Fix this by validating the number - this requires moving the define for the maximum out to a visible header file.

Reported-by [email protected]
Fixes b59ec8dd4394 ("mac80211_hwsim: fix number of channels in interface combinations")

[bwh: Backported to 3.16:
- Test chans intead of param.channels
- GENL_SET_ERR_MSG() is not available
- Adjust context]

---

drivers/net/wireless/mac80211_hwsim.c +3 -0
@@ -2439,6 +2439,9 @@ static int hwsim_create_radio_nl(struct
 	if (info->attrs[HWSIM_ATTR_CHANNELS])
 		chans = nla_get_u32(info->attrs[HWSIM_ATTR_CHANNELS]);
 
+	if (chans > CFG80211_MAX_NUM_DIFFERENT_CHANNELS)
+		return -EINVAL;
+
 	if (info->attrs[HWSIM_ATTR_USE_CHANCTX])
 		use_chanctx = true;
 	else
include/net/cfg80211.h +2 -0
@@ -716,6 +716,8 @@ struct cfg80211_csa_settings {
 	u8 count;
 };
 
+#define CFG80211_MAX_NUM_DIFFERENT_CHANNELS 10
+
 /**
  * enum station_parameters_apply_mask - station parameter values to apply
  * @STATION_PARAM_APPLY_UAPSD: apply new uAPSD parameters (uapsd_queues, max_sp)
net/wireless/core.h +0 -2
@@ -456,8 +456,6 @@ void cfg80211_leave(struct cfg80211_regi
 void cfg80211_stop_p2p_device(struct cfg80211_registered_device *rdev,
 			      struct wireless_dev *wdev);
 
-#define CFG80211_MAX_NUM_DIFFERENT_CHANNELS 10
-
 #ifdef CONFIG_CFG80211_DEVELOPER_WARNINGS
 #define CFG80211_DEV_WARN_ON(cond)	WARN_ON(cond)
 #else
#211 lore raw

PATCH 207/254 cfg80211: check dev_set_name() return value

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Johannes Berg <[email protected]>

commit 59b179b48ce2a6076448a44531242ac2b3f6cef2 upstream.

syzbot reported a warning from rfkill_alloc(), and after a while I think that the reason is that it was doing fault injection and the dev_set_name() failed, leaving the name NULL, and we didn't check the return value and got to rfkill_alloc() with a NULL name. Since we really don't want a NULL name, we ought to check the return value.

Fixes fb28ad35906a ("net: struct device - replace bus_id with dev_name(), dev_set_name()")
Reported-by [email protected]
[bwh: Backported to 3.16: adjust context]

---

net/wireless/core.c +6 -1
@@ -315,6 +315,7 @@ struct wiphy *wiphy_new(const struct cfg
 
 	struct cfg80211_registered_device *rdev;
 	int alloc_size;
+	int rv;
 
 	WARN_ON(ops->add_key && (!ops->del_key || !ops->set_default_key));
 	WARN_ON(ops->auth && (!ops->assoc || !ops->deauth || !ops->disassoc));
@@ -346,7 +347,11 @@ struct wiphy *wiphy_new(const struct cfg
 	rdev->wiphy_idx--;
 
 	/* give it a proper name */
-	dev_set_name(&rdev->wiphy.dev, PHY_NAME "%d", rdev->wiphy_idx);
+	rv = dev_set_name(&rdev->wiphy.dev, PHY_NAME "%d", rdev->wiphy_idx);
+	if (rv < 0) {
+		kfree(rdev);
+		return NULL;
+	}
 
 	INIT_LIST_HEAD(&rdev->wdev_list);
 	INIT_LIST_HEAD(&rdev->beacon_registrations);
#212 lore raw

PATCH 208/254 arm64: KVM: Fix SMCCC handling of unimplemented SMC/HVC calls

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Marc Zyngier <[email protected]>

commit acfb3b883f6d6a4b5d27ad7fdded11f6a09ae6dd upstream.

KVM doesn't follow the SMCCC when it comes to unimplemented calls, and inject an UNDEF instead of returning an error. Since firmware calls are now used for security mitigation, they are becoming more common, and the undef is counter productive.

Instead, let's follow the SMCCC which states that -1 must be returned to the caller when getting an unknown function number.

[bwh: Backported to 3.16: use vcpu_reg()]
---
 arch/arm64/kvm/handle_exit.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
arch/arm64/kvm/handle_exit.c +2 -2
@@ -34,7 +34,7 @@ static int handle_hvc(struct kvm_vcpu *v
 
 	ret = kvm_psci_call(vcpu);
 	if (ret < 0) {
-		kvm_inject_undefined(vcpu);
+		*vcpu_reg(vcpu, 0) = ~0UL;
 		return 1;
 	}
 
@@ -43,7 +43,7 @@ static int handle_hvc(struct kvm_vcpu *v
 
 static int handle_smc(struct kvm_vcpu *vcpu, struct kvm_run *run)
 {
-	kvm_inject_undefined(vcpu);
+	*vcpu_reg(vcpu, 0) = ~0UL;
 	return 1;
 }
 
#213 lore raw

PATCH 209/254 sctp: use the right sk after waking up from wait_buf sleep

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Xin Long <[email protected]>

commit cea0cc80a6777beb6eb643d4ad53690e1ad1d4ff upstream.

Commit dfcb9f4f99f1 ("sctp: deny peeloff operation on asocs with threads sleeping on it") fixed the race between peeloff and wait sndbuf by checking waitqueue_active(&asoc->wait) in sctp_do_peeloff().

But it actually doesn't work, as even if waitqueue_active returns false the waiting sndbuf thread may still not yet hold sk lock. After asoc is peeled off, sk is not asoc->base.sk any more, then to hold the old sk lock couldn't make assoc safe to access.

This patch is to fix this by changing to hold the new sk lock if sk is not asoc->base.sk, meanwhile, also set the sk in sctp_sendmsg with the new sk.

With this fix, there is no more race between peeloff and waitbuf, the check 'waitqueue_active' in sctp_do_peeloff can be removed.

Thanks Marcelo and Neil for making this clear.

v1->v2:
fix it by changing to lock the new sock instead of adding a flag in asoc.

---
 net/sctp/socket.c | 21 +++++++++++----------
 1 file changed, 11 insertions(+), 10 deletions(-)
net/sctp/socket.c +11 -10
@@ -82,8 +82,8 @@
 /* Forward declarations for internal helper functions. */
 static int sctp_writeable(struct sock *sk);
 static void sctp_wfree(struct sk_buff *skb);
-static int sctp_wait_for_sndbuf(struct sctp_association *, long *timeo_p,
-				size_t msg_len);
+static int sctp_wait_for_sndbuf(struct sctp_association *asoc, long *timeo_p,
+				size_t msg_len, struct sock **orig_sk);
 static int sctp_wait_for_packet(struct sock *sk, int *err, long *timeo_p);
 static int sctp_wait_for_connect(struct sctp_association *, long *timeo_p);
 static int sctp_wait_for_accept(struct sock *sk, long timeo);
@@ -1905,7 +1905,8 @@ static int sctp_sendmsg(struct kiocb *io
 
 	timeo = sock_sndtimeo(sk, msg->msg_flags & MSG_DONTWAIT);
 	if (!sctp_wspace(asoc)) {
-		err = sctp_wait_for_sndbuf(asoc, &timeo, msg_len);
+		/* sk can be changed by peel off when waiting for buf. */
+		err = sctp_wait_for_sndbuf(asoc, &timeo, msg_len, &sk);
 		if (err)
 			goto out_free;
 	}
@@ -4335,12 +4336,6 @@ int sctp_do_peeloff(struct sock *sk, sct
 	if (!asoc)
 		return -EINVAL;
 
-	/* If there is a thread waiting on more sndbuf space for
-	 * sending on this asoc, it cannot be peeled.
-	 */
-	if (waitqueue_active(&asoc->wait))
-		return -EBUSY;
-
 	/* An association cannot be branched off from an already peeled-off
 	 * socket, nor is this supported for tcp style sockets.
 	 */
@@ -6737,7 +6732,7 @@ void sctp_sock_rfree(struct sk_buff *skb
 
 /* Helper function to wait for space in the sndbuf.  */
 static int sctp_wait_for_sndbuf(struct sctp_association *asoc, long *timeo_p,
-				size_t msg_len)
+				size_t msg_len, struct sock **orig_sk)
 {
 	struct sock *sk = asoc->base.sk;
 	int err = 0;
@@ -6770,11 +6765,17 @@ static int sctp_wait_for_sndbuf(struct s
 		release_sock(sk);
 		current_timeo = schedule_timeout(current_timeo);
 		lock_sock(sk);
+		if (sk != asoc->base.sk) {
+			release_sock(sk);
+			sk = asoc->base.sk;
+			lock_sock(sk);
+		}
 
 		*timeo_p = current_timeo;
 	}
 
 out:
+	*orig_sk = sk;
 	finish_wait(&asoc->wait, &wait);
 
 	/* Release the association's refcnt.  */
#214 lore raw

PATCH 210/254 sctp: return error if the asoc has been peeled off in sctp_wait_for_sndbuf

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Xin Long <[email protected]>

commit a0ff660058b88d12625a783ce9e5c1371c87951f upstream.

After commit cea0cc80a677 ("sctp: use the right sk after waking up from wait_buf sleep"), it may change to lock another sk if the asoc has been peeled off in sctp_wait_for_sndbuf.

However, the asoc's new sk could be already closed elsewhere, as it's in the sendmsg context of the old sk that can't avoid the new sk's closing. If the sk's last one refcnt is held by this asoc, later on after putting this asoc, the new sk will be freed, while under it's own lock.

This patch is to revert that commit, but fix the old issue by returning error under the old sk's lock.

Fixes cea0cc80a677 ("sctp: use the right sk after waking up from wait_buf sleep")
Reported-by [email protected]
[bwh: Backported to 3.16: adjust context]
---
 net/sctp/socket.c | 16 ++++++----------
 1 file changed, 6 insertions(+), 10 deletions(-)
net/sctp/socket.c +6 -10
@@ -83,7 +83,7 @@
 static int sctp_writeable(struct sock *sk);
 static void sctp_wfree(struct sk_buff *skb);
 static int sctp_wait_for_sndbuf(struct sctp_association *asoc, long *timeo_p,
-				size_t msg_len, struct sock **orig_sk);
+				size_t msg_len);
 static int sctp_wait_for_packet(struct sock *sk, int *err, long *timeo_p);
 static int sctp_wait_for_connect(struct sctp_association *, long *timeo_p);
 static int sctp_wait_for_accept(struct sock *sk, long timeo);
@@ -1906,7 +1906,7 @@ static int sctp_sendmsg(struct kiocb *io
 	timeo = sock_sndtimeo(sk, msg->msg_flags & MSG_DONTWAIT);
 	if (!sctp_wspace(asoc)) {
 		/* sk can be changed by peel off when waiting for buf. */
-		err = sctp_wait_for_sndbuf(asoc, &timeo, msg_len, &sk);
+		err = sctp_wait_for_sndbuf(asoc, &timeo, msg_len);
 		if (err)
 			goto out_free;
 	}
@@ -6732,12 +6732,12 @@ void sctp_sock_rfree(struct sk_buff *skb
 
 /* Helper function to wait for space in the sndbuf.  */
 static int sctp_wait_for_sndbuf(struct sctp_association *asoc, long *timeo_p,
-				size_t msg_len, struct sock **orig_sk)
+				size_t msg_len)
 {
 	struct sock *sk = asoc->base.sk;
-	int err = 0;
 	long current_timeo = *timeo_p;
 	DEFINE_WAIT(wait);
+	int err = 0;
 
 	pr_debug("%s: asoc:%p, timeo:%ld, msg_len:%zu\n", __func__, asoc,
 		 *timeo_p, msg_len);
@@ -6765,17 +6765,13 @@ static int sctp_wait_for_sndbuf(struct s
 		release_sock(sk);
 		current_timeo = schedule_timeout(current_timeo);
 		lock_sock(sk);
-		if (sk != asoc->base.sk) {
-			release_sock(sk);
-			sk = asoc->base.sk;
-			lock_sock(sk);
-		}
+		if (sk != asoc->base.sk)
+			goto do_error;
 
 		*timeo_p = current_timeo;
 	}
 
 out:
-	*orig_sk = sk;
 	finish_wait(&asoc->wait, &wait);
 
 	/* Release the association's refcnt.  */
#215 lore raw

PATCH 211/254 sctp: do not allow the v4 socket to bind a v4mapped v6 address

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Xin Long <[email protected]>

commit c5006b8aa74599ce19104b31d322d2ea9ff887cc upstream.

The check in sctp_sockaddr_af is not robust enough to forbid binding a v4mapped v6 addr on a v4 socket.

The worse thing is that v4 socket's bind_verify would not convert this v4mapped v6 addr to a v4 addr. syzbot even reported a crash as the v4 socket bound a v6 addr.

This patch is to fix it by doing the common sa.sa_family check first, then AF_INET check for v4mapped v6 addrs.

Fixes 7dab83de50c7 ("sctp: Support ipv6only AF_INET6 sockets.")
Reported-by [email protected]
---
 net/sctp/socket.c | 14 ++++++--------
 1 file changed, 6 insertions(+), 8 deletions(-)
net/sctp/socket.c +6 -8
@@ -303,16 +303,14 @@ static struct sctp_af *sctp_sockaddr_af(
 	if (len < sizeof (struct sockaddr))
 		return NULL;
 
+	if (!opt->pf->af_supported(addr->sa.sa_family, opt))
+		return NULL;
+
 	/* V4 mapped address are really of AF_INET family */
 	if (addr->sa.sa_family == AF_INET6 &&
-	    ipv6_addr_v4mapped(&addr->v6.sin6_addr)) {
-		if (!opt->pf->af_supported(AF_INET, opt))
-			return NULL;
-	} else {
-		/* Does this PF support this AF? */
-		if (!opt->pf->af_supported(addr->sa.sa_family, opt))
-			return NULL;
-	}
+	    ipv6_addr_v4mapped(&addr->v6.sin6_addr) &&
+	    !opt->pf->af_supported(AF_INET, opt))
+		return NULL;
 
 	/* If we get this far, af is valid. */
 	af = sctp_get_af_specific(addr->sa.sa_family);
#216 lore raw

PATCH 212/254 KVM/x86: Check input paging mode when cs.l is set

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Lan Tianyu <[email protected]>

commit f29810335965ac1f7bcb501ee2af5f039f792416 upstream.

Reported by syzkaller:

    WARNING: CPU: 0 PID: 27962 at arch/x86/kvm/emulate.c:5631 x86_emulate_insn+0x557/0x15f0 [kvm]
    Modules linked in: kvm_intel kvm [last unloaded: kvm]
    CPU: 0 PID: 27962 Comm: syz-executor Tainted: G    B   W        4.15.0-rc2-next-20171208+ #32
    Hardware name: Intel Corporation S1200SP/S1200SP, BIOS S1200SP.86B.01.03.0006.040720161253 04/07/2016
    RIP: 0010:x86_emulate_insn+0x557/0x15f0 [kvm]
    RSP: 0018:ffff8807234476d0 EFLAGS: 00010282
    RAX: 0000000000000000 RBX: ffff88072d0237a0 RCX: ffffffffa0065c4d
    RDX: 1ffff100e5a046f9 RSI: 0000000000000003 RDI: ffff88072d0237c8
    RBP: ffff880723447728 R08: ffff88072d020000 R09: ffffffffa008d240
    R10: 0000000000000002 R11: ffffed00e7d87db3 R12: ffff88072d0237c8
    R13: ffff88072d023870 R14: ffff88072d0238c2 R15: ffffffffa008d080
    FS:  00007f8a68666700(0000) GS:ffff880802200000(0000) knlGS:0000000000000000
    CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
    CR2: 000000002009506c CR3: 000000071fec4005 CR4: 00000000003626f0
    Call Trace:
     x86_emulate_instruction+0x3bc/0xb70 [kvm]
     ? reexecute_instruction.part.162+0x130/0x130 [kvm]
     vmx_handle_exit+0x46d/0x14f0 [kvm_intel]
     ? trace_event_raw_event_kvm_entry+0xe7/0x150 [kvm]
     ? handle_vmfunc+0x2f0/0x2f0 [kvm_intel]
     ? wait_lapic_expire+0x25/0x270 [kvm]
     vcpu_enter_guest+0x720/0x1ef0 [kvm]
     ...

When CS.L is set, vcpu should run in the 64 bit paging mode. Current kvm set_sregs function doesn't have such check when userspace inputs sreg values. This will lead unexpected behavior. This patch is to add checks for CS.L, EFER.LME, EFER.LMA and CR4.PAE when get SREG inputs from userspace in order to avoid unexpected behavior.

Cc Paolo Bonzini <[email protected]>
Cc Radim Krčmář <[email protected]>
Cc Dmitry Vyukov <[email protected]>
Cc Jim Mattson <[email protected]>
[bwh: Backported to 3.16: adjust context]
---
 arch/x86/kvm/x86.c | 26 ++++++++++++++++++++++++++
 1 file changed, 26 insertions(+)
arch/x86/kvm/x86.c +26 -0
@@ -6579,6 +6579,29 @@ int kvm_task_switch(struct kvm_vcpu *vcp
 }
 EXPORT_SYMBOL_GPL(kvm_task_switch);
 
+int kvm_valid_sregs(struct kvm_vcpu *vcpu, struct kvm_sregs *sregs)
+{
+	if ((sregs->efer & EFER_LME) && (sregs->cr0 & X86_CR0_PG_BIT)) {
+		/*
+		 * When EFER.LME and CR0.PG are set, the processor is in
+		 * 64-bit mode (though maybe in a 32-bit code segment).
+		 * CR4.PAE and EFER.LMA must be set.
+		 */
+		if (!(sregs->cr4 & X86_CR4_PAE_BIT)
+		    || !(sregs->efer & EFER_LMA))
+			return -EINVAL;
+	} else {
+		/*
+		 * Not in 64-bit mode: EFER.LMA is clear and the code
+		 * segment cannot be 64-bit.
+		 */
+		if (sregs->efer & EFER_LMA || sregs->cs.l)
+			return -EINVAL;
+	}
+
+	return 0;
+}
+
 int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 				  struct kvm_sregs *sregs)
 {
@@ -6590,6 +6613,9 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct
 	if (!guest_cpuid_has_xsave(vcpu) && (sregs->cr4 & X86_CR4_OSXSAVE))
 		return -EINVAL;
 
+	if (kvm_valid_sregs(vcpu, sregs))
+		return -EINVAL;
+
 	dt.size = sregs->idt.limit;
 	dt.address = sregs->idt.base;
 	kvm_x86_ops->set_idt(vcpu, &dt);
#217 lore raw

PATCH 213/254 KVM/x86: Fix wrong macro references of X86_CR0_PG_BIT and X86_CR4_PAE_BIT in kvm_valid_sregs()

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Tianyu Lan <[email protected]>

commit 37b95951c58fdf08dc10afa9d02066ed9f176fb5 upstream.

kvm_valid_sregs() should use X86_CR0_PG and X86_CR4_PAE to check bit status rather than X86_CR0_PG_BIT and X86_CR4_PAE_BIT. This patch is to fix it.

Fixes f29810335965a(KVM/x86: Check input paging mode when cs.l is set)
Cc Paolo Bonzini <[email protected]>
Cc Radim Krčmář <[email protected]>
---
 arch/x86/kvm/x86.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
arch/x86/kvm/x86.c +2 -2
@@ -6581,13 +6581,13 @@ EXPORT_SYMBOL_GPL(kvm_task_switch);
 
 int kvm_valid_sregs(struct kvm_vcpu *vcpu, struct kvm_sregs *sregs)
 {
-	if ((sregs->efer & EFER_LME) && (sregs->cr0 & X86_CR0_PG_BIT)) {
+	if ((sregs->efer & EFER_LME) && (sregs->cr0 & X86_CR0_PG)) {
 		/*
 		 * When EFER.LME and CR0.PG are set, the processor is in
 		 * 64-bit mode (though maybe in a 32-bit code segment).
 		 * CR4.PAE and EFER.LMA must be set.
 		 */
-		if (!(sregs->cr4 & X86_CR4_PAE_BIT)
+		if (!(sregs->cr4 & X86_CR4_PAE)
 		    || !(sregs->efer & EFER_LMA))
 			return -EINVAL;
 	} else {
#218 lore raw

PATCH 214/254 dm thin metadata: THIN_MAX_CONCURRENT_LOCKS should be 6

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Dennis Yang <[email protected]>

commit 490ae017f54e55bde382d45ea24bddfb6d1a0aaf upstream.

For btree removal, there is a corner case that a single thread could takes 6 locks which is more than THIN_MAX_CONCURRENT_LOCKS(5) and leads to deadlock.

A btree removal might eventually call rebalance_children()->rebalance3() to rebalance entries of three neighbor child nodes when shadow_spine has already acquired two write locks. In rebalance3(), it tries to shadow and acquire the write locks of all three child nodes. However, shadowing a child node requires acquiring a read lock of the original child node and a write lock of the new block. Although the read lock will be released after block shadowing, shadowing the third child node in rebalance3() could still take the sixth lock.
(2 write locks for shadow_spine +
2 write locks for the first two child nodes's shadow + 1 write lock for the last child node's shadow +
1 read lock for the last child node)

---
 drivers/md/dm-thin-metadata.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/md/dm-thin-metadata.c +5 -1
@@ -81,10 +81,14 @@
 #define SECTOR_TO_BLOCK_SHIFT 3
 
 /*
+ * For btree insert:
  *  3 for btree insert +
  *  2 for btree lookup used within space map
+ * For btree remove:
+ *  2 for shadow spine +
+ *  4 for rebalance 3 child node
  */
-#define THIN_MAX_CONCURRENT_LOCKS 5
+#define THIN_MAX_CONCURRENT_LOCKS 6
 
 /* This should be plenty */
 #define SPACE_MAP_ROOT_SIZE 128
#219 lore raw

PATCH 215/254 dm btree: fix serious bug in btree_split_beneath()

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Joe Thornber <[email protected]>

commit bc68d0a43560e950850fc69b58f0f8254b28f6d6 upstream.

When inserting a new key/value pair into a btree we walk down the spine of btree nodes performing the following 2 operations:

i) space for a new entry
ii) adjusting the first key entry if the new key is lower than any in the node.

If the _root_ node is full, the function btree_split_beneath() allocates 2 new nodes, and redistibutes the root nodes entries between them. The root node is left with 2 entries corresponding to the 2 new nodes.

btree_split_beneath() then adjusts the spine to point to one of the two new children. This means the first key is never adjusted if the new key was lower, ie. operation (ii) gets missed out. This can result in the new key being 'lost' for a period; until another low valued key is inserted that will uncover it.

This is a serious bug, and quite hard to make trigger in normal use. A reproducing test case ("thin create devices-in-reverse-order") is available as part of the thin-provision-tools project:
https://github.com/jthornber/thin-provisioning-tools/blob/master/functional-tests/device-mapper/dm-tests.scm#L593

Fix the issue by changing btree_split_beneath() so it no longer adjusts the spine. Instead it unlocks both the new nodes, and lets the main loop in btree_insert_raw() relock the appropriate one and make any neccessary adjustments.

---
 drivers/md/persistent-data/dm-btree.c | 19 ++-----------------
 1 file changed, 2 insertions(+), 17 deletions(-)
drivers/md/persistent-data/dm-btree.c +2 -17
@@ -588,23 +588,8 @@ static int btree_split_beneath(struct sh
 	pn->keys[1] = rn->keys[0];
 	memcpy_disk(value_ptr(pn, 1), &val, sizeof(__le64));
 
-	/*
-	 * rejig the spine.  This is ugly, since it knows too
-	 * much about the spine
-	 */
-	if (s->nodes[0] != new_parent) {
-		unlock_block(s->info, s->nodes[0]);
-		s->nodes[0] = new_parent;
-	}
-	if (key < le64_to_cpu(rn->keys[0])) {
-		unlock_block(s->info, right);
-		s->nodes[1] = left;
-	} else {
-		unlock_block(s->info, left);
-		s->nodes[1] = right;
-	}
-	s->count = 2;
-
+	unlock_block(s->info, left);
+	unlock_block(s->info, right);
 	return 0;
 }
 
#220 lore raw

PATCH 216/254 i2c: core: decrease reference count of device node in i2c_unregister_device

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Lixin Wang <[email protected]>

commit e0638fa400eaccf9fa8060f67140264c4e276552 upstream.

Reference count of device node was increased in of_i2c_register_device, but without decreasing it in i2c_unregister_device. Then the added device node will never be released. Fix this by adding the of_node_put.

[bwh: Backported to 3.16: adjust context]
---
 drivers/i2c/i2c-core.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/i2c/i2c-core.c +4 -0
@@ -740,6 +740,10 @@ EXPORT_SYMBOL_GPL(i2c_new_device);
  */
 void i2c_unregister_device(struct i2c_client *client)
 {
+	if (client->dev.of_node) {
+		of_node_put(client->dev.of_node);
+	}
+
 	device_unregister(&client->dev);
 }
 EXPORT_SYMBOL_GPL(i2c_unregister_device);
#221 lore raw

PATCH 217/254 i2c: core-smbus: prevent stack corruption on read I2C_BLOCK_DATA

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Jeremy Compostella <[email protected]>

commit 89c6efa61f5709327ecfa24bff18e57a4e80c7fa upstream.

On a I2C_SMBUS_I2C_BLOCK_DATA read request, if data->block[0] is greater than I2C_SMBUS_BLOCK_MAX + 1, the underlying I2C driver writes data out of the msgbuf1 array boundary.

It is possible from a user application to run into that issue by calling the I2C_SMBUS ioctl with data.block[0] greater than I2C_SMBUS_BLOCK_MAX + 1.

This patch makes the code compliant with Documentation/i2c/dev-interface by raising an error when the requested size is larger than 32 bytes.

Call Trace:
 [<ffffffff8139f695>] dump_stack+0x67/0x92
 [<ffffffff811802a4>] panic+0xc5/0x1eb
 [<ffffffff810ecb5f>] ? vprintk_default+0x1f/0x30
 [<ffffffff817456d3>] ? i2cdev_ioctl_smbus+0x303/0x320
 [<ffffffff8109a68b>] __stack_chk_fail+0x1b/0x20
 [<ffffffff817456d3>] i2cdev_ioctl_smbus+0x303/0x320
 [<ffffffff81745aed>] i2cdev_ioctl+0x4d/0x1e0
 [<ffffffff811f761a>] do_vfs_ioctl+0x2ba/0x490
 [<ffffffff81336e43>] ? security_file_ioctl+0x43/0x60
 [<ffffffff811f7869>] SyS_ioctl+0x79/0x90
 [<ffffffff81a22e97>] entry_SYSCALL_64_fastpath+0x12/0x6a
[bwh: Backported to 3.16: adjust filename]
---
 drivers/i2c/i2c-core.c | 13 +++++++------
 1 file changed, 7 insertions(+), 6 deletions(-)
drivers/i2c/i2c-core.c +7 -6
@@ -2481,16 +2481,17 @@ static s32 i2c_smbus_xfer_emulated(struc
 				   the underlying bus driver */
 		break;
 	case I2C_SMBUS_I2C_BLOCK_DATA:
+		if (data->block[0] > I2C_SMBUS_BLOCK_MAX) {
+			dev_err(&adapter->dev, "Invalid block %s size %d\n",
+				read_write == I2C_SMBUS_READ ? "read" : "write",
+				data->block[0]);
+			return -EINVAL;
+		}
+
 		if (read_write == I2C_SMBUS_READ) {
 			msg[1].len = data->block[0];
 		} else {
 			msg[0].len = data->block[0] + 1;
-			if (msg[0].len > I2C_SMBUS_BLOCK_MAX + 1) {
-				dev_err(&adapter->dev,
-					"Invalid block write size %d\n",
-					data->block[0]);
-				return -EINVAL;
-			}
 			for (i = 1; i <= data->block[0]; i++)
 				msgbuf0[i] = data->block[i];
 		}
#222 lore raw

PATCH 218/254 net: fs_enet: do not call phy_stop() in interrupts

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Christophe Leroy <[email protected]>

commit f8b39039cbf2a15f2b8c9f081e1cbd5dee00aaf5 upstream.

In case of TX timeout, fs_timeout() calls phy_stop(), which triggers the following BUG_ON() as we are in interrupt.

[92708.199889] kernel BUG at drivers/net/phy/mdio_bus.c:482!
[92708.204985] Oops: Exception in kernel mode, sig: 5 [#1]
[92708.210119] PREEMPT
[92708.212107] CMPC885
[92708.214216] CPU: 0 PID: 3 Comm: ksoftirqd/0 Tainted: G        W       4.9.61 #39
[92708.223227] task: c60f0a40 task.stack: c6104000
[92708.227697] NIP: c02a84bc LR: c02a947c CTR: c02a93d8
[92708.232614] REGS: c6105c70 TRAP: 0700   Tainted: G        W        (4.9.61)
[92708.241193] MSR: 00021032 <ME,IR,DR,RI>[92708.244818]   CR: 24000822  XER: 20000000
[92708.248767]

GPR00: c02a947c c6105d20 c60f0a40 c62b4c00 00000005 0000001f c069aad8 0001a688
GPR08: 00000007 00000100 c02a93d8 00000000 000005fc 00000000 c6213240 c06338e4
GPR16: 00000001 c06330d4 c0633094 00000000 c0680000 c6104000 c6104000 00000000
GPR24: 00000200 00000000 ffffffff 00000004 00000078 00009032 00000000 c62b4c00 NIP [c02a84bc] mdiobus_read+0x20/0x74

[92708.281517] LR [c02a947c] kszphy_config_intr+0xa4/0xc4
[92708.286547] Call Trace:
[92708.288980] [c6105d20] [c6104000] 0xc6104000 (unreliable)
[92708.294339] [c6105d40] [c02a947c] kszphy_config_intr+0xa4/0xc4
[92708.300098] [c6105d50] [c02a5330] phy_stop+0x60/0x9c
[92708.305007] [c6105d60] [c02c84d0] fs_timeout+0xdc/0x110
[92708.310197] [c6105d80] [c035cd48] dev_watchdog+0x268/0x2a0
[92708.315593] [c6105db0] [c0060288] call_timer_fn+0x34/0x17c
[92708.321014] [c6105dd0] [c00605f0] run_timer_softirq+0x21c/0x2e4
[92708.326887] [c6105e50] [c001e19c] __do_softirq+0xf4/0x2f4
[92708.332207] [c6105eb0] [c001e3c8] run_ksoftirqd+0x2c/0x40
[92708.337560] [c6105ec0] [c003b420] smpboot_thread_fn+0x1f0/0x258
[92708.343405] [c6105ef0] [c003745c] kthread+0xbc/0xd0
[92708.348217] [c6105f40] [c000c400] ret_from_kernel_thread+0x5c/0x64
[92708.354275] Instruction dump:
[92708.357207] 7c0803a6 bbc10018 38210020 4e800020 7c0802a6 9421ffe0 54290024 bfc10018
[92708.364865] 90010024 7c7f1b78 81290008 552902ee <0f090000> 3bc3002c 7fc3f378 90810008
[92708.372711] ---[ end trace 42b05441616fafd7 ]---

This patch moves fs_timeout() actions into an async worker.

Fixes commit 48257c4f168e5 ("Add fs_enet ethernet network driver, for several embedded platforms")
[bwh: Backported to 3.16: adjust context]
---
 drivers/net/ethernet/freescale/fs_enet/fs_enet-main.c | 16 +++++++++++++---
 drivers/net/ethernet/freescale/fs_enet/fs_enet.h      |  1 +
 2 files changed, 14 insertions(+), 3 deletions(-)
drivers/net/ethernet/freescale/fs_enet/fs_enet-main.c +13 -3
@@ -696,9 +696,11 @@ static int fs_enet_start_xmit(struct sk_
 	return NETDEV_TX_OK;
 }
 
-static void fs_timeout(struct net_device *dev)
+static void fs_timeout_work(struct work_struct *work)
 {
-	struct fs_enet_private *fep = netdev_priv(dev);
+	struct fs_enet_private *fep = container_of(work, struct fs_enet_private,
+						   timeout_work);
+	struct net_device *dev = fep->ndev;
 	unsigned long flags;
 	int wake = 0;
 
@@ -710,7 +712,6 @@ static void fs_timeout(struct net_device
 		phy_stop(fep->phydev);
 		(*fep->ops->stop)(dev);
 		(*fep->ops->restart)(dev);
-		phy_start(fep->phydev);
 	}
 
 	phy_start(fep->phydev);
@@ -721,6 +722,13 @@ static void fs_timeout(struct net_device
 		netif_wake_queue(dev);
 }
 
+static void fs_timeout(struct net_device *dev)
+{
+	struct fs_enet_private *fep = netdev_priv(dev);
+
+	schedule_work(&fep->timeout_work);
+}
+
 /*-----------------------------------------------------------------------------
  *  generic link-change handler - should be sufficient for most cases
  *-----------------------------------------------------------------------------*/
@@ -847,6 +855,7 @@ static int fs_enet_close(struct net_devi
 	netif_carrier_off(dev);
 	if (fep->fpi->use_napi)
 		napi_disable(&fep->napi);
+	cancel_work_sync(&fep->timeout_work);
 	phy_stop(fep->phydev);
 
 	spin_lock_irqsave(&fep->lock, flags);
@@ -1102,6 +1111,7 @@ static int fs_enet_probe(struct platform
 
 	ndev->netdev_ops = &fs_enet_netdev_ops;
 	ndev->watchdog_timeo = 2 * HZ;
+	INIT_WORK(&fep->timeout_work, fs_timeout_work);
 	if (fpi->use_napi)
 		netif_napi_add(ndev, &fep->napi, fs_enet_rx_napi,
 		               fpi->napi_weight);
drivers/net/ethernet/freescale/fs_enet/fs_enet.h +1 -0
@@ -124,6 +124,7 @@ struct fs_enet_private {
 	spinlock_t lock;	/* during all ops except TX pckt processing */
 	spinlock_t tx_lock;	/* during fs_start_xmit and fs_tx         */
 	struct fs_platform_info *fpi;
+	struct work_struct timeout_work;
 	const struct fs_ops *ops;
 	int rx_ring, tx_ring;
 	dma_addr_t ring_mem_addr;
#223 lore raw

PATCH 219/254 can: af_can: can_rcv(): replace WARN_ONCE by pr_warn_once

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Marc Kleine-Budde <[email protected]>

commit 8cb68751c115d176ec851ca56ecfbb411568c9e8 upstream.

If an invalid CAN frame is received, from a driver or from a tun interface, a Kernel warning is generated.

This patch replaces the WARN_ONCE by a simple pr_warn_once, so that a kernel, bootet with panic_on_warn, does not panic. A printk seems to be more appropriate here.

[bwh: Backported to 3.16:
- Keep using the 'drop' label, as it has another user
- Adjust context]

---

net/can/af_can.c +5 -6
@@ -719,13 +719,12 @@ static int can_rcv(struct sk_buff *skb,
 	if (unlikely(!net_eq(dev_net(dev), &init_net)))
 		goto drop;
 
-	if (WARN_ONCE(dev->type != ARPHRD_CAN ||
-		      skb->len != CAN_MTU ||
-		      cfd->len > CAN_MAX_DLEN,
-		      "PF_CAN: dropped non conform CAN skbuf: "
-		      "dev type %d, len %d, datalen %d\n",
-		      dev->type, skb->len, cfd->len))
+	if (unlikely(dev->type != ARPHRD_CAN || skb->len != CAN_MTU ||
+		     cfd->len > CAN_MAX_DLEN)) {
+		pr_warn_once("PF_CAN: dropped non conform CAN skbuf: dev type %d, len %d, datalen %d\n",
+			     dev->type, skb->len, cfd->len);
 		goto drop;
+	}
 
 	can_receive(skb, dev);
 	return NET_RX_SUCCESS;
#224 lore raw

PATCH 220/254 can: af_can: canfd_rcv(): replace WARN_ONCE by pr_warn_once

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Marc Kleine-Budde <[email protected]>

commit d4689846881d160a4d12a514e991a740bcb5d65a upstream.

If an invalid CANFD frame is received, from a driver or from a tun interface, a Kernel warning is generated.

This patch replaces the WARN_ONCE by a simple pr_warn_once, so that a kernel, bootet with panic_on_warn, does not panic. A printk seems to be more appropriate here.

[bwh: Backported to 3.16:
- Keep using the 'drop' label, as it has another user
- Adjust context]

---

net/can/af_can.c +5 -6
@@ -742,13 +742,12 @@ static int canfd_rcv(struct sk_buff *skb
 	if (unlikely(!net_eq(dev_net(dev), &init_net)))
 		goto drop;
 
-	if (WARN_ONCE(dev->type != ARPHRD_CAN ||
-		      skb->len != CANFD_MTU ||
-		      cfd->len > CANFD_MAX_DLEN,
-		      "PF_CAN: dropped non conform CAN FD skbuf: "
-		      "dev type %d, len %d, datalen %d\n",
-		      dev->type, skb->len, cfd->len))
+	if (unlikely(dev->type != ARPHRD_CAN || skb->len != CANFD_MTU ||
+		     cfd->len > CANFD_MAX_DLEN)) {
+		pr_warn_once("PF_CAN: dropped non conform CAN FD skbuf: dev type %d, len %d, datalen %d\n",
+			     dev->type, skb->len, cfd->len);
 		goto drop;
+	}
 
 	can_receive(skb, dev);
 	return NET_RX_SUCCESS;
#225 lore raw

PATCH 221/254 cfg80211: fix station info handling bugs

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Johannes Berg <[email protected]>

commit 5762d7d3eda25c03cc2d9d45227be3f5ab6bec9e upstream.

Fix two places where the structure isn't initialized to zero, and thus can't be filled properly by the driver.

Fixes 4a4b8169501b ("cfg80211: Accept multiple RSSI thresholds for CQM")
Fixes 9930380f0bd8 ("cfg80211: implement IWRATE")
[bwh: Backported to 3.16: drop change in cfg80211_cqm_rssi_update()]

---

net/wireless/wext-compat.c +1 -2
@@ -1273,8 +1273,7 @@ static int cfg80211_wext_giwrate(struct
 {
 	struct wireless_dev *wdev = dev->ieee80211_ptr;
 	struct cfg80211_registered_device *rdev = wiphy_to_rdev(wdev->wiphy);
-	/* we are under RTNL - globally locked - so can use a static struct */
-	static struct station_info sinfo;
+	struct station_info sinfo = {};
 	u8 addr[ETH_ALEN];
 	int err;
 
#226 lore raw

PATCH 222/254 x86/mce: Make machine check speculation protected

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Thomas Gleixner <[email protected]>

commit 6f41c34d69eb005e7848716bbcafc979b35037d5 upstream.

The machine check idtentry uses an indirect branch directly from the low level code. This evades the speculation protection.

Replace it by a direct call into C code and issue the indirect call there so the compiler can apply the proper speculation protection.

[bwh: Backported to 3.16
- #include <asm/traps.h> in mce.c
- Adjust filename, context]

---

arch/x86/include/asm/traps.h +1 -0
@@ -91,6 +91,7 @@ dotraplinkage void do_simd_coprocessor_e
 #ifdef CONFIG_X86_32
 dotraplinkage void do_iret_error(struct pt_regs *, long);
 #endif
+dotraplinkage void do_mce(struct pt_regs *, long);
 
 static inline int get_si_code(unsigned long condition)
 {
arch/x86/kernel/cpu/mcheck/mce.c +6 -0
@@ -46,6 +46,7 @@
 #include <asm/tlbflush.h>
 #include <asm/mce.h>
 #include <asm/msr.h>
+#include <asm/traps.h>
 
 #include "mce-internal.h"
 
@@ -1693,6 +1694,11 @@ static void unexpected_machine_check(str
 void (*machine_check_vector)(struct pt_regs *, long error_code) =
 						unexpected_machine_check;
 
+dotraplinkage void do_mce(struct pt_regs *regs, long error_code)
+{
+	machine_check_vector(regs, error_code);
+}
+
 /*
  * Called for each booted CPU to set up machine checks.
  * Must be called with preempt off:
arch/x86/kernel/entry_64.S +1 -1
@@ -1320,7 +1320,7 @@ trace_idtentry page_fault do_page_fault
 idtentry async_page_fault do_async_page_fault has_error_code=1
 #endif
 #ifdef CONFIG_X86_MCE
-idtentry machine_check has_error_code=0 paranoid=1 do_sym=*machine_check_vector(%rip)
+idtentry machine_check		do_mce			has_error_code=0	paranoid=1
 #endif
 
 	/*
#227 lore raw

PATCH 223/254 net: igmp: Use correct source address on IGMPv3 reports

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Kevin Cernekee <[email protected]>

commit a46182b00290839fa3fa159d54fd3237bd8669f0 upstream.

Closing a multicast socket after the final IPv4 address is deleted from an interface can generate a membership report that uses the source IP from a different interface. The following test script, run from an isolated netns, reproduces the issue:

    #!/bin/bash

    ip link add dummy0 type dummy
    ip link add dummy1 type dummy
    ip link set dummy0 up
    ip link set dummy1 up
    ip addr add 10.1.1.1/24 dev dummy0
    ip addr add 192.168.99.99/24 dev dummy1

    tcpdump -U -i dummy0 &
    socat EXEC:"sleep 2" \
        UDP4-DATAGRAM:239.101.1.68:8889,ip-add-membership=239.0.1.68:10.1.1.1 &

    sleep 1
    ip addr del 10.1.1.1/24 dev dummy0
    sleep 5
    kill %tcpdump

RFC 3376 specifies that the report must be sent with a valid IP source address from the destination subnet, or from address 0.0.0.0. Add an extra check to make sure this is the case.

---
 net/ipv4/igmp.c | 20 +++++++++++++++++++-
 1 file changed, 19 insertions(+), 1 deletion(-)
net/ipv4/igmp.c +19 -1
@@ -89,6 +89,7 @@
 #include <linux/rtnetlink.h>
 #include <linux/times.h>
 #include <linux/pkt_sched.h>
+#include <linux/byteorder/generic.h>
 
 #include <net/net_namespace.h>
 #include <net/arp.h>
@@ -323,6 +324,23 @@ igmp_scount(struct ip_mc_list *pmc, int
 	return scount;
 }
 
+/* source address selection per RFC 3376 section 4.2.13 */
+static __be32 igmpv3_get_srcaddr(struct net_device *dev,
+				 const struct flowi4 *fl4)
+{
+	struct in_device *in_dev = __in_dev_get_rcu(dev);
+
+	if (!in_dev)
+		return htonl(INADDR_ANY);
+
+	for_ifa(in_dev) {
+		if (inet_ifa_match(fl4->saddr, ifa))
+			return fl4->saddr;
+	} endfor_ifa(in_dev);
+
+	return htonl(INADDR_ANY);
+}
+
 static struct sk_buff *igmpv3_newpack(struct net_device *dev, unsigned int mtu)
 {
 	struct sk_buff *skb;
@@ -370,7 +388,7 @@ static struct sk_buff *igmpv3_newpack(st
 	pip->frag_off = htons(IP_DF);
 	pip->ttl      = 1;
 	pip->daddr    = fl4.daddr;
-	pip->saddr    = fl4.saddr;
+	pip->saddr    = igmpv3_get_srcaddr(dev, &fl4);
 	pip->protocol = IPPROTO_IGMP;
 	pip->tot_len  = 0;	/* filled in later */
 	ip_select_ident(skb, NULL);
#228 lore raw

PATCH 224/254 net: igmp: fix source address check for IGMPv3 reports

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Felix Fietkau <[email protected]>

commit ad23b750933ea7bf962678972a286c78a8fa36aa upstream.

Commit "net: igmp: Use correct source address on IGMPv3 reports" introduced a check to validate the source address of locally generated IGMPv3 packets.
Instead of checking the local interface address directly, it uses inet_ifa_match(fl4->saddr, ifa), which checks if the address is on the local subnet (or equal to the point-to-point address if used).

This breaks for point-to-point interfaces, so check against ifa->ifa_local directly.

Cc Kevin Cernekee <[email protected]>
Fixes a46182b00290 ("net: igmp: Use correct source address on IGMPv3 reports")
---
 net/ipv4/igmp.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/ipv4/igmp.c +1 -1
@@ -334,7 +334,7 @@ static __be32 igmpv3_get_srcaddr(struct
 		return htonl(INADDR_ANY);
 
 	for_ifa(in_dev) {
-		if (inet_ifa_match(fl4->saddr, ifa))
+		if (fl4->saddr == ifa->ifa_local)
 			return fl4->saddr;
 	} endfor_ifa(in_dev);
 
#229 lore raw

PATCH 225/254 Input: trackpoint - assume 3 buttons when buttons detection fails

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Oscar Campos <[email protected]>

commit 293b915fd9bebf33cdc906516fb28d54649a25ac upstream.

Trackpoint buttons detection fails on ThinkPad 570 and 470 series, this makes the middle button of the trackpoint to not being recogized. As I don't believe there is any trackpoint with less than 3 buttons this patch just assumes three buttons when the extended button information read fails.

[bwh: Backported to 3.16: adjust context]
---
 drivers/input/mouse/trackpoint.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/input/mouse/trackpoint.c +2 -2
@@ -377,8 +377,8 @@ int trackpoint_detect(struct psmouse *ps
 		return 0;
 
 	if (trackpoint_read(&psmouse->ps2dev, TP_EXT_BTN, &button_info)) {
-		psmouse_warn(psmouse, "failed to get extended button data\n");
-		button_info = 0;
+		psmouse_warn(psmouse, "failed to get extended button data, assuming 3 buttons\n");
+		button_info = 0x33;
 	}
 
 	psmouse->private = kzalloc(sizeof(struct trackpoint_data), GFP_KERNEL);
#230 lore raw

PATCH 226/254 Input: trackpoint - force 3 buttons if 0 button is reported

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Aaron Ma <[email protected]>

commit f5d07b9e98022d50720e38aa936fc11c67868ece upstream.

Lenovo introduced trackpoint compatible sticks with minimum PS/2 commands. They supposed to reply with 0x02, 0x03, or 0x04 in response to the "Read Extended ID" command, so we would know not to try certain extended commands. Unfortunately even some trackpoints reporting the original IBM version (0x01 firmware 0x0e) now respond with incorrect data to the "Get Extended Buttons" command:

thinkpad_acpi: ThinkPad BIOS R0DET87W (1.87 ), EC unknown
thinkpad_acpi: Lenovo ThinkPad E470, model 20H1004SGE

psmouse serio2: trackpoint: IBM TrackPoint firmware: 0x0e, buttons: 0/0

Since there are no trackpoints without buttons, let's assume the trackpoint has 3 buttons when we get 0 response to the extended buttons query.

---
 drivers/input/mouse/trackpoint.c | 3 +++
 1 file changed, 3 insertions(+)
drivers/input/mouse/trackpoint.c +3 -0
@@ -379,6 +379,9 @@ int trackpoint_detect(struct psmouse *ps
 	if (trackpoint_read(&psmouse->ps2dev, TP_EXT_BTN, &button_info)) {
 		psmouse_warn(psmouse, "failed to get extended button data, assuming 3 buttons\n");
 		button_info = 0x33;
+	} else if (!button_info) {
+		psmouse_warn(psmouse, "got 0 in extended button data, assuming 3 buttons\n");
+		button_info = 0x33;
 	}
 
 	psmouse->private = kzalloc(sizeof(struct trackpoint_data), GFP_KERNEL);
#231 lore raw

PATCH 227/254 pppoe: take ->needed_headroom of lower device into account on xmit

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Guillaume Nault <[email protected]>

commit 02612bb05e51df8489db5e94d0cf8d1c81f87b0c upstream.

In pppoe_sendmsg(), reserving dev->hard_header_len bytes of headroom was probably fine before the introduction of ->needed_headroom in commit f5184d267c1a ("net: Allow netdevices to specify needed head/tailroom").

But now, virtual devices typically advertise the size of their overhead in dev->needed_headroom, so we must also take it into account in skb_reserve().
Allocation size of skb is also updated to take dev->needed_tailroom into account and replace the arbitrary 32 bytes with the real size of a PPPoE header.

This issue was discovered by syzbot, who connected a pppoe socket to a gre device which had dev->header_ops->create == ipgre_header and dev->hard_header_len == 0. Therefore, PPPoE didn't reserve any headroom, and dev_hard_header() crashed when ipgre_header() tried to prepend its header to skb->data.

skbuff: skb_under_panic: text:000000001d390b3a len:31 put:24 head:00000000d8ed776f data:000000008150e823 tail:0x7 end:0xc0 dev:gre0

------------[ cut here ]------------
kernel BUG at net/core/skbuff.c:104!
invalid opcode: 0000 [#1] SMP KASAN
Dumping ftrace buffer:
    (ftrace buffer empty)
Modules linked in:
CPU: 1 PID: 3670 Comm: syzkaller801466 Not tainted
4.15.0-rc7-next-20180115+ #97
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS
Google 01/01/2011
RIP: 0010:skb_panic+0x162/0x1f0 net/core/skbuff.c:100
RSP: 0018:ffff8801d9bd7840 EFLAGS: 00010282
RAX: 0000000000000083 RBX: ffff8801d4f083c0 RCX: 0000000000000000
RDX: 0000000000000083 RSI: 1ffff1003b37ae92 RDI: ffffed003b37aefc
RBP: ffff8801d9bd78a8 R08: 1ffff1003b37ae8a R09: 0000000000000000
R10: 0000000000000001 R11: 0000000000000000 R12: ffffffff86200de0
R13: ffffffff84a981ad R14: 0000000000000018 R15: ffff8801d2d34180
FS:  00000000019c4880(0000) GS:ffff8801db300000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00000000208bc000 CR3: 00000001d9111001 CR4: 00000000001606e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
  skb_under_panic net/core/skbuff.c:114 [inline]
  skb_push+0xce/0xf0 net/core/skbuff.c:1714
  ipgre_header+0x6d/0x4e0 net/ipv4/ip_gre.c:879
  dev_hard_header include/linux/netdevice.h:2723 [inline]
  pppoe_sendmsg+0x58e/0x8b0 drivers/net/ppp/pppoe.c:890
  sock_sendmsg_nosec net/socket.c:630 [inline]
  sock_sendmsg+0xca/0x110 net/socket.c:640
  sock_write_iter+0x31a/0x5d0 net/socket.c:909
  call_write_iter include/linux/fs.h:1775 [inline]
  do_iter_readv_writev+0x525/0x7f0 fs/read_write.c:653
  do_iter_write+0x154/0x540 fs/read_write.c:932
  vfs_writev+0x18a/0x340 fs/read_write.c:977
  do_writev+0xfc/0x2a0 fs/read_write.c:1012
  SYSC_writev fs/read_write.c:1085 [inline]
  SyS_writev+0x27/0x30 fs/read_write.c:1082
  entry_SYSCALL_64_fastpath+0x29/0xa0

Admittedly PPPoE shouldn't be allowed to run on non Ethernet-like interfaces, but reserving space for ->needed_headroom is a more fundamental issue that needs to be addressed first.

Same problem exists for __pppoe_xmit(), which also needs to take dev->needed_headroom into account in skb_cow_head().

Fixes f5184d267c1a ("net: Allow netdevices to specify needed head/tailroom")
---
 drivers/net/ppp/pppoe.c | 11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)
drivers/net/ppp/pppoe.c +6 -5
@@ -832,6 +832,7 @@ static int pppoe_sendmsg(struct kiocb *i
 	struct pppoe_hdr *ph;
 	struct net_device *dev;
 	char *start;
+	int hlen;
 
 	lock_sock(sk);
 	if (sock_flag(sk, SOCK_DEAD) || !(sk->sk_state & PPPOX_CONNECTED)) {
@@ -850,16 +851,16 @@ static int pppoe_sendmsg(struct kiocb *i
 	if (total_len > (dev->mtu + dev->hard_header_len))
 		goto end;
 
-
-	skb = sock_wmalloc(sk, total_len + dev->hard_header_len + 32,
-			   0, GFP_KERNEL);
+	hlen = LL_RESERVED_SPACE(dev);
+	skb = sock_wmalloc(sk, hlen + sizeof(*ph) + total_len +
+			   dev->needed_tailroom, 0, GFP_KERNEL);
 	if (!skb) {
 		error = -ENOMEM;
 		goto end;
 	}
 
 	/* Reserve space for headers. */
-	skb_reserve(skb, dev->hard_header_len);
+	skb_reserve(skb, hlen);
 	skb_reset_network_header(skb);
 
 	skb->dev = dev;
@@ -920,7 +921,7 @@ static int __pppoe_xmit(struct sock *sk,
 	/* Copy the data if there is no space for the header or if it's
 	 * read-only.
 	 */
-	if (skb_cow_head(skb, sizeof(*ph) + dev->hard_header_len))
+	if (skb_cow_head(skb, LL_RESERVED_SPACE(dev) + sizeof(*ph)))
 		goto abort;
 
 	__skb_push(skb, sizeof(*ph));
#232 lore raw

PATCH 228/254 x86/microcode/intel: Extend BDW late-loading further with LLC size check

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Jia Zhang <[email protected]>

commit 7e702d17ed138cf4ae7c00e8c00681ed464587c7 upstream.

Commit b94b73733171 ("x86/microcode/intel: Extend BDW late-loading with a revision check") reduced the impact of erratum BDF90 for Broadwell model 79.

The impact can be reduced further by checking the size of the last level cache portion per core.

Tony: "The erratum says the problem only occurs on the large-cache SKUs. So we only need to avoid the update if we are on a big cache SKU that is also running old microcode."

For more details, see erratum BDF90 in document #334165 (Intel Xeon Processor E7-8800/4800 v4 Product Family Specification Update) from September 2017.

Fixes b94b73733171 ("x86/microcode/intel: Extend BDW late-loading with a revision check")
[bwh: Backported to 3.16: adjust context]
---
 arch/x86/kernel/cpu/microcode/intel.c | 20 ++++++++++++++++++--
 1 file changed, 18 insertions(+), 2 deletions(-)
arch/x86/kernel/cpu/microcode/intel.c +18 -2
@@ -87,6 +87,9 @@ MODULE_DESCRIPTION("Microcode Update Dri
 MODULE_AUTHOR("Tigran Aivazian <[email protected]>");
 MODULE_LICENSE("GPL");
 
+/* last level cache size per core */
+static int llc_size_per_core;
+
 static int collect_cpu_info(int cpu_num, struct cpu_signature *csig)
 {
 	struct cpuinfo_x86 *c = &cpu_data(cpu_num);
@@ -273,12 +276,14 @@ static bool is_blacklisted(unsigned int
 
 	/*
 	 * Late loading on model 79 with microcode revision less than 0x0b000021
-	 * may result in a system hang. This behavior is documented in item
-	 * BDF90, #334165 (Intel Xeon Processor E7-8800/4800 v4 Product Family).
+	 * and LLC size per core bigger than 2.5MB may result in a system hang.
+	 * This behavior is documented in item BDF90, #334165 (Intel Xeon
+	 * Processor E7-8800/4800 v4 Product Family).
 	 */
 	if (c->x86 == 6 &&
 	    c->x86_model == 0x4F &&
 	    c->x86_mask == 0x01 &&
+	    llc_size_per_core > 2621440 &&
 	    c->microcode < 0x0b000021) {
 		pr_err_once("Erratum BDF90: late loading with revision < 0x0b000021 (0x%x) disabled.\n", c->microcode);
 		pr_err_once("Please consider either early loading through initrd/built-in or a potential BIOS update.\n");
@@ -345,6 +350,15 @@ static struct microcode_ops microcode_in
 	.microcode_fini_cpu               = microcode_fini_cpu,
 };
 
+static int __init calc_llc_size_per_core(struct cpuinfo_x86 *c)
+{
+	u64 llc_size = c->x86_cache_size * 1024;
+
+	do_div(llc_size, c->x86_max_cores);
+
+	return (int)llc_size;
+}
+
 struct microcode_ops * __init init_intel_microcode(void)
 {
 	struct cpuinfo_x86 *c = &cpu_data(0);
@@ -355,6 +369,8 @@ struct microcode_ops * __init init_intel
 		return NULL;
 	}
 
+	llc_size_per_core = calc_llc_size_per_core(c);
+
 	return &microcode_intel_ops;
 }
 
#233 lore raw

PATCH 229/254 dccp: don't restart ccid2_hc_tx_rto_expire() if sk in closed state

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Alexey Kodanev <[email protected]>

commit dd5684ecae3bd8e44b644f50e2c12c7e57fdfef5 upstream.

ccid2_hc_tx_rto_expire() timer callback always restarts the timer again and can run indefinitely (unless it is stopped outside), and after commit 120e9dabaf55 ("dccp: defer ccid_hc_tx_delete() at dismantle time"), which moved ccid_hc_tx_delete() (also includes sk_stop_timer()) from dccp_destroy_sock() to sk_destruct(), this started to happen quite often. The timer prevents releasing the socket, as a result, sk_destruct() won't be called.

Found with LTP/dccp_ipsec tests running on the bonding device, which later couldn't be unloaded after the tests were completed:

unregister_netdevice: waiting for bond0 to become free. Usage count = 148

Fixes 2a91aa396739 ("[DCCP] CCID2: Initial CCID2 (TCP-Like) implementation")
---
 net/dccp/ccids/ccid2.c | 3 +++
 1 file changed, 3 insertions(+)
net/dccp/ccids/ccid2.c +3 -0
@@ -140,6 +140,9 @@ static void ccid2_hc_tx_rto_expire(unsig
 
 	ccid2_pr_debug("RTO_EXPIRE\n");
 
+	if (sk->sk_state == DCCP_CLOSED)
+		goto out;
+
 	/* back-off timer */
 	hc->tx_rto <<= 1;
 	if (hc->tx_rto > DCCP_RTO_MAX)
#234 lore raw

PATCH 230/254 hrtimer: Reset hrtimer cpu base proper on CPU hotplug

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Thomas Gleixner <[email protected]>

commit d5421ea43d30701e03cadc56a38854c36a8b4433 upstream.

The hrtimer interrupt code contains a hang detection and mitigation mechanism, which prevents that a long delayed hrtimer interrupt causes a continous retriggering of interrupts which prevent the system from making progress. If a hang is detected then the timer hardware is programmed with a certain delay into the future and a flag is set in the hrtimer cpu base which prevents newly enqueued timers from reprogramming the timer hardware prior to the chosen delay. The subsequent hrtimer interrupt after the delay clears the flag and resumes normal operation.

If such a hang happens in the last hrtimer interrupt before a CPU is unplugged then the hang_detected flag is set and stays that way when the CPU is plugged in again. At that point the timer hardware is not armed and it cannot be armed because the hang_detected flag is still active, so nothing clears that flag. As a consequence the CPU does not receive hrtimer interrupts and no timers expire on that CPU which results in RCU stalls and other malfunctions.

Clear the flag along with some other less critical members of the hrtimer cpu base to ensure starting from a clean state when a CPU is plugged in.

Thanks to Paul, Sebastian and Anna-Maria for their help to get down to the root cause of that hard to reproduce heisenbug. Once understood it's trivial and certainly justifies a brown paperbag.

Fixes 41d2e4949377 ("hrtimer: Tune hrtimer_interrupt hang logic")
Cc Peter Zijlstra <[email protected]>
Cc Sebastian Sewior <[email protected]>
Cc Anna-Maria Gleixner <[email protected]>

[bwh: Backported to 3.16:
- There's no next_timer field to reset
- Adjust filename, context]

---
 kernel/hrtimer.c | 3 +++
 1 file changed, 3 insertions(+)
kernel/hrtimer.c +2 -0
@@ -659,6 +659,7 @@ static int hrtimer_reprogram(struct hrti
 static inline void hrtimer_init_hres(struct hrtimer_cpu_base *base)
 {
 	base->expires_next.tv64 = KTIME_MAX;
+	base->hang_detected = 0;
 	base->hres_active = 0;
 }
 
@@ -1680,6 +1681,7 @@ static void init_hrtimers_cpu(int cpu)
 		timerqueue_init_head(&cpu_base->clock_base[i].active);
 	}
 
+	cpu_base->active_bases = 0;
 	hrtimer_init_hres(cpu_base);
 }
 
#235 lore raw

PATCH 231/254 of: fdt: Fix return with value in void function

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Ben Hutchings <[email protected]>

Commit 49e67dd17649 "of: fdt: add missing allocation-failure check" added a "return NULL" statement in __unflatten_device_tree(). When applied to the 3.16-stable branch, this introduced a compiler warning (not an error!) because the function returns void here.

---

drivers/of/fdt.c +1 -1
@@ -381,7 +381,7 @@ static void __unflatten_device_tree(void
 	/* Allocate memory for the expanded device tree */
 	mem = dt_alloc(size + 4, __alignof__(struct device_node));
 	if (!mem)
-		return NULL;
+		return;
 
 	memset(mem, 0, size);
 
#236 lore raw

PATCH 232/254 net: bcmgenet: fix bcmgenet_open()

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Petri Gynther <[email protected]>

commit fac25940c5e0d6f31d56bb2a704fadad6d5e382a upstream.

If bcmgenet_init_dma() fails, it cleans up after itself. Rx and Tx DMAs are off, and NAPI instances haven't been netif_napi_add()'ed. Therefore, we need to skip calling bcmgenet_fini_dma() on the error handling path. bcmgenet_resume() already does this correctly.

---
 drivers/net/ethernet/broadcom/genet/bcmgenet.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/broadcom/genet/bcmgenet.c +1 -1
@@ -2058,7 +2058,7 @@ static int bcmgenet_open(struct net_devi
 	ret = bcmgenet_init_dma(priv);
 	if (ret) {
 		netdev_err(dev, "failed to initialize DMA\n");
-		goto err_fini_dma;
+		goto err_clk_disable;
 	}
 
 	/* Always enable ring 16 - descriptor ring */
#237 lore raw

PATCH 233/254 mac80211_hwsim: fix compiler warning on MIPS

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Andrew Bresticker <[email protected]>

commit 5d26b50813ea6206a7bbab2e645e68044f101ac5 upstream.

The dividend in do_div() is expected to be an unsigned 64-bit integer, which leads to the following warning when building for 32-bit MIPS:

drivers/net/wireless/mac80211_hwsim.c: In function 'mac80211_hwsim_set_tsf':
drivers/net/wireless/mac80211_hwsim.c:664:98: warning: comparison of distinct pointer types lacks a cast [enabled by default]

    data->bcn_delta = do_div(delta, bcn_int);

Since we care about the signedness of delta when adjusting tsf_offset and bcm_delta, use the absolute value for the division and compare the two timestamps to determine the sign.

---
 drivers/net/wireless/mac80211_hwsim.c | 11 ++++++++---
 1 file changed, 8 insertions(+), 3 deletions(-)
drivers/net/wireless/mac80211_hwsim.c +8 -3
@@ -685,11 +685,16 @@ static void mac80211_hwsim_set_tsf(struc
 	struct mac80211_hwsim_data *data = hw->priv;
 	u64 now = mac80211_hwsim_get_tsf(hw, vif);
 	u32 bcn_int = data->beacon_int;
-	s64 delta = tsf - now;
+	u64 delta = abs64(tsf - now);
 
-	data->tsf_offset += delta;
 	/* adjust after beaconing with new timestamp at old TBTT */
-	data->bcn_delta = do_div(delta, bcn_int);
+	if (tsf > now) {
+		data->tsf_offset += delta;
+		data->bcn_delta = do_div(delta, bcn_int);
+	} else {
+		data->tsf_offset -= delta;
+		data->bcn_delta = -do_div(delta, bcn_int);
+	}
 }
 
 static void mac80211_hwsim_monitor_rx(struct ieee80211_hw *hw,
#238 lore raw

PATCH 234/254 blk-mq: fix race between timeout and freeing request

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Ming Lei <[email protected]>

commit 0048b4837affd153897ed1222283492070027aa9 upstream.

Inside timeout handler, blk_mq_tag_to_rq() is called to retrieve the request from one tag. This way is obviously wrong because the request can be freed any time and some fiedds of the request can't be trusted, then kernel oops might be triggered[1].

Currently wrt. blk_mq_tag_to_rq(), the only special case is that the flush request can share same tag with the request cloned from, and the two requests can't be active at the same time, so this patch fixes the above issue by updating tags->rqs[tag] with the active request(either flush rq or the request cloned from) of the tag.

Also blk_mq_tag_to_rq() gets much simplified with this patch.

Given blk_mq_tag_to_rq() is mainly for drivers and the caller must make sure the request can't be freed, so in bt_for_each() this helper is replaced with tags->rqs[tag].

[1] kernel oops log

[  439.696220] BUG: unable to handle kernel NULL pointer dereference at 0000000000000158^M
[  439.697162] IP: [<ffffffff812d89ba>] blk_mq_tag_to_rq+0x21/0x6e^M
[  439.700653] PGD 7ef765067 PUD 7ef764067 PMD 0 ^M
[  439.700653] Oops: 0000 [#1] PREEMPT SMP DEBUG_PAGEALLOC ^M
[  439.700653] Dumping ftrace buffer:^M
[  439.700653]    (ftrace buffer empty)^M
[  439.700653] Modules linked in: nbd ipv6 kvm_intel kvm serio_raw^M
[  439.700653] CPU: 6 PID: 2779 Comm: stress-ng-sigfd Not tainted 4.2.0-rc5-next-20150805+ #265^M
[  439.730500] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011^M
[  439.730500] task: ffff880605308000 ti: ffff88060530c000 task.ti: ffff88060530c000^M
[  439.730500] RIP: 0010:[<ffffffff812d89ba>]  [<ffffffff812d89ba>] blk_mq_tag_to_rq+0x21/0x6e^M
[  439.730500] RSP: 0018:ffff880819203da0  EFLAGS: 00010283^M
[  439.730500] RAX: ffff880811b0e000 RBX: ffff8800bb465f00 RCX: 0000000000000002^M
[  439.730500] RDX: 0000000000000000 RSI: 0000000000000202 RDI: 0000000000000000^M
[  439.730500] RBP: ffff880819203db0 R08: 0000000000000002 R09: 0000000000000000^M
[  439.730500] R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000000202^M
[  439.730500] R13: ffff880814104800 R14: 0000000000000002 R15: ffff880811a2ea00^M
[  439.730500] FS:  00007f165b3f5740(0000) GS:ffff880819200000(0000) knlGS:0000000000000000^M
[  439.730500] CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b^M
[  439.730500] CR2: 0000000000000158 CR3: 00000007ef766000 CR4: 00000000000006e0^M
[  439.730500] Stack:^M
[  439.730500]  0000000000000008 ffff8808114eed90 ffff880819203e00 ffffffff812dc104^M
[  439.755663]  ffff880819203e40 ffffffff812d9f5e 0000020000000000 ffff8808114eed80^M
[  439.755663] Call Trace:^M
[  439.755663]  <IRQ> ^M
[  439.755663]  [<ffffffff812dc104>] bt_for_each+0x6e/0xc8^M
[  439.755663]  [<ffffffff812d9f5e>] ? blk_mq_rq_timed_out+0x6a/0x6a^M
[  439.755663]  [<ffffffff812d9f5e>] ? blk_mq_rq_timed_out+0x6a/0x6a^M
[  439.755663]  [<ffffffff812dc1b3>] blk_mq_tag_busy_iter+0x55/0x5e^M
[  439.755663]  [<ffffffff812d88b4>] ? blk_mq_bio_to_request+0x38/0x38^M
[  439.755663]  [<ffffffff812d8911>] blk_mq_rq_timer+0x5d/0xd4^M
[  439.755663]  [<ffffffff810a3e10>] call_timer_fn+0xf7/0x284^M
[  439.755663]  [<ffffffff810a3d1e>] ? call_timer_fn+0x5/0x284^M
[  439.755663]  [<ffffffff812d88b4>] ? blk_mq_bio_to_request+0x38/0x38^M
[  439.755663]  [<ffffffff810a46d6>] run_timer_softirq+0x1ce/0x1f8^M
[  439.755663]  [<ffffffff8104c367>] __do_softirq+0x181/0x3a4^M
[  439.755663]  [<ffffffff8104c76e>] irq_exit+0x40/0x94^M
[  439.755663]  [<ffffffff81031482>] smp_apic_timer_interrupt+0x33/0x3e^M
[  439.755663]  [<ffffffff815559a4>] apic_timer_interrupt+0x84/0x90^M
[  439.755663]  <EOI> ^M
[  439.755663]  [<ffffffff81554350>] ? _raw_spin_unlock_irq+0x32/0x4a^M
[  439.755663]  [<ffffffff8106a98b>] finish_task_switch+0xe0/0x163^M
[  439.755663]  [<ffffffff8106a94d>] ? finish_task_switch+0xa2/0x163^M
[  439.755663]  [<ffffffff81550066>] __schedule+0x469/0x6cd^M
[  439.755663]  [<ffffffff8155039b>] schedule+0x82/0x9a^M
[  439.789267]  [<ffffffff8119b28b>] signalfd_read+0x186/0x49a^M
[  439.790911]  [<ffffffff8106d86a>] ? wake_up_q+0x47/0x47^M
[  439.790911]  [<ffffffff811618c2>] __vfs_read+0x28/0x9f^M
[  439.790911]  [<ffffffff8117a289>] ? __fget_light+0x4d/0x74^M
[  439.790911]  [<ffffffff811620a7>] vfs_read+0x7a/0xc6^M
[  439.790911]  [<ffffffff8116292b>] SyS_read+0x49/0x7f^M
[  439.790911]  [<ffffffff81554c17>] entry_SYSCALL_64_fastpath+0x12/0x6f^M
[  439.790911] Code: 48 89 e5 e8 a9 b8 e7 ff 5d c3 0f 1f 44 00 00 55 89
f2 48 89 e5 41 54 41 89 f4 53 48 8b 47 60 48 8b 1c d0 48 8b 7b 30 48 8b
53 38 <48> 8b 87 58 01 00 00 48 85 c0 75 09 48 8b 97 88 0c 00 00 eb 10
^M
[  439.790911] RIP  [<ffffffff812d89ba>] blk_mq_tag_to_rq+0x21/0x6e^M
[  439.790911]  RSP <ffff880819203da0>^M
[  439.790911] CR2: 0000000000000158^M
[  439.790911] ---[ end trace d40af58949325661 ]---^M

[bwh: Backported to 3.16:
- Flush state is in struct request_queue, not struct blk_flush_queue
- Flush request cloning is done in blk_mq_clone_flush_request() rather than blk_kick_flush()
- Drop changes in bt{,_tags}_for_each()
- Adjust filename, context]

---
 block/blk-flush.c  | 15 ++++++++++++++-
 block/blk-mq-tag.c |  4 ++--
 block/blk-mq-tag.h | 12 ++++++++++++
 block/blk-mq.c     | 16 +---------------
 block/blk.h        |  6 ++++++
 5 files changed, 35 insertions(+), 18 deletions(-)
block/blk-flush.c +6 -0
@@ -73,6 +73,7 @@
 
 #include "blk.h"
 #include "blk-mq.h"
+#include "blk-mq-tag.h"
 
 /* FLUSH/FUA sequences */
 enum {
@@ -224,7 +225,12 @@ static void flush_end_io(struct request
 	unsigned long flags = 0;
 
 	if (q->mq_ops) {
+		struct blk_mq_hw_ctx *hctx;
+
+		/* release the tag's ownership to the req cloned from */
 		spin_lock_irqsave(&q->mq_flush_lock, flags);
+		hctx = q->mq_ops->map_queue(q, q->flush_rq->mq_ctx->cpu);
+		blk_mq_tag_set_rq(hctx, q->flush_rq->tag, q->orig_rq);
 		q->flush_rq->tag = -1;
 	}
 
block/blk-mq-tag.h +12 -0
@@ -85,4 +85,16 @@ static inline void blk_mq_tag_idle(struc
 	__blk_mq_tag_idle(hctx);
 }
 
+/*
+ * This helper should only be used for flush request to share tag
+ * with the request cloned from, and both the two requests can't be
+ * in flight at the same time. The caller has to make sure the tag
+ * can't be freed.
+ */
+static inline void blk_mq_tag_set_rq(struct blk_mq_hw_ctx *hctx,
+		unsigned int tag, struct request *rq)
+{
+	hctx->tags->rqs[tag] = rq;
+}
+
 #endif
block/blk-mq.c +4 -12
@@ -310,6 +310,9 @@ void blk_mq_clone_flush_request(struct r
 	flush_rq->tag = orig_rq->tag;
 	memcpy(blk_mq_rq_to_pdu(flush_rq), blk_mq_rq_to_pdu(orig_rq),
 		hctx->cmd_size);
+	orig_rq->q->orig_rq = orig_rq;
+
+	blk_mq_tag_set_rq(hctx, orig_rq->tag, flush_rq);
 }
 
 inline void __blk_mq_end_io(struct request *rq, int error)
@@ -520,20 +523,9 @@ void blk_mq_kick_requeue_list(struct req
 }
 EXPORT_SYMBOL(blk_mq_kick_requeue_list);
 
-static inline bool is_flush_request(struct request *rq, unsigned int tag)
-{
-	return ((rq->cmd_flags & REQ_FLUSH_SEQ) &&
-			rq->q->flush_rq->tag == tag);
-}
-
 struct request *blk_mq_tag_to_rq(struct blk_mq_tags *tags, unsigned int tag)
 {
-	struct request *rq = tags->rqs[tag];
-
-	if (!is_flush_request(rq, tag))
-		return rq;
-
-	return rq->q->flush_rq;
+	return tags->rqs[tag];
 }
 EXPORT_SYMBOL(blk_mq_tag_to_rq);
 
include/linux/blkdev.h +6 -0
@@ -462,6 +462,12 @@ struct request_queue {
 	struct list_head	flush_queue[2];
 	struct list_head	flush_data_in_flight;
 	struct request		*flush_rq;
+
+	/*
+	 * flush_rq shares tag with this rq, both can't be active
+	 * at the same time
+	 */
+	struct request		*orig_rq;
 	spinlock_t		mq_flush_lock;
 
 	struct list_head	requeue_list;
#239 lore raw

PATCH 235/254 v4l2-compat-ioctl32: fix sparse warnings

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Hans Verkuil <[email protected]>

commit 8ae632b11775254c5e555ee8c42b7d19baeb1473 upstream.

A lot of these warnings are caused by the fact that we don't generally use __user in videodev2.h. Normally the video_usercopy function will copy anything pointed to by pointers into kernel space, so having __user in the struct will only cause lots of warnings in the drivers. But the flip side of that is that you need to add __force casts here.

drivers/media/v4l2-core/v4l2-compat-ioctl32.c:337:26: warning: incorrect type in argument 1 (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:337:30: warning: incorrect type in argument 2 (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:338:31: warning: incorrect type in argument 1 (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:338:49: warning: incorrect type in argument 2 (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:343:21: warning: incorrect type in argument 1 (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:346:21: warning: incorrect type in argument 1 (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:349:35: warning: incorrect type in argument 1 (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:349:46: warning: incorrect type in argument 2 (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:352:35: warning: incorrect type in argument 1 (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:352:54: warning: incorrect type in argument 2 (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:363:26: warning: incorrect type in argument 1 (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:363:32: warning: incorrect type in argument 2 (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:364:31: warning: incorrect type in argument 1 (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:364:51: warning: incorrect type in argument 2 (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:371:35: warning: incorrect type in argument 1 (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:371:56: warning: incorrect type in argument 2 (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:376:35: warning: incorrect type in argument 1 (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:376:48: warning: incorrect type in argument 2 (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:430:30: warning: incorrect type in assignment (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:433:48: warning: incorrect type in argument 1 (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:433:56: warning: incorrect type in argument 2 (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:501:24: warning: incorrect type in assignment (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:507:48: warning: incorrect type in argument 1 (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:507:56: warning: incorrect type in argument 2 (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:565:18: warning: incorrect type in assignment (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:670:22: warning: incorrect type in assignment (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:680:29: warning: incorrect type in assignment (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:692:55: warning: incorrect type in initializer (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:773:18: warning: incorrect type in assignment (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:786:30: warning: incorrect type in argument 1 (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:786:44: warning: incorrect type in argument 2 (different address spaces) drivers/media/v4l2-core/v4l2-compat-ioctl32.c:674:37: warning: dereference of noderef expression drivers/media/v4l2-core/v4l2-compat-ioctl32.c:718:37: warning: dereference of noderef expression

[bwh: Backported to 3.16: adjust context]
---
 drivers/media/v4l2-core/v4l2-compat-ioctl32.c | 30 +++++++++++++++++----------
 1 file changed, 19 insertions(+), 11 deletions(-)
drivers/media/v4l2-core/v4l2-compat-ioctl32.c +19 -11
@@ -328,7 +328,7 @@ struct v4l2_buffer32 {
 	__u32			reserved;
 };
 
-static int get_v4l2_plane32(struct v4l2_plane *up, struct v4l2_plane32 *up32,
+static int get_v4l2_plane32(struct v4l2_plane __user *up, struct v4l2_plane32 __user *up32,
 				enum v4l2_memory memory)
 {
 	void __user *up_pln;
@@ -357,7 +357,7 @@ static int get_v4l2_plane32(struct v4l2_
 	return 0;
 }
 
-static int put_v4l2_plane32(struct v4l2_plane *up, struct v4l2_plane32 *up32,
+static int put_v4l2_plane32(struct v4l2_plane __user *up, struct v4l2_plane32 __user *up32,
 				enum v4l2_memory memory)
 {
 	if (copy_in_user(up32, up, 2 * sizeof(__u32)) ||
@@ -425,7 +425,7 @@ static int get_v4l2_buffer32(struct v4l2
 		 * by passing a very big num_planes value */
 		uplane = compat_alloc_user_space(num_planes *
 						sizeof(struct v4l2_plane));
-		kp->m.planes = uplane;
+		kp->m.planes = (__force struct v4l2_plane *)uplane;
 
 		while (--num_planes >= 0) {
 			ret = get_v4l2_plane32(uplane, uplane32, kp->memory);
@@ -495,7 +495,7 @@ static int put_v4l2_buffer32(struct v4l2
 		if (num_planes == 0)
 			return 0;
 
-		uplane = kp->m.planes;
+		uplane = (__force struct v4l2_plane __user *)kp->m.planes;
 		if (get_user(p, &up->m.planes))
 			return -EFAULT;
 		uplane32 = compat_ptr(p);
@@ -547,7 +547,7 @@ static int get_v4l2_framebuffer32(struct
 		get_user(kp->capability, &up->capability) ||
 		get_user(kp->flags, &up->flags))
 			return -EFAULT;
-	kp->base = compat_ptr(tmp);
+	kp->base = (__force void *)compat_ptr(tmp);
 	get_v4l2_pix_format(&kp->fmt, &up->fmt);
 	return 0;
 }
@@ -653,11 +653,15 @@ static int get_v4l2_ext_controls32(struc
 			n * sizeof(struct v4l2_ext_control32)))
 		return -EFAULT;
 	kcontrols = compat_alloc_user_space(n * sizeof(struct v4l2_ext_control));
-	kp->controls = kcontrols;
+	kp->controls = (__force struct v4l2_ext_control *)kcontrols;
 	while (--n >= 0) {
+		u32 id;
+
 		if (copy_in_user(kcontrols, ucontrols, sizeof(*ucontrols)))
 			return -EFAULT;
-		if (ctrl_is_pointer(kcontrols->id)) {
+		if (get_user(id, &kcontrols->id))
+			return -EFAULT;
+		if (ctrl_is_pointer(id)) {
 			void __user *s;
 
 			if (get_user(p, &ucontrols->string))
@@ -675,7 +679,8 @@ static int get_v4l2_ext_controls32(struc
 static int put_v4l2_ext_controls32(struct v4l2_ext_controls *kp, struct v4l2_ext_controls32 __user *up)
 {
 	struct v4l2_ext_control32 __user *ucontrols;
-	struct v4l2_ext_control __user *kcontrols = kp->controls;
+	struct v4l2_ext_control __user *kcontrols =
+		(__force struct v4l2_ext_control __user *)kp->controls;
 	int n = kp->count;
 	compat_caddr_t p;
 
@@ -697,11 +702,14 @@ static int put_v4l2_ext_controls32(struc
 
 	while (--n >= 0) {
 		unsigned size = sizeof(*ucontrols);
+		u32 id;
 
+		if (get_user(id, &kcontrols->id))
+			return -EFAULT;
 		/* Do not modify the pointer when copying a pointer control.
 		   The contents of the pointer was changed, not the pointer
 		   itself. */
-		if (ctrl_is_pointer(kcontrols->id))
+		if (ctrl_is_pointer(id))
 			size -= sizeof(ucontrols->value64);
 		if (copy_in_user(ucontrols, kcontrols, size))
 			return -EFAULT;
@@ -758,7 +766,7 @@ static int get_v4l2_edid32(struct v4l2_e
 		get_user(tmp, &up->edid) ||
 		copy_from_user(kp->reserved, up->reserved, sizeof(kp->reserved)))
 			return -EFAULT;
-	kp->edid = compat_ptr(tmp);
+	kp->edid = (__force u8 *)compat_ptr(tmp);
 	return 0;
 }
 
@@ -771,7 +779,7 @@ static int put_v4l2_edid32(struct v4l2_e
 		put_user(kp->start_block, &up->start_block) ||
 		put_user(kp->blocks, &up->blocks) ||
 		put_user(tmp, &up->edid) ||
-		copy_to_user(kp->reserved, up->reserved, sizeof(kp->reserved)))
+		copy_to_user(up->reserved, kp->reserved, sizeof(up->reserved)))
 			return -EFAULT;
 	return 0;
 }
#240 lore raw

PATCH 236/254 V4L2: fix VIDIOC_CREATE_BUFS 32-bit compatibility mode data copy-back

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Guennadi Liakhovetski <[email protected]>

commit 6ed9b28504326f8cf542e6b68245b2f7ce009216 upstream.

Similar to an earlier patch, fixing reading user-space data for the VIDIOC_CREATE_BUFS ioctl() in 32-bit compatibility mode, this patch fixes writing back of the possibly modified struct to the user. However, unlike the former bug, this one is much less harmful, because it only results in the kernel failing to write the .type field back to the user, but in fact this is likely unneeded, because the kernel will hardly want to change that field. Therefore this bug is more of a theoretical nature.

---
 drivers/media/v4l2-core/v4l2-compat-ioctl32.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)
drivers/media/v4l2-core/v4l2-compat-ioctl32.c +6 -4
@@ -222,6 +222,9 @@ static int get_v4l2_create32(struct v4l2
 
 static int __put_v4l2_format32(struct v4l2_format *kp, struct v4l2_format32 __user *up)
 {
+	if (put_user(kp->type, &up->type))
+		return -EFAULT;
+
 	switch (kp->type) {
 	case V4L2_BUF_TYPE_VIDEO_CAPTURE:
 	case V4L2_BUF_TYPE_VIDEO_OUTPUT:
@@ -248,8 +251,7 @@ static int __put_v4l2_format32(struct v4
 
 static int put_v4l2_format32(struct v4l2_format *kp, struct v4l2_format32 __user *up)
 {
-	if (!access_ok(VERIFY_WRITE, up, sizeof(struct v4l2_format32)) ||
-		put_user(kp->type, &up->type))
+	if (!access_ok(VERIFY_WRITE, up, sizeof(struct v4l2_format32)))
 		return -EFAULT;
 	return __put_v4l2_format32(kp, up);
 }
@@ -257,8 +259,8 @@ static int put_v4l2_format32(struct v4l2
 static int put_v4l2_create32(struct v4l2_create_buffers *kp, struct v4l2_create_buffers32 __user *up)
 {
 	if (!access_ok(VERIFY_WRITE, up, sizeof(struct v4l2_create_buffers32)) ||
-	    copy_to_user(up, kp, offsetof(struct v4l2_create_buffers32, format.fmt)))
-			return -EFAULT;
+	    copy_to_user(up, kp, offsetof(struct v4l2_create_buffers32, format)))
+		return -EFAULT;
 	return __put_v4l2_format32(&kp->format, &up->format);
 }
 
#241 lore raw

PATCH 237/254 media: v4l2-compat-ioctl32: fix missing reserved field copy in put_v4l2_create32

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Tiffany Lin <[email protected]>

commit baf43c6eace43868e490f18560287fa3481b2159 upstream.

In v4l2-compliance utility, test VIDIOC_CREATE_BUFS will check whether reserved filed of v4l2_create_buffers filled with zero
Reserved field is filled with zero in v4l_create_bufs.
This patch copy reserved field of v4l2_create_buffer from kernel space to user space

---
 drivers/media/v4l2-core/v4l2-compat-ioctl32.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/media/v4l2-core/v4l2-compat-ioctl32.c +2 -1
@@ -259,7 +259,8 @@ static int put_v4l2_format32(struct v4l2
 static int put_v4l2_create32(struct v4l2_create_buffers *kp, struct v4l2_create_buffers32 __user *up)
 {
 	if (!access_ok(VERIFY_WRITE, up, sizeof(struct v4l2_create_buffers32)) ||
-	    copy_to_user(up, kp, offsetof(struct v4l2_create_buffers32, format)))
+	    copy_to_user(up, kp, offsetof(struct v4l2_create_buffers32, format)) ||
+	    copy_to_user(up->reserved, kp->reserved, sizeof(kp->reserved)))
 		return -EFAULT;
 	return __put_v4l2_format32(&kp->format, &up->format);
 }
#242 lore raw

PATCH 238/254 media: v4l2-compat-ioctl32.c: add capabilities field to, v4l2_input32

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Hans Verkuil <[email protected]>

commit 037e0865c2ecbaa4558feba239ece08d7e457ec0 upstream.

The v4l2_input32 struct wasn't updated when this field was added. It didn't cause a failure in the compat code, but it is better to keep it in sync with v4l2_input to avoid confusion.

---
 drivers/media/v4l2-core/v4l2-compat-ioctl32.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/media/v4l2-core/v4l2-compat-ioctl32.c +2 -1
@@ -576,7 +576,8 @@ struct v4l2_input32 {
 	__u32        tuner;             /*  Associated tuner */
 	compat_u64   std;
 	__u32	     status;
-	__u32	     reserved[4];
+	__u32	     capabilities;
+	__u32	     reserved[3];
 };
 
 /* The 64-bit v4l2_input struct has extra padding at the end of the struct.
#243 lore raw

PATCH 239/254 adv7604: use correct drive strength defines

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Hans Verkuil <[email protected]>

The prefix is ADV7604_, not ADV76XX.

Fixes f31b62e14a ("adv7604: add hdmi driver strength adjustment")
---
 drivers/media/i2c/adv7604.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/media/i2c/adv7604.c +3 -3
index af8a99716de5..9e0e592f50ab 100644
@@ -2735,9 +2735,9 @@ static int adv7604_parse_dt(struct adv7604_state *state)
 	state->pdata.alt_data_sat = 1;
 	state->pdata.op_format_mode_sel = ADV7604_OP_FORMAT_MODE0;
 	state->pdata.bus_order = ADV7604_BUS_ORDER_RGB;
-	state->pdata.dr_str_data = ADV76XX_DR_STR_MEDIUM_HIGH;
-	state->pdata.dr_str_clk = ADV76XX_DR_STR_MEDIUM_HIGH;
-	state->pdata.dr_str_sync = ADV76XX_DR_STR_MEDIUM_HIGH;
+	state->pdata.dr_str_data = ADV7604_DR_STR_MEDIUM_HIGH;
+	state->pdata.dr_str_clk = ADV7604_DR_STR_MEDIUM_HIGH;
+	state->pdata.dr_str_sync = ADV7604_DR_STR_MEDIUM_HIGH;
 
 	return 0;
 }
#244 lore raw

PATCH 240/254 media: v4l2-ioctl.c: don't copy back the result for -ENOTTY

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Hans Verkuil <[email protected]>

commit 181a4a2d5a0a7b43cab08a70710d727e7764ccdd upstream.

If the ioctl returned -ENOTTY, then don't bother copying back the result as there is no point.

---
 drivers/media/v4l2-core/v4l2-ioctl.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
drivers/media/v4l2-core/v4l2-ioctl.c +4 -1
index 16bffd851bf9..e2f71def945a 100644
@@ -2402,8 +2402,11 @@ video_usercopy(struct file *file, unsigned int cmd, unsigned long arg,
 
 	/* Handles IOCTL */
 	err = func(file, cmd, parg);
-	if (err == -ENOIOCTLCMD)
+	if (err == -ENOTTY || err == -ENOIOCTLCMD) {
 		err = -ENOTTY;
+		goto out;
+	}
+
 	if (err == 0) {
 		if (cmd == VIDIOC_DQBUF)
 			trace_v4l2_dqbuf(video_devdata(file)->minor, parg);
#245 lore raw

PATCH 241/254 vb2: V4L2_BUF_FLAG_DONE is set after DQBUF

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Ricardo Ribalda <[email protected]>

commit 3171cc2b4eb9831ab4df1d80d0410a945b8bc84e upstream.

According to the doc, V4L2_BUF_FLAG_DONE is cleared after DQBUF:

V4L2_BUF_FLAG_DONE 0x00000004 ... After calling the VIDIOC_QBUF or VIDIOC_DQBUF it is always cleared ...

Unfortunately, it seems that videobuf2 keeps it set after DQBUF. This can be tested with vivid and dev_debug:

[257604.338082] video1: VIDIOC_DQBUF: 71:33:25.00260479 index=3,

type=vid-cap, flags=0x00002004, field=none, sequence=163, memory=userptr, bytesused=460800, offset/userptr=0x344b000, length=460800

This patch forces FLAG_DONE to 0 after calling DQBUF.

---
 drivers/media/v4l2-core/videobuf2-core.c | 5 +++++
 1 file changed, 5 insertions(+)
drivers/media/v4l2-core/videobuf2-core.c +5 -0
index e63a7904bf5b..ee8b697972bb 100644
@@ -2069,6 +2069,11 @@ static int vb2_internal_dqbuf(struct vb2_queue *q, struct v4l2_buffer *b, bool n
 	dprintk(1, "dqbuf of buffer %d, with state %d\n",
 			vb->v4l2_buf.index, vb->state);
 
+	/*
+	 * After calling the VIDIOC_DQBUF V4L2_BUF_FLAG_DONE must be
+	 * cleared.
+	 */
+	b->flags &= ~V4L2_BUF_FLAG_DONE;
 	return 0;
 }
 
#246 lore raw

PATCH 242/254 media: v4l2-compat-ioctl32.c: add missing VIDIOC_PREPARE_BUF

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Hans Verkuil <[email protected]>

commit 3ee6d040719ae09110e5cdf24d5386abe5d1b776 upstream.

The result of the VIDIOC_PREPARE_BUF ioctl was never copied back to userspace since it was missing in the switch.

---
 drivers/media/v4l2-core/v4l2-compat-ioctl32.c | 1 +
 1 file changed, 1 insertion(+)
drivers/media/v4l2-core/v4l2-compat-ioctl32.c +1 -0
@@ -992,6 +992,7 @@ static long do_video_ioctl(struct file *
 		err = put_v4l2_create32(&karg.v2crt, up);
 		break;
 
+	case VIDIOC_PREPARE_BUF:
 	case VIDIOC_QUERYBUF:
 	case VIDIOC_QBUF:
 	case VIDIOC_DQBUF:
#247 lore raw

PATCH 243/254 media: v4l2-compat-ioctl32.c: fix the indentation

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Hans Verkuil <[email protected]>

commit b7b957d429f601d6d1942122b339474f31191d75 upstream.

The indentation of this source is all over the place. Fix this. This patch only changes whitespace.

---

drivers/media/v4l2-core/v4l2-compat-ioctl32.c +106 -106
@@ -48,11 +48,11 @@ struct v4l2_window32 {
 static int get_v4l2_window32(struct v4l2_window *kp, struct v4l2_window32 __user *up)
 {
 	if (!access_ok(VERIFY_READ, up, sizeof(struct v4l2_window32)) ||
-		copy_from_user(&kp->w, &up->w, sizeof(up->w)) ||
-		get_user(kp->field, &up->field) ||
-		get_user(kp->chromakey, &up->chromakey) ||
-		get_user(kp->clipcount, &up->clipcount))
-			return -EFAULT;
+	    copy_from_user(&kp->w, &up->w, sizeof(up->w)) ||
+	    get_user(kp->field, &up->field) ||
+	    get_user(kp->chromakey, &up->chromakey) ||
+	    get_user(kp->clipcount, &up->clipcount))
+		return -EFAULT;
 	if (kp->clipcount > 2048)
 		return -EINVAL;
 	if (kp->clipcount) {
@@ -82,10 +82,10 @@ static int get_v4l2_window32(struct v4l2
 static int put_v4l2_window32(struct v4l2_window *kp, struct v4l2_window32 __user *up)
 {
 	if (copy_to_user(&up->w, &kp->w, sizeof(kp->w)) ||
-		put_user(kp->field, &up->field) ||
-		put_user(kp->chromakey, &up->chromakey) ||
-		put_user(kp->clipcount, &up->clipcount))
-			return -EFAULT;
+	    put_user(kp->field, &up->field) ||
+	    put_user(kp->chromakey, &up->chromakey) ||
+	    put_user(kp->clipcount, &up->clipcount))
+		return -EFAULT;
 	return 0;
 }
 
@@ -97,7 +97,7 @@ static inline int get_v4l2_pix_format(st
 }
 
 static inline int get_v4l2_pix_format_mplane(struct v4l2_pix_format_mplane *kp,
-				struct v4l2_pix_format_mplane __user *up)
+					     struct v4l2_pix_format_mplane __user *up)
 {
 	if (copy_from_user(kp, up, sizeof(struct v4l2_pix_format_mplane)))
 		return -EFAULT;
@@ -112,7 +112,7 @@ static inline int put_v4l2_pix_format(st
 }
 
 static inline int put_v4l2_pix_format_mplane(struct v4l2_pix_format_mplane *kp,
-				struct v4l2_pix_format_mplane __user *up)
+					     struct v4l2_pix_format_mplane __user *up)
 {
 	if (copy_to_user(up, kp, sizeof(struct v4l2_pix_format_mplane)))
 		return -EFAULT;
@@ -200,7 +200,7 @@ static int __get_v4l2_format32(struct v4
 		return get_v4l2_sliced_vbi_format(&kp->fmt.sliced, &up->fmt.sliced);
 	default:
 		printk(KERN_INFO "compat_ioctl32: unexpected VIDIOC_FMT type %d\n",
-								kp->type);
+		       kp->type);
 		return -EINVAL;
 	}
 }
@@ -244,7 +244,7 @@ static int __put_v4l2_format32(struct v4
 		return put_v4l2_sliced_vbi_format(&kp->fmt.sliced, &up->fmt.sliced);
 	default:
 		printk(KERN_INFO "compat_ioctl32: unexpected VIDIOC_FMT type %d\n",
-								kp->type);
+		       kp->type);
 		return -EINVAL;
 	}
 }
@@ -278,7 +278,7 @@ static int get_v4l2_standard32(struct v4
 {
 	/* other fields are not set by the user, nor used by the driver */
 	if (!access_ok(VERIFY_READ, up, sizeof(struct v4l2_standard32)) ||
-		get_user(kp->index, &up->index))
+	    get_user(kp->index, &up->index))
 		return -EFAULT;
 	return 0;
 }
@@ -286,13 +286,13 @@ static int get_v4l2_standard32(struct v4
 static int put_v4l2_standard32(struct v4l2_standard *kp, struct v4l2_standard32 __user *up)
 {
 	if (!access_ok(VERIFY_WRITE, up, sizeof(struct v4l2_standard32)) ||
-		put_user(kp->index, &up->index) ||
-		put_user(kp->id, &up->id) ||
-		copy_to_user(up->name, kp->name, 24) ||
-		copy_to_user(&up->frameperiod, &kp->frameperiod, sizeof(kp->frameperiod)) ||
-		put_user(kp->framelines, &up->framelines) ||
-		copy_to_user(up->reserved, kp->reserved, 4 * sizeof(__u32)))
-			return -EFAULT;
+	    put_user(kp->index, &up->index) ||
+	    put_user(kp->id, &up->id) ||
+	    copy_to_user(up->name, kp->name, 24) ||
+	    copy_to_user(&up->frameperiod, &kp->frameperiod, sizeof(kp->frameperiod)) ||
+	    put_user(kp->framelines, &up->framelines) ||
+	    copy_to_user(up->reserved, kp->reserved, 4 * sizeof(__u32)))
+		return -EFAULT;
 	return 0;
 }
 
@@ -332,14 +332,14 @@ struct v4l2_buffer32 {
 };
 
 static int get_v4l2_plane32(struct v4l2_plane __user *up, struct v4l2_plane32 __user *up32,
-				enum v4l2_memory memory)
+			    enum v4l2_memory memory)
 {
 	void __user *up_pln;
 	compat_long_t p;
 
 	if (copy_in_user(up, up32, 2 * sizeof(__u32)) ||
-		copy_in_user(&up->data_offset, &up32->data_offset,
-				sizeof(__u32)))
+	    copy_in_user(&up->data_offset, &up32->data_offset,
+			 sizeof(__u32)))
 		return -EFAULT;
 
 	if (memory == V4L2_MEMORY_USERPTR) {
@@ -353,7 +353,7 @@ static int get_v4l2_plane32(struct v4l2_
 			return -EFAULT;
 	} else {
 		if (copy_in_user(&up->m.mem_offset, &up32->m.mem_offset,
-					sizeof(__u32)))
+				 sizeof(__u32)))
 			return -EFAULT;
 	}
 
@@ -361,23 +361,23 @@ static int get_v4l2_plane32(struct v4l2_
 }
 
 static int put_v4l2_plane32(struct v4l2_plane __user *up, struct v4l2_plane32 __user *up32,
-				enum v4l2_memory memory)
+			    enum v4l2_memory memory)
 {
 	if (copy_in_user(up32, up, 2 * sizeof(__u32)) ||
-		copy_in_user(&up32->data_offset, &up->data_offset,
-				sizeof(__u32)))
+	    copy_in_user(&up32->data_offset, &up->data_offset,
+			 sizeof(__u32)))
 		return -EFAULT;
 
 	/* For MMAP, driver might've set up the offset, so copy it back.
 	 * USERPTR stays the same (was userspace-provided), so no copying. */
 	if (memory == V4L2_MEMORY_MMAP)
 		if (copy_in_user(&up32->m.mem_offset, &up->m.mem_offset,
-					sizeof(__u32)))
+				 sizeof(__u32)))
 			return -EFAULT;
 	/* For DMABUF, driver might've set up the fd, so copy it back. */
 	if (memory == V4L2_MEMORY_DMABUF)
 		if (copy_in_user(&up32->m.fd, &up->m.fd,
-					sizeof(int)))
+				 sizeof(int)))
 			return -EFAULT;
 
 	return 0;
@@ -392,19 +392,19 @@ static int get_v4l2_buffer32(struct v4l2
 	int ret;
 
 	if (!access_ok(VERIFY_READ, up, sizeof(struct v4l2_buffer32)) ||
-		get_user(kp->index, &up->index) ||
-		get_user(kp->type, &up->type) ||
-		get_user(kp->flags, &up->flags) ||
-		get_user(kp->memory, &up->memory) ||
-		get_user(kp->length, &up->length))
-			return -EFAULT;
+	    get_user(kp->index, &up->index) ||
+	    get_user(kp->type, &up->type) ||
+	    get_user(kp->flags, &up->flags) ||
+	    get_user(kp->memory, &up->memory) ||
+	    get_user(kp->length, &up->length))
+		return -EFAULT;
 
 	if (V4L2_TYPE_IS_OUTPUT(kp->type))
 		if (get_user(kp->bytesused, &up->bytesused) ||
-			get_user(kp->field, &up->field) ||
-			get_user(kp->timestamp.tv_sec, &up->timestamp.tv_sec) ||
-			get_user(kp->timestamp.tv_usec,
-					&up->timestamp.tv_usec))
+		    get_user(kp->field, &up->field) ||
+		    get_user(kp->timestamp.tv_sec, &up->timestamp.tv_sec) ||
+		    get_user(kp->timestamp.tv_usec,
+			     &up->timestamp.tv_usec))
 			return -EFAULT;
 
 	if (V4L2_TYPE_IS_MULTIPLANAR(kp->type)) {
@@ -421,13 +421,13 @@ static int get_v4l2_buffer32(struct v4l2
 
 		uplane32 = compat_ptr(p);
 		if (!access_ok(VERIFY_READ, uplane32,
-				num_planes * sizeof(struct v4l2_plane32)))
+			       num_planes * sizeof(struct v4l2_plane32)))
 			return -EFAULT;
 
 		/* We don't really care if userspace decides to kill itself
 		 * by passing a very big num_planes value */
 		uplane = compat_alloc_user_space(num_planes *
-						sizeof(struct v4l2_plane));
+						 sizeof(struct v4l2_plane));
 		kp->m.planes = (__force struct v4l2_plane *)uplane;
 
 		while (--num_planes >= 0) {
@@ -445,12 +445,12 @@ static int get_v4l2_buffer32(struct v4l2
 			break;
 		case V4L2_MEMORY_USERPTR:
 			{
-			compat_long_t tmp;
+				compat_long_t tmp;
 
-			if (get_user(tmp, &up->m.userptr))
-				return -EFAULT;
+				if (get_user(tmp, &up->m.userptr))
+					return -EFAULT;
 
-			kp->m.userptr = (unsigned long)compat_ptr(tmp);
+				kp->m.userptr = (unsigned long)compat_ptr(tmp);
 			}
 			break;
 		case V4L2_MEMORY_OVERLAY:
@@ -476,22 +476,22 @@ static int put_v4l2_buffer32(struct v4l2
 	int ret;
 
 	if (!access_ok(VERIFY_WRITE, up, sizeof(struct v4l2_buffer32)) ||
-		put_user(kp->index, &up->index) ||
-		put_user(kp->type, &up->type) ||
-		put_user(kp->flags, &up->flags) ||
-		put_user(kp->memory, &up->memory))
-			return -EFAULT;
+	    put_user(kp->index, &up->index) ||
+	    put_user(kp->type, &up->type) ||
+	    put_user(kp->flags, &up->flags) ||
+	    put_user(kp->memory, &up->memory))
+		return -EFAULT;
 
 	if (put_user(kp->bytesused, &up->bytesused) ||
-		put_user(kp->field, &up->field) ||
-		put_user(kp->timestamp.tv_sec, &up->timestamp.tv_sec) ||
-		put_user(kp->timestamp.tv_usec, &up->timestamp.tv_usec) ||
-		copy_to_user(&up->timecode, &kp->timecode, sizeof(struct v4l2_timecode)) ||
-		put_user(kp->sequence, &up->sequence) ||
-		put_user(kp->reserved2, &up->reserved2) ||
-		put_user(kp->reserved, &up->reserved) ||
-		put_user(kp->length, &up->length))
-			return -EFAULT;
+	    put_user(kp->field, &up->field) ||
+	    put_user(kp->timestamp.tv_sec, &up->timestamp.tv_sec) ||
+	    put_user(kp->timestamp.tv_usec, &up->timestamp.tv_usec) ||
+	    copy_to_user(&up->timecode, &kp->timecode, sizeof(struct v4l2_timecode)) ||
+	    put_user(kp->sequence, &up->sequence) ||
+	    put_user(kp->reserved2, &up->reserved2) ||
+	    put_user(kp->reserved, &up->reserved) ||
+	    put_user(kp->length, &up->length))
+		return -EFAULT;
 
 	if (V4L2_TYPE_IS_MULTIPLANAR(kp->type)) {
 		num_planes = kp->length;
@@ -546,10 +546,10 @@ static int get_v4l2_framebuffer32(struct
 	u32 tmp;
 
 	if (!access_ok(VERIFY_READ, up, sizeof(struct v4l2_framebuffer32)) ||
-		get_user(tmp, &up->base) ||
-		get_user(kp->capability, &up->capability) ||
-		get_user(kp->flags, &up->flags))
-			return -EFAULT;
+	    get_user(tmp, &up->base) ||
+	    get_user(kp->capability, &up->capability) ||
+	    get_user(kp->flags, &up->flags))
+		return -EFAULT;
 	kp->base = (__force void *)compat_ptr(tmp);
 	get_v4l2_pix_format(&kp->fmt, &up->fmt);
 	return 0;
@@ -560,10 +560,10 @@ static int put_v4l2_framebuffer32(struct
 	u32 tmp = (u32)((unsigned long)kp->base);
 
 	if (!access_ok(VERIFY_WRITE, up, sizeof(struct v4l2_framebuffer32)) ||
-		put_user(tmp, &up->base) ||
-		put_user(kp->capability, &up->capability) ||
-		put_user(kp->flags, &up->flags))
-			return -EFAULT;
+	    put_user(tmp, &up->base) ||
+	    put_user(kp->capability, &up->capability) ||
+	    put_user(kp->flags, &up->flags))
+		return -EFAULT;
 	put_v4l2_pix_format(&kp->fmt, &up->fmt);
 	return 0;
 }
@@ -597,11 +597,11 @@ static inline int put_v4l2_input32(struc
 }
 
 struct v4l2_ext_controls32 {
-       __u32 ctrl_class;
-       __u32 count;
-       __u32 error_idx;
-       __u32 reserved[2];
-       compat_caddr_t controls; /* actually struct v4l2_ext_control32 * */
+	__u32 ctrl_class;
+	__u32 count;
+	__u32 error_idx;
+	__u32 reserved[2];
+	compat_caddr_t controls; /* actually struct v4l2_ext_control32 * */
 };
 
 struct v4l2_ext_control32 {
@@ -640,11 +640,11 @@ static int get_v4l2_ext_controls32(struc
 	compat_caddr_t p;
 
 	if (!access_ok(VERIFY_READ, up, sizeof(struct v4l2_ext_controls32)) ||
-		get_user(kp->ctrl_class, &up->ctrl_class) ||
-		get_user(kp->count, &up->count) ||
-		get_user(kp->error_idx, &up->error_idx) ||
-		copy_from_user(kp->reserved, up->reserved, sizeof(kp->reserved)))
-			return -EFAULT;
+	    get_user(kp->ctrl_class, &up->ctrl_class) ||
+	    get_user(kp->count, &up->count) ||
+	    get_user(kp->error_idx, &up->error_idx) ||
+	    copy_from_user(kp->reserved, up->reserved, sizeof(kp->reserved)))
+		return -EFAULT;
 	n = kp->count;
 	if (n == 0) {
 		kp->controls = NULL;
@@ -654,7 +654,7 @@ static int get_v4l2_ext_controls32(struc
 		return -EFAULT;
 	ucontrols = compat_ptr(p);
 	if (!access_ok(VERIFY_READ, ucontrols,
-			n * sizeof(struct v4l2_ext_control32)))
+		       n * sizeof(struct v4l2_ext_control32)))
 		return -EFAULT;
 	kcontrols = compat_alloc_user_space(n * sizeof(struct v4l2_ext_control));
 	kp->controls = (__force struct v4l2_ext_control *)kcontrols;
@@ -689,11 +689,11 @@ static int put_v4l2_ext_controls32(struc
 	compat_caddr_t p;
 
 	if (!access_ok(VERIFY_WRITE, up, sizeof(struct v4l2_ext_controls32)) ||
-		put_user(kp->ctrl_class, &up->ctrl_class) ||
-		put_user(kp->count, &up->count) ||
-		put_user(kp->error_idx, &up->error_idx) ||
-		copy_to_user(up->reserved, kp->reserved, sizeof(up->reserved)))
-			return -EFAULT;
+	    put_user(kp->ctrl_class, &up->ctrl_class) ||
+	    put_user(kp->count, &up->count) ||
+	    put_user(kp->error_idx, &up->error_idx) ||
+	    copy_to_user(up->reserved, kp->reserved, sizeof(up->reserved)))
+		return -EFAULT;
 	if (!kp->count)
 		return 0;
 
@@ -701,7 +701,7 @@ static int put_v4l2_ext_controls32(struc
 		return -EFAULT;
 	ucontrols = compat_ptr(p);
 	if (!access_ok(VERIFY_WRITE, ucontrols,
-			n * sizeof(struct v4l2_ext_control32)))
+		       n * sizeof(struct v4l2_ext_control32)))
 		return -EFAULT;
 
 	while (--n >= 0) {
@@ -739,15 +739,15 @@ struct v4l2_event32 {
 static int put_v4l2_event32(struct v4l2_event *kp, struct v4l2_event32 __user *up)
 {
 	if (!access_ok(VERIFY_WRITE, up, sizeof(struct v4l2_event32)) ||
-		put_user(kp->type, &up->type) ||
-		copy_to_user(&up->u, &kp->u, sizeof(kp->u)) ||
-		put_user(kp->pending, &up->pending) ||
-		put_user(kp->sequence, &up->sequence) ||
-		put_user(kp->timestamp.tv_sec, &up->timestamp.tv_sec) ||
-		put_user(kp->timestamp.tv_nsec, &up->timestamp.tv_nsec) ||
-		put_user(kp->id, &up->id) ||
-		copy_to_user(up->reserved, kp->reserved, 8 * sizeof(__u32)))
-			return -EFAULT;
+	    put_user(kp->type, &up->type) ||
+	    copy_to_user(&up->u, &kp->u, sizeof(kp->u)) ||
+	    put_user(kp->pending, &up->pending) ||
+	    put_user(kp->sequence, &up->sequence) ||
+	    put_user(kp->timestamp.tv_sec, &up->timestamp.tv_sec) ||
+	    put_user(kp->timestamp.tv_nsec, &up->timestamp.tv_nsec) ||
+	    put_user(kp->id, &up->id) ||
+	    copy_to_user(up->reserved, kp->reserved, 8 * sizeof(__u32)))
+		return -EFAULT;
 	return 0;
 }
 
@@ -764,12 +764,12 @@ static int get_v4l2_edid32(struct v4l2_e
 	u32 tmp;
 
 	if (!access_ok(VERIFY_READ, up, sizeof(struct v4l2_edid32)) ||
-		get_user(kp->pad, &up->pad) ||
-		get_user(kp->start_block, &up->start_block) ||
-		get_user(kp->blocks, &up->blocks) ||
-		get_user(tmp, &up->edid) ||
-		copy_from_user(kp->reserved, up->reserved, sizeof(kp->reserved)))
-			return -EFAULT;
+	    get_user(kp->pad, &up->pad) ||
+	    get_user(kp->start_block, &up->start_block) ||
+	    get_user(kp->blocks, &up->blocks) ||
+	    get_user(tmp, &up->edid) ||
+	    copy_from_user(kp->reserved, up->reserved, sizeof(kp->reserved)))
+		return -EFAULT;
 	kp->edid = (__force u8 *)compat_ptr(tmp);
 	return 0;
 }
@@ -779,12 +779,12 @@ static int put_v4l2_edid32(struct v4l2_e
 	u32 tmp = (u32)((unsigned long)kp->edid);
 
 	if (!access_ok(VERIFY_WRITE, up, sizeof(struct v4l2_edid32)) ||
-		put_user(kp->pad, &up->pad) ||
-		put_user(kp->start_block, &up->start_block) ||
-		put_user(kp->blocks, &up->blocks) ||
-		put_user(tmp, &up->edid) ||
-		copy_to_user(up->reserved, kp->reserved, sizeof(up->reserved)))
-			return -EFAULT;
+	    put_user(kp->pad, &up->pad) ||
+	    put_user(kp->start_block, &up->start_block) ||
+	    put_user(kp->blocks, &up->blocks) ||
+	    put_user(tmp, &up->edid) ||
+	    copy_to_user(up->reserved, kp->reserved, sizeof(up->reserved)))
+		return -EFAULT;
 	return 0;
 }
 
#248 lore raw

PATCH 244/254 media: v4l2-compat-ioctl32.c: move 'helper' functions to __get/put_v4l2_format32

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Hans Verkuil <[email protected]>

commit 486c521510c44a04cd756a9267e7d1e271c8a4ba upstream.

These helper functions do not really help. Move the code to the __get/put_v4l2_format32 functions.

---

drivers/media/v4l2-core/v4l2-compat-ioctl32.c +20 -72
@@ -89,64 +89,6 @@ static int put_v4l2_window32(struct v4l2
 	return 0;
 }
 
-static inline int get_v4l2_pix_format(struct v4l2_pix_format *kp, struct v4l2_pix_format __user *up)
-{
-	if (copy_from_user(kp, up, sizeof(struct v4l2_pix_format)))
-		return -EFAULT;
-	return 0;
-}
-
-static inline int get_v4l2_pix_format_mplane(struct v4l2_pix_format_mplane *kp,
-					     struct v4l2_pix_format_mplane __user *up)
-{
-	if (copy_from_user(kp, up, sizeof(struct v4l2_pix_format_mplane)))
-		return -EFAULT;
-	return 0;
-}
-
-static inline int put_v4l2_pix_format(struct v4l2_pix_format *kp, struct v4l2_pix_format __user *up)
-{
-	if (copy_to_user(up, kp, sizeof(struct v4l2_pix_format)))
-		return -EFAULT;
-	return 0;
-}
-
-static inline int put_v4l2_pix_format_mplane(struct v4l2_pix_format_mplane *kp,
-					     struct v4l2_pix_format_mplane __user *up)
-{
-	if (copy_to_user(up, kp, sizeof(struct v4l2_pix_format_mplane)))
-		return -EFAULT;
-	return 0;
-}
-
-static inline int get_v4l2_vbi_format(struct v4l2_vbi_format *kp, struct v4l2_vbi_format __user *up)
-{
-	if (copy_from_user(kp, up, sizeof(struct v4l2_vbi_format)))
-		return -EFAULT;
-	return 0;
-}
-
-static inline int put_v4l2_vbi_format(struct v4l2_vbi_format *kp, struct v4l2_vbi_format __user *up)
-{
-	if (copy_to_user(up, kp, sizeof(struct v4l2_vbi_format)))
-		return -EFAULT;
-	return 0;
-}
-
-static inline int get_v4l2_sliced_vbi_format(struct v4l2_sliced_vbi_format *kp, struct v4l2_sliced_vbi_format __user *up)
-{
-	if (copy_from_user(kp, up, sizeof(struct v4l2_sliced_vbi_format)))
-		return -EFAULT;
-	return 0;
-}
-
-static inline int put_v4l2_sliced_vbi_format(struct v4l2_sliced_vbi_format *kp, struct v4l2_sliced_vbi_format __user *up)
-{
-	if (copy_to_user(up, kp, sizeof(struct v4l2_sliced_vbi_format)))
-		return -EFAULT;
-	return 0;
-}
-
 struct v4l2_format32 {
 	__u32	type;	/* enum v4l2_buf_type */
 	union {
@@ -184,20 +126,23 @@ static int __get_v4l2_format32(struct v4
 	switch (kp->type) {
 	case V4L2_BUF_TYPE_VIDEO_CAPTURE:
 	case V4L2_BUF_TYPE_VIDEO_OUTPUT:
-		return get_v4l2_pix_format(&kp->fmt.pix, &up->fmt.pix);
+		return copy_from_user(&kp->fmt.pix, &up->fmt.pix,
+				      sizeof(kp->fmt.pix)) ? -EFAULT : 0;
 	case V4L2_BUF_TYPE_VIDEO_CAPTURE_MPLANE:
 	case V4L2_BUF_TYPE_VIDEO_OUTPUT_MPLANE:
-		return get_v4l2_pix_format_mplane(&kp->fmt.pix_mp,
-						  &up->fmt.pix_mp);
+		return copy_from_user(&kp->fmt.pix_mp, &up->fmt.pix_mp,
+				      sizeof(kp->fmt.pix_mp)) ? -EFAULT : 0;
 	case V4L2_BUF_TYPE_VIDEO_OVERLAY:
 	case V4L2_BUF_TYPE_VIDEO_OUTPUT_OVERLAY:
 		return get_v4l2_window32(&kp->fmt.win, &up->fmt.win);
 	case V4L2_BUF_TYPE_VBI_CAPTURE:
 	case V4L2_BUF_TYPE_VBI_OUTPUT:
-		return get_v4l2_vbi_format(&kp->fmt.vbi, &up->fmt.vbi);
+		return copy_from_user(&kp->fmt.vbi, &up->fmt.vbi,
+				      sizeof(kp->fmt.vbi)) ? -EFAULT : 0;
 	case V4L2_BUF_TYPE_SLICED_VBI_CAPTURE:
 	case V4L2_BUF_TYPE_SLICED_VBI_OUTPUT:
-		return get_v4l2_sliced_vbi_format(&kp->fmt.sliced, &up->fmt.sliced);
+		return copy_from_user(&kp->fmt.sliced, &up->fmt.sliced,
+				      sizeof(kp->fmt.sliced)) ? -EFAULT : 0;
 	default:
 		printk(KERN_INFO "compat_ioctl32: unexpected VIDIOC_FMT type %d\n",
 		       kp->type);
@@ -228,20 +173,23 @@ static int __put_v4l2_format32(struct v4
 	switch (kp->type) {
 	case V4L2_BUF_TYPE_VIDEO_CAPTURE:
 	case V4L2_BUF_TYPE_VIDEO_OUTPUT:
-		return put_v4l2_pix_format(&kp->fmt.pix, &up->fmt.pix);
+		return copy_to_user(&up->fmt.pix, &kp->fmt.pix,
+				    sizeof(kp->fmt.pix)) ? -EFAULT : 0;
 	case V4L2_BUF_TYPE_VIDEO_CAPTURE_MPLANE:
 	case V4L2_BUF_TYPE_VIDEO_OUTPUT_MPLANE:
-		return put_v4l2_pix_format_mplane(&kp->fmt.pix_mp,
-						  &up->fmt.pix_mp);
+		return copy_to_user(&up->fmt.pix_mp, &kp->fmt.pix_mp,
+				    sizeof(kp->fmt.pix_mp)) ? -EFAULT : 0;
 	case V4L2_BUF_TYPE_VIDEO_OVERLAY:
 	case V4L2_BUF_TYPE_VIDEO_OUTPUT_OVERLAY:
 		return put_v4l2_window32(&kp->fmt.win, &up->fmt.win);
 	case V4L2_BUF_TYPE_VBI_CAPTURE:
 	case V4L2_BUF_TYPE_VBI_OUTPUT:
-		return put_v4l2_vbi_format(&kp->fmt.vbi, &up->fmt.vbi);
+		return copy_to_user(&up->fmt.vbi, &kp->fmt.vbi,
+				    sizeof(kp->fmt.vbi)) ? -EFAULT : 0;
 	case V4L2_BUF_TYPE_SLICED_VBI_CAPTURE:
 	case V4L2_BUF_TYPE_SLICED_VBI_OUTPUT:
-		return put_v4l2_sliced_vbi_format(&kp->fmt.sliced, &up->fmt.sliced);
+		return copy_to_user(&up->fmt.sliced, &kp->fmt.sliced,
+				    sizeof(kp->fmt.sliced)) ? -EFAULT : 0;
 	default:
 		printk(KERN_INFO "compat_ioctl32: unexpected VIDIOC_FMT type %d\n",
 		       kp->type);
@@ -548,10 +496,10 @@ static int get_v4l2_framebuffer32(struct
 	if (!access_ok(VERIFY_READ, up, sizeof(struct v4l2_framebuffer32)) ||
 	    get_user(tmp, &up->base) ||
 	    get_user(kp->capability, &up->capability) ||
-	    get_user(kp->flags, &up->flags))
+	    get_user(kp->flags, &up->flags) ||
+	    copy_from_user(&kp->fmt, &up->fmt, sizeof(up->fmt)))
 		return -EFAULT;
 	kp->base = (__force void *)compat_ptr(tmp);
-	get_v4l2_pix_format(&kp->fmt, &up->fmt);
 	return 0;
 }
 
@@ -562,9 +510,9 @@ static int put_v4l2_framebuffer32(struct
 	if (!access_ok(VERIFY_WRITE, up, sizeof(struct v4l2_framebuffer32)) ||
 	    put_user(tmp, &up->base) ||
 	    put_user(kp->capability, &up->capability) ||
-	    put_user(kp->flags, &up->flags))
+	    put_user(kp->flags, &up->flags) ||
+	    copy_to_user(&up->fmt, &kp->fmt, sizeof(up->fmt)))
 		return -EFAULT;
-	put_v4l2_pix_format(&kp->fmt, &up->fmt);
 	return 0;
 }
 
#249 lore raw

PATCH 245/254 media: v4l2-compat-ioctl32.c: avoid sizeof(type)

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Hans Verkuil <[email protected]>

commit 333b1e9f96ce05f7498b581509bb30cde03018bf upstream.

Instead of doing sizeof(struct foo) use sizeof(*up). There even were cases where 4 * sizeof(__u32) was used instead of sizeof(kp->reserved), which is very dangerous when the size of the reserved array changes.

---

drivers/media/v4l2-core/v4l2-compat-ioctl32.c +35 -38
@@ -47,7 +47,7 @@ struct v4l2_window32 {
 
 static int get_v4l2_window32(struct v4l2_window *kp, struct v4l2_window32 __user *up)
 {
-	if (!access_ok(VERIFY_READ, up, sizeof(struct v4l2_window32)) ||
+	if (!access_ok(VERIFY_READ, up, sizeof(*up)) ||
 	    copy_from_user(&kp->w, &up->w, sizeof(up->w)) ||
 	    get_user(kp->field, &up->field) ||
 	    get_user(kp->chromakey, &up->chromakey) ||
@@ -64,7 +64,7 @@ static int get_v4l2_window32(struct v4l2
 		if (get_user(p, &up->clips))
 			return -EFAULT;
 		uclips = compat_ptr(p);
-		kclips = compat_alloc_user_space(n * sizeof(struct v4l2_clip));
+		kclips = compat_alloc_user_space(n * sizeof(*kclips));
 		kp->clips = kclips;
 		while (--n >= 0) {
 			if (copy_in_user(&kclips->c, &uclips->c, sizeof(uclips->c)))
@@ -152,14 +152,14 @@ static int __get_v4l2_format32(struct v4
 
 static int get_v4l2_format32(struct v4l2_format *kp, struct v4l2_format32 __user *up)
 {
-	if (!access_ok(VERIFY_READ, up, sizeof(struct v4l2_format32)))
+	if (!access_ok(VERIFY_READ, up, sizeof(*up)))
 		return -EFAULT;
 	return __get_v4l2_format32(kp, up);
 }
 
 static int get_v4l2_create32(struct v4l2_create_buffers *kp, struct v4l2_create_buffers32 __user *up)
 {
-	if (!access_ok(VERIFY_READ, up, sizeof(struct v4l2_create_buffers32)) ||
+	if (!access_ok(VERIFY_READ, up, sizeof(*up)) ||
 	    copy_from_user(kp, up, offsetof(struct v4l2_create_buffers32, format)))
 		return -EFAULT;
 	return __get_v4l2_format32(&kp->format, &up->format);
@@ -199,14 +199,14 @@ static int __put_v4l2_format32(struct v4
 
 static int put_v4l2_format32(struct v4l2_format *kp, struct v4l2_format32 __user *up)
 {
-	if (!access_ok(VERIFY_WRITE, up, sizeof(struct v4l2_format32)))
+	if (!access_ok(VERIFY_WRITE, up, sizeof(*up)))
 		return -EFAULT;
 	return __put_v4l2_format32(kp, up);
 }
 
 static int put_v4l2_create32(struct v4l2_create_buffers *kp, struct v4l2_create_buffers32 __user *up)
 {
-	if (!access_ok(VERIFY_WRITE, up, sizeof(struct v4l2_create_buffers32)) ||
+	if (!access_ok(VERIFY_WRITE, up, sizeof(*up)) ||
 	    copy_to_user(up, kp, offsetof(struct v4l2_create_buffers32, format)) ||
 	    copy_to_user(up->reserved, kp->reserved, sizeof(kp->reserved)))
 		return -EFAULT;
@@ -225,7 +225,7 @@ struct v4l2_standard32 {
 static int get_v4l2_standard32(struct v4l2_standard *kp, struct v4l2_standard32 __user *up)
 {
 	/* other fields are not set by the user, nor used by the driver */
-	if (!access_ok(VERIFY_READ, up, sizeof(struct v4l2_standard32)) ||
+	if (!access_ok(VERIFY_READ, up, sizeof(*up)) ||
 	    get_user(kp->index, &up->index))
 		return -EFAULT;
 	return 0;
@@ -233,13 +233,13 @@ static int get_v4l2_standard32(struct v4
 
 static int put_v4l2_standard32(struct v4l2_standard *kp, struct v4l2_standard32 __user *up)
 {
-	if (!access_ok(VERIFY_WRITE, up, sizeof(struct v4l2_standard32)) ||
+	if (!access_ok(VERIFY_WRITE, up, sizeof(*up)) ||
 	    put_user(kp->index, &up->index) ||
 	    put_user(kp->id, &up->id) ||
-	    copy_to_user(up->name, kp->name, 24) ||
+	    copy_to_user(up->name, kp->name, sizeof(up->name)) ||
 	    copy_to_user(&up->frameperiod, &kp->frameperiod, sizeof(kp->frameperiod)) ||
 	    put_user(kp->framelines, &up->framelines) ||
-	    copy_to_user(up->reserved, kp->reserved, 4 * sizeof(__u32)))
+	    copy_to_user(up->reserved, kp->reserved, sizeof(kp->reserved)))
 		return -EFAULT;
 	return 0;
 }
@@ -287,7 +287,7 @@ static int get_v4l2_plane32(struct v4l2_
 
 	if (copy_in_user(up, up32, 2 * sizeof(__u32)) ||
 	    copy_in_user(&up->data_offset, &up32->data_offset,
-			 sizeof(__u32)))
+			 sizeof(up->data_offset)))
 		return -EFAULT;
 
 	if (memory == V4L2_MEMORY_USERPTR) {
@@ -297,11 +297,11 @@ static int get_v4l2_plane32(struct v4l2_
 		if (put_user((unsigned long)up_pln, &up->m.userptr))
 			return -EFAULT;
 	} else if (memory == V4L2_MEMORY_DMABUF) {
-		if (copy_in_user(&up->m.fd, &up32->m.fd, sizeof(int)))
+		if (copy_in_user(&up->m.fd, &up32->m.fd, sizeof(up32->m.fd)))
 			return -EFAULT;
 	} else {
 		if (copy_in_user(&up->m.mem_offset, &up32->m.mem_offset,
-				 sizeof(__u32)))
+				 sizeof(up32->m.mem_offset)))
 			return -EFAULT;
 	}
 
@@ -313,19 +313,19 @@ static int put_v4l2_plane32(struct v4l2_
 {
 	if (copy_in_user(up32, up, 2 * sizeof(__u32)) ||
 	    copy_in_user(&up32->data_offset, &up->data_offset,
-			 sizeof(__u32)))
+			 sizeof(up->data_offset)))
 		return -EFAULT;
 
 	/* For MMAP, driver might've set up the offset, so copy it back.
 	 * USERPTR stays the same (was userspace-provided), so no copying. */
 	if (memory == V4L2_MEMORY_MMAP)
 		if (copy_in_user(&up32->m.mem_offset, &up->m.mem_offset,
-				 sizeof(__u32)))
+				 sizeof(up->m.mem_offset)))
 			return -EFAULT;
 	/* For DMABUF, driver might've set up the fd, so copy it back. */
 	if (memory == V4L2_MEMORY_DMABUF)
 		if (copy_in_user(&up32->m.fd, &up->m.fd,
-				 sizeof(int)))
+				 sizeof(up->m.fd)))
 			return -EFAULT;
 
 	return 0;
@@ -339,7 +339,7 @@ static int get_v4l2_buffer32(struct v4l2
 	int num_planes;
 	int ret;
 
-	if (!access_ok(VERIFY_READ, up, sizeof(struct v4l2_buffer32)) ||
+	if (!access_ok(VERIFY_READ, up, sizeof(*up)) ||
 	    get_user(kp->index, &up->index) ||
 	    get_user(kp->type, &up->type) ||
 	    get_user(kp->flags, &up->flags) ||
@@ -351,8 +351,7 @@ static int get_v4l2_buffer32(struct v4l2
 		if (get_user(kp->bytesused, &up->bytesused) ||
 		    get_user(kp->field, &up->field) ||
 		    get_user(kp->timestamp.tv_sec, &up->timestamp.tv_sec) ||
-		    get_user(kp->timestamp.tv_usec,
-			     &up->timestamp.tv_usec))
+		    get_user(kp->timestamp.tv_usec, &up->timestamp.tv_usec))
 			return -EFAULT;
 
 	if (V4L2_TYPE_IS_MULTIPLANAR(kp->type)) {
@@ -369,13 +368,13 @@ static int get_v4l2_buffer32(struct v4l2
 
 		uplane32 = compat_ptr(p);
 		if (!access_ok(VERIFY_READ, uplane32,
-			       num_planes * sizeof(struct v4l2_plane32)))
+			       num_planes * sizeof(*uplane32)))
 			return -EFAULT;
 
 		/* We don't really care if userspace decides to kill itself
 		 * by passing a very big num_planes value */
 		uplane = compat_alloc_user_space(num_planes *
-						 sizeof(struct v4l2_plane));
+						 sizeof(*uplane));
 		kp->m.planes = (__force struct v4l2_plane *)uplane;
 
 		while (--num_planes >= 0) {
@@ -423,7 +422,7 @@ static int put_v4l2_buffer32(struct v4l2
 	int num_planes;
 	int ret;
 
-	if (!access_ok(VERIFY_WRITE, up, sizeof(struct v4l2_buffer32)) ||
+	if (!access_ok(VERIFY_WRITE, up, sizeof(*up)) ||
 	    put_user(kp->index, &up->index) ||
 	    put_user(kp->type, &up->type) ||
 	    put_user(kp->flags, &up->flags) ||
@@ -434,7 +433,7 @@ static int put_v4l2_buffer32(struct v4l2
 	    put_user(kp->field, &up->field) ||
 	    put_user(kp->timestamp.tv_sec, &up->timestamp.tv_sec) ||
 	    put_user(kp->timestamp.tv_usec, &up->timestamp.tv_usec) ||
-	    copy_to_user(&up->timecode, &kp->timecode, sizeof(struct v4l2_timecode)) ||
+	    copy_to_user(&up->timecode, &kp->timecode, sizeof(kp->timecode)) ||
 	    put_user(kp->sequence, &up->sequence) ||
 	    put_user(kp->reserved2, &up->reserved2) ||
 	    put_user(kp->reserved, &up->reserved) ||
@@ -493,7 +492,7 @@ static int get_v4l2_framebuffer32(struct
 {
 	u32 tmp;
 
-	if (!access_ok(VERIFY_READ, up, sizeof(struct v4l2_framebuffer32)) ||
+	if (!access_ok(VERIFY_READ, up, sizeof(*up)) ||
 	    get_user(tmp, &up->base) ||
 	    get_user(kp->capability, &up->capability) ||
 	    get_user(kp->flags, &up->flags) ||
@@ -507,7 +506,7 @@ static int put_v4l2_framebuffer32(struct
 {
 	u32 tmp = (u32)((unsigned long)kp->base);
 
-	if (!access_ok(VERIFY_WRITE, up, sizeof(struct v4l2_framebuffer32)) ||
+	if (!access_ok(VERIFY_WRITE, up, sizeof(*up)) ||
 	    put_user(tmp, &up->base) ||
 	    put_user(kp->capability, &up->capability) ||
 	    put_user(kp->flags, &up->flags) ||
@@ -532,14 +531,14 @@ struct v4l2_input32 {
    Otherwise it is identical to the 32-bit version. */
 static inline int get_v4l2_input32(struct v4l2_input *kp, struct v4l2_input32 __user *up)
 {
-	if (copy_from_user(kp, up, sizeof(struct v4l2_input32)))
+	if (copy_from_user(kp, up, sizeof(*up)))
 		return -EFAULT;
 	return 0;
 }
 
 static inline int put_v4l2_input32(struct v4l2_input *kp, struct v4l2_input32 __user *up)
 {
-	if (copy_to_user(up, kp, sizeof(struct v4l2_input32)))
+	if (copy_to_user(up, kp, sizeof(*up)))
 		return -EFAULT;
 	return 0;
 }
@@ -587,7 +586,7 @@ static int get_v4l2_ext_controls32(struc
 	int n;
 	compat_caddr_t p;
 
-	if (!access_ok(VERIFY_READ, up, sizeof(struct v4l2_ext_controls32)) ||
+	if (!access_ok(VERIFY_READ, up, sizeof(*up)) ||
 	    get_user(kp->ctrl_class, &up->ctrl_class) ||
 	    get_user(kp->count, &up->count) ||
 	    get_user(kp->error_idx, &up->error_idx) ||
@@ -601,10 +600,9 @@ static int get_v4l2_ext_controls32(struc
 	if (get_user(p, &up->controls))
 		return -EFAULT;
 	ucontrols = compat_ptr(p);
-	if (!access_ok(VERIFY_READ, ucontrols,
-		       n * sizeof(struct v4l2_ext_control32)))
+	if (!access_ok(VERIFY_READ, ucontrols, n * sizeof(*ucontrols)))
 		return -EFAULT;
-	kcontrols = compat_alloc_user_space(n * sizeof(struct v4l2_ext_control));
+	kcontrols = compat_alloc_user_space(n * sizeof(*kcontrols));
 	kp->controls = (__force struct v4l2_ext_control *)kcontrols;
 	while (--n >= 0) {
 		u32 id;
@@ -636,7 +634,7 @@ static int put_v4l2_ext_controls32(struc
 	int n = kp->count;
 	compat_caddr_t p;
 
-	if (!access_ok(VERIFY_WRITE, up, sizeof(struct v4l2_ext_controls32)) ||
+	if (!access_ok(VERIFY_WRITE, up, sizeof(*up)) ||
 	    put_user(kp->ctrl_class, &up->ctrl_class) ||
 	    put_user(kp->count, &up->count) ||
 	    put_user(kp->error_idx, &up->error_idx) ||
@@ -648,8 +646,7 @@ static int put_v4l2_ext_controls32(struc
 	if (get_user(p, &up->controls))
 		return -EFAULT;
 	ucontrols = compat_ptr(p);
-	if (!access_ok(VERIFY_WRITE, ucontrols,
-		       n * sizeof(struct v4l2_ext_control32)))
+	if (!access_ok(VERIFY_WRITE, ucontrols, n * sizeof(*ucontrols)))
 		return -EFAULT;
 
 	while (--n >= 0) {
@@ -686,7 +683,7 @@ struct v4l2_event32 {
 
 static int put_v4l2_event32(struct v4l2_event *kp, struct v4l2_event32 __user *up)
 {
-	if (!access_ok(VERIFY_WRITE, up, sizeof(struct v4l2_event32)) ||
+	if (!access_ok(VERIFY_WRITE, up, sizeof(*up)) ||
 	    put_user(kp->type, &up->type) ||
 	    copy_to_user(&up->u, &kp->u, sizeof(kp->u)) ||
 	    put_user(kp->pending, &up->pending) ||
@@ -694,7 +691,7 @@ static int put_v4l2_event32(struct v4l2_
 	    put_user(kp->timestamp.tv_sec, &up->timestamp.tv_sec) ||
 	    put_user(kp->timestamp.tv_nsec, &up->timestamp.tv_nsec) ||
 	    put_user(kp->id, &up->id) ||
-	    copy_to_user(up->reserved, kp->reserved, 8 * sizeof(__u32)))
+	    copy_to_user(up->reserved, kp->reserved, sizeof(kp->reserved)))
 		return -EFAULT;
 	return 0;
 }
@@ -711,7 +708,7 @@ static int get_v4l2_edid32(struct v4l2_e
 {
 	u32 tmp;
 
-	if (!access_ok(VERIFY_READ, up, sizeof(struct v4l2_edid32)) ||
+	if (!access_ok(VERIFY_READ, up, sizeof(*up)) ||
 	    get_user(kp->pad, &up->pad) ||
 	    get_user(kp->start_block, &up->start_block) ||
 	    get_user(kp->blocks, &up->blocks) ||
@@ -726,7 +723,7 @@ static int put_v4l2_edid32(struct v4l2_e
 {
 	u32 tmp = (u32)((unsigned long)kp->edid);
 
-	if (!access_ok(VERIFY_WRITE, up, sizeof(struct v4l2_edid32)) ||
+	if (!access_ok(VERIFY_WRITE, up, sizeof(*up)) ||
 	    put_user(kp->pad, &up->pad) ||
 	    put_user(kp->start_block, &up->start_block) ||
 	    put_user(kp->blocks, &up->blocks) ||
#250 lore raw

PATCH 246/254 media: v4l2-compat-ioctl32.c: copy m.userptr in put_v4l2_plane32

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Hans Verkuil <[email protected]>

commit 8ed5a59dcb47a6f76034ee760b36e089f3e82529 upstream.

The struct v4l2_plane32 should set m.userptr as well. The same happens in v4l2_buffer32 and v4l2-compliance tests for this.

---
 drivers/media/v4l2-core/v4l2-compat-ioctl32.c | 47 ++++++++++++++++-----------
 1 file changed, 28 insertions(+), 19 deletions(-)
drivers/media/v4l2-core/v4l2-compat-ioctl32.c +28 -19
@@ -290,19 +290,24 @@ static int get_v4l2_plane32(struct v4l2_
 			 sizeof(up->data_offset)))
 		return -EFAULT;
 
-	if (memory == V4L2_MEMORY_USERPTR) {
+	switch (memory) {
+	case V4L2_MEMORY_MMAP:
+	case V4L2_MEMORY_OVERLAY:
+		if (copy_in_user(&up->m.mem_offset, &up32->m.mem_offset,
+				 sizeof(up32->m.mem_offset)))
+			return -EFAULT;
+		break;
+	case V4L2_MEMORY_USERPTR:
 		if (get_user(p, &up32->m.userptr))
 			return -EFAULT;
 		up_pln = compat_ptr(p);
 		if (put_user((unsigned long)up_pln, &up->m.userptr))
 			return -EFAULT;
-	} else if (memory == V4L2_MEMORY_DMABUF) {
+		break;
+	case V4L2_MEMORY_DMABUF:
 		if (copy_in_user(&up->m.fd, &up32->m.fd, sizeof(up32->m.fd)))
 			return -EFAULT;
-	} else {
-		if (copy_in_user(&up->m.mem_offset, &up32->m.mem_offset,
-				 sizeof(up32->m.mem_offset)))
-			return -EFAULT;
+		break;
 	}
 
 	return 0;
@@ -311,22 +316,32 @@ static int get_v4l2_plane32(struct v4l2_
 static int put_v4l2_plane32(struct v4l2_plane __user *up, struct v4l2_plane32 __user *up32,
 			    enum v4l2_memory memory)
 {
+	unsigned long p;
+
 	if (copy_in_user(up32, up, 2 * sizeof(__u32)) ||
 	    copy_in_user(&up32->data_offset, &up->data_offset,
 			 sizeof(up->data_offset)))
 		return -EFAULT;
 
-	/* For MMAP, driver might've set up the offset, so copy it back.
-	 * USERPTR stays the same (was userspace-provided), so no copying. */
-	if (memory == V4L2_MEMORY_MMAP)
+	switch (memory) {
+	case V4L2_MEMORY_MMAP:
+	case V4L2_MEMORY_OVERLAY:
 		if (copy_in_user(&up32->m.mem_offset, &up->m.mem_offset,
 				 sizeof(up->m.mem_offset)))
 			return -EFAULT;
-	/* For DMABUF, driver might've set up the fd, so copy it back. */
-	if (memory == V4L2_MEMORY_DMABUF)
+		break;
+	case V4L2_MEMORY_USERPTR:
+		if (get_user(p, &up->m.userptr) ||
+		    put_user((compat_ulong_t)ptr_to_compat((__force void *)p),
+			     &up32->m.userptr))
+			return -EFAULT;
+		break;
+	case V4L2_MEMORY_DMABUF:
 		if (copy_in_user(&up32->m.fd, &up->m.fd,
 				 sizeof(up->m.fd)))
 			return -EFAULT;
+		break;
+	}
 
 	return 0;
 }
@@ -387,6 +402,7 @@ static int get_v4l2_buffer32(struct v4l2
 	} else {
 		switch (kp->memory) {
 		case V4L2_MEMORY_MMAP:
+		case V4L2_MEMORY_OVERLAY:
 			if (get_user(kp->m.offset, &up->m.offset))
 				return -EFAULT;
 			break;
@@ -400,10 +416,6 @@ static int get_v4l2_buffer32(struct v4l2
 				kp->m.userptr = (unsigned long)compat_ptr(tmp);
 			}
 			break;
-		case V4L2_MEMORY_OVERLAY:
-			if (get_user(kp->m.offset, &up->m.offset))
-				return -EFAULT;
-			break;
 		case V4L2_MEMORY_DMABUF:
 			if (get_user(kp->m.fd, &up->m.fd))
 				return -EFAULT;
@@ -460,6 +472,7 @@ static int put_v4l2_buffer32(struct v4l2
 	} else {
 		switch (kp->memory) {
 		case V4L2_MEMORY_MMAP:
+		case V4L2_MEMORY_OVERLAY:
 			if (put_user(kp->m.offset, &up->m.offset))
 				return -EFAULT;
 			break;
@@ -467,10 +480,6 @@ static int put_v4l2_buffer32(struct v4l2
 			if (put_user(kp->m.userptr, &up->m.userptr))
 				return -EFAULT;
 			break;
-		case V4L2_MEMORY_OVERLAY:
-			if (put_user(kp->m.offset, &up->m.offset))
-				return -EFAULT;
-			break;
 		case V4L2_MEMORY_DMABUF:
 			if (put_user(kp->m.fd, &up->m.fd))
 				return -EFAULT;
#251 lore raw

PATCH 247/254 media: v4l2-compat-ioctl32.c: fix ctrl_is_pointer

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Hans Verkuil <[email protected]>

commit b8c601e8af2d08f733d74defa8465303391bb930 upstream.

ctrl_is_pointer just hardcoded two known string controls, but that caused problems when using e.g. custom controls that use a pointer for the payload.

Reimplement this function: it now finds the v4l2_ctrl (if the driver uses the control framework) or it calls vidioc_query_ext_ctrl (if the driver implements that directly).

In both cases it can now check if the control is a pointer control or not.

---

drivers/media/v4l2-core/v4l2-compat-ioctl32.c +32 -20
@@ -18,6 +18,8 @@
 #include <linux/videodev2.h>
 #include <linux/v4l2-subdev.h>
 #include <media/v4l2-dev.h>
+#include <media/v4l2-fh.h>
+#include <media/v4l2-ctrls.h>
 #include <media/v4l2-ioctl.h>
 
 static long native_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
@@ -571,24 +573,32 @@ struct v4l2_ext_control32 {
 	};
 } __attribute__ ((packed));
 
-/* The following function really belong in v4l2-common, but that causes
-   a circular dependency between modules. We need to think about this, but
-   for now this will do. */
-
-/* Return non-zero if this control is a pointer type. Currently only
-   type STRING is a pointer type. */
-static inline int ctrl_is_pointer(u32 id)
-{
-	switch (id) {
-	case V4L2_CID_RDS_TX_PS_NAME:
-	case V4L2_CID_RDS_TX_RADIO_TEXT:
-		return 1;
-	default:
-		return 0;
+/* Return true if this control is a pointer type. */
+static inline bool ctrl_is_pointer(struct file *file, u32 id)
+{
+	struct video_device *vdev = video_devdata(file);
+	struct v4l2_fh *fh = NULL;
+	struct v4l2_ctrl_handler *hdl = NULL;
+
+	if (test_bit(V4L2_FL_USES_V4L2_FH, &vdev->flags))
+		fh = file->private_data;
+
+	if (fh && fh->ctrl_handler)
+		hdl = fh->ctrl_handler;
+	else if (vdev->ctrl_handler)
+		hdl = vdev->ctrl_handler;
+
+	if (hdl) {
+		struct v4l2_ctrl *ctrl = v4l2_ctrl_find(hdl, id);
+
+		return ctrl && ctrl->type == V4L2_CTRL_TYPE_STRING;
 	}
+	return false;
 }
 
-static int get_v4l2_ext_controls32(struct v4l2_ext_controls *kp, struct v4l2_ext_controls32 __user *up)
+static int get_v4l2_ext_controls32(struct file *file,
+				   struct v4l2_ext_controls *kp,
+				   struct v4l2_ext_controls32 __user *up)
 {
 	struct v4l2_ext_control32 __user *ucontrols;
 	struct v4l2_ext_control __user *kcontrols;
@@ -620,7 +630,7 @@ static int get_v4l2_ext_controls32(struc
 			return -EFAULT;
 		if (get_user(id, &kcontrols->id))
 			return -EFAULT;
-		if (ctrl_is_pointer(id)) {
+		if (ctrl_is_pointer(file, id)) {
 			void __user *s;
 
 			if (get_user(p, &ucontrols->string))
@@ -635,7 +645,9 @@ static int get_v4l2_ext_controls32(struc
 	return 0;
 }
 
-static int put_v4l2_ext_controls32(struct v4l2_ext_controls *kp, struct v4l2_ext_controls32 __user *up)
+static int put_v4l2_ext_controls32(struct file *file,
+				   struct v4l2_ext_controls *kp,
+				   struct v4l2_ext_controls32 __user *up)
 {
 	struct v4l2_ext_control32 __user *ucontrols;
 	struct v4l2_ext_control __user *kcontrols =
@@ -667,7 +679,7 @@ static int put_v4l2_ext_controls32(struc
 		/* Do not modify the pointer when copying a pointer control.
 		   The contents of the pointer was changed, not the pointer
 		   itself. */
-		if (ctrl_is_pointer(id))
+		if (ctrl_is_pointer(file, id))
 			size -= sizeof(ucontrols->value64);
 		if (copy_in_user(ucontrols, kcontrols, size))
 			return -EFAULT;
@@ -881,7 +893,7 @@ static long do_video_ioctl(struct file *
 	case VIDIOC_G_EXT_CTRLS:
 	case VIDIOC_S_EXT_CTRLS:
 	case VIDIOC_TRY_EXT_CTRLS:
-		err = get_v4l2_ext_controls32(&karg.v2ecs, up);
+		err = get_v4l2_ext_controls32(file, &karg.v2ecs, up);
 		compatible_arg = 0;
 		break;
 	case VIDIOC_DQEVENT:
@@ -908,7 +920,7 @@ static long do_video_ioctl(struct file *
 	case VIDIOC_G_EXT_CTRLS:
 	case VIDIOC_S_EXT_CTRLS:
 	case VIDIOC_TRY_EXT_CTRLS:
-		if (put_v4l2_ext_controls32(&karg.v2ecs, up))
+		if (put_v4l2_ext_controls32(file, &karg.v2ecs, up))
 			err = -EFAULT;
 		break;
 	}
#252 lore raw

PATCH 248/254 media: v4l2-compat-ioctl32: Copy v4l2_window->global_alpha

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Daniel Mentz <[email protected]>

commit 025a26fa14f8fd55d50ab284a30c016a5be953d0 upstream.

Commit b2787845fb91 ("V4L/DVB (5289): Add support for video output overlays.") added the field global_alpha to struct v4l2_window but did not update the compat layer accordingly. This change adds global_alpha to struct v4l2_window32 and copies the value for global_alpha back and forth.

---
 drivers/media/v4l2-core/v4l2-compat-ioctl32.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)
drivers/media/v4l2-core/v4l2-compat-ioctl32.c +5 -2
@@ -45,6 +45,7 @@ struct v4l2_window32 {
 	compat_caddr_t		clips; /* actually struct v4l2_clip32 * */
 	__u32			clipcount;
 	compat_caddr_t		bitmap;
+	__u8                    global_alpha;
 };
 
 static int get_v4l2_window32(struct v4l2_window *kp, struct v4l2_window32 __user *up)
@@ -53,7 +54,8 @@ static int get_v4l2_window32(struct v4l2
 	    copy_from_user(&kp->w, &up->w, sizeof(up->w)) ||
 	    get_user(kp->field, &up->field) ||
 	    get_user(kp->chromakey, &up->chromakey) ||
-	    get_user(kp->clipcount, &up->clipcount))
+	    get_user(kp->clipcount, &up->clipcount) ||
+	    get_user(kp->global_alpha, &up->global_alpha))
 		return -EFAULT;
 	if (kp->clipcount > 2048)
 		return -EINVAL;
@@ -86,7 +88,8 @@ static int put_v4l2_window32(struct v4l2
 	if (copy_to_user(&up->w, &kp->w, sizeof(kp->w)) ||
 	    put_user(kp->field, &up->field) ||
 	    put_user(kp->chromakey, &up->chromakey) ||
-	    put_user(kp->clipcount, &up->clipcount))
+	    put_user(kp->clipcount, &up->clipcount) ||
+	    put_user(kp->global_alpha, &up->global_alpha))
 		return -EFAULT;
 	return 0;
 }
#253 lore raw

PATCH 249/254 media: v4l2-compat-ioctl32.c: copy clip list in put_v4l2_window32

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Hans Verkuil <[email protected]>

commit a751be5b142ef6bcbbb96d9899516f4d9c8d0ef4 upstream.

put_v4l2_window32() didn't copy back the clip list to userspace. Drivers can update the clip rectangles, so this should be done.

---
 drivers/media/v4l2-core/v4l2-compat-ioctl32.c | 59 ++++++++++++++++++---------
 1 file changed, 40 insertions(+), 19 deletions(-)
drivers/media/v4l2-core/v4l2-compat-ioctl32.c +40 -19
@@ -50,6 +50,11 @@ struct v4l2_window32 {
 
 static int get_v4l2_window32(struct v4l2_window *kp, struct v4l2_window32 __user *up)
 {
+	struct v4l2_clip32 __user *uclips;
+	struct v4l2_clip __user *kclips;
+	compat_caddr_t p;
+	u32 n;
+
 	if (!access_ok(VERIFY_READ, up, sizeof(*up)) ||
 	    copy_from_user(&kp->w, &up->w, sizeof(up->w)) ||
 	    get_user(kp->field, &up->field) ||
@@ -59,38 +64,54 @@ static int get_v4l2_window32(struct v4l2
 		return -EFAULT;
 	if (kp->clipcount > 2048)
 		return -EINVAL;
-	if (kp->clipcount) {
-		struct v4l2_clip32 __user *uclips;
-		struct v4l2_clip __user *kclips;
-		int n = kp->clipcount;
-		compat_caddr_t p;
+	if (!kp->clipcount) {
+		kp->clips = NULL;
+		return 0;
+	}
 
-		if (get_user(p, &up->clips))
+	n = kp->clipcount;
+	if (get_user(p, &up->clips))
+		return -EFAULT;
+	uclips = compat_ptr(p);
+	kclips = compat_alloc_user_space(n * sizeof(*kclips));
+	kp->clips = kclips;
+	while (n--) {
+		if (copy_in_user(&kclips->c, &uclips->c, sizeof(uclips->c)))
 			return -EFAULT;
-		uclips = compat_ptr(p);
-		kclips = compat_alloc_user_space(n * sizeof(*kclips));
-		kp->clips = kclips;
-		while (--n >= 0) {
-			if (copy_in_user(&kclips->c, &uclips->c, sizeof(uclips->c)))
-				return -EFAULT;
-			if (put_user(n ? kclips + 1 : NULL, &kclips->next))
-				return -EFAULT;
-			uclips += 1;
-			kclips += 1;
-		}
-	} else
-		kp->clips = NULL;
+		if (put_user(n ? kclips + 1 : NULL, &kclips->next))
+			return -EFAULT;
+		uclips++;
+		kclips++;
+	}
 	return 0;
 }
 
 static int put_v4l2_window32(struct v4l2_window *kp, struct v4l2_window32 __user *up)
 {
+	struct v4l2_clip __user *kclips = kp->clips;
+	struct v4l2_clip32 __user *uclips;
+	u32 n = kp->clipcount;
+	compat_caddr_t p;
+
 	if (copy_to_user(&up->w, &kp->w, sizeof(kp->w)) ||
 	    put_user(kp->field, &up->field) ||
 	    put_user(kp->chromakey, &up->chromakey) ||
 	    put_user(kp->clipcount, &up->clipcount) ||
 	    put_user(kp->global_alpha, &up->global_alpha))
 		return -EFAULT;
+
+	if (!kp->clipcount)
+		return 0;
+
+	if (get_user(p, &up->clips))
+		return -EFAULT;
+	uclips = compat_ptr(p);
+	while (n--) {
+		if (copy_in_user(&uclips->c, &kclips->c, sizeof(uclips->c)))
+			return -EFAULT;
+		uclips++;
+		kclips++;
+	}
 	return 0;
 }
 
#254 lore raw

PATCH 250/254 media: v4l2-compat-ioctl32.c: drop pr_info for unknown buffer type

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Hans Verkuil <[email protected]>

commit 169f24ca68bf0f247d111aef07af00dd3a02ae88 upstream.

There is nothing wrong with using an unknown buffer type. So stop spamming the kernel log whenever this happens. The kernel will just return -EINVAL to signal this.

---
 drivers/media/v4l2-core/v4l2-compat-ioctl32.c | 4 ----
 1 file changed, 4 deletions(-)
drivers/media/v4l2-core/v4l2-compat-ioctl32.c +0 -4
@@ -170,8 +170,6 @@ static int __get_v4l2_format32(struct v4
 		return copy_from_user(&kp->fmt.sliced, &up->fmt.sliced,
 				      sizeof(kp->fmt.sliced)) ? -EFAULT : 0;
 	default:
-		printk(KERN_INFO "compat_ioctl32: unexpected VIDIOC_FMT type %d\n",
-		       kp->type);
 		return -EINVAL;
 	}
 }
@@ -217,8 +215,6 @@ static int __put_v4l2_format32(struct v4
 		return copy_to_user(&up->fmt.sliced, &kp->fmt.sliced,
 				    sizeof(kp->fmt.sliced)) ? -EFAULT : 0;
 	default:
-		printk(KERN_INFO "compat_ioctl32: unexpected VIDIOC_FMT type %d\n",
-		       kp->type);
 		return -EINVAL;
 	}
 }
#255 lore raw

PATCH 251/254 media: v4l2-compat-ioctl32.c: don't copy back the result for certain errors

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Hans Verkuil <[email protected]>

commit d83a8243aaefe62ace433e4384a4f077bed86acb upstream.

Some ioctls need to copy back the result even if the ioctl returned an error. However, don't do this for the error code -ENOTTY. It makes no sense in that cases.

---
 drivers/media/v4l2-core/v4l2-compat-ioctl32.c | 3 +++
 1 file changed, 3 insertions(+)
drivers/media/v4l2-core/v4l2-compat-ioctl32.c +3 -0
@@ -933,6 +933,9 @@ static long do_video_ioctl(struct file *
 		set_fs(old_fs);
 	}
 
+	if (err == -ENOTTY)
+		return err;
+
 	/* Special case: even after an error we need to put the
 	   results back for these ioctls since the error_idx will
 	   contain information on which control failed. */
#256 lore raw

PATCH 252/254 media: v4l2-compat-ioctl32.c: refactor compat ioctl32 logic

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Daniel Mentz <[email protected]>

commit a1dfb4c48cc1e64eeb7800a27c66a6f7e88d075a upstream.

The 32-bit compat v4l2 ioctl handling is implemented based on its 64-bit equivalent. It converts 32-bit data structures into its 64-bit equivalents and needs to provide the data to the 64-bit ioctl in user space memory which is commonly allocated using compat_alloc_user_space().

However, due to how that function is implemented, it can only be called a single time for every syscall invocation.

Supposedly to avoid this limitation, the existing code uses a mix of memory from the kernel stack and memory allocated through compat_alloc_user_space().

Under normal circumstances, this would not work, because the 64-bit ioctl expects all pointers to point to user space memory. As a workaround, set_fs(KERNEL_DS) is called to temporarily disable this extra safety check and allow kernel pointers. However, this might introduce a security vulnerability: The result of the 32-bit to 64-bit conversion is writeable by user space because the output buffer has been allocated via compat_alloc_user_space(). A malicious user space process could then manipulate pointers inside this output buffer, and due to the previous set_fs(KERNEL_DS) call, functions like get_user() or put_user() no longer prevent kernel memory access.

The new approach is to pre-calculate the total amount of user space memory that is needed, allocate it using compat_alloc_user_space() and then divide up the allocated memory to accommodate all data structures that need to be converted.

An alternative approach would have been to retain the union type karg that they allocated on the kernel stack in do_video_ioctl(), copy all data from user space into karg and then back to user space. However, we decided against this approach because it does not align with other compat syscall implementations. Instead, we tried to replicate the get_user/put_user pairs as found in other places in the kernel:

    if (get_user(clipcount, &up->clipcount) ||
        put_user(clipcount, &kp->clipcount)) return -EFAULT;

Notes from [email protected]:

This patch was taken from:

    https://github.com/LineageOS/android_kernel_samsung_apq8084/commit/97b733953c06e4f0398ade18850f0817778255f7

Clearly nobody could be bothered to upstream this patch or at minimum tell us :-( We only heard about this a week ago.

This patch was rebased and cleaned up. Compared to the original I also swapped the order of the convert_in_user arguments so that they matched copy_in_user. It was hard to review otherwise. I also replaced the ALLOC_USER_SPACE/ALLOC_AND_GET by a normal function.

Fixes 6b5a9492ca ("v4l: introduce string control support.")

---

drivers/media/v4l2-core/v4l2-compat-ioctl32.c +492 -263
@@ -22,6 +22,14 @@
 #include <media/v4l2-ctrls.h>
 #include <media/v4l2-ioctl.h>
 
+/* Use the same argument order as copy_in_user */
+#define assign_in_user(to, from)					\
+({									\
+	typeof(*from) __assign_tmp;					\
+									\
+	get_user(__assign_tmp, from) || put_user(__assign_tmp, to);	\
+})
+
 static long native_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
 {
 	long ret = -ENOIOCTLCMD;
@@ -35,12 +43,12 @@ static long native_ioctl(struct file *fi
 
 struct v4l2_clip32 {
 	struct v4l2_rect        c;
-	compat_caddr_t 		next;
+	compat_caddr_t		next;
 };
 
 struct v4l2_window32 {
 	struct v4l2_rect        w;
-	__u32		  	field;	/* enum v4l2_field */
+	__u32			field;	/* enum v4l2_field */
 	__u32			chromakey;
 	compat_caddr_t		clips; /* actually struct v4l2_clip32 * */
 	__u32			clipcount;
@@ -48,37 +56,41 @@ struct v4l2_window32 {
 	__u8                    global_alpha;
 };
 
-static int get_v4l2_window32(struct v4l2_window *kp, struct v4l2_window32 __user *up)
+static int get_v4l2_window32(struct v4l2_window __user *kp,
+			     struct v4l2_window32 __user *up,
+			     void __user *aux_buf, u32 aux_space)
 {
 	struct v4l2_clip32 __user *uclips;
 	struct v4l2_clip __user *kclips;
 	compat_caddr_t p;
-	u32 n;
+	u32 clipcount;
 
 	if (!access_ok(VERIFY_READ, up, sizeof(*up)) ||
-	    copy_from_user(&kp->w, &up->w, sizeof(up->w)) ||
-	    get_user(kp->field, &up->field) ||
-	    get_user(kp->chromakey, &up->chromakey) ||
-	    get_user(kp->clipcount, &up->clipcount) ||
-	    get_user(kp->global_alpha, &up->global_alpha))
+	    copy_in_user(&kp->w, &up->w, sizeof(up->w)) ||
+	    assign_in_user(&kp->field, &up->field) ||
+	    assign_in_user(&kp->chromakey, &up->chromakey) ||
+	    assign_in_user(&kp->global_alpha, &up->global_alpha) ||
+	    get_user(clipcount, &up->clipcount) ||
+	    put_user(clipcount, &kp->clipcount))
 		return -EFAULT;
-	if (kp->clipcount > 2048)
+	if (clipcount > 2048)
 		return -EINVAL;
-	if (!kp->clipcount) {
-		kp->clips = NULL;
-		return 0;
-	}
+	if (!clipcount)
+		return put_user(NULL, &kp->clips);
 
-	n = kp->clipcount;
 	if (get_user(p, &up->clips))
 		return -EFAULT;
 	uclips = compat_ptr(p);
-	kclips = compat_alloc_user_space(n * sizeof(*kclips));
-	kp->clips = kclips;
-	while (n--) {
+	if (aux_space < clipcount * sizeof(*kclips))
+		return -EFAULT;
+	kclips = aux_buf;
+	if (put_user(kclips, &kp->clips))
+		return -EFAULT;
+
+	while (clipcount--) {
 		if (copy_in_user(&kclips->c, &uclips->c, sizeof(uclips->c)))
 			return -EFAULT;
-		if (put_user(n ? kclips + 1 : NULL, &kclips->next))
+		if (put_user(clipcount ? kclips + 1 : NULL, &kclips->next))
 			return -EFAULT;
 		uclips++;
 		kclips++;
@@ -86,27 +98,28 @@ static int get_v4l2_window32(struct v4l2
 	return 0;
 }
 
-static int put_v4l2_window32(struct v4l2_window *kp, struct v4l2_window32 __user *up)
+static int put_v4l2_window32(struct v4l2_window __user *kp,
+			     struct v4l2_window32 __user *up)
 {
 	struct v4l2_clip __user *kclips = kp->clips;
 	struct v4l2_clip32 __user *uclips;
-	u32 n = kp->clipcount;
 	compat_caddr_t p;
+	u32 clipcount;
 
-	if (copy_to_user(&up->w, &kp->w, sizeof(kp->w)) ||
-	    put_user(kp->field, &up->field) ||
-	    put_user(kp->chromakey, &up->chromakey) ||
-	    put_user(kp->clipcount, &up->clipcount) ||
-	    put_user(kp->global_alpha, &up->global_alpha))
+	if (copy_in_user(&up->w, &kp->w, sizeof(kp->w)) ||
+	    assign_in_user(&up->field, &kp->field) ||
+	    assign_in_user(&up->chromakey, &kp->chromakey) ||
+	    assign_in_user(&up->global_alpha, &kp->global_alpha) ||
+	    get_user(clipcount, &kp->clipcount) ||
+	    put_user(clipcount, &up->clipcount))
 		return -EFAULT;
-
-	if (!kp->clipcount)
+	if (!clipcount)
 		return 0;
 
 	if (get_user(p, &up->clips))
 		return -EFAULT;
 	uclips = compat_ptr(p);
-	while (n--) {
+	while (clipcount--) {
 		if (copy_in_user(&uclips->c, &kclips->c, sizeof(uclips->c)))
 			return -EFAULT;
 		uclips++;
@@ -144,93 +157,150 @@ struct v4l2_create_buffers32 {
 	__u32			reserved[8];
 };
 
-static int __get_v4l2_format32(struct v4l2_format *kp, struct v4l2_format32 __user *up)
+static int __bufsize_v4l2_format(struct v4l2_format32 __user *up, u32 *size)
+{
+	u32 type;
+
+	if (get_user(type, &up->type))
+		return -EFAULT;
+
+	switch (type) {
+	case V4L2_BUF_TYPE_VIDEO_OVERLAY:
+	case V4L2_BUF_TYPE_VIDEO_OUTPUT_OVERLAY: {
+		u32 clipcount;
+
+		if (get_user(clipcount, &up->fmt.win.clipcount))
+			return -EFAULT;
+		if (clipcount > 2048)
+			return -EINVAL;
+		*size = clipcount * sizeof(struct v4l2_clip);
+		return 0;
+	}
+	default:
+		*size = 0;
+		return 0;
+	}
+}
+
+static int bufsize_v4l2_format(struct v4l2_format32 __user *up, u32 *size)
 {
-	if (get_user(kp->type, &up->type))
+	if (!access_ok(VERIFY_READ, up, sizeof(*up)))
 		return -EFAULT;
+	return __bufsize_v4l2_format(up, size);
+}
 
-	switch (kp->type) {
+static int __get_v4l2_format32(struct v4l2_format __user *kp,
+			       struct v4l2_format32 __user *up,
+			       void __user *aux_buf, u32 aux_space)
+{
+	u32 type;
+
+	if (get_user(type, &up->type) || put_user(type, &kp->type))
+		return -EFAULT;
+
+	switch (type) {
 	case V4L2_BUF_TYPE_VIDEO_CAPTURE:
 	case V4L2_BUF_TYPE_VIDEO_OUTPUT:
-		return copy_from_user(&kp->fmt.pix, &up->fmt.pix,
-				      sizeof(kp->fmt.pix)) ? -EFAULT : 0;
+		return copy_in_user(&kp->fmt.pix, &up->fmt.pix,
+				    sizeof(kp->fmt.pix)) ? -EFAULT : 0;
 	case V4L2_BUF_TYPE_VIDEO_CAPTURE_MPLANE:
 	case V4L2_BUF_TYPE_VIDEO_OUTPUT_MPLANE:
-		return copy_from_user(&kp->fmt.pix_mp, &up->fmt.pix_mp,
-				      sizeof(kp->fmt.pix_mp)) ? -EFAULT : 0;
+		return copy_in_user(&kp->fmt.pix_mp, &up->fmt.pix_mp,
+				    sizeof(kp->fmt.pix_mp)) ? -EFAULT : 0;
 	case V4L2_BUF_TYPE_VIDEO_OVERLAY:
 	case V4L2_BUF_TYPE_VIDEO_OUTPUT_OVERLAY:
-		return get_v4l2_window32(&kp->fmt.win, &up->fmt.win);
+		return get_v4l2_window32(&kp->fmt.win, &up->fmt.win,
+					 aux_buf, aux_space);
 	case V4L2_BUF_TYPE_VBI_CAPTURE:
 	case V4L2_BUF_TYPE_VBI_OUTPUT:
-		return copy_from_user(&kp->fmt.vbi, &up->fmt.vbi,
-				      sizeof(kp->fmt.vbi)) ? -EFAULT : 0;
+		return copy_in_user(&kp->fmt.vbi, &up->fmt.vbi,
+				    sizeof(kp->fmt.vbi)) ? -EFAULT : 0;
 	case V4L2_BUF_TYPE_SLICED_VBI_CAPTURE:
 	case V4L2_BUF_TYPE_SLICED_VBI_OUTPUT:
-		return copy_from_user(&kp->fmt.sliced, &up->fmt.sliced,
-				      sizeof(kp->fmt.sliced)) ? -EFAULT : 0;
+		return copy_in_user(&kp->fmt.sliced, &up->fmt.sliced,
+				    sizeof(kp->fmt.sliced)) ? -EFAULT : 0;
 	default:
 		return -EINVAL;
 	}
 }
 
-static int get_v4l2_format32(struct v4l2_format *kp, struct v4l2_format32 __user *up)
+static int get_v4l2_format32(struct v4l2_format __user *kp,
+			     struct v4l2_format32 __user *up,
+			     void __user *aux_buf, u32 aux_space)
+{
+	if (!access_ok(VERIFY_READ, up, sizeof(*up)))
+		return -EFAULT;
+	return __get_v4l2_format32(kp, up, aux_buf, aux_space);
+}
+
+static int bufsize_v4l2_create(struct v4l2_create_buffers32 __user *up,
+			       u32 *size)
 {
 	if (!access_ok(VERIFY_READ, up, sizeof(*up)))
 		return -EFAULT;
-	return __get_v4l2_format32(kp, up);
+	return __bufsize_v4l2_format(&up->format, size);
 }
 
-static int get_v4l2_create32(struct v4l2_create_buffers *kp, struct v4l2_create_buffers32 __user *up)
+static int get_v4l2_create32(struct v4l2_create_buffers __user *kp,
+			     struct v4l2_create_buffers32 __user *up,
+			     void __user *aux_buf, u32 aux_space)
 {
 	if (!access_ok(VERIFY_READ, up, sizeof(*up)) ||
-	    copy_from_user(kp, up, offsetof(struct v4l2_create_buffers32, format)))
+	    copy_in_user(kp, up,
+			 offsetof(struct v4l2_create_buffers32, format)))
 		return -EFAULT;
-	return __get_v4l2_format32(&kp->format, &up->format);
+	return __get_v4l2_format32(&kp->format, &up->format,
+				   aux_buf, aux_space);
 }
 
-static int __put_v4l2_format32(struct v4l2_format *kp, struct v4l2_format32 __user *up)
+static int __put_v4l2_format32(struct v4l2_format __user *kp,
+			       struct v4l2_format32 __user *up)
 {
-	if (put_user(kp->type, &up->type))
+	u32 type;
+
+	if (get_user(type, &kp->type))
 		return -EFAULT;
 
-	switch (kp->type) {
+	switch (type) {
 	case V4L2_BUF_TYPE_VIDEO_CAPTURE:
 	case V4L2_BUF_TYPE_VIDEO_OUTPUT:
-		return copy_to_user(&up->fmt.pix, &kp->fmt.pix,
+		return copy_in_user(&up->fmt.pix, &kp->fmt.pix,
 				    sizeof(kp->fmt.pix)) ? -EFAULT : 0;
 	case V4L2_BUF_TYPE_VIDEO_CAPTURE_MPLANE:
 	case V4L2_BUF_TYPE_VIDEO_OUTPUT_MPLANE:
-		return copy_to_user(&up->fmt.pix_mp, &kp->fmt.pix_mp,
+		return copy_in_user(&up->fmt.pix_mp, &kp->fmt.pix_mp,
 				    sizeof(kp->fmt.pix_mp)) ? -EFAULT : 0;
 	case V4L2_BUF_TYPE_VIDEO_OVERLAY:
 	case V4L2_BUF_TYPE_VIDEO_OUTPUT_OVERLAY:
 		return put_v4l2_window32(&kp->fmt.win, &up->fmt.win);
 	case V4L2_BUF_TYPE_VBI_CAPTURE:
 	case V4L2_BUF_TYPE_VBI_OUTPUT:
-		return copy_to_user(&up->fmt.vbi, &kp->fmt.vbi,
+		return copy_in_user(&up->fmt.vbi, &kp->fmt.vbi,
 				    sizeof(kp->fmt.vbi)) ? -EFAULT : 0;
 	case V4L2_BUF_TYPE_SLICED_VBI_CAPTURE:
 	case V4L2_BUF_TYPE_SLICED_VBI_OUTPUT:
-		return copy_to_user(&up->fmt.sliced, &kp->fmt.sliced,
+		return copy_in_user(&up->fmt.sliced, &kp->fmt.sliced,
 				    sizeof(kp->fmt.sliced)) ? -EFAULT : 0;
 	default:
 		return -EINVAL;
 	}
 }
 
-static int put_v4l2_format32(struct v4l2_format *kp, struct v4l2_format32 __user *up)
+static int put_v4l2_format32(struct v4l2_format __user *kp,
+			     struct v4l2_format32 __user *up)
 {
 	if (!access_ok(VERIFY_WRITE, up, sizeof(*up)))
 		return -EFAULT;
 	return __put_v4l2_format32(kp, up);
 }
 
-static int put_v4l2_create32(struct v4l2_create_buffers *kp, struct v4l2_create_buffers32 __user *up)
+static int put_v4l2_create32(struct v4l2_create_buffers __user *kp,
+			     struct v4l2_create_buffers32 __user *up)
 {
 	if (!access_ok(VERIFY_WRITE, up, sizeof(*up)) ||
-	    copy_to_user(up, kp, offsetof(struct v4l2_create_buffers32, format)) ||
-	    copy_to_user(up->reserved, kp->reserved, sizeof(kp->reserved)))
+	    copy_in_user(up, kp,
+			 offsetof(struct v4l2_create_buffers32, format)) ||
+	    copy_in_user(up->reserved, kp->reserved, sizeof(kp->reserved)))
 		return -EFAULT;
 	return __put_v4l2_format32(&kp->format, &up->format);
 }
@@ -244,24 +314,27 @@ struct v4l2_standard32 {
 	__u32		     reserved[4];
 };
 
-static int get_v4l2_standard32(struct v4l2_standard *kp, struct v4l2_standard32 __user *up)
+static int get_v4l2_standard32(struct v4l2_standard __user *kp,
+			       struct v4l2_standard32 __user *up)
 {
 	/* other fields are not set by the user, nor used by the driver */
 	if (!access_ok(VERIFY_READ, up, sizeof(*up)) ||
-	    get_user(kp->index, &up->index))
+	    assign_in_user(&kp->index, &up->index))
 		return -EFAULT;
 	return 0;
 }
 
-static int put_v4l2_standard32(struct v4l2_standard *kp, struct v4l2_standard32 __user *up)
+static int put_v4l2_standard32(struct v4l2_standard __user *kp,
+			       struct v4l2_standard32 __user *up)
 {
 	if (!access_ok(VERIFY_WRITE, up, sizeof(*up)) ||
-	    put_user(kp->index, &up->index) ||
-	    put_user(kp->id, &up->id) ||
-	    copy_to_user(up->name, kp->name, sizeof(up->name)) ||
-	    copy_to_user(&up->frameperiod, &kp->frameperiod, sizeof(kp->frameperiod)) ||
-	    put_user(kp->framelines, &up->framelines) ||
-	    copy_to_user(up->reserved, kp->reserved, sizeof(kp->reserved)))
+	    assign_in_user(&up->index, &kp->index) ||
+	    assign_in_user(&up->id, &kp->id) ||
+	    copy_in_user(up->name, kp->name, sizeof(up->name)) ||
+	    copy_in_user(&up->frameperiod, &kp->frameperiod,
+			 sizeof(up->frameperiod)) ||
+	    assign_in_user(&up->framelines, &kp->framelines) ||
+	    copy_in_user(up->reserved, kp->reserved, sizeof(up->reserved)))
 		return -EFAULT;
 	return 0;
 }
@@ -301,11 +374,11 @@ struct v4l2_buffer32 {
 	__u32			reserved;
 };
 
-static int get_v4l2_plane32(struct v4l2_plane __user *up, struct v4l2_plane32 __user *up32,
+static int get_v4l2_plane32(struct v4l2_plane __user *up,
+			    struct v4l2_plane32 __user *up32,
 			    enum v4l2_memory memory)
 {
-	void __user *up_pln;
-	compat_long_t p;
+	compat_ulong_t p;
 
 	if (copy_in_user(up, up32, 2 * sizeof(__u32)) ||
 	    copy_in_user(&up->data_offset, &up32->data_offset,
@@ -320,10 +393,8 @@ static int get_v4l2_plane32(struct v4l2_
 			return -EFAULT;
 		break;
 	case V4L2_MEMORY_USERPTR:
-		if (get_user(p, &up32->m.userptr))
-			return -EFAULT;
-		up_pln = compat_ptr(p);
-		if (put_user((unsigned long)up_pln, &up->m.userptr))
+		if (get_user(p, &up32->m.userptr) ||
+		    put_user((unsigned long)compat_ptr(p), &up->m.userptr))
 			return -EFAULT;
 		break;
 	case V4L2_MEMORY_DMABUF:
@@ -335,7 +406,8 @@ static int get_v4l2_plane32(struct v4l2_
 	return 0;
 }
 
-static int put_v4l2_plane32(struct v4l2_plane __user *up, struct v4l2_plane32 __user *up32,
+static int put_v4l2_plane32(struct v4l2_plane __user *up,
+			    struct v4l2_plane32 __user *up32,
 			    enum v4l2_memory memory)
 {
 	unsigned long p;
@@ -359,8 +431,7 @@ static int put_v4l2_plane32(struct v4l2_
 			return -EFAULT;
 		break;
 	case V4L2_MEMORY_DMABUF:
-		if (copy_in_user(&up32->m.fd, &up->m.fd,
-				 sizeof(up->m.fd)))
+		if (copy_in_user(&up32->m.fd, &up->m.fd, sizeof(up->m.fd)))
 			return -EFAULT;
 		break;
 	}
@@ -368,37 +439,75 @@ static int put_v4l2_plane32(struct v4l2_
 	return 0;
 }
 
-static int get_v4l2_buffer32(struct v4l2_buffer *kp, struct v4l2_buffer32 __user *up)
+static int bufsize_v4l2_buffer(struct v4l2_buffer32 __user *up, u32 *size)
 {
+	u32 type;
+	u32 length;
+
+	if (!access_ok(VERIFY_READ, up, sizeof(*up)) ||
+	    get_user(type, &up->type) ||
+	    get_user(length, &up->length))
+		return -EFAULT;
+
+	if (V4L2_TYPE_IS_MULTIPLANAR(type)) {
+		if (length > VIDEO_MAX_PLANES)
+			return -EINVAL;
+
+		/*
+		 * We don't really care if userspace decides to kill itself
+		 * by passing a very big length value
+		 */
+		*size = length * sizeof(struct v4l2_plane);
+	} else {
+		*size = 0;
+	}
+	return 0;
+}
+
+static int get_v4l2_buffer32(struct v4l2_buffer __user *kp,
+			     struct v4l2_buffer32 __user *up,
+			     void __user *aux_buf, u32 aux_space)
+{
+	u32 type;
+	u32 length;
+	enum v4l2_memory memory;
 	struct v4l2_plane32 __user *uplane32;
 	struct v4l2_plane __user *uplane;
 	compat_caddr_t p;
-	int num_planes;
 	int ret;
 
 	if (!access_ok(VERIFY_READ, up, sizeof(*up)) ||
-	    get_user(kp->index, &up->index) ||
-	    get_user(kp->type, &up->type) ||
-	    get_user(kp->flags, &up->flags) ||
-	    get_user(kp->memory, &up->memory) ||
-	    get_user(kp->length, &up->length))
-		return -EFAULT;
-
-	if (V4L2_TYPE_IS_OUTPUT(kp->type))
-		if (get_user(kp->bytesused, &up->bytesused) ||
-		    get_user(kp->field, &up->field) ||
-		    get_user(kp->timestamp.tv_sec, &up->timestamp.tv_sec) ||
-		    get_user(kp->timestamp.tv_usec, &up->timestamp.tv_usec))
+	    assign_in_user(&kp->index, &up->index) ||
+	    get_user(type, &up->type) ||
+	    put_user(type, &kp->type) ||
+	    assign_in_user(&kp->flags, &up->flags) ||
+	    get_user(memory, &up->memory) ||
+	    put_user(memory, &kp->memory) ||
+	    get_user(length, &up->length) ||
+	    put_user(length, &kp->length))
+		return -EFAULT;
+
+	if (V4L2_TYPE_IS_OUTPUT(type))
+		if (assign_in_user(&kp->bytesused, &up->bytesused) ||
+		    assign_in_user(&kp->field, &up->field) ||
+		    assign_in_user(&kp->timestamp.tv_sec,
+				   &up->timestamp.tv_sec) ||
+		    assign_in_user(&kp->timestamp.tv_usec,
+				   &up->timestamp.tv_usec))
 			return -EFAULT;
 
-	if (V4L2_TYPE_IS_MULTIPLANAR(kp->type)) {
-		num_planes = kp->length;
+	if (V4L2_TYPE_IS_MULTIPLANAR(type)) {
+		u32 num_planes = length;
+
 		if (num_planes == 0) {
-			kp->m.planes = NULL;
-			/* num_planes == 0 is legal, e.g. when userspace doesn't
-			 * need planes array on DQBUF*/
-			return 0;
+			/*
+			 * num_planes == 0 is legal, e.g. when userspace doesn't
+			 * need planes array on DQBUF
+			 */
+			return put_user(NULL, &kp->m.planes);
 		}
+		if (num_planes > VIDEO_MAX_PLANES)
+			return -EINVAL;
 
 		if (get_user(p, &up->m.planes))
 			return -EFAULT;
@@ -408,38 +517,43 @@ static int get_v4l2_buffer32(struct v4l2
 			       num_planes * sizeof(*uplane32)))
 			return -EFAULT;
 
-		/* We don't really care if userspace decides to kill itself
-		 * by passing a very big num_planes value */
-		uplane = compat_alloc_user_space(num_planes *
-						 sizeof(*uplane));
-		kp->m.planes = (__force struct v4l2_plane *)uplane;
+		/*
+		 * We don't really care if userspace decides to kill itself
+		 * by passing a very big num_planes value
+		 */
+		if (aux_space < num_planes * sizeof(*uplane))
+			return -EFAULT;
+
+		uplane = aux_buf;
+		if (put_user((__force struct v4l2_plane *)uplane,
+			     &kp->m.planes))
+			return -EFAULT;
 
-		while (--num_planes >= 0) {
-			ret = get_v4l2_plane32(uplane, uplane32, kp->memory);
+		while (num_planes--) {
+			ret = get_v4l2_plane32(uplane, uplane32, memory);
 			if (ret)
 				return ret;
-			++uplane;
-			++uplane32;
+			uplane++;
+			uplane32++;
 		}
 	} else {
-		switch (kp->memory) {
+		switch (memory) {
 		case V4L2_MEMORY_MMAP:
 		case V4L2_MEMORY_OVERLAY:
-			if (get_user(kp->m.offset, &up->m.offset))
+			if (assign_in_user(&kp->m.offset, &up->m.offset))
 				return -EFAULT;
 			break;
-		case V4L2_MEMORY_USERPTR:
-			{
-				compat_long_t tmp;
+		case V4L2_MEMORY_USERPTR: {
+			compat_ulong_t userptr;
 
-				if (get_user(tmp, &up->m.userptr))
-					return -EFAULT;
-
-				kp->m.userptr = (unsigned long)compat_ptr(tmp);
-			}
+			if (get_user(userptr, &up->m.userptr) ||
+			    put_user((unsigned long)compat_ptr(userptr),
+				     &kp->m.userptr))
+				return -EFAULT;
 			break;
+		}
 		case V4L2_MEMORY_DMABUF:
-			if (get_user(kp->m.fd, &up->m.fd))
+			if (assign_in_user(&kp->m.fd, &up->m.fd))
 				return -EFAULT;
 			break;
 		}
@@ -448,62 +562,70 @@ static int get_v4l2_buffer32(struct v4l2
 	return 0;
 }
 
-static int put_v4l2_buffer32(struct v4l2_buffer *kp, struct v4l2_buffer32 __user *up)
+static int put_v4l2_buffer32(struct v4l2_buffer __user *kp,
+			     struct v4l2_buffer32 __user *up)
 {
+	u32 type;
+	u32 length;
+	enum v4l2_memory memory;
 	struct v4l2_plane32 __user *uplane32;
 	struct v4l2_plane __user *uplane;
 	compat_caddr_t p;
-	int num_planes;
 	int ret;
 
 	if (!access_ok(VERIFY_WRITE, up, sizeof(*up)) ||
-	    put_user(kp->index, &up->index) ||
-	    put_user(kp->type, &up->type) ||
-	    put_user(kp->flags, &up->flags) ||
-	    put_user(kp->memory, &up->memory))
-		return -EFAULT;
-
-	if (put_user(kp->bytesused, &up->bytesused) ||
-	    put_user(kp->field, &up->field) ||
-	    put_user(kp->timestamp.tv_sec, &up->timestamp.tv_sec) ||
-	    put_user(kp->timestamp.tv_usec, &up->timestamp.tv_usec) ||
-	    copy_to_user(&up->timecode, &kp->timecode, sizeof(kp->timecode)) ||
-	    put_user(kp->sequence, &up->sequence) ||
-	    put_user(kp->reserved2, &up->reserved2) ||
-	    put_user(kp->reserved, &up->reserved) ||
-	    put_user(kp->length, &up->length))
+	    assign_in_user(&up->index, &kp->index) ||
+	    get_user(type, &kp->type) ||
+	    put_user(type, &up->type) ||
+	    assign_in_user(&up->flags, &kp->flags) ||
+	    get_user(memory, &kp->memory) ||
+	    put_user(memory, &up->memory))
+		return -EFAULT;
+
+	if (assign_in_user(&up->bytesused, &kp->bytesused) ||
+	    assign_in_user(&up->field, &kp->field) ||
+	    assign_in_user(&up->timestamp.tv_sec, &kp->timestamp.tv_sec) ||
+	    assign_in_user(&up->timestamp.tv_usec, &kp->timestamp.tv_usec) ||
+	    copy_in_user(&up->timecode, &kp->timecode, sizeof(kp->timecode)) ||
+	    assign_in_user(&up->sequence, &kp->sequence) ||
+	    assign_in_user(&up->reserved2, &kp->reserved2) ||
+	    assign_in_user(&up->reserved, &kp->reserved) ||
+	    get_user(length, &kp->length) ||
+	    put_user(length, &up->length))
 		return -EFAULT;
 
-	if (V4L2_TYPE_IS_MULTIPLANAR(kp->type)) {
-		num_planes = kp->length;
+	if (V4L2_TYPE_IS_MULTIPLANAR(type)) {
+		u32 num_planes = length;
+
 		if (num_planes == 0)
 			return 0;
 
-		uplane = (__force struct v4l2_plane __user *)kp->m.planes;
+		if (get_user(uplane, ((__force struct v4l2_plane __user **)&kp->m.planes)))
+			return -EFAULT;
 		if (get_user(p, &up->m.planes))
 			return -EFAULT;
 		uplane32 = compat_ptr(p);
 
-		while (--num_planes >= 0) {
-			ret = put_v4l2_plane32(uplane, uplane32, kp->memory);
+		while (num_planes--) {
+			ret = put_v4l2_plane32(uplane, uplane32, memory);
 			if (ret)
 				return ret;
 			++uplane;
 			++uplane32;
 		}
 	} else {
-		switch (kp->memory) {
+		switch (memory) {
 		case V4L2_MEMORY_MMAP:
 		case V4L2_MEMORY_OVERLAY:
-			if (put_user(kp->m.offset, &up->m.offset))
+			if (assign_in_user(&up->m.offset, &kp->m.offset))
 				return -EFAULT;
 			break;
 		case V4L2_MEMORY_USERPTR:
-			if (put_user(kp->m.userptr, &up->m.userptr))
+			if (assign_in_user(&up->m.userptr, &kp->m.userptr))
 				return -EFAULT;
 			break;
 		case V4L2_MEMORY_DMABUF:
-			if (put_user(kp->m.fd, &up->m.fd))
+			if (assign_in_user(&up->m.fd, &kp->m.fd))
 				return -EFAULT;
 			break;
 		}
@@ -515,33 +637,45 @@ static int put_v4l2_buffer32(struct v4l2
 struct v4l2_framebuffer32 {
 	__u32			capability;
 	__u32			flags;
-	compat_caddr_t 		base;
-	struct v4l2_pix_format	fmt;
+	compat_caddr_t		base;
+	struct {
+		__u32		width;
+		__u32		height;
+		__u32		pixelformat;
+		__u32		field;
+		__u32		bytesperline;
+		__u32		sizeimage;
+		__u32		colorspace;
+		__u32		priv;
+	} fmt;
 };
 
-static int get_v4l2_framebuffer32(struct v4l2_framebuffer *kp, struct v4l2_framebuffer32 __user *up)
+static int get_v4l2_framebuffer32(struct v4l2_framebuffer __user *kp,
+				  struct v4l2_framebuffer32 __user *up)
 {
-	u32 tmp;
+	compat_caddr_t tmp;
 
 	if (!access_ok(VERIFY_READ, up, sizeof(*up)) ||
 	    get_user(tmp, &up->base) ||
-	    get_user(kp->capability, &up->capability) ||
-	    get_user(kp->flags, &up->flags) ||
-	    copy_from_user(&kp->fmt, &up->fmt, sizeof(up->fmt)))
+	    put_user((__force void *)compat_ptr(tmp), &kp->base) ||
+	    assign_in_user(&kp->capability, &up->capability) ||
+	    assign_in_user(&kp->flags, &up->flags) ||
+	    copy_in_user(&kp->fmt, &up->fmt, sizeof(kp->fmt)))
 		return -EFAULT;
-	kp->base = (__force void *)compat_ptr(tmp);
 	return 0;
 }
 
-static int put_v4l2_framebuffer32(struct v4l2_framebuffer *kp, struct v4l2_framebuffer32 __user *up)
+static int put_v4l2_framebuffer32(struct v4l2_framebuffer __user *kp,
+				  struct v4l2_framebuffer32 __user *up)
 {
-	u32 tmp = (u32)((unsigned long)kp->base);
+	void *base;
 
 	if (!access_ok(VERIFY_WRITE, up, sizeof(*up)) ||
-	    put_user(tmp, &up->base) ||
-	    put_user(kp->capability, &up->capability) ||
-	    put_user(kp->flags, &up->flags) ||
-	    copy_to_user(&up->fmt, &kp->fmt, sizeof(up->fmt)))
+	    get_user(base, &kp->base) ||
+	    put_user(ptr_to_compat(base), &up->base) ||
+	    assign_in_user(&up->capability, &kp->capability) ||
+	    assign_in_user(&up->flags, &kp->flags) ||
+	    copy_in_user(&up->fmt, &kp->fmt, sizeof(kp->fmt)))
 		return -EFAULT;
 	return 0;
 }
@@ -558,18 +692,22 @@ struct v4l2_input32 {
 	__u32	     reserved[3];
 };
 
-/* The 64-bit v4l2_input struct has extra padding at the end of the struct.
-   Otherwise it is identical to the 32-bit version. */
-static inline int get_v4l2_input32(struct v4l2_input *kp, struct v4l2_input32 __user *up)
+/*
+ * The 64-bit v4l2_input struct has extra padding at the end of the struct.
+ * Otherwise it is identical to the 32-bit version.
+ */
+static inline int get_v4l2_input32(struct v4l2_input __user *kp,
+				   struct v4l2_input32 __user *up)
 {
-	if (copy_from_user(kp, up, sizeof(*up)))
+	if (copy_in_user(kp, up, sizeof(*up)))
 		return -EFAULT;
 	return 0;
 }
 
-static inline int put_v4l2_input32(struct v4l2_input *kp, struct v4l2_input32 __user *up)
+static inline int put_v4l2_input32(struct v4l2_input __user *kp,
+				   struct v4l2_input32 __user *up)
 {
-	if (copy_to_user(up, kp, sizeof(*up)))
+	if (copy_in_user(up, kp, sizeof(*up)))
 		return -EFAULT;
 	return 0;
 }
@@ -616,40 +754,64 @@ static inline bool ctrl_is_pointer(struc
 	return false;
 }
 
+static int bufsize_v4l2_ext_controls(struct v4l2_ext_controls32 __user *up,
+				     u32 *size)
+{
+	u32 count;
+
+	if (!access_ok(VERIFY_READ, up, sizeof(*up)) ||
+	    get_user(count, &up->count))
+		return -EFAULT;
+	if (count > V4L2_CID_MAX_CTRLS)
+		return -EINVAL;
+	*size = count * sizeof(struct v4l2_ext_control);
+	return 0;
+}
+
 static int get_v4l2_ext_controls32(struct file *file,
-				   struct v4l2_ext_controls *kp,
-				   struct v4l2_ext_controls32 __user *up)
+				   struct v4l2_ext_controls __user *kp,
+				   struct v4l2_ext_controls32 __user *up,
+				   void __user *aux_buf, u32 aux_space)
 {
 	struct v4l2_ext_control32 __user *ucontrols;
 	struct v4l2_ext_control __user *kcontrols;
-	int n;
+	u32 count;
+	u32 n;
 	compat_caddr_t p;
 
 	if (!access_ok(VERIFY_READ, up, sizeof(*up)) ||
-	    get_user(kp->ctrl_class, &up->ctrl_class) ||
-	    get_user(kp->count, &up->count) ||
-	    get_user(kp->error_idx, &up->error_idx) ||
-	    copy_from_user(kp->reserved, up->reserved, sizeof(kp->reserved)))
-		return -EFAULT;
-	n = kp->count;
-	if (n == 0) {
-		kp->controls = NULL;
-		return 0;
-	}
+	    assign_in_user(&kp->ctrl_class, &up->ctrl_class) ||
+	    get_user(count, &up->count) ||
+	    put_user(count, &kp->count) ||
+	    assign_in_user(&kp->error_idx, &up->error_idx) ||
+	    copy_in_user(kp->reserved, up->reserved, sizeof(kp->reserved)))
+		return -EFAULT;
+
+	if (count == 0)
+		return put_user(NULL, &kp->controls);
+	if (count > V4L2_CID_MAX_CTRLS)
+		return -EINVAL;
 	if (get_user(p, &up->controls))
 		return -EFAULT;
 	ucontrols = compat_ptr(p);
-	if (!access_ok(VERIFY_READ, ucontrols, n * sizeof(*ucontrols)))
+	if (!access_ok(VERIFY_READ, ucontrols, count * sizeof(*ucontrols)))
+		return -EFAULT;
+	if (aux_space < count * sizeof(*kcontrols))
 		return -EFAULT;
-	kcontrols = compat_alloc_user_space(n * sizeof(*kcontrols));
-	kp->controls = (__force struct v4l2_ext_control *)kcontrols;
-	while (--n >= 0) {
+	kcontrols = aux_buf;
+	if (put_user((__force struct v4l2_ext_control *)kcontrols,
+		     &kp->controls))
+		return -EFAULT;
+
+	for (n = 0; n < count; n++) {
 		u32 id;
 
 		if (copy_in_user(kcontrols, ucontrols, sizeof(*ucontrols)))
 			return -EFAULT;
+
 		if (get_user(id, &kcontrols->id))
 			return -EFAULT;
+
 		if (ctrl_is_pointer(file, id)) {
 			void __user *s;
 
@@ -666,43 +828,54 @@ static int get_v4l2_ext_controls32(struc
 }
 
 static int put_v4l2_ext_controls32(struct file *file,
-				   struct v4l2_ext_controls *kp,
+				   struct v4l2_ext_controls __user *kp,
 				   struct v4l2_ext_controls32 __user *up)
 {
 	struct v4l2_ext_control32 __user *ucontrols;
-	struct v4l2_ext_control __user *kcontrols =
-		(__force struct v4l2_ext_control __user *)kp->controls;
-	int n = kp->count;
+	struct v4l2_ext_control __user *kcontrols;
+	u32 count;
+	u32 n;
 	compat_caddr_t p;
 
 	if (!access_ok(VERIFY_WRITE, up, sizeof(*up)) ||
-	    put_user(kp->ctrl_class, &up->ctrl_class) ||
-	    put_user(kp->count, &up->count) ||
-	    put_user(kp->error_idx, &up->error_idx) ||
-	    copy_to_user(up->reserved, kp->reserved, sizeof(up->reserved)))
+	    assign_in_user(&up->ctrl_class, &kp->ctrl_class) ||
+	    get_user(count, &kp->count) ||
+	    put_user(count, &up->count) ||
+	    assign_in_user(&up->error_idx, &kp->error_idx) ||
+	    copy_in_user(up->reserved, kp->reserved, sizeof(up->reserved)) ||
+	    get_user(kcontrols, &kp->controls))
 		return -EFAULT;
-	if (!kp->count)
-		return 0;
 
+	if (!count)
+		return 0;
 	if (get_user(p, &up->controls))
 		return -EFAULT;
 	ucontrols = compat_ptr(p);
-	if (!access_ok(VERIFY_WRITE, ucontrols, n * sizeof(*ucontrols)))
+	if (!access_ok(VERIFY_WRITE, ucontrols, count * sizeof(*ucontrols)))
 		return -EFAULT;
 
-	while (--n >= 0) {
-		unsigned size = sizeof(*ucontrols);
+	for (n = 0; n < count; n++) {
+		unsigned int size = sizeof(*ucontrols);
 		u32 id;
 
-		if (get_user(id, &kcontrols->id))
+		if (get_user(id, &kcontrols->id) ||
+		    put_user(id, &ucontrols->id) ||
+		    assign_in_user(&ucontrols->size, &kcontrols->size) ||
+		    copy_in_user(&ucontrols->reserved2, &kcontrols->reserved2,
+				 sizeof(ucontrols->reserved2)))
 			return -EFAULT;
-		/* Do not modify the pointer when copying a pointer control.
-		   The contents of the pointer was changed, not the pointer
-		   itself. */
+
+		/*
+		 * Do not modify the pointer when copying a pointer control.
+		 * The contents of the pointer was changed, not the pointer
+		 * itself.
+		 */
 		if (ctrl_is_pointer(file, id))
 			size -= sizeof(ucontrols->value64);
+
 		if (copy_in_user(ucontrols, kcontrols, size))
 			return -EFAULT;
+
 		ucontrols++;
 		kcontrols++;
 	}
@@ -722,17 +895,18 @@ struct v4l2_event32 {
 	__u32				reserved[8];
 };
 
-static int put_v4l2_event32(struct v4l2_event *kp, struct v4l2_event32 __user *up)
+static int put_v4l2_event32(struct v4l2_event __user *kp,
+			    struct v4l2_event32 __user *up)
 {
 	if (!access_ok(VERIFY_WRITE, up, sizeof(*up)) ||
-	    put_user(kp->type, &up->type) ||
-	    copy_to_user(&up->u, &kp->u, sizeof(kp->u)) ||
-	    put_user(kp->pending, &up->pending) ||
-	    put_user(kp->sequence, &up->sequence) ||
-	    put_user(kp->timestamp.tv_sec, &up->timestamp.tv_sec) ||
-	    put_user(kp->timestamp.tv_nsec, &up->timestamp.tv_nsec) ||
-	    put_user(kp->id, &up->id) ||
-	    copy_to_user(up->reserved, kp->reserved, sizeof(kp->reserved)))
+	    assign_in_user(&up->type, &kp->type) ||
+	    copy_in_user(&up->u, &kp->u, sizeof(kp->u)) ||
+	    assign_in_user(&up->pending, &kp->pending) ||
+	    assign_in_user(&up->sequence, &kp->sequence) ||
+	    assign_in_user(&up->timestamp.tv_sec, &kp->timestamp.tv_sec) ||
+	    assign_in_user(&up->timestamp.tv_nsec, &kp->timestamp.tv_nsec) ||
+	    assign_in_user(&up->id, &kp->id) ||
+	    copy_in_user(up->reserved, kp->reserved, sizeof(up->reserved)))
 		return -EFAULT;
 	return 0;
 }
@@ -745,31 +919,34 @@ struct v4l2_edid32 {
 	compat_caddr_t edid;
 };
 
-static int get_v4l2_edid32(struct v4l2_edid *kp, struct v4l2_edid32 __user *up)
+static int get_v4l2_edid32(struct v4l2_edid __user *kp,
+			   struct v4l2_edid32 __user *up)
 {
-	u32 tmp;
+	compat_uptr_t tmp;
 
 	if (!access_ok(VERIFY_READ, up, sizeof(*up)) ||
-	    get_user(kp->pad, &up->pad) ||
-	    get_user(kp->start_block, &up->start_block) ||
-	    get_user(kp->blocks, &up->blocks) ||
+	    assign_in_user(&kp->pad, &up->pad) ||
+	    assign_in_user(&kp->start_block, &up->start_block) ||
+	    assign_in_user(&kp->blocks, &up->blocks) ||
 	    get_user(tmp, &up->edid) ||
-	    copy_from_user(kp->reserved, up->reserved, sizeof(kp->reserved)))
+	    put_user(compat_ptr(tmp), &kp->edid) ||
+	    copy_in_user(kp->reserved, up->reserved, sizeof(kp->reserved)))
 		return -EFAULT;
-	kp->edid = (__force u8 *)compat_ptr(tmp);
 	return 0;
 }
 
-static int put_v4l2_edid32(struct v4l2_edid *kp, struct v4l2_edid32 __user *up)
+static int put_v4l2_edid32(struct v4l2_edid __user *kp,
+			   struct v4l2_edid32 __user *up)
 {
-	u32 tmp = (u32)((unsigned long)kp->edid);
+	void *edid;
 
 	if (!access_ok(VERIFY_WRITE, up, sizeof(*up)) ||
-	    put_user(kp->pad, &up->pad) ||
-	    put_user(kp->start_block, &up->start_block) ||
-	    put_user(kp->blocks, &up->blocks) ||
-	    put_user(tmp, &up->edid) ||
-	    copy_to_user(up->reserved, kp->reserved, sizeof(up->reserved)))
+	    assign_in_user(&up->pad, &kp->pad) ||
+	    assign_in_user(&up->start_block, &kp->start_block) ||
+	    assign_in_user(&up->blocks, &kp->blocks) ||
+	    get_user(edid, &kp->edid) ||
+	    put_user(ptr_to_compat(edid), &up->edid) ||
+	    copy_in_user(up->reserved, kp->reserved, sizeof(up->reserved)))
 		return -EFAULT;
 	return 0;
 }
@@ -786,7 +963,7 @@ static int put_v4l2_edid32(struct v4l2_e
 #define VIDIOC_ENUMINPUT32	_IOWR('V', 26, struct v4l2_input32)
 #define VIDIOC_G_EDID32		_IOWR('V', 40, struct v4l2_edid32)
 #define VIDIOC_S_EDID32		_IOWR('V', 41, struct v4l2_edid32)
-#define VIDIOC_TRY_FMT32      	_IOWR('V', 64, struct v4l2_format32)
+#define VIDIOC_TRY_FMT32	_IOWR('V', 64, struct v4l2_format32)
 #define VIDIOC_G_EXT_CTRLS32    _IOWR('V', 71, struct v4l2_ext_controls32)
 #define VIDIOC_S_EXT_CTRLS32    _IOWR('V', 72, struct v4l2_ext_controls32)
 #define VIDIOC_TRY_EXT_CTRLS32  _IOWR('V', 73, struct v4l2_ext_controls32)
@@ -802,22 +979,23 @@ static int put_v4l2_edid32(struct v4l2_e
 #define VIDIOC_G_OUTPUT32	_IOR ('V', 46, s32)
 #define VIDIOC_S_OUTPUT32	_IOWR('V', 47, s32)
 
+static int alloc_userspace(unsigned int size, u32 aux_space,
+			   void __user **up_native)
+{
+	*up_native = compat_alloc_user_space(size + aux_space);
+	if (!*up_native)
+		return -ENOMEM;
+	if (clear_user(*up_native, size))
+		return -EFAULT;
+	return 0;
+}
+
 static long do_video_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
 {
-	union {
-		struct v4l2_format v2f;
-		struct v4l2_buffer v2b;
-		struct v4l2_framebuffer v2fb;
-		struct v4l2_input v2i;
-		struct v4l2_standard v2s;
-		struct v4l2_ext_controls v2ecs;
-		struct v4l2_event v2ev;
-		struct v4l2_create_buffers v2crt;
-		struct v4l2_edid v2edid;
-		unsigned long vx;
-		int vi;
-	} karg;
 	void __user *up = compat_ptr(arg);
+	void __user *up_native = NULL;
+	void __user *aux_buf;
+	u32 aux_space;
 	int compatible_arg = 1;
 	long err = 0;
 
@@ -856,30 +1034,52 @@ static long do_video_ioctl(struct file *
 	case VIDIOC_STREAMOFF:
 	case VIDIOC_S_INPUT:
 	case VIDIOC_S_OUTPUT:
-		err = get_user(karg.vi, (s32 __user *)up);
+		err = alloc_userspace(sizeof(unsigned int), 0, &up_native);
+		if (!err && assign_in_user((unsigned int __user *)up_native,
+					   (compat_uint_t __user *)up))
+			err = -EFAULT;
 		compatible_arg = 0;
 		break;
 
 	case VIDIOC_G_INPUT:
 	case VIDIOC_G_OUTPUT:
+		err = alloc_userspace(sizeof(unsigned int), 0, &up_native);
 		compatible_arg = 0;
 		break;
 
 	case VIDIOC_G_EDID:
 	case VIDIOC_S_EDID:
-		err = get_v4l2_edid32(&karg.v2edid, up);
+		err = alloc_userspace(sizeof(struct v4l2_edid), 0, &up_native);
+		if (!err)
+			err = get_v4l2_edid32(up_native, up);
 		compatible_arg = 0;
 		break;
 
 	case VIDIOC_G_FMT:
 	case VIDIOC_S_FMT:
 	case VIDIOC_TRY_FMT:
-		err = get_v4l2_format32(&karg.v2f, up);
+		err = bufsize_v4l2_format(up, &aux_space);
+		if (!err)
+			err = alloc_userspace(sizeof(struct v4l2_format),
+					      aux_space, &up_native);
+		if (!err) {
+			aux_buf = up_native + sizeof(struct v4l2_format);
+			err = get_v4l2_format32(up_native, up,
+						aux_buf, aux_space);
+		}
 		compatible_arg = 0;
 		break;
 
 	case VIDIOC_CREATE_BUFS:
-		err = get_v4l2_create32(&karg.v2crt, up);
+		err = bufsize_v4l2_create(up, &aux_space);
+		if (!err)
+			err = alloc_userspace(sizeof(struct v4l2_create_buffers),
+					      aux_space, &up_native);
+		if (!err) {
+			aux_buf = up_native + sizeof(struct v4l2_create_buffers);
+			err = get_v4l2_create32(up_native, up,
+						aux_buf, aux_space);
+		}
 		compatible_arg = 0;
 		break;
 
@@ -887,36 +1087,63 @@ static long do_video_ioctl(struct file *
 	case VIDIOC_QUERYBUF:
 	case VIDIOC_QBUF:
 	case VIDIOC_DQBUF:
-		err = get_v4l2_buffer32(&karg.v2b, up);
+		err = bufsize_v4l2_buffer(up, &aux_space);
+		if (!err)
+			err = alloc_userspace(sizeof(struct v4l2_buffer),
+					      aux_space, &up_native);
+		if (!err) {
+			aux_buf = up_native + sizeof(struct v4l2_buffer);
+			err = get_v4l2_buffer32(up_native, up,
+						aux_buf, aux_space);
+		}
 		compatible_arg = 0;
 		break;
 
 	case VIDIOC_S_FBUF:
-		err = get_v4l2_framebuffer32(&karg.v2fb, up);
+		err = alloc_userspace(sizeof(struct v4l2_framebuffer), 0,
+				      &up_native);
+		if (!err)
+			err = get_v4l2_framebuffer32(up_native, up);
 		compatible_arg = 0;
 		break;
 
 	case VIDIOC_G_FBUF:
+		err = alloc_userspace(sizeof(struct v4l2_framebuffer), 0,
+				      &up_native);
 		compatible_arg = 0;
 		break;
 
 	case VIDIOC_ENUMSTD:
-		err = get_v4l2_standard32(&karg.v2s, up);
+		err = alloc_userspace(sizeof(struct v4l2_standard), 0,
+				      &up_native);
+		if (!err)
+			err = get_v4l2_standard32(up_native, up);
 		compatible_arg = 0;
 		break;
 
 	case VIDIOC_ENUMINPUT:
-		err = get_v4l2_input32(&karg.v2i, up);
+		err = alloc_userspace(sizeof(struct v4l2_input), 0, &up_native);
+		if (!err)
+			err = get_v4l2_input32(up_native, up);
 		compatible_arg = 0;
 		break;
 
 	case VIDIOC_G_EXT_CTRLS:
 	case VIDIOC_S_EXT_CTRLS:
 	case VIDIOC_TRY_EXT_CTRLS:
-		err = get_v4l2_ext_controls32(file, &karg.v2ecs, up);
+		err = bufsize_v4l2_ext_controls(up, &aux_space);
+		if (!err)
+			err = alloc_userspace(sizeof(struct v4l2_ext_controls),
+					      aux_space, &up_native);
+		if (!err) {
+			aux_buf = up_native + sizeof(struct v4l2_ext_controls);
+			err = get_v4l2_ext_controls32(file, up_native, up,
+						      aux_buf, aux_space);
+		}
 		compatible_arg = 0;
 		break;
 	case VIDIOC_DQEVENT:
+		err = alloc_userspace(sizeof(struct v4l2_event), 0, &up_native);
 		compatible_arg = 0;
 		break;
 	}
@@ -925,25 +1152,26 @@ static long do_video_ioctl(struct file *
 
 	if (compatible_arg)
 		err = native_ioctl(file, cmd, (unsigned long)up);
-	else {
-		mm_segment_t old_fs = get_fs();
-
-		set_fs(KERNEL_DS);
-		err = native_ioctl(file, cmd, (unsigned long)&karg);
-		set_fs(old_fs);
-	}
+	else
+		err = native_ioctl(file, cmd, (unsigned long)up_native);
 
 	if (err == -ENOTTY)
 		return err;
 
-	/* Special case: even after an error we need to put the
-	   results back for these ioctls since the error_idx will
-	   contain information on which control failed. */
+	/*
+	 * Special case: even after an error we need to put the
+	 * results back for these ioctls since the error_idx will
+	 * contain information on which control failed.
+	 */
 	switch (cmd) {
 	case VIDIOC_G_EXT_CTRLS:
 	case VIDIOC_S_EXT_CTRLS:
 	case VIDIOC_TRY_EXT_CTRLS:
-		if (put_v4l2_ext_controls32(file, &karg.v2ecs, up))
+		if (put_v4l2_ext_controls32(file, up_native, up))
+			err = -EFAULT;
+		break;
+	case VIDIOC_S_EDID:
+		if (put_v4l2_edid32(up_native, up))
 			err = -EFAULT;
 		break;
 	}
@@ -955,45 +1183,46 @@ static long do_video_ioctl(struct file *
 	case VIDIOC_S_OUTPUT:
 	case VIDIOC_G_INPUT:
 	case VIDIOC_G_OUTPUT:
-		err = put_user(((s32)karg.vi), (s32 __user *)up);
+		if (assign_in_user((compat_uint_t __user *)up,
+				   ((unsigned int __user *)up_native)))
+			err = -EFAULT;
 		break;
 
 	case VIDIOC_G_FBUF:
-		err = put_v4l2_framebuffer32(&karg.v2fb, up);
+		err = put_v4l2_framebuffer32(up_native, up);
 		break;
 
 	case VIDIOC_DQEVENT:
-		err = put_v4l2_event32(&karg.v2ev, up);
+		err = put_v4l2_event32(up_native, up);
 		break;
 
 	case VIDIOC_G_EDID:
-	case VIDIOC_S_EDID:
-		err = put_v4l2_edid32(&karg.v2edid, up);
+		err = put_v4l2_edid32(up_native, up);
 		break;
 
 	case VIDIOC_G_FMT:
 	case VIDIOC_S_FMT:
 	case VIDIOC_TRY_FMT:
-		err = put_v4l2_format32(&karg.v2f, up);
+		err = put_v4l2_format32(up_native, up);
 		break;
 
 	case VIDIOC_CREATE_BUFS:
-		err = put_v4l2_create32(&karg.v2crt, up);
+		err = put_v4l2_create32(up_native, up);
 		break;
 
 	case VIDIOC_PREPARE_BUF:
 	case VIDIOC_QUERYBUF:
 	case VIDIOC_QBUF:
 	case VIDIOC_DQBUF:
-		err = put_v4l2_buffer32(&karg.v2b, up);
+		err = put_v4l2_buffer32(up_native, up);
 		break;
 
 	case VIDIOC_ENUMSTD:
-		err = put_v4l2_standard32(&karg.v2s, up);
+		err = put_v4l2_standard32(up_native, up);
 		break;
 
 	case VIDIOC_ENUMINPUT:
-		err = put_v4l2_input32(&karg.v2i, up);
+		err = put_v4l2_input32(up_native, up);
 		break;
 	}
 	return err;
#257 lore raw

PATCH 253/254 ACPI: sbshc: remove raw pointer from printk() message

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

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

commit 43cdd1b716b26f6af16da4e145b6578f98798bf6 upstream.

There's no need to be printing a raw kernel pointer to the kernel log at every boot. So just remove it, and change the whole message to use the correct dev_info() call at the same time.

---
 drivers/acpi/sbshc.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/acpi/sbshc.c +2 -2
@@ -287,8 +287,8 @@ static int acpi_smbus_hc_add(struct acpi
 	device->driver_data = hc;
 
 	acpi_ec_add_query_handler(hc->ec, hc->query_bit, NULL, smbus_alarm, hc);
-	printk(KERN_INFO PREFIX "SBS HC: EC = 0x%p, offset = 0x%0x, query_bit = 0x%0x\n",
-		hc->ec, hc->offset, hc->query_bit);
+	dev_info(&device->dev, "SBS HC: offset = 0x%0x, query_bit = 0x%0x\n",
+		 hc->offset, hc->query_bit);
 
 	return 0;
 }
#258 lore raw

PATCH 254/254 rds: Fix NULL pointer dereference in __rds_rdma_map

In reply to Ben Hutchings

3.16.55-rc1 review patch. If anyone has any objections, please let me know.

------------------

From: Håkon Bugge <[email protected]>

commit f3069c6d33f6ae63a1668737bc78aaaa51bff7ca upstream.

This is a fix for syzkaller719569, where memory registration was attempted without any underlying transport being loaded.

Analysis of the case reveals that it is the setsockopt() RDS_GET_MR
(2) and RDS_GET_MR_FOR_DEST (7) that are vulnerable.

Here is an example stack trace when the bug is hit:

BUG: unable to handle kernel NULL pointer dereference at 00000000000000c0
IP: __rds_rdma_map+0x36/0x440 [rds]
PGD 2f93d03067 P4D 2f93d03067 PUD 2f93d02067 PMD 0
Oops: 0000 [#1] SMP
Modules linked in: bridge stp llc tun rpcsec_gss_krb5 nfsv4

dns_resolver nfs fscache rds binfmt_misc sb_edac intel_powerclamp coretemp kvm_intel kvm irqbypass crct10dif_pclmul c rc32_pclmul ghash_clmulni_intel pcbc aesni_intel crypto_simd glue_helper cryptd iTCO_wdt mei_me sg iTCO_vendor_support ipmi_si mei ipmi_devintf nfsd shpchp pcspkr i2c_i801 ioatd ma ipmi_msghandler wmi lpc_ich mfd_core auth_rpcgss nfs_acl lockd grace sunrpc ip_tables ext4 mbcache jbd2 mgag200 i2c_algo_bit drm_kms_helper ixgbe syscopyarea ahci sysfillrect sysimgblt libahci mdio fb_sys_fops ttm ptp libata sd_mod mlx4_core drm crc32c_intel pps_core megaraid_sas i2c_core dca dm_mirror dm_region_hash dm_log dm_mod

CPU: 48 PID: 45787 Comm: repro_set2 Not tainted 4.14.2-3.el7uek.x86_64 #2
Hardware name: Oracle Corporation ORACLE SERVER X5-2L/ASM,MOBO TRAY,2U, BIOS 31110000 03/03/2017
task: ffff882f9190db00 task.stack: ffffc9002b994000
RIP: 0010:__rds_rdma_map+0x36/0x440 [rds]
RSP: 0018:ffffc9002b997df0 EFLAGS: 00010202
RAX: 0000000000000000 RBX: ffff882fa2182580 RCX: 0000000000000000
RDX: 0000000000000000 RSI: ffffc9002b997e40 RDI: ffff882fa2182580
RBP: ffffc9002b997e30 R08: 0000000000000000 R09: 0000000000000002
R10: ffff885fb29e3838 R11: 0000000000000000 R12: ffff882fa2182580
R13: ffff882fa2182580 R14: 0000000000000002 R15: 0000000020000ffc
FS:  00007fbffa20b700(0000) GS:ffff882fbfb80000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00000000000000c0 CR3: 0000002f98a66006 CR4: 00000000001606e0
Call Trace:
 rds_get_mr+0x56/0x80 [rds]
 rds_setsockopt+0x172/0x340 [rds]
 ? __fget_light+0x25/0x60
 ? __fdget+0x13/0x20
 SyS_setsockopt+0x80/0xe0
 do_syscall_64+0x67/0x1b0
 entry_SYSCALL64_slow_path+0x25/0x25
RIP: 0033:0x7fbff9b117f9
RSP: 002b:00007fbffa20aed8 EFLAGS: 00000293 ORIG_RAX: 0000000000000036
RAX: ffffffffffffffda RBX: 00000000000c84a4 RCX: 00007fbff9b117f9
RDX: 0000000000000002 RSI: 0000400000000114 RDI: 000000000000109b
RBP: 00007fbffa20af10 R08: 0000000000000020 R09: 00007fbff9dd7860
R10: 0000000020000ffc R11: 0000000000000293 R12: 0000000000000000
R13: 00007fbffa20b9c0 R14: 00007fbffa20b700 R15: 0000000000000021

Code: 41 56 41 55 49 89 fd 41 54 53 48 83 ec 18 8b 87 f0 02 00 00 48

89 55 d0 48 89 4d c8 85 c0 0f 84 2d 03 00 00 48 8b 87 00 03 00 00 <48> 83 b8 c0 00 00 00 00 0f 84 25 03 00 0 0 48 8b 06 48 8b 56 08

The fix is to check the existence of an underlying transport in __rds_rdma_map().

---
 net/rds/rdma.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/rds/rdma.c +1 -1
@@ -184,7 +184,7 @@ static int __rds_rdma_map(struct rds_soc
 	long i;
 	int ret;
 
-	if (rs->rs_bound_addr == 0) {
+	if (rs->rs_bound_addr == 0 || !rs->rs_transport) {
 		ret = -ENOTCONN; /* XXX not a great errno */
 		goto out;
 	}
#259 lore raw
In reply to Ben Hutchings
Quoted 8 lines · Ben Hutchings

This is the start of the stable review cycle for the 3.16.55 release. There are 254 patches in this series, which will be posted as responses to this one. If anyone has any issues with these being applied, please let me know.

Responses should be made by Fri Mar 02 18:00:00 UTC 2018. Anything received after that time might be too late.

Build results:

	total: 136 pass: 136 fail: 0
Qemu test results:
	total: 108 pass: 108 fail: 0

Details are available at http://kerneltests.org/builders.

Guenter

#260 lore raw
Quoted 10 lines · Guenter Roeck
Quoted 8 lines · Ben Hutchings

This is the start of the stable review cycle for the 3.16.55 release. There are 254 patches in this series, which will be posted as responses to this one. If anyone has any issues with these being applied, please let me know.

Responses should be made by Fri Mar 02 18:00:00 UTC 2018. Anything received after that time might be too late.

Quoted 6 lines

Build results:

	total: 136 pass: 136 fail: 0
Qemu test results:
	total: 108 pass: 108 fail: 0

Details are available at http://kerneltests.org/builders.

Thanks for testing both of these.

Ben.

Signature · 2 lines
Ben Hutchings
If the facts do not conform to your theory, they must be disposed of.