2.6.38.4-rc1-stable review

This is the start of the stable review cycle for the 2.6.38.4 release. There are 70 patches in this series, all will be posted as a response to this one. If anyone has any issues with these being applied, please let us know. If anyone is a maintainer of the proper subsystem, and wants to add a Signed-off-by: line to the patch, please respond with it.

Responses should be made by Thursday, April 21, 2011, 20:00:00 UTC. Anything received after that time might be too late.

The whole patch series can be found in one patch at:

	kernel.org/pub/linux/kernel/v2.6/stable-review/patch-2.6.38.4-rc1.gz

and the diffstat can be found below.

thanks,

greg k-h

 MAINTAINERS                                        |    1 -
 Makefile                                           |    2 +-
 arch/arm/kernel/hw_breakpoint.c                    |    7 +
 arch/ia64/kernel/mca.c                             |    3 +-
 arch/ia64/sn/pci/tioca_provider.c                  |    2 +-
 arch/powerpc/kernel/perf_event.c                   |   37 +-
 arch/powerpc/kernel/time.c                         |    3 +
 arch/sparc/kernel/entry.S                          |    2 +-
 arch/sparc/kernel/iommu.c                          |    3 -
 arch/sparc/kernel/ldc.c                            |   28 +-
 arch/sparc/kernel/pci.c                            |    1 +
 arch/sparc/kernel/pci_common.c                     |   11 +-
 arch/sparc/kernel/pci_fire.c                       |    2 -
 arch/sparc/kernel/pci_schizo.c                     |    4 +-
 arch/sparc/kernel/pci_sun4v.c                      |    3 +-
 arch/sparc/kernel/pcr.c                            |    2 +-
 arch/sparc/kernel/ptrace_64.c                      |    3 +-
 arch/sparc/kernel/smp_64.c                         |   11 +-
 arch/sparc/kernel/traps_64.c                       |    3 +-
 arch/sparc/kernel/una_asm_64.S                     |    2 +-
 arch/sparc/mm/fault_32.c                           |    3 +-
 arch/x86/include/asm/msr-index.h                   |    4 +
 arch/x86/kernel/cpu/amd.c                          |   23 +
 arch/x86/lib/semaphore_32.S                        |    2 +-
 drivers/gpu/drm/radeon/atom.c                      |    6 +-
 drivers/gpu/drm/radeon/atombios_crtc.c             |    6 +
 drivers/gpu/drm/radeon/radeon_fence.c              |    2 +-
 drivers/gpu/drm/radeon/radeon_ring.c               |    2 +-
 drivers/i2c/algos/i2c-algo-bit.c                   |   22 +-
 drivers/media/video/sn9c102/sn9c102_core.c         |    6 +-
 drivers/net/gianfar.c                              |   16 +-
 drivers/net/gianfar.h                              |    1 +
 drivers/net/pppoe.c                                |    2 +-
 drivers/net/usb/cdc-phonet.c                       |    9 +-
 drivers/rtc/rtc-bfin.c                             |    2 +
 drivers/usb/core/devices.c                         |   10 +-
 drivers/usb/core/hcd.c                             |    2 +-
 drivers/usb/host/ehci-q.c                          |   15 +-
 drivers/usb/host/xhci-mem.c                        |   90 ++-
 drivers/usb/host/xhci.c                            |   16 +-
 drivers/usb/host/xhci.h                            |    6 +-
 drivers/usb/musb/Kconfig                           |    6 +-
 drivers/usb/serial/ftdi_sio.c                      |    5 +
 drivers/usb/serial/ftdi_sio_ids.h                  |   12 +
 drivers/usb/serial/option.c                        |    5 +
 drivers/usb/serial/qcserial.c                      |   31 +-
 drivers/video/via/viafbdev.h                       |    3 -
 fs/binfmt_elf.c                                    |    6 +-
 fs/cifs/cifsencrypt.c                              |   15 +-
 fs/cifs/cifsfs.c                                   |    1 +
 fs/cifs/connect.c                                  |    2 +-
 fs/cifs/file.c                                     |    6 +-
 fs/namei.c                                         |    1 +
 fs/proc/base.c                                     |    9 +-
 fs/ramfs/file-nommu.c                              |    1 +
 fs/ubifs/debug.c                                   |    6 +-
 fs/ubifs/file.c                                    |    3 +
 fs/ubifs/super.c                                   |    3 +-
 include/linux/kernel.h                             |   70 ++-
 include/linux/perf_event.h                         |    2 +-
 include/linux/pid.h                                |    2 +-
 include/linux/sched.h                              |    3 +
 include/net/bluetooth/hci.h                        |    2 +
 kernel/futex.c                                     |    2 +-
 kernel/pid.c                                       |    5 +-
 kernel/sched_fair.c                                |   11 +-
 lib/Kconfig.debug                                  |    3 +
 lib/Makefile                                       |    2 +
 lib/kstrtox.c                                      |  227 ++++++
 lib/test-kstrtox.c                                 |  739 ++++++++++++++++++++
 lib/vsprintf.c                                     |  141 ----
 mm/huge_memory.c                                   |   24 +-
 mm/memory.c                                        |   26 +-
 mm/mlock.c                                         |   13 -
 mm/mmap.c                                          |   15 +-
 mm/oom_kill.c                                      |   27 -
 mm/vmscan.c                                        |   24 +-
 net/8021q/vlan_dev.c                               |    1 +
 net/bluetooth/hci_core.c                           |    5 +-
 net/bluetooth/hci_event.c                          |    4 +-
 net/bridge/br_multicast.c                          |    2 +-
 net/bridge/br_netfilter.c                          |    3 +
 net/ipv4/fib_frontend.c                            |    2 +
 net/ipv4/route.c                                   |    2 +-
 net/ipv4/tcp_input.c                               |   12 +-
 net/ipv6/sysctl_net_ipv6.c                         |   18 +-
 net/l2tp/l2tp_eth.c                                |    2 +-
 net/sctp/protocol.c                                |    4 +-
 net/xfrm/xfrm_policy.c                             |    2 +
 scripts/checkpatch.pl                              |    4 +-
 sound/pci/hda/patch_realtek.c                      |   21 +-
 tools/perf/bench/sched-pipe.c                      |    2 +-
 tools/perf/builtin-sched.c                         |   12 +-
 tools/perf/builtin-top.c                           |    5 +-
 tools/perf/util/header.c                           |    2 +-
 tools/perf/util/hist.c                             |    3 -
 .../util/scripting-engines/trace-event-python.c    |    3 +-
 tools/perf/util/symbol.c                           |    4 +-
 tools/perf/util/trace-event-parse.c                |    2 +-
 tools/perf/util/ui/browsers/map.c                  |    2 +-
 100 files changed, 1522 insertions(+), 423 deletions(-)

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Linus Torvalds <[email protected]>

commit a626ca6a656450e9f4df91d0dda238fff23285f4 upstream.

Commit 982134ba6261 ("mm: avoid wrapping vm_pgoff in mremap()") fixed the case of a expanding mapping causing vm_pgoff wrapping when you used mremap. But there was another case where we expand mappings hiding in plain sight: the automatic stack expansion.

This fixes that case too.

This one also found by Robert Święcki, using his nasty system call fuzzer tool. Good job.

---
 mm/mmap.c |   13 ++++++++-----
 1 file changed, 8 insertions(+), 5 deletions(-)

@@ -1814,11 +1814,14 @@ static int expand_downwards(struct vm_ar
 		size = vma->vm_end - address;
 		grow = (vma->vm_start - address) >> PAGE_SHIFT;
 
-		error = acct_stack_growth(vma, size, grow);
-		if (!error) {
-			vma->vm_start = address;
-			vma->vm_pgoff -= grow;
-			perf_event_mmap(vma);
+		error = -ENOMEM;
+		if (grow <= vma->vm_pgoff) {
+			error = acct_stack_growth(vma, size, grow);
+			if (!error) {
+				vma->vm_start = address;
+				vma->vm_pgoff -= grow;
+				perf_event_mmap(vma);
+			}
 		}
 	}
 	vma_unlock_anon_vma(vma);

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Alex Deucher <[email protected]>

commit 9bb09fa1b5b07459279301ac6220d575f307597b upstream.

Prefer minm over maxp.

Fixes:
https://bugs.freedesktop.org/show_bug.cgi?id=35994

---
 drivers/gpu/drm/radeon/atombios_crtc.c |    6 ++++++
 1 file changed, 6 insertions(+)

@@ -531,6 +531,12 @@ static u32 atombios_adjust_pll(struct dr
 			pll->flags |= RADEON_PLL_PREFER_HIGH_FB_DIV;
 		else
 			pll->flags |= RADEON_PLL_PREFER_LOW_REF_DIV;
+
+		if ((rdev->family == CHIP_R600) ||
+		    (rdev->family == CHIP_RV610) ||
+		    (rdev->family == CHIP_RV630) ||
+		    (rdev->family == CHIP_RV670))
+			pll->flags |= RADEON_PLL_PREFER_MINM_OVER_MAXP;
 	} else {
 		pll->flags |= RADEON_PLL_LEGACY;
 

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Alex Deucher <[email protected]>

commit 71e16bfbd2b1c63d4d97cc5059694c9346aee340 upstream.

Apparently only rv515 asics need the workaround added in f24d86f1a49505cdea56728b853a5d0a3f8e3d11 (drm/radeon/kms: fix resume regression for some r5xx laptops).

Fixes:
https://bugs.freedesktop.org/show_bug.cgi?id=34709

---
 drivers/gpu/drm/radeon/atom.c |    6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

@@ -32,6 +32,7 @@
 #include "atom.h"
 #include "atom-names.h"
 #include "atom-bits.h"
+#include "radeon.h"
 
 #define ATOM_COND_ABOVE		0
 #define ATOM_COND_ABOVEOREQUAL	1

@@ -101,7 +102,9 @@ static void debug_print_spaces(int n)
 static uint32_t atom_iio_execute(struct atom_context *ctx, int base,
 				 uint32_t index, uint32_t data)
 {
+	struct radeon_device *rdev = ctx->card->dev->dev_private;
 	uint32_t temp = 0xCDCDCDCD;
+
 	while (1)
 		switch (CU8(base)) {
 		case ATOM_IIO_NOP:

@@ -112,7 +115,8 @@ static uint32_t atom_iio_execute(struct
 			base += 3;
 			break;
 		case ATOM_IIO_WRITE:
-			(void)ctx->card->ioreg_read(ctx->card, CU16(base + 1));
+			if (rdev->family == CHIP_RV515)
+				(void)ctx->card->ioreg_read(ctx->card, CU16(base + 1));
 			ctx->card->ioreg_write(ctx->card, CU16(base + 1), temp);
 			base += 3;
 			break;

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Jeff Layton <[email protected]>

commit 70945643722ffeac779d2529a348f99567fa5c33 upstream.

Currently, we skip doing the is_path_accessible check in cifs_mount if there is no prefixpath. I have a report of at least one server however that allows a TREE_CONNECT to a share that has a DFS referral at its root. The reporter in this case was using a UNC that had no prefixpath, so the is_path_accessible check was not triggered and the box later hit a BUG() because we were chasing a DFS referral on the root dentry for the mount.

This patch fixes this by removing the check for a zero-length prefixpath. That should make the is_path_accessible check be done in this situation and should allow the client to chase the DFS referral at mount time instead.

---
 fs/cifs/connect.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -2826,7 +2826,7 @@ try_mount_again:
 
 remote_path_check:
 	/* check if a whole path (including prepath) is not remote */
-	if (!rc && cifs_sb->prepathlen && tcon) {
+	if (!rc && tcon) {
 		/* build_path_to_root works only when we have a valid tcon */
 		full_path = cifs_build_path_to_root(cifs_sb, tcon);
 		if (full_path == NULL) {

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Jeff Layton <[email protected]>

commit 7797069305d13252fd66cf722aa8f2cbeb3c95cd upstream.

cifs_close doesn't check that the filp->private_data is non-NULL before trying to put it. That can cause an oops in certain error conditions that can occur on open or lookup before the private_data is set.

---
 fs/cifs/file.c |    6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

@@ -575,8 +575,10 @@ reopen_error_exit:
 
 int cifs_close(struct inode *inode, struct file *file)
 {
-	cifsFileInfo_put(file->private_data);
-	file->private_data = NULL;
+	if (file->private_data != NULL) {
+		cifsFileInfo_put(file->private_data);
+		file->private_data = NULL;
+	}
 
 	/* return code from the ->release op is always ignored */
 	return 0;

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Jeff Layton <[email protected]>

commit 2b6c26a0a62cc0bab0ad487533d5581d7c293fef upstream.

Commit 522440ed made cifs set backing_dev_info on the mapping attached to new inodes. This change caused a fairly significant read performance regression, as cifs started doing page-sized reads exclusively.

By virtue of the fact that they're allocated as part of cifs_sb_info by kzalloc, the ra_pages on cifs BDIs get set to 0, which prevents any readahead. This forces the normal read codepaths to use readpage instead of readpages causing a four-fold increase in the number of read calls with the default rsize.

Fix it by setting ra_pages in the BDI to the same value as that in the default_backing_dev_info.

Fixes https://bugzilla.kernel.org/show_bug.cgi?id=31662

---
 fs/cifs/cifsfs.c |    1 +
 1 file changed, 1 insertion(+)

@@ -127,6 +127,7 @@ cifs_read_super(struct super_block *sb,
 		kfree(cifs_sb);
 		return rc;
 	}
+	cifs_sb->bdi.ra_pages = default_backing_dev_info.ra_pages;
 
 #ifdef CONFIG_CIFS_DFS_UPCALL
 	/* copy mount params to sb for use in submounts */

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Jeff Layton <[email protected]>

commit 157c249114508aa71daa308a426e15d81a4eed00 upstream.

While testing my patchset to fix asynchronous writes, I hit a bunch of signature problems when testing with signing on. The problem seems to be that signature checks on receive can be running at the same time as a process that is sending, or even that multiple receives can be checking signatures at the same time, clobbering the same data structures.

While we're at it, clean up the comments over cifs_calculate_signature and add a note that the srv_mutex should be held when calling this function.

This patch seems to fix the problems for me, but I'm not clear on whether it's the best approach. If it is, then this should probably go to stable too.

---
 fs/cifs/cifsencrypt.c |   15 +++++++++------
 1 file changed, 9 insertions(+), 6 deletions(-)

@@ -30,12 +30,13 @@
 #include <linux/ctype.h>
 #include <linux/random.h>
 
-/* Calculate and return the CIFS signature based on the mac key and SMB PDU */
-/* the 16 byte signature must be allocated by the caller  */
-/* Note we only use the 1st eight bytes */
-/* Note that the smb header signature field on input contains the
-	sequence number before this function is called */
-
+/*
+ * Calculate and return the CIFS signature based on the mac key and SMB PDU.
+ * The 16 byte signature must be allocated by the caller. Note we only use the
+ * 1st eight bytes and that the smb header signature field on input contains
+ * the sequence number before this function is called. Also, this function
+ * should be called with the server->srv_mutex held.
+ */
 static int cifs_calculate_signature(const struct smb_hdr *cifs_pdu,
 				struct TCP_Server_Info *server, char *signature)
 {

@@ -209,8 +210,10 @@ int cifs_verify_signature(struct smb_hdr
 					cpu_to_le32(expected_sequence_number);
 	cifs_pdu->Signature.Sequence.Reserved = 0;
 
+	mutex_lock(&server->srv_mutex);
 	rc = cifs_calculate_signature(cifs_pdu, server,
 		what_we_think_sig_should_be);
+	mutex_unlock(&server->srv_mutex);
 
 	if (rc)
 		return rc;

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Vasiliy Kulikov <[email protected]>

commit 14ddc3188d50855ae2a419a6aced995e2834e5d4 upstream.

Don't allow everybody to change video settings.

---
 drivers/media/video/sn9c102/sn9c102_core.c |    6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

@@ -1430,9 +1430,9 @@ static DEVICE_ATTR(i2c_reg, S_IRUGO | S_
 		   sn9c102_show_i2c_reg, sn9c102_store_i2c_reg);
 static DEVICE_ATTR(i2c_val, S_IRUGO | S_IWUSR,
 		   sn9c102_show_i2c_val, sn9c102_store_i2c_val);
-static DEVICE_ATTR(green, S_IWUGO, NULL, sn9c102_store_green);
-static DEVICE_ATTR(blue, S_IWUGO, NULL, sn9c102_store_blue);
-static DEVICE_ATTR(red, S_IWUGO, NULL, sn9c102_store_red);
+static DEVICE_ATTR(green, S_IWUSR, NULL, sn9c102_store_green);
+static DEVICE_ATTR(blue, S_IWUSR, NULL, sn9c102_store_blue);
+static DEVICE_ATTR(red, S_IWUSR, NULL, sn9c102_store_red);
 static DEVICE_ATTR(frame_header, S_IRUGO, sn9c102_show_frame_header, NULL);
 
 

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Vasiliy Kulikov <[email protected]>

commit 8c559d30b4e59cf6994215ada1fe744928f494bf upstream.

Don't allow everybody to dump sensitive information about filesystems.

---
 fs/ubifs/debug.c |    6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

@@ -2844,19 +2844,19 @@ int dbg_debugfs_init_fs(struct ubifs_inf
 	}
 
 	fname = "dump_lprops";
-	dent = debugfs_create_file(fname, S_IWUGO, d->dfs_dir, c, &dfs_fops);
+	dent = debugfs_create_file(fname, S_IWUSR, d->dfs_dir, c, &dfs_fops);
 	if (IS_ERR(dent))
 		goto out_remove;
 	d->dfs_dump_lprops = dent;
 
 	fname = "dump_budg";
-	dent = debugfs_create_file(fname, S_IWUGO, d->dfs_dir, c, &dfs_fops);
+	dent = debugfs_create_file(fname, S_IWUSR, d->dfs_dir, c, &dfs_fops);
 	if (IS_ERR(dent))
 		goto out_remove;
 	d->dfs_dump_budg = dent;
 
 	fname = "dump_tnc";
-	dent = debugfs_create_file(fname, S_IWUGO, d->dfs_dir, c, &dfs_fops);
+	dent = debugfs_create_file(fname, S_IWUSR, d->dfs_dir, c, &dfs_fops);
 	if (IS_ERR(dent))
 		goto out_remove;
 	d->dfs_dump_tnc = dent;

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Takashi Iwai <[email protected]>

commit c6b358748e19ce7e230b0926ac42696bc485a562 upstream.

Use pin-fix instead of the static quirk for Gigabyte mobos 1458:a002.

Bugzilla: https://bugzilla.novell.com/show_bug.cgi?id=677256

---
 sound/pci/hda/patch_realtek.c |   21 ++++++++++++++++++---
 1 file changed, 18 insertions(+), 3 deletions(-)

@@ -9932,7 +9932,6 @@ static struct snd_pci_quirk alc882_cfg_t
 	SND_PCI_QUIRK(0x1071, 0x8258, "Evesham Voyaeger", ALC883_LAPTOP_EAPD),
 	SND_PCI_QUIRK(0x10f1, 0x2350, "TYAN-S2350", ALC888_6ST_DELL),
 	SND_PCI_QUIRK(0x108e, 0x534d, NULL, ALC883_3ST_6ch),
-	SND_PCI_QUIRK(0x1458, 0xa002, "Gigabyte P35 DS3R", ALC882_6ST_DIG),
 
 	SND_PCI_QUIRK(0x1462, 0x0349, "MSI", ALC883_TARGA_2ch_DIG),
 	SND_PCI_QUIRK(0x1462, 0x040d, "MSI", ALC883_TARGA_2ch_DIG),

@@ -10769,6 +10768,7 @@ enum {
 	PINFIX_LENOVO_Y530,
 	PINFIX_PB_M5210,
 	PINFIX_ACER_ASPIRE_7736,
+	PINFIX_GIGABYTE_880GM,
 };
 
 static const struct alc_fixup alc882_fixups[] = {

@@ -10800,6 +10800,13 @@ static const struct alc_fixup alc882_fix
 		.type = ALC_FIXUP_SKU,
 		.v.sku = ALC_FIXUP_SKU_IGNORE,
 	},
+	[PINFIX_GIGABYTE_880GM] = {
+		.type = ALC_FIXUP_PINS,
+		.v.pins = (const struct alc_pincfg[]) {
+			{ 0x14, 0x1114410 }, /* set as speaker */
+			{ }
+		}
+	},
 };
 
 static struct snd_pci_quirk alc882_fixup_tbl[] = {

@@ -10807,6 +10814,7 @@ static struct snd_pci_quirk alc882_fixup
 	SND_PCI_QUIRK(0x17aa, 0x3a0d, "Lenovo Y530", PINFIX_LENOVO_Y530),
 	SND_PCI_QUIRK(0x147b, 0x107a, "Abit AW9D-MAX", PINFIX_ABIT_AW9D_MAX),
 	SND_PCI_QUIRK(0x1025, 0x0296, "Acer Aspire 7736z", PINFIX_ACER_ASPIRE_7736),
+	SND_PCI_QUIRK(0x1458, 0xa002, "Gigabyte", PINFIX_GIGABYTE_880GM),
 	{}
 };
 

@@ -18851,8 +18859,6 @@ static struct snd_pci_quirk alc662_cfg_t
 		      ALC662_3ST_6ch_DIG),
 	SND_PCI_QUIRK(0x1179, 0xff6e, "Toshiba NB20x", ALC662_AUTO),
 	SND_PCI_QUIRK(0x144d, 0xca00, "Samsung NC10", ALC272_SAMSUNG_NC10),
-	SND_PCI_QUIRK(0x1458, 0xa002, "Gigabyte 945GCM-S2L",
-		      ALC662_3ST_6ch_DIG),
 	SND_PCI_QUIRK(0x152d, 0x2304, "Quanta WH1", ALC663_ASUS_H13),
 	SND_PCI_QUIRK(0x1565, 0x820f, "Biostar TA780G M2+", ALC662_3ST_6ch_DIG),
 	SND_PCI_QUIRK(0x1631, 0xc10c, "PB RS65", ALC663_ASUS_M51VA),

@@ -19526,6 +19532,7 @@ enum {
 	ALC662_FIXUP_IDEAPAD,
 	ALC272_FIXUP_MARIO,
 	ALC662_FIXUP_CZC_P10T,
+	ALC662_FIXUP_GIGABYTE,
 };
 
 static const struct alc_fixup alc662_fixups[] = {

@@ -19554,12 +19561,20 @@ static const struct alc_fixup alc662_fix
 			{}
 		}
 	},
+	[ALC662_FIXUP_GIGABYTE] = {
+		.type = ALC_FIXUP_PINS,
+		.v.pins = (const struct alc_pincfg[]) {
+			{ 0x14, 0x1114410 }, /* set as speaker */
+			{ }
+		}
+	},
 };
 
 static struct snd_pci_quirk alc662_fixup_tbl[] = {
 	SND_PCI_QUIRK(0x1025, 0x0308, "Acer Aspire 8942G", ALC662_FIXUP_ASPIRE),
 	SND_PCI_QUIRK(0x1025, 0x038b, "Acer Aspire 8943G", ALC662_FIXUP_ASPIRE),
 	SND_PCI_QUIRK(0x144d, 0xc051, "Samsung R720", ALC662_FIXUP_IDEAPAD),
+	SND_PCI_QUIRK(0x1458, 0xa002, "Gigabyte", ALC662_FIXUP_GIGABYTE),
 	SND_PCI_QUIRK(0x17aa, 0x38af, "Lenovo Ideapad Y550P", ALC662_FIXUP_IDEAPAD),
 	SND_PCI_QUIRK(0x17aa, 0x3a0d, "Lenovo Ideapad Y550", ALC662_FIXUP_IDEAPAD),
 	SND_PCI_QUIRK(0x1b35, 0x2206, "CZC P10T", ALC662_FIXUP_CZC_P10T),

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Jiri Slaby <[email protected]>

commit 468c3f924f043cad7a04f4f4d5224a2c9bc886c1 upstream.

Currently, for N 5800 XM I get:
cdc_phonet: probe of 1-6:1.10 failed with error -22

It's because phonet_header is empty. Extra altsetting looks like there:
E 05 24 00 01 10 03 24 ab 05 24 06 0a 0b 04 24 fd .$....$..$....$.
E 00 .

I don't see the header used anywhere so just check if the phonet descriptor is there, not the structure itself.

---
 drivers/net/usb/cdc-phonet.c |    9 +++------
 1 file changed, 3 insertions(+), 6 deletions(-)

@@ -328,13 +328,13 @@ int usbpn_probe(struct usb_interface *in
 {
 	static const char ifname[] = "usbpn%d";
 	const struct usb_cdc_union_desc *union_header = NULL;
-	const struct usb_cdc_header_desc *phonet_header = NULL;
 	const struct usb_host_interface *data_desc;
 	struct usb_interface *data_intf;
 	struct usb_device *usbdev = interface_to_usbdev(intf);
 	struct net_device *dev;
 	struct usbpn_dev *pnd;
 	u8 *data;
+	int phonet = 0;
 	int len, err;
 
 	data = intf->altsetting->extra;

@@ -355,10 +355,7 @@ int usbpn_probe(struct usb_interface *in
 					(struct usb_cdc_union_desc *)data;
 				break;
 			case 0xAB:
-				if (phonet_header || dlen < 5)
-					break;
-				phonet_header =
-					(struct usb_cdc_header_desc *)data;
+				phonet = 1;
 				break;
 			}
 		}

@@ -366,7 +363,7 @@ int usbpn_probe(struct usb_interface *in
 		len -= dlen;
 	}
 
-	if (!union_header || !phonet_header)
+	if (!union_header || !phonet)
 		return -EINVAL;
 
 	data_intf = usb_ifnum_to_if(usbdev, union_header->bSlaveInterface0);

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Jan Beulich <[email protected]>

commit e938c287ea8d977e079f07464ac69923412663ce upstream.

'simple' would have required specifying current frame address and return address location manually, but that's obviously not the case (and not necessary) here.

LKML-Reference: <[email protected]>

---
 arch/x86/lib/semaphore_32.S |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -36,7 +36,7 @@
  */
 #ifdef CONFIG_SMP
 ENTRY(__write_lock_failed)
-	CFI_STARTPROC simple
+	CFI_STARTPROC
 	FRAME
 2: 	LOCK_PREFIX
 	addl	$ RW_LOCK_BIAS,(%eax)

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Jeff Mahoney <[email protected]>

commit b4a6b3436531f6c5256e6d60d388c3c28ff1a0e9 upstream.

The prototype for sn_pci_provider->{dma_map,dma_map_consistent} expects an unsigned long instead of a u64.

---
 arch/ia64/sn/pci/tioca_provider.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -509,7 +509,7 @@ tioca_dma_unmap(struct pci_dev *pdev, dm
  * use the GART mapped mode.
  */
 static u64
-tioca_dma_map(struct pci_dev *pdev, u64 paddr, size_t byte_count, int dma_flags)
+tioca_dma_map(struct pci_dev *pdev, unsigned long paddr, size_t byte_count, int dma_flags)
 {
 	u64 mapaddr;
 

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Jeff Mahoney <[email protected]>

commit c1d036c4d1cb00b7e8473a2ad0a78f13e13a8183 upstream.

ia64_mca_cpu_init has a void *data local variable that is assigned the value from either __get_free_pages() or mca_bootmem(). The problem is that __get_free_pages returns an unsigned long and mca_bootmem, via alloc_bootmem(), returns a void *. format_mca_init_stack takes the void *, and it's also used with __pa(), but that casts it to long anyway.

This results in the following build warning:

arch/ia64/kernel/mca.c:1898: warning: assignment makes pointer from integer without a cast

Cast the return of __get_free_pages to a void * to avoid the warning.

---
 arch/ia64/kernel/mca.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

@@ -1859,7 +1859,8 @@ ia64_mca_cpu_init(void *cpu_data)
 			data = mca_bootmem();
 			first_time = 0;
 		} else
-			data = __get_free_pages(GFP_KERNEL, get_order(sz));
+			data = (void *)__get_free_pages(GFP_KERNEL,
+							get_order(sz));
 		if (!data)
 			panic("Could not allocate MCA memory for cpu %d\n",
 					cpu);

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Linus Torvalds <[email protected]>

commit 95042f9eb78a8d9a17455e2ef263f2f310ecef15 upstream.

Commit 53a7706d5ed8 ("mlock: do not hold mmap_sem for extended periods of time") changed mlock() to care about the exact number of pages that __get_user_pages() had brought it. Before, it would only care about errors.

And that doesn't work, because we also handled one page specially in __mlock_vma_pages_range(), namely the stack guard page. So when that case was handled, the number of pages that the function returned was off by one. In particular, it could be zero, and then the caller would end up not making any progress at all.

Rather than try to fix up that off-by-one error for the mlock case specially, this just moves the logic to handle the stack guard page into__get_user_pages() itself, thus making all the counts come out right automatically.

---
 mm/memory.c |   26 ++++++++++++++++++--------
 mm/mlock.c  |   13 -------------
 2 files changed, 18 insertions(+), 21 deletions(-)

@@ -1410,6 +1410,13 @@ no_page_table:
 	return page;
 }
 
+static inline int stack_guard_page(struct vm_area_struct *vma, unsigned long addr)
+{
+	return (vma->vm_flags & VM_GROWSDOWN) &&
+		(vma->vm_start == addr) &&
+		!vma_stack_continue(vma->vm_prev, addr);
+}
+
 int __get_user_pages(struct task_struct *tsk, struct mm_struct *mm,
 		     unsigned long start, int nr_pages, unsigned int gup_flags,
 		     struct page **pages, struct vm_area_struct **vmas,

@@ -1439,7 +1446,6 @@ int __get_user_pages(struct task_struct
 		vma = find_extend_vma(mm, start);
 		if (!vma && in_gate_area(tsk, start)) {
 			unsigned long pg = start & PAGE_MASK;
-			struct vm_area_struct *gate_vma = get_gate_vma(tsk);
 			pgd_t *pgd;
 			pud_t *pud;
 			pmd_t *pmd;

@@ -1464,10 +1470,11 @@ int __get_user_pages(struct task_struct
 				pte_unmap(pte);
 				return i ? : -EFAULT;
 			}
+			vma = get_gate_vma(tsk);
 			if (pages) {
 				struct page *page;
 
-				page = vm_normal_page(gate_vma, start, *pte);
+				page = vm_normal_page(vma, start, *pte);
 				if (!page) {
 					if (!(gup_flags & FOLL_DUMP) &&
 					     is_zero_pfn(pte_pfn(*pte)))

@@ -1481,12 +1488,7 @@ int __get_user_pages(struct task_struct
 				get_page(page);
 			}
 			pte_unmap(pte);
-			if (vmas)
-				vmas[i] = gate_vma;
-			i++;
-			start += PAGE_SIZE;
-			nr_pages--;
-			continue;
+			goto next_page;
 		}
 
 		if (!vma ||

@@ -1500,6 +1502,13 @@ int __get_user_pages(struct task_struct
 			continue;
 		}
 
+		/*
+		 * If we don't actually want the page itself,
+		 * and it's the stack guard page, just skip it.
+		 */
+		if (!pages && stack_guard_page(vma, start))
+			goto next_page;
+
 		do {
 			struct page *page;
 			unsigned int foll_flags = gup_flags;

@@ -1569,6 +1578,7 @@ int __get_user_pages(struct task_struct
 				flush_anon_page(vma, page, start);
 				flush_dcache_page(page);
 			}
+next_page:
 			if (vmas)
 				vmas[i] = vma;
 			i++;

@@ -135,13 +135,6 @@ void munlock_vma_page(struct page *page)
 	}
 }
 
-static inline int stack_guard_page(struct vm_area_struct *vma, unsigned long addr)
-{
-	return (vma->vm_flags & VM_GROWSDOWN) &&
-		(vma->vm_start == addr) &&
-		!vma_stack_continue(vma->vm_prev, addr);
-}
-
 /**
  * __mlock_vma_pages_range() -  mlock a range of pages in the vma.
  * @vma:   target vma

@@ -188,12 +181,6 @@ static long __mlock_vma_pages_range(stru
 	if (vma->vm_flags & VM_LOCKED)
 		gup_flags |= FOLL_MLOCK;
 
-	/* We don't try to access the guard page of a stack vma */
-	if (stack_guard_page(vma, start)) {
-		addr += PAGE_SIZE;
-		nr_pages--;
-	}
-
 	return __get_user_pages(current, mm, addr, nr_pages, gup_flags,
 				NULL, NULL, nonblocking);
 }

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Artem Bityutskiy <[email protected]>

commit c88ac00c5af70c2a0741da14b22cdcf8507ddd92 upstream.

This patch fixes UBIFS assertion warnings like:

UBIFS assert failed in ubifs_leb_unmap at 135 (pid 29365)
Pid: 29365, comm: integck Tainted: G I 2.6.37-ubi-2.6+ #34

Call Trace:
 [<ffffffffa047c663>] ubifs_lpt_init+0x95e/0x9ee [ubifs]
 [<ffffffffa04623a7>] ubifs_remount_fs+0x2c7/0x762 [ubifs]
 [<ffffffff810f066e>] do_remount_sb+0xb6/0x101
 [<ffffffff81106ff4>] ? do_mount+0x191/0x78e
 [<ffffffff811070bb>] do_mount+0x258/0x78e
 [<ffffffff810da1e8>] ? alloc_pages_current+0xa2/0xc5
 [<ffffffff81107674>] sys_mount+0x83/0xbd
 [<ffffffff81009a12>] system_call_fastpath+0x16/0x1b

They happen when we re-mount from R/O mode to R/W mode. While re-mounting, we write to the media, but we still have the c->ro_mount flag set. The fix is very simple - just clear the flag before starting re-mounting R/W.

These warnings are caused by the following commit:
2ef13294d29bcfb306e0d360f1b97f37b647b0c0

For -stable guys: this bug was introduced in 2.6.38, this is materieal for 2.6.38-stable.

---
 fs/ubifs/super.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

@@ -1543,6 +1543,7 @@ static int ubifs_remount_rw(struct ubifs
 	mutex_lock(&c->umount_mutex);
 	dbg_save_space_info(c);
 	c->remounting_rw = 1;
+	c->ro_mount = 0;
 	c->always_chk_crc = 1;
 
 	err = check_free_space(c);

@@ -1648,7 +1649,6 @@ static int ubifs_remount_rw(struct ubifs
 	}
 
 	dbg_gen("re-mounted read-write");
-	c->ro_mount = 0;
 	c->remounting_rw = 0;
 	c->always_chk_crc = 0;
 	err = dbg_check_space_info(c);

@@ -1656,6 +1656,7 @@ static int ubifs_remount_rw(struct ubifs
 	return err;
 
 out:
+	c->ro_mount = 1;
 	vfree(c->orph_buf);
 	c->orph_buf = NULL;
 	if (c->bgt) {

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Peter Zijlstra <[email protected]>

commit ab711fe08297de1485fff0a366e6db8828cafd6a upstream.

Jiri reported:

 |
 | - once an event is created by sys_perf_event_open, task context
 |   is created and it stays even if the event is closed, until the
 |   task is finished ... thats what I see in code and I assume it's
 |   correct
 |
 | - when the task opens event, perf_sched_events jump label is
 |   incremented and following callbacks are started from scheduler
 |
 |         __perf_event_task_sched_in
 |         __perf_event_task_sched_out
 |
 |   These callback *in/out set/unset cpuctx->task_ctx value to the
 |   task context.
 |
 | - close is called on event on CPU 0:
 |         - the task is scheduled on CPU 0
 |         - __perf_event_task_sched_in is called
 |         - cpuctx->task_ctx is set
 |         - perf_sched_events jump label is decremented and == 0
 |         - __perf_event_task_sched_out is not called
 |         - cpuctx->task_ctx on CPU 0 stays set
 |
 | - exit is called on CPU 1:
 |         - the task is scheduled on CPU 1
 |         - perf_event_exit_task is called
 |         - task_ctx_sched_out unsets cpuctx->task_ctx on CPU 1
 |         - put_ctx destroys the context
 |
 | - another call of perf_rotate_context on CPU 0 will use invalid
 |   task_ctx pointer, and eventualy panic.
 |

Cure this the simplest possibly way by partially reverting the jump_label optimization for the sched_out case.

LKML-Reference: <1301520405.4859.213.camel@twins>

---
 include/linux/perf_event.h |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -1052,7 +1052,7 @@ void perf_event_task_sched_out(struct ta
 {
 	perf_sw_event(PERF_COUNT_SW_CONTEXT_SWITCHES, 1, 1, NULL, 0);
 
-	COND_STMT(&perf_task_events, __perf_event_task_sched_out(task, next));
+	__perf_event_task_sched_out(task, next);
 }
 
 extern void perf_event_mmap(struct vm_area_struct *vma);

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Linus L�ssing <[email protected]>

[ Upstream commit a7bff75b087e7a355838a32efe61707cfa73c194 ]

The ipv6_dev_get_saddr() is currently called with an uninitialized destination address. Although in tests it usually seemed to nevertheless always fetch the right source address, there seems to be a possible race condition.

Therefore this commit changes this, first setting the destination address and only after that fetching the source address.

---
 net/bridge/br_multicast.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -445,9 +445,9 @@ static struct sk_buff *br_ip6_multicast_
 	ip6h->payload_len = htons(8 + sizeof(*mldq));
 	ip6h->nexthdr = IPPROTO_HOPOPTS;
 	ip6h->hop_limit = 1;
+	ipv6_addr_set(&ip6h->daddr, htonl(0xff020000), 0, 0, htonl(1));
 	ipv6_dev_get_saddr(dev_net(br->dev), br->dev, &ip6h->daddr, 0,
 			   &ip6h->saddr);
-	ipv6_addr_set(&ip6h->daddr, htonl(0xff020000), 0, 0, htonl(1));
 	ipv6_eth_mc_map(&ip6h->daddr, eth->h_dest);
 
 	hopopt = (u8 *)(ip6h + 1);

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Eric Dumazet <[email protected]>

[ Upstream commit e2666f84958adb3a034b98e99699b55705117e01 ]

Daniel J Blueman reported a lockdep splat in trie_firstleaf(), caused by RTNL being not locked before a call to fib_table_flush()

---
 net/ipv4/fib_frontend.c |    2 ++
 1 file changed, 2 insertions(+)

@@ -1041,6 +1041,7 @@ static void ip_fib_net_exit(struct net *
 	fib4_rules_exit(net);
 #endif
 
+	rtnl_lock();
 	for (i = 0; i < FIB_TABLE_HASHSZ; i++) {
 		struct fib_table *tb;
 		struct hlist_head *head;

@@ -1053,6 +1054,7 @@ static void ip_fib_net_exit(struct net *
 			fib_free_table(tb);
 		}
 	}
+	rtnl_unlock();
 	kfree(net->ipv4.fib_table_hash);
 }
 

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Alex Dubov <[email protected]>

[ Upstream commit 4363c2fddb1399b728ef21ee8101c148a311ea45 ]

As specified by errata eTSEC49 of MPC8548 and errata eTSEC12 of MPC83xx, older revisions of gianfar controllers will be unable to calculate a TCP/UDP packet checksum for some alignments of the appropriate FCB. This patch checks for FCB alignment on such controllers and falls back to software checksumming if the alignment is known to be bad.

---
 drivers/net/gianfar.c |   16 ++++++++++++++--
 drivers/net/gianfar.h |    1 +
 2 files changed, 15 insertions(+), 2 deletions(-)

@@ -950,6 +950,11 @@ static void gfar_detect_errata(struct gf
 			(pvr == 0x80861010 && (mod & 0xfff9) == 0x80c0))
 		priv->errata |= GFAR_ERRATA_A002;
 
+	/* MPC8313 Rev < 2.0, MPC8548 rev 2.0 */
+	if ((pvr == 0x80850010 && mod == 0x80b0 && rev < 0x0020) ||
+			(pvr == 0x80210020 && mod == 0x8030 && rev == 0x0020))
+		priv->errata |= GFAR_ERRATA_12;
+
 	if (priv->errata)
 		dev_info(dev, "enabled errata workarounds, flags: 0x%x\n",
 			 priv->errata);

@@ -2156,8 +2161,15 @@ static int gfar_start_xmit(struct sk_buf
 	/* Set up checksumming */
 	if (CHECKSUM_PARTIAL == skb->ip_summed) {
 		fcb = gfar_add_fcb(skb);
-		lstatus |= BD_LFLAG(TXBD_TOE);
-		gfar_tx_checksum(skb, fcb);
+		/* as specified by errata */
+		if (unlikely(gfar_has_errata(priv, GFAR_ERRATA_12)
+			     && ((unsigned long)fcb % 0x20) > 0x18)) {
+			__skb_pull(skb, GMAC_FCB_LEN);
+			skb_checksum_help(skb);
+		} else {
+			lstatus |= BD_LFLAG(TXBD_TOE);
+			gfar_tx_checksum(skb, fcb);
+		}
 	}
 
 	if (vlan_tx_tag_present(skb)) {

@@ -1039,6 +1039,7 @@ enum gfar_errata {
 	GFAR_ERRATA_74		= 0x01,
 	GFAR_ERRATA_76		= 0x02,
 	GFAR_ERRATA_A002	= 0x04,
+	GFAR_ERRATA_12		= 0x08, /* a.k.a errata eTSEC49 */
 };
 
 /* Struct stolen almost completely (and shamelessly) from the FCC enet source

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: James Chapman <[email protected]>

[ Upstream commit 8aa525a9340da4227797a06221ca08399006635f ]

A struct used in the l2tp_eth driver for registering network namespace ops was incorrectly marked as __net_initdata, leading to oops when module unloaded.

BUG: unable to handle kernel paging request at ffffffffa00ec098
IP: [<ffffffff8123dbd8>] ops_exit_list+0x7/0x4b
PGD 142d067 PUD 1431063 PMD 195da8067 PTE 0
Oops: 0000 [#1] SMP
last sysfs file: /sys/module/l2tp_eth/refcnt
Call Trace:
 [<ffffffff8123dc94>] ? unregister_pernet_operations+0x32/0x93
 [<ffffffff8123dd20>] ? unregister_pernet_device+0x2b/0x38
 [<ffffffff81068b6e>] ? sys_delete_module+0x1b8/0x222
 [<ffffffff810c7300>] ? do_munmap+0x254/0x318
 [<ffffffff812c64e5>] ? page_fault+0x25/0x30
 [<ffffffff812c6952>] ? system_call_fastpath+0x16/0x1b

---
 net/l2tp/l2tp_eth.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -283,7 +283,7 @@ static __net_init int l2tp_eth_init_net(
 	return 0;
 }
 
-static __net_initdata struct pernet_operations l2tp_eth_net_ops = {
+static struct pernet_operations l2tp_eth_net_ops = {
 	.init = l2tp_eth_init_net,
 	.id   = &l2tp_eth_net_id,
 	.size = sizeof(struct l2tp_eth_net),

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Eric W. Biederman <[email protected]>

[ Upstream commit 9d2a8fa96a44ba242de3a6f56acaef7a40a97b97 ]

When I was fixing issues with unregisgtering tables under /proc/sys/net/ipv6/neigh by adding a mount point it appears I missed a critical ordering issue, in the ipv6 initialization. I had not realized that ipv6_sysctl_register is called at the very end of the ipv6 initialization and in particular after we call neigh_sysctl_register from ndisc_init.

"neigh" needs to be initialized in ipv6_static_sysctl_register which is the first ipv6 table to initialized, and definitely before ndisc_init. This removes the weirdness of duplicate tables while still providing a "neigh" mount point which prevents races in sysctl unregistering.

This was initially reported at https://bugzilla.kernel.org/show_bug.cgi?id=31232

---
 net/ipv6/sysctl_net_ipv6.c |   18 +++++++++++-------
 1 file changed, 11 insertions(+), 7 deletions(-)

@@ -17,6 +17,16 @@
 
 static struct ctl_table empty[1];
 
+static ctl_table ipv6_static_skeleton[] = {
+	{
+		.procname	= "neigh",
+		.maxlen		= 0,
+		.mode		= 0555,
+		.child		= empty,
+	},
+	{ }
+};
+
 static ctl_table ipv6_table_template[] = {
 	{
 		.procname	= "route",

@@ -37,12 +47,6 @@ static ctl_table ipv6_table_template[] =
 		.mode		= 0644,
 		.proc_handler	= proc_dointvec
 	},
-	{
-		.procname	= "neigh",
-		.maxlen		= 0,
-		.mode		= 0555,
-		.child		= empty,
-	},
 	{ }
 };
 

@@ -160,7 +164,7 @@ static struct ctl_table_header *ip6_base
 
 int ipv6_static_sysctl_register(void)
 {
-	ip6_base = register_sysctl_paths(net_ipv6_ctl_path, empty);
+	ip6_base = register_sysctl_paths(net_ipv6_ctl_path, ipv6_static_skeleton);
 	if (ip6_base == NULL)
 		return -ENOMEM;
 	return 0;

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Dan Siemon <[email protected]>

[ Upstream commit 4a2b9c3756077c05dd8666e458a751d2248b61b6 ]

ECN support incorrectly maps ECN BESTEFFORT packets to TC_PRIO_FILLER
(1) instead of TC_PRIO_BESTEFFORT (0)

This means ECN enabled flows are placed in pfifo_fast/prio low priority band, giving ECN enabled flows [ECT(0) and CE codepoints] higher drop probabilities.

This is rather unfortunate, given we would like ECN being more widely used.

Ref : http://www.coverfire.com/archives/2011/03/13/pfifo_fast-and-ecn/

---
 net/ipv4/route.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -171,7 +171,7 @@ static struct dst_ops ipv4_dst_ops = {
 
 const __u8 ip_tos2prio[16] = {
 	TC_PRIO_BESTEFFORT,
-	ECN_OR_COST(FILLER),
+	ECN_OR_COST(BESTEFFORT),
 	TC_PRIO_BESTEFFORT,
 	ECN_OR_COST(BESTEFFORT),
 	TC_PRIO_BULK,

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Ulrich Weber <[email protected]>

[ Upstream commit ae07b0b221b6ab2edf9e3abd518aec6cd3f1ba66 ]

otherwise we loop forever if a PPPoE socket was set to PPPOX_ZOMBIE state by a PADT message when the ethernet device is going down afterwards.

---
 drivers/net/pppoe.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -317,7 +317,7 @@ static void pppoe_flush_dev(struct net_d
 			lock_sock(sk);
 
 			if (po->pppoe_dev == dev &&
-			    sk->sk_state & (PPPOX_CONNECTED | PPPOX_BOUND)) {
+			    sk->sk_state & (PPPOX_CONNECTED | PPPOX_BOUND | PPPOX_ZOMBIE)) {
 				pppox_unbind_sock(sk);
 				sk->sk_state = PPPOX_ZOMBIE;
 				sk->sk_state_change(sk);

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: David S. Miller <[email protected]>

[ Upstream commit a84b50ceb7d640437d0dc28a2bef0d0de054de89 ]

Like DCCP and other similar pieces of code, there are mechanisms here to try allocating smaller hash tables if the allocation fails. So pass in __GFP_NOWARN like the others do instead of emitting a scary message.

---
 net/sctp/protocol.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

@@ -1204,7 +1204,7 @@ SCTP_STATIC __init int sctp_init(void)
 		if ((sctp_assoc_hashsize > (64 * 1024)) && order > 0)
 			continue;
 		sctp_assoc_hashtable = (struct sctp_hashbucket *)
-					__get_free_pages(GFP_ATOMIC, order);
+			__get_free_pages(GFP_ATOMIC|__GFP_NOWARN, order);
 	} while (!sctp_assoc_hashtable && --order > 0);
 	if (!sctp_assoc_hashtable) {
 		pr_err("Failed association hash alloc\n");

@@ -1237,7 +1237,7 @@ SCTP_STATIC __init int sctp_init(void)
 		if ((sctp_port_hashsize > (64 * 1024)) && order > 0)
 			continue;
 		sctp_port_hashtable = (struct sctp_bind_hashbucket *)
-					__get_free_pages(GFP_ATOMIC, order);
+			__get_free_pages(GFP_ATOMIC|__GFP_NOWARN, order);
 	} while (!sctp_port_hashtable && --order > 0);
 	if (!sctp_port_hashtable) {
 		pr_err("Failed bind hash alloc\n");

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Yuchung Cheng <[email protected]>

[ Upstream commit 67d4120a1793138bc9f4a6eb61d0fc5298ed97e0 ]

In the current undo logic, cwnd is moderated after it was restored to the value prior entering fast-recovery. It was moderated first in tcp_try_undo_recovery then again in tcp_complete_cwr.

Since the undo indicates recovery was false, these moderations are not necessary. If the undo is triggered when most of the outstanding data have been acknowledged, the (restored) cwnd is falsely pulled down to a small value.

This patch removes these cwnd moderations if cwnd is undone
a) during fast-recovery
b) by receiving DSACKs past fast-recovery

---
 net/ipv4/tcp_input.c |   12 +++++++-----
 1 file changed, 7 insertions(+), 5 deletions(-)

@@ -2659,7 +2659,7 @@ static void DBGUNDO(struct sock *sk, con
 #define DBGUNDO(x...) do { } while (0)
 #endif
 
-static void tcp_undo_cwr(struct sock *sk, const int undo)
+static void tcp_undo_cwr(struct sock *sk, const int undo_ssthresh)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 

@@ -2671,14 +2671,13 @@ static void tcp_undo_cwr(struct sock *sk
 		else
 			tp->snd_cwnd = max(tp->snd_cwnd, tp->snd_ssthresh << 1);
 
-		if (undo && tp->prior_ssthresh > tp->snd_ssthresh) {
+		if (undo_ssthresh && tp->prior_ssthresh > tp->snd_ssthresh) {
 			tp->snd_ssthresh = tp->prior_ssthresh;
 			TCP_ECN_withdraw_cwr(tp);
 		}
 	} else {
 		tp->snd_cwnd = max(tp->snd_cwnd, tp->snd_ssthresh);
 	}
-	tcp_moderate_cwnd(tp);
 	tp->snd_cwnd_stamp = tcp_time_stamp;
 }
 

@@ -2822,8 +2821,11 @@ static int tcp_try_undo_loss(struct sock
 static inline void tcp_complete_cwr(struct sock *sk)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
-	tp->snd_cwnd = min(tp->snd_cwnd, tp->snd_ssthresh);
-	tp->snd_cwnd_stamp = tcp_time_stamp;
+	/* Do not moderate cwnd if it's already undone in cwr or recovery */
+	if (tp->undo_marker && tp->snd_cwnd > tp->snd_ssthresh) {
+		tp->snd_cwnd = tp->snd_ssthresh;
+		tp->snd_cwnd_stamp = tcp_time_stamp;
+	}
 	tcp_ca_event(sk, CA_EVENT_COMPLETE_CWR);
 }
 

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Steffen Klassert <[email protected]>

[ Upstream commit fbd5060875d25f7764fd1c3d35b83a8ed1d88d7b ]

We return a destination entry without refcount if a socket policy is found in xfrm_lookup. This triggers a warning on a negative refcount when freeeing this dst entry. So take a refcount in this case to fix it.

This refcount was forgotten when xfrm changed to cache bundles instead of policies for outgoing flows.

---
 net/xfrm/xfrm_policy.c |    2 ++
 1 file changed, 2 insertions(+)

@@ -1778,6 +1778,8 @@ restart:
 				goto no_transform;
 			}
 
+			dst_hold(&xdst->u.dst);
+
 			spin_lock_bh(&xfrm_policy_sk_bundle_lock);
 			xdst->u.dst.next = xfrm_policy_sk_bundles;
 			xfrm_policy_sk_bundles = &xdst->u.dst;

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Eric Dumazet <[email protected]>

[ Upstream commit d870bfb9d366c5d466c0f5419a4ec95a3f71ea8a ]

Commit c95b819ad7 (gre: Use needed_headroom) made gre use needed_headroom instead of hard_header_len

This uncover a bug in vlan code.

We should make sure vlan devices take into account their real_dev->needed_headroom or we risk a crash in ipgre_header(), because we dont have enough room to push IP header in skb.

---
 net/8021q/vlan_dev.c |    1 +
 1 file changed, 1 insertion(+)

@@ -707,6 +707,7 @@ static int vlan_dev_init(struct net_devi
 	dev->fcoe_ddp_xid = real_dev->fcoe_ddp_xid;
 #endif
 
+	dev->needed_headroom = real_dev->needed_headroom;
 	if (real_dev->features & NETIF_F_HW_VLAN_TX) {
 		dev->header_ops      = real_dev->header_ops;
 		dev->hard_header_len = real_dev->hard_header_len;

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Herbert Xu <[email protected]>

[ Upstream commit 6b1e960fdbd75dcd9bcc3ba5ff8898ff1ad30b6e ]

Whenever we enter the IP stack proper from bridge netfilter we need to ensure that the skb is in a form the IP stack expects it to be in.

The entry point on NF_FORWARD did not meet the requirements of the IP stack, therefore leading to potential crashes/panics.

This patch fixes the problem.

---
 net/bridge/br_netfilter.c |    3 +++
 1 file changed, 3 insertions(+)

@@ -741,6 +741,9 @@ static unsigned int br_nf_forward_ip(uns
 		nf_bridge->mask |= BRNF_PKT_TYPE;
 	}
 
+	if (br_parse_ip_options(skb))
+		return NF_DROP;
+
 	/* The physdev module checks on this */
 	nf_bridge->mask |= BRNF_BRIDGED;
 	nf_bridge->physoutdev = skb->dev;

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Ben Hutchings <[email protected]>

[ Upstream commit 35043c428f1fcb92feb5792f5878a8852ee00771 ]

gas used to accept (and ignore?) .size directives which referred to undefined symbols, as this does. In binutils 2.21 these are treated as errors.

---
 arch/sparc/kernel/una_asm_64.S |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -127,7 +127,7 @@ do_int_load:
 	wr	%o5, 0x0, %asi
 	retl
 	 mov	0, %o0
-	.size	__do_int_load, .-__do_int_load
+	.size	do_int_load, .-do_int_load
 
 	.section	__ex_table,"a"
 	.word		4b, __retl_efault

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: David S. Miller <[email protected]>

[ Upstream commit c816be7b5f24585baa9eba1f2413935f771d6ad6 ]

When we try to handle vmalloc faults, we can take a code path which uses "code" before we actually set it.

Amusingly gcc-3.3 notices this yet gcc-4.x does not.

---
 arch/sparc/mm/fault_32.c |    3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

@@ -240,11 +240,10 @@ asmlinkage void do_sparc_fault(struct pt
 	 * only copy the information from the master page table,
 	 * nothing more.
 	 */
+	code = SEGV_MAPERR;
 	if (!ARCH_SUN4C && address >= TASK_SIZE)
 		goto vmalloc_fault;
 
-	code = SEGV_MAPERR;
-
 	/*
 	 * If we're in an interrupt or have no user
 	 * context, we must not take the fault..

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Tkhai Kirill <[email protected]>

[ Upstream commit 47c7c97a93a5b8f719093dbf83555090b3b8228b ]

We have to pass task_struct of previous process to function schedule_tail(). Currently in ret_from_fork previous thread_info is passed:

switch_to: mov %g6, %g3 /* previous thread_info in g6 */

ret_from_fork: call schedule_tail
mov %g3, %o0 /* previous thread_info is passed */

void schedule_tail(struct task_struct *prev);

---
 arch/sparc/kernel/entry.S |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -1283,7 +1283,7 @@ linux_syscall_trace:
 	.globl	ret_from_fork
 ret_from_fork:
 	call	schedule_tail
-	 mov	%g3, %o0
+	 ld	[%g3 + TI_TASK], %o0
 	b	ret_sys_call
 	 ld	[%sp + STACKFRAME_SZ + PT_I0], %o0
 

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: David S. Miller <[email protected]>

[ Upstream commit c6fee0810df4e0f4cf9c4834d2569ca01c02cffc ]

Most of the warnings emitted (we fail arch/sparc file builds with -Werror) were legitimate but harmless, however one case (n2_pcr_write) was a genuine bug.

Based almost entirely upon a patch by Sam Ravnborg.

---
 arch/sparc/kernel/iommu.c      |    3 ---
 arch/sparc/kernel/ldc.c        |   28 ++++++++++++++++++----------
 arch/sparc/kernel/pci.c        |    1 +
 arch/sparc/kernel/pci_common.c |   11 +++++++----
 arch/sparc/kernel/pci_fire.c   |    2 --
 arch/sparc/kernel/pci_schizo.c |    4 +---
 arch/sparc/kernel/pci_sun4v.c  |    3 +--
 arch/sparc/kernel/pcr.c        |    2 +-
 arch/sparc/kernel/ptrace_64.c  |    3 ++-
 arch/sparc/kernel/smp_64.c     |   11 ++++-------
 arch/sparc/kernel/traps_64.c   |    3 +--
 11 files changed, 36 insertions(+), 35 deletions(-)

@@ -333,13 +333,10 @@ static void dma_4u_free_coherent(struct
 				 void *cpu, dma_addr_t dvma)
 {
 	struct iommu *iommu;
-	iopte_t *iopte;
 	unsigned long flags, order, npages;
 
 	npages = IO_PAGE_ALIGN(size) >> IO_PAGE_SHIFT;
 	iommu = dev->archdata.iommu;
-	iopte = iommu->page_table +
-		((dvma - iommu->page_table_map_base) >> IO_PAGE_SHIFT);
 
 	spin_lock_irqsave(&iommu->lock, flags);
 

@@ -790,16 +790,20 @@ static void send_events(struct ldc_chann
 static irqreturn_t ldc_rx(int irq, void *dev_id)
 {
 	struct ldc_channel *lp = dev_id;
-	unsigned long orig_state, hv_err, flags;
+	unsigned long orig_state, flags;
 	unsigned int event_mask;
 
 	spin_lock_irqsave(&lp->lock, flags);
 
 	orig_state = lp->chan_state;
-	hv_err = sun4v_ldc_rx_get_state(lp->id,
-					&lp->rx_head,
-					&lp->rx_tail,
-					&lp->chan_state);
+
+	/* We should probably check for hypervisor errors here and
+	 * reset the LDC channel if we get one.
+	 */
+	sun4v_ldc_rx_get_state(lp->id,
+			       &lp->rx_head,
+			       &lp->rx_tail,
+			       &lp->chan_state);
 
 	ldcdbg(RX, "RX state[0x%02lx:0x%02lx] head[0x%04lx] tail[0x%04lx]\n",
 	       orig_state, lp->chan_state, lp->rx_head, lp->rx_tail);

@@ -904,16 +908,20 @@ out:
 static irqreturn_t ldc_tx(int irq, void *dev_id)
 {
 	struct ldc_channel *lp = dev_id;
-	unsigned long flags, hv_err, orig_state;
+	unsigned long flags, orig_state;
 	unsigned int event_mask = 0;
 
 	spin_lock_irqsave(&lp->lock, flags);
 
 	orig_state = lp->chan_state;
-	hv_err = sun4v_ldc_tx_get_state(lp->id,
-					&lp->tx_head,
-					&lp->tx_tail,
-					&lp->chan_state);
+
+	/* We should probably check for hypervisor errors here and
+	 * reset the LDC channel if we get one.
+	 */
+	sun4v_ldc_tx_get_state(lp->id,
+			       &lp->tx_head,
+			       &lp->tx_tail,
+			       &lp->chan_state);
 
 	ldcdbg(TX, " TX state[0x%02lx:0x%02lx] head[0x%04lx] tail[0x%04lx]\n",
 	       orig_state, lp->chan_state, lp->tx_head, lp->tx_tail);

@@ -675,6 +675,7 @@ static void __devinit pci_bus_register_o
 		 * humanoid.
 		 */
 		err = sysfs_create_file(&dev->dev.kobj, &dev_attr_obppath.attr);
+		(void) err;
 	}
 	list_for_each_entry(child_bus, &bus->children, node)
 		pci_bus_register_of_sysfs(child_bus);

@@ -295,14 +295,17 @@ static int sun4v_write_pci_cfg(struct pc
 	unsigned int bus = bus_dev->number;
 	unsigned int device = PCI_SLOT(devfn);
 	unsigned int func = PCI_FUNC(devfn);
-	unsigned long ret;
 
 	if (config_out_of_range(pbm, bus, devfn, where)) {
 		/* Do nothing. */
 	} else {
-		ret = pci_sun4v_config_put(devhandle,
-				HV_PCI_DEVICE_BUILD(bus, device, func),
-				where, size, value);
+		/* We don't check for hypervisor errors here, but perhaps
+		 * we should and influence our return value depending upon
+		 * what kind of error is thrown.
+		 */
+		pci_sun4v_config_put(devhandle,
+				     HV_PCI_DEVICE_BUILD(bus, device, func),
+				     where, size, value);
 	}
 	return PCIBIOS_SUCCESSFUL;
 }

@@ -214,11 +214,9 @@ static int pci_fire_msi_setup(struct pci
 
 static int pci_fire_msi_teardown(struct pci_pbm_info *pbm, unsigned long msi)
 {
-	unsigned long msiqid;
 	u64 val;
 
 	val = upa_readq(pbm->pbm_regs + MSI_MAP(msi));
-	msiqid = (val & MSI_MAP_EQNUM);
 
 	val &= ~MSI_MAP_VALID;
 

@@ -1313,7 +1313,7 @@ static int __devinit schizo_pbm_init(str
 	const struct linux_prom64_registers *regs;
 	struct device_node *dp = op->dev.of_node;
 	const char *chipset_name;
-	int is_pbm_a, err;
+	int err;
 
 	switch (chip_type) {
 	case PBM_CHIP_TYPE_TOMATILLO:

@@ -1343,8 +1343,6 @@ static int __devinit schizo_pbm_init(str
 	 */
 	regs = of_get_property(dp, "reg", NULL);
 
-	is_pbm_a = ((regs[0].phys_addr & 0x00700000) == 0x00600000);
-
 	pbm->next = pci_pbm_root;
 	pci_pbm_root = pbm;
 

@@ -580,7 +580,7 @@ static int __devinit pci_sun4v_iommu_ini
 {
 	static const u32 vdma_default[] = { 0x80000000, 0x80000000 };
 	struct iommu *iommu = pbm->iommu;
-	unsigned long num_tsb_entries, sz, tsbsize;
+	unsigned long num_tsb_entries, sz;
 	u32 dma_mask, dma_offset;
 	const u32 *vdma;
 

@@ -596,7 +596,6 @@ static int __devinit pci_sun4v_iommu_ini
 
 	dma_mask = (roundup_pow_of_two(vdma[1]) - 1UL);
 	num_tsb_entries = vdma[1] / IO_PAGE_SIZE;
-	tsbsize = num_tsb_entries * sizeof(iopte_t);
 
 	dma_offset = vdma[0];
 

@@ -81,7 +81,7 @@ static void n2_pcr_write(u64 val)
 	unsigned long ret;
 
 	ret = sun4v_niagara2_setperf(HV_N2_PERF_SPARC_CTL, val);
-	if (val != HV_EOK)
+	if (ret != HV_EOK)
 		write_pcr(val);
 }
 

@@ -1086,6 +1086,7 @@ asmlinkage int syscall_trace_enter(struc
 
 asmlinkage void syscall_trace_leave(struct pt_regs *regs)
 {
+#ifdef CONFIG_AUDITSYSCALL
 	if (unlikely(current->audit_context)) {
 		unsigned long tstate = regs->tstate;
 		int result = AUDITSC_SUCCESS;

@@ -1095,7 +1096,7 @@ asmlinkage void syscall_trace_leave(stru
 
 		audit_syscall_exit(result, regs->u_regs[UREG_I0]);
 	}
-
+#endif
 	if (unlikely(test_thread_flag(TIF_SYSCALL_TRACEPOINT)))
 		trace_sys_exit(regs, regs->u_regs[UREG_G1]);
 

@@ -189,7 +189,7 @@ static inline long get_delta (long *rt,
 void smp_synchronize_tick_client(void)
 {
 	long i, delta, adj, adjust_latency = 0, done = 0;
-	unsigned long flags, rt, master_time_stamp, bound;
+	unsigned long flags, rt, master_time_stamp;
 #if DEBUG_TICK_SYNC
 	struct {
 		long rt;	/* roundtrip time */

@@ -208,10 +208,8 @@ void smp_synchronize_tick_client(void)
 	{
 		for (i = 0; i < NUM_ROUNDS; i++) {
 			delta = get_delta(&rt, &master_time_stamp);
-			if (delta == 0) {
+			if (delta == 0)
 				done = 1;	/* let's lock on to this... */
-				bound = rt;
-			}
 
 			if (!done) {
 				if (i > 0) {

@@ -933,13 +931,12 @@ void smp_flush_dcache_page_impl(struct p
 void flush_dcache_page_all(struct mm_struct *mm, struct page *page)
 {
 	void *pg_addr;
-	int this_cpu;
 	u64 data0;
 
 	if (tlb_type == hypervisor)
 		return;
 
-	this_cpu = get_cpu();
+	preempt_disable();
 
 #ifdef CONFIG_DEBUG_DCFLUSH
 	atomic_inc(&dcpage_flushes);

@@ -964,7 +961,7 @@ void flush_dcache_page_all(struct mm_str
 	}
 	__local_flush_dcache_page(page);
 
-	put_cpu();
+	preempt_enable();
 }
 
 void __irq_entry smp_new_mmu_context_version_client(int irq, struct pt_regs *regs)

@@ -2152,7 +2152,7 @@ static void user_instruction_dump(unsign
 
 void show_stack(struct task_struct *tsk, unsigned long *_ksp)
 {
-	unsigned long fp, thread_base, ksp;
+	unsigned long fp, ksp;
 	struct thread_info *tp;
 	int count = 0;
 #ifdef CONFIG_FUNCTION_GRAPH_TRACER

@@ -2173,7 +2173,6 @@ void show_stack(struct task_struct *tsk,
 		flushw_all();
 
 	fp = ksp + STACK_BIAS;
-	thread_base = (unsigned long) tp;
 
 	printk("Call Trace:\n");
 	do {

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Darren Hart <[email protected]>

commit 0cd9c6494ee5c19aef085152bc37f3a4e774a9e1 upstream.

The FLAGS_HAS_TIMEOUT flag was not getting set, causing the restart_block to restart futex_wait() without a timeout after a signal.

Commit b41277dc7a18ee332d in 2.6.38 introduced the regression by accidentally removing the the FLAGS_HAS_TIMEOUT assignment from futex_wait() during the setup of the restart block. Restore the originaly behavior.

---
 kernel/futex.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -1886,7 +1886,7 @@ retry:
 	restart->futex.val = val;
 	restart->futex.time = abs_time->tv64;
 	restart->futex.bitset = bitset;
-	restart->futex.flags = flags;
+	restart->futex.flags = flags | FLAGS_HAS_TIMEOUT;
 
 	ret = -ERESTART_RESTARTBLOCK;
 

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Alexey Dobriyan <[email protected]>

commit 33ee3b2e2eb9b4b6c64dcf9ed66e2ac3124e748c upstream.

1. simple_strto*() do not contain overflow checks and crufty, libc way to indicate failure.
2. strict_strto*() also do not have overflow checks but the name and comments pretend they do.
3. Both families have only "long long" and "long" variants, but users want strtou8()
4. Both "simple" and "strict" prefixes are wrong:
Simple doesn't exactly say what's so simple, strict should not exist because conversion should be strict by default.

The solution is to use "k" prefix and add convertors for more types. Enter

	kstrtoull()
	kstrtoll()
	kstrtoul()
	kstrtol()
	kstrtouint()
	kstrtoint()

	kstrtou64()
	kstrtos64()
	kstrtou32()
	kstrtos32()
	kstrtou16()
	kstrtos16()
	kstrtou8()
	kstrtos8()

Include runtime testsuite (somewhat incomplete) as well.

strict_strto*() become deprecated, stubbed to kstrto*() and eventually will be removed altogether.

Use kstrto*() in code today!

Note: on some archs _kstrtoul() and _kstrtol() are left in tree, even if
they'll be unused at runtime. This is temporarily solution, because I don't want to hardcode list of archs where these functions aren't needed. Current solution with sizeof() and __alignof__ at least always works.

---
 drivers/video/via/viafbdev.h |    3 
 include/linux/kernel.h       |   70 +++-
 lib/Kconfig.debug            |    3 
 lib/Makefile                 |    2 
 lib/kstrtox.c                |  227 +++++++++++++
 lib/test-kstrtox.c           |  739 +++++++++++++++++++++++++++++++++++++++++++
 lib/vsprintf.c               |  141 --------
 scripts/checkpatch.pl        |    4 
 8 files changed, 1039 insertions(+), 150 deletions(-)

@@ -94,9 +94,6 @@ extern int viafb_LCD_ON;
 extern int viafb_DVI_ON;
 extern int viafb_hotplug;
 
-extern int strict_strtoul(const char *cp, unsigned int base,
-	unsigned long *res);
-
 u8 viafb_gpio_i2c_read_lvds(struct lvds_setting_information
 	*plvds_setting_info, struct lvds_chip_information
 	*plvds_chip_info, u8 index);

@@ -1236,3 +1236,6 @@ source "samples/Kconfig"
 source "lib/Kconfig.kgdb"
 
 source "lib/Kconfig.kmemcheck"
+
+config TEST_KSTRTOX
+	tristate "Test kstrto*() family of functions at runtime"

@@ -22,6 +22,8 @@ lib-y	+= kobject.o kref.o klist.o
 obj-y += bcd.o div64.o sort.o parser.o halfmd4.o debug_locks.o random32.o \
 	 bust_spinlocks.o hexdump.o kasprintf.o bitmap.o scatterlist.o \
 	 string_helpers.o gcd.o lcm.o list_sort.o uuid.o flex_array.o
+obj-y += kstrtox.o
+obj-$(CONFIG_TEST_KSTRTOX) += test-kstrtox.o
 
 ifeq ($(CONFIG_DEBUG_KOBJECT),y)
 CFLAGS_kobject.o += -DDEBUG

@@ -2809,9 +2809,9 @@ sub process {
 			WARN("consider using a completion\n" . $herecurr);
 
 		}
-# recommend strict_strto* over simple_strto*
+# recommend kstrto* over simple_strto*
 		if ($line =~ /\bsimple_(strto.*?)\s*\(/) {
-			WARN("consider using strict_$1 in preference to simple_$1\n" . $herecurr);
+			WARN("consider using kstrto* in preference to simple_$1\n" . $herecurr);
 		}
 # check for __initcall(), use device_initcall() explicitly please
 		if ($line =~ /^.\s*__initcall\s*\(/) {

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Ben Hutchings <[email protected]>

commit e27e6151b154ff6e5e8162efa291bc60196d29ea upstream.

The conventional format for boolean attributes in sysfs is numeric ("0" or "1" followed by new-line). Any boolean attribute can then be read and written using a generic function. Using the strings "yes [no]", "[yes] no" (read), "yes" and "no" (write) will frustrate this.

---
 mm/huge_memory.c |   24 ++++++++++++++----------
 1 file changed, 14 insertions(+), 10 deletions(-)

@@ -244,24 +244,28 @@ static ssize_t single_flag_show(struct k
 				struct kobj_attribute *attr, char *buf,
 				enum transparent_hugepage_flag flag)
 {
-	if (test_bit(flag, &transparent_hugepage_flags))
-		return sprintf(buf, "[yes] no\n");
-	else
-		return sprintf(buf, "yes [no]\n");
+	return sprintf(buf, "%d\n",
+		       !!test_bit(flag, &transparent_hugepage_flags));
 }
+
 static ssize_t single_flag_store(struct kobject *kobj,
 				 struct kobj_attribute *attr,
 				 const char *buf, size_t count,
 				 enum transparent_hugepage_flag flag)
 {
-	if (!memcmp("yes", buf,
-		    min(sizeof("yes")-1, count))) {
+	unsigned long value;
+	int ret;
+
+	ret = kstrtoul(buf, 10, &value);
+	if (ret < 0)
+		return ret;
+	if (value > 1)
+		return -EINVAL;
+
+	if (value)
 		set_bit(flag, &transparent_hugepage_flags);
-	} else if (!memcmp("no", buf,
-			   min(sizeof("no")-1, count))) {
+	else
 		clear_bit(flag, &transparent_hugepage_flags);
-	} else
-		return -EINVAL;
 
 	return count;
 }

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Bob Liu <[email protected]>

commit b836aec53e2bce71de1d5415313380688c851477 upstream.

On no-mmu arch, there is a memleak during shmem test. The cause of this memleak is ramfs_nommu_expand_for_mapping() added page refcount to 2 which makes iput() can't free that pages.

The simple test file is like this:

int main(void)

  {
	int i;
	key_t k = ftok("/etc", 42);

	for ( i=0; i<100; ++i) {
		int id = shmget(k, 10000, 0644|IPC_CREAT);
		if (id == -1) {
			printf("shmget error\n");
		}
		if(shmctl(id, IPC_RMID, NULL ) == -1) {
			printf("shm  rm error\n");
			return -1;
		}
	}
	printf("run ok...\n");
	return 0;

}

And the result:

root:/> free
total used free shared buffers

  Mem:         60320        17912        42408            0            0
  -/+ buffers:              17912        42408
  root:/> shmem
  run ok...
  root:/> free
               total         used         free       shared      buffers
  Mem:         60320        19096        41224            0            0
  -/+ buffers:              19096        41224
  root:/> shmem
  run ok...
  root:/> free
               total         used         free       shared      buffers
  Mem:         60320        20296        40024            0            0
  -/+ buffers:              20296        40024

...

After this patch the test result is:(no memleak anymore)

root:/> free
total used free shared buffers

  Mem:         60320        16668        43652            0            0
  -/+ buffers:              16668        43652
  root:/> shmem
  run ok...
  root:/> free
               total         used         free       shared      buffers
  Mem:         60320        16668        43652            0            0
  -/+ buffers:              16668        43652

---
 fs/ramfs/file-nommu.c |    1 +
 1 file changed, 1 insertion(+)

@@ -112,6 +112,7 @@ int ramfs_nommu_expand_for_mapping(struc
 		SetPageDirty(page);
 
 		unlock_page(page);
+		put_page(page);
 	}
 
 	return 0;

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: KOSAKI Motohiro <[email protected]>

commit 341aea2bc48bf652777fb015cc2b3dfa9a451817 upstream.

This is an almost-revert of commit 93b43fa ("oom: give the dying task a higher priority").

That commit dramatically improved oom killer logic when a fork-bomb occurs. But I've found that it has nasty corner case. Now cpu cgroup has strange default RT runtime. It's 0! That said, if a process under cpu cgroup promote RT scheduling class, the process never run at all.

If an admin inserts a !RT process into a cpu cgroup by setting rtruntime=0, usually it runs perfectly because a !RT task isn't affected by the rtruntime knob. But if it promotes an RT task via an explicit setscheduler() syscall or an OOM, the task can't run at all. In short, the oom killer doesn't work at all if admins are using cpu cgroup and don't touch the rtruntime knob.

Eventually, kernel may hang up when oom kill occur. I and the original author Luis agreed to disable this logic.

---
 mm/oom_kill.c |   27 ---------------------------
 1 file changed, 27 deletions(-)

@@ -84,24 +84,6 @@ static bool has_intersects_mems_allowed(
 #endif /* CONFIG_NUMA */
 
 /*
- * If this is a system OOM (not a memcg OOM) and the task selected to be
- * killed is not already running at high (RT) priorities, speed up the
- * recovery by boosting the dying task to the lowest FIFO priority.
- * That helps with the recovery and avoids interfering with RT tasks.
- */
-static void boost_dying_task_prio(struct task_struct *p,
-				  struct mem_cgroup *mem)
-{
-	struct sched_param param = { .sched_priority = 1 };
-
-	if (mem)
-		return;
-
-	if (!rt_task(p))
-		sched_setscheduler_nocheck(p, SCHED_FIFO, &param);
-}
-
-/*
  * The process p may have detached its own ->mm while exiting or through
  * use_mm(), but one or more of its subthreads may still have a valid
  * pointer.  Return p, or any of its subthreads with a valid ->mm, with

@@ -452,13 +434,6 @@ static int oom_kill_task(struct task_str
 	set_tsk_thread_flag(p, TIF_MEMDIE);
 	force_sig(SIGKILL, p);
 
-	/*
-	 * We give our sacrificial lamb high priority and access to
-	 * all the memory it needs. That way it should be able to
-	 * exit() and clear out its resources quickly...
-	 */
-	boost_dying_task_prio(p, mem);
-
 	return 0;
 }
 #undef K

@@ -482,7 +457,6 @@ static int oom_kill_process(struct task_
 	 */
 	if (p->flags & PF_EXITING) {
 		set_tsk_thread_flag(p, TIF_MEMDIE);
-		boost_dying_task_prio(p, mem);
 		return 0;
 	}
 

@@ -701,7 +675,6 @@ void out_of_memory(struct zonelist *zone
 	 */
 	if (fatal_signal_pending(current)) {
 		set_thread_flag(TIF_MEMDIE);
-		boost_dying_task_prio(current, NULL);
 		return;
 	}
 

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Randy Dunlap <[email protected]>

commit d00ebeac5f24f290636f7a895dafc124b2930a08 upstream.

Drop Chris Wright from STABLE maintainers. He hasn't done STABLE release work for quite some time.

---
 MAINTAINERS |    1 -
 1 file changed, 1 deletion(-)

@@ -5944,7 +5944,6 @@ F:	arch/alpha/kernel/srm_env.c
 
 STABLE BRANCH
 M:	Greg Kroah-Hartman <[email protected]>
-M:	Chris Wright <[email protected]>
 L:	[email protected]
 S:	Maintained
 

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Artem Bityutskiy <[email protected]>

commit 78530bf7f2559b317c04991b52217c1608d5a58d upstream.

This patch fixes severe UBIFS bug: UBIFS oopses when we 'fsync()' an file on R/O-mounter file-system. We (the UBIFS authors) incorrectly thought that VFS would not propagate 'fsync()' down to the file-system if it is read-only, but this is not the case.

It is easy to exploit this bug using the following simple perl script:

use strict;
use File::Sync qw(fsync sync);

die "File path is not specified" if not defined $ARGV[0]; my $path = $ARGV[0];

open FILE, "<", "$path" or die "Cannot open $path: $!"; fsync(\*FILE) or die "cannot fsync $path: $!"; close FILE or die "Cannot close $path: $!";

Thanks to Reuben Dowle <[email protected]> for reporting about this issue.

---
 fs/ubifs/file.c |    3 +++
 1 file changed, 3 insertions(+)

@@ -1309,6 +1309,9 @@ int ubifs_fsync(struct file *file, int d
 
 	dbg_gen("syncing inode %lu", inode->i_ino);
 
+	if (inode->i_sb->s_flags & MS_RDONLY)
+		return 0;
+
 	/*
 	 * VFS has already synchronized dirty pages for this inode. Synchronize
 	 * the inode unless this is a 'datasync()' call.

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Boris Ostrovsky <[email protected]>

commit b87cf80af3ba4b4c008b4face3c68d604e1715c6 upstream.

Support for Always Running APIC timer (ARAT) was introduced in commit db954b5898dd3ef3ef93f4144158ea8f97deb058. This feature allows us to avoid switching timers from LAPIC to something else (e.g. HPET) and go into timer broadcasts when entering deep C-states.

AMD processors don't provide a CPUID bit for that feature but they also keep APIC timers running in deep C-states (except for cases when the processor is affected by erratum 400). Therefore we should set ARAT feature bit on AMD CPUs.

LKML-Reference: <[email protected]>

---
 arch/x86/kernel/cpu/amd.c |    4 ++++
 1 file changed, 4 insertions(+)

@@ -594,6 +594,10 @@ static void __cpuinit init_amd(struct cp
 		}
 	}
 #endif
+
+	/* As a rule processors have APIC timer running in deep C states */
+	if (c->x86 >= 0xf && !cpu_has_amd_erratum(amd_erratum_400))
+		set_cpu_cap(c, X86_FEATURE_ARAT);
 }
 
 #ifdef CONFIG_X86_32

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Joerg Roedel <[email protected]>

commit 5bbc097d890409d8eff4e3f1d26f11a9d6b7c07e upstream.

This patch disables GartTlbWlk errors on AMD Fam10h CPUs if the BIOS forgets to do is (or is just too old). Letting these errors enabled can cause a sync-flood on the CPU causing a reboot.

The AMD BKDG recommends disabling GART TLB Wlk Error completely.

This patch is the fix for

	https://bugzilla.kernel.org/show_bug.cgi?id=33012

on my machine.

---
 arch/x86/include/asm/msr-index.h |    4 ++++
 arch/x86/kernel/cpu/amd.c        |   19 +++++++++++++++++++
 2 files changed, 23 insertions(+)

@@ -92,11 +92,15 @@
 #define MSR_IA32_MC0_ADDR		0x00000402
 #define MSR_IA32_MC0_MISC		0x00000403
 
+#define MSR_AMD64_MC0_MASK		0xc0010044
+
 #define MSR_IA32_MCx_CTL(x)		(MSR_IA32_MC0_CTL + 4*(x))
 #define MSR_IA32_MCx_STATUS(x)		(MSR_IA32_MC0_STATUS + 4*(x))
 #define MSR_IA32_MCx_ADDR(x)		(MSR_IA32_MC0_ADDR + 4*(x))
 #define MSR_IA32_MCx_MISC(x)		(MSR_IA32_MC0_MISC + 4*(x))
 
+#define MSR_AMD64_MCx_MASK(x)		(MSR_AMD64_MC0_MASK + (x))
+
 /* These are consecutive and not in the normal 4er MCE bank block */
 #define MSR_IA32_MC0_CTL2		0x00000280
 #define MSR_IA32_MCx_CTL2(x)		(MSR_IA32_MC0_CTL2 + (x))

@@ -598,6 +598,25 @@ static void __cpuinit init_amd(struct cp
 	/* As a rule processors have APIC timer running in deep C states */
 	if (c->x86 >= 0xf && !cpu_has_amd_erratum(amd_erratum_400))
 		set_cpu_cap(c, X86_FEATURE_ARAT);
+
+	/*
+	 * Disable GART TLB Walk Errors on Fam10h. We do this here
+	 * because this is always needed when GART is enabled, even in a
+	 * kernel which has no MCE support built in.
+	 */
+	if (c->x86 == 0x10) {
+		/*
+		 * BIOS should disable GartTlbWlk Errors themself. If
+		 * it doesn't do it here as suggested by the BKDG.
+		 *
+		 * Fixes: https://bugzilla.kernel.org/show_bug.cgi?id=33012
+		 */
+		u64 mask;
+
+		rdmsrl(MSR_AMD64_MCx_MASK(4), mask);
+		mask |= (1 << 10);
+		wrmsrl(MSR_AMD64_MCx_MASK(4), mask);
+	}
 }
 
 #ifdef CONFIG_X86_32

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Tim Chen <[email protected]>

commit c1530019e311c91d14b24d8e74d233152d806e45 upstream.

During RCU walk in path_lookupat and path_openat, the rcu lookup frequently failed if looking up an absolute path, because when root directory was looked up, seq number was not properly set in nameidata.

We dropped out of RCU walk in nameidata_drop_rcu due to mismatch in directory entry's seq number. We reverted to slow path walk that need to take references.

With the following patch, I saw a 50% increase in an exim mail server benchmark throughput on a 4-socket Nehalem-EX system.

---
 fs/namei.c |    1 +
 1 file changed, 1 insertion(+)

@@ -712,6 +712,7 @@ static __always_inline void set_root_rcu
 		do {
 			seq = read_seqcount_begin(&fs->seq);
 			nd->root = fs->root;
+			nd->seq = __read_seqcount_begin(&nd->root.dentry->d_seq);
 		} while (read_seqcount_retry(&fs->seq, seq));
 	}
 }

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Will Deacon <[email protected]>

commit e89c0d7090c54d7b11b9b091e495a1ae345dd3ff upstream.

The DBGVCR, used for configuring vector catch debug events, is UNKNOWN out of reset on ARMv7. When enabling monitor mode, this must be zeroed to avoid UNPREDICTABLE behaviour.

This patch adds the zeroing code to the debug reset path.

---
 arch/arm/kernel/hw_breakpoint.c |    7 +++++++
 1 file changed, 7 insertions(+)

@@ -868,6 +868,13 @@ static void reset_ctrl_regs(void *info)
 		 */
 		asm volatile("mcr p14, 0, %0, c1, c0, 4" : : "r" (0));
 		isb();
+
+		/*
+		 * Clear any configured vector-catch events before
+		 * enabling monitor mode.
+		 */
+		asm volatile("mcr p14, 0, %0, c0, c7, 0" : : "r" (0));
+		isb();
 	}
 
 	if (enable_monitor_mode())

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Alex Deucher <[email protected]>

commit d3b3e15da14ded61c9654db05863b04a2435f4cc upstream.

Apparently some distros set i2c-algo-bit.bit_test to 1 by default. In some cases this causes i2c_bit_add_bus to fail and prevents the i2c bus from being added. In the radeon case, we fail to add the ddc i2c buses which prevents the driver from being able to detect attached monitors. The i2c bus works fine even if bit_test fails. This is likely due to gpio switching that is required and handled in the pre/post_xfer hooks, so call the pre/post_xfer hooks in the bit test as well.

Fixes:
https://bugs.freedesktop.org/show_bug.cgi?id=36221

---
 drivers/i2c/algos/i2c-algo-bit.c |   22 +++++++++++++++++++---
 1 file changed, 19 insertions(+), 3 deletions(-)

@@ -232,9 +232,17 @@ static int i2c_inb(struct i2c_adapter *i
  * Sanity check for the adapter hardware - check the reaction of
  * the bus lines only if it seems to be idle.
  */
-static int test_bus(struct i2c_algo_bit_data *adap, char *name)
+static int test_bus(struct i2c_adapter *i2c_adap)
 {
-	int scl, sda;
+	struct i2c_algo_bit_data *adap = i2c_adap->algo_data;
+	const char *name = i2c_adap->name;
+	int scl, sda, ret;
+
+	if (adap->pre_xfer) {
+		ret = adap->pre_xfer(i2c_adap);
+		if (ret < 0)
+			return -ENODEV;
+	}
 
 	if (adap->getscl == NULL)
 		pr_info("%s: Testing SDA only, SCL is not readable\n", name);

@@ -297,11 +305,19 @@ static int test_bus(struct i2c_algo_bit_
 		       "while pulling SCL high!\n", name);
 		goto bailout;
 	}
+
+	if (adap->post_xfer)
+		adap->post_xfer(i2c_adap);
+
 	pr_info("%s: Test OK\n", name);
 	return 0;
 bailout:
 	sdahi(adap);
 	sclhi(adap);
+
+	if (adap->post_xfer)
+		adap->post_xfer(i2c_adap);
+
 	return -ENODEV;
 }
 

@@ -607,7 +623,7 @@ static int __i2c_bit_add_bus(struct i2c_
 	int ret;
 
 	if (bit_test) {
-		ret = test_bus(bit_adap, adap->name);
+		ret = test_bus(adap);
 		if (ret < 0)
 			return -ENODEV;
 	}

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Mike Frysinger <[email protected]>

commit 8c122b96866580c99e44f3f07ac93a993d964ec3 upstream.

The new bfin_rtc_alarm_irq_enable function forgot to add a "return 0" to the end leading to the build warning:
drivers/rtc/rtc-bfin.c: In function 'bfin_rtc_alarm_irq_enable':
drivers/rtc/rtc-bfin.c:253: warning: control reaches end of non-void function

---
 drivers/rtc/rtc-bfin.c |    2 ++
 1 file changed, 2 insertions(+)

@@ -276,6 +276,8 @@ static int bfin_rtc_alarm_irq_enable(str
 		bfin_rtc_int_set_alarm(rtc);
 	else
 		bfin_rtc_int_clear(~(RTC_ISTAT_ALARM | RTC_ISTAT_ALARM_DAY));
+
+	return 0;
 }
 
 static int bfin_rtc_read_time(struct device *dev, struct rtc_time *tm)

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Ken Chen <[email protected]>

commit b30aef17f71cf9e24b10c11cbb5e5f0ebe8a85ab upstream.

The scheduler load balancer has specific code to deal with cases of unbalanced system due to lots of unmovable tasks (for example because of hard CPU affinity). In those situation, it excludes the busiest CPU that has pinned tasks for load balance consideration such that it can perform second 2nd load balance pass on the rest of the system.

This all works as designed if there is only one cgroup in the system.

However, when we have multiple cgroups, this logic has false positives and triggers multiple load balance passes despite there are actually no pinned tasks at all.

The reason it has false positives is that the all pinned logic is deep in the lowest function of can_migrate_task() and is too low level:

load_balance_fair() iterates each task group and calls balance_tasks() to migrate target load. Along the way, balance_tasks() will also set a all_pinned variable. Given that task-groups are iterated, this all_pinned variable is essentially the status of last group in the scanning process. Task group can have number of reasons that no load being migrated, none due to cpu affinity. However, this status bit is being propagated back up to the higher level load_balance(), which incorrectly think that no tasks were moved. It kick off the all pinned logic and start multiple passes attempt to move load onto puller CPU.

To fix this, move the all_pinned aggregation up at the iterator level. This ensures that the status is aggregated over all task-groups, not just last one in the list.

---
 kernel/sched_fair.c |   11 ++++-------
 1 file changed, 4 insertions(+), 7 deletions(-)

@@ -2043,21 +2043,20 @@ balance_tasks(struct rq *this_rq, int th
 	      enum cpu_idle_type idle, int *all_pinned,
 	      int *this_best_prio, struct cfs_rq *busiest_cfs_rq)
 {
-	int loops = 0, pulled = 0, pinned = 0;
+	int loops = 0, pulled = 0;
 	long rem_load_move = max_load_move;
 	struct task_struct *p, *n;
 
 	if (max_load_move == 0)
 		goto out;
 
-	pinned = 1;
-
 	list_for_each_entry_safe(p, n, &busiest_cfs_rq->tasks, se.group_node) {
 		if (loops++ > sysctl_sched_nr_migrate)
 			break;
 
 		if ((p->se.load.weight >> 1) > rem_load_move ||
-		    !can_migrate_task(p, busiest, this_cpu, sd, idle, &pinned))
+		    !can_migrate_task(p, busiest, this_cpu, sd, idle,
+				      all_pinned))
 			continue;
 
 		pull_task(busiest, p, this_rq, this_cpu);

@@ -2092,9 +2091,6 @@ out:
 	 */
 	schedstat_add(sd, lb_gained[idle], pulled);
 
-	if (all_pinned)
-		*all_pinned = pinned;
-
 	return max_load_move - rem_load_move;
 }
 

@@ -3297,6 +3293,7 @@ redo:
 		 * still unbalanced. ld_moved simply stays zero, so it is
 		 * correctly treated as an imbalance.
 		 */
+		all_pinned = 1;
 		local_irq_save(flags);
 		double_rq_lock(this_rq, busiest);
 		ld_moved = move_tasks(this_rq, this_cpu, busiest,

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: KOSAKI Motohiro <[email protected]>

commit 929bea7c714220fc76ce3f75bef9056477c28e74 upstream.

all_unreclaimable check in direct reclaim has been introduced at 2.6.19 by following commit.

	2006 Sep 25; commit 408d8544; oom: use unreclaimable info

And it went through strange history. firstly, following commit broke the logic unintentionally.

2008 Apr 29; commit a41f24ea; page allocator: smarter retry of
costly-order allocations

Two years later, I've found obvious meaningless code fragment and restored original intention by following commit.

	2010 Jun 04; commit bb21c7ce; vmscan: fix do_try_to_free_pages()
				      return value when priority==0

But, the logic didn't works when 32bit highmem system goes hibernation and Minchan slightly changed the algorithm and fixed it .

	2010 Sep 22: commit d1908362: vmscan: check all_unreclaimable
				      in direct reclaim path

But, recently, Andrey Vagin found the new corner case. Look,

	struct zone {
	  ..
	        int                     all_unreclaimable;
	  ..
	        unsigned long           pages_scanned;
	  ..
	}

zone->all_unreclaimable and zone->pages_scanned are neigher atomic variables nor protected by lock. Therefore zones can become a state of zone->page_scanned=0 and zone->all_unreclaimable=1. In this case, current all_unreclaimable() return false even though zone->all_unreclaimabe=1.

This resulted in the kernel hanging up when executing a loop of the form

1. fork
2. mmap
3. touch memory
4. read memory
5. munmmap

as described in
http://www.gossamer-threads.com/lists/linux/kernel/1348725#1348725

Is this ignorable minor issue? No. Unfortunately, x86 has very small dma zone and it become zone->all_unreclamble=1 easily. and if it become all_unreclaimable=1, it never restore all_unreclaimable=0. Why? if all_unreclaimable=1, vmscan only try DEF_PRIORITY reclaim and a-few-lru-pages>>DEF_PRIORITY always makes 0. that mean no page scan at all!

Eventually, oom-killer never works on such systems. That said, we can't use zone->pages_scanned for this purpose. This patch restore all_unreclaimable() use zone->all_unreclaimable as old. and in addition, to add oom_killer_disabled check to avoid reintroduce the issue of commit d1908362 ("vmscan: check all_unreclaimable in direct reclaim path").

---
 mm/vmscan.c |   24 +++++++++++++-----------
 1 file changed, 13 insertions(+), 11 deletions(-)

@@ -41,6 +41,7 @@
 #include <linux/memcontrol.h>
 #include <linux/delayacct.h>
 #include <linux/sysctl.h>
+#include <linux/oom.h>
 
 #include <asm/tlbflush.h>
 #include <asm/div64.h>

@@ -1988,17 +1989,12 @@ static bool zone_reclaimable(struct zone
 	return zone->pages_scanned < zone_reclaimable_pages(zone) * 6;
 }
 
-/*
- * As hibernation is going on, kswapd is freezed so that it can't mark
- * the zone into all_unreclaimable. It can't handle OOM during hibernation.
- * So let's check zone's unreclaimable in direct reclaim as well as kswapd.
- */
+/* All zones in zonelist are unreclaimable? */
 static bool all_unreclaimable(struct zonelist *zonelist,
 		struct scan_control *sc)
 {
 	struct zoneref *z;
 	struct zone *zone;
-	bool all_unreclaimable = true;
 
 	for_each_zone_zonelist_nodemask(zone, z, zonelist,
 			gfp_zone(sc->gfp_mask), sc->nodemask) {

@@ -2006,13 +2002,11 @@ static bool all_unreclaimable(struct zon
 			continue;
 		if (!cpuset_zone_allowed_hardwall(zone, GFP_KERNEL))
 			continue;
-		if (zone_reclaimable(zone)) {
-			all_unreclaimable = false;
-			break;
-		}
+		if (!zone->all_unreclaimable)
+			return false;
 	}
 
-	return all_unreclaimable;
+	return true;
 }
 
 /*

@@ -2108,6 +2102,14 @@ out:
 	if (sc->nr_reclaimed)
 		return sc->nr_reclaimed;
 
+	/*
+	 * As hibernation is going on, kswapd is freezed so that it can't mark
+	 * the zone into all_unreclaimable. Thus bypassing all_unreclaimable
+	 * check.
+	 */
+	if (oom_killer_disabled)
+		return 0;
+
 	/* top priority shrink_zones still had more to do? don't OOM, then */
 	if (scanning_global_lru(sc) && !all_unreclaimable(zonelist, sc))
 		return 1;

[...]

So presumably this needs to be fixed in 2.6.32.y and other longterm series as well. Though there seems to be a whole series of fixes required in 2.6.32.y!

Are you going to look after that, or should someone else prepare backports? (I'm certainly not volunteering - I don't have the VM knowledge to work out what needs doing.)

Ben.

Ben Hutchings
Once a job is fouled up, anything done to improve it makes it worse.

Hi Ben

Honestly, I didn't prepare. If my remember is correct, you are debian guy. So, Can I think the backport 2.6.32.y help debian people? If so, it's good thing to increase my priority to do this.

Thanks.

Most of the 'enterprise' and long-term supported distributions (Debian, Ubuntu, SLE, OEL and RHEL) have kernels based on 2.6.32. RH seem to be doing their own thing but the rest of us are using 2.6.32.y as a basis.

Ben.

Ben Hutchings
Once a job is fouled up, anything done to improve it makes it worse.

I've finished see current head of longterm-2.6.32. It has

2006 Sep 25; commit 408d8544; oom: use unreclaimable info 2008 Apr 29; commit a41f24ea; page allocator: smarter retry of
costly-order allocations

and, doesn't have

        2010 Jun 04; commit bb21c7ce; vmscan: fix do_try_to_free_pages()
                                      return value when priority==0
        2010 Sep 22: commit d1908362: vmscan: check all_unreclaimable
                                      in direct reclaim path
        2011 Apr 14 commit 929bea7c7: vmscan: all_unreclaimable() use 
                                      zone->all_unreclaimable as a name

Then, the code mean
1) commit 408d8544 doesn't works as well.
2) But, there is no hangup risk as commit 929bea7c7 described.

So, I think there is no worth to backport. Two years no bug report mean it's no big matter. And we can't make a patch which include the above three patch and its dependencies smaller than 100 lines.

Thanks.

On Thu, 2011-04-21 at 15:10 +0900, KOSAKI Motohiro wrote:
[...]

OK. Thanks a lot for checking.

Ben.

Ben Hutchings
Once a job is fouled up, anything done to improve it makes it worse.

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Jiri Kosina <[email protected]>

commit 4471a675dfc7ca676c165079e91c712b09dc9ce4 upstream.

5520e89 ("brk: fix min_brk lower bound computation for COMPAT_BRK") tried to get the whole logic of brk randomization for legacy (libc5-based) applications finally right.

It turns out that the way to detect whether brk has actually been randomized in the end or not introduced by that patch still doesn't work for those binaries, as reported by Geert:

: /sbin/init from my old m68k ramdisk exists prematurely. :
: Before the patch:
:
: | brk(0x80005c8e) = 0x80006000
:
: After the patch:
:
: | brk(0x80005c8e) = 0x80005c8e
:
: Old libc5 considers brk() to have failed if the return value is not : identical to the requested value.

I don't like it, but currently see no better option than a bit flag in task_struct to catch the CONFIG_COMPAT_BRK && randomize_va_space == 2 case.

---
 fs/binfmt_elf.c       |    6 +++++-
 include/linux/sched.h |    3 +++
 mm/mmap.c             |    2 +-
 3 files changed, 9 insertions(+), 2 deletions(-)

@@ -941,9 +941,13 @@ static int load_elf_binary(struct linux_
 	current->mm->start_stack = bprm->p;
 
 #ifdef arch_randomize_brk
-	if ((current->flags & PF_RANDOMIZE) && (randomize_va_space > 1))
+	if ((current->flags & PF_RANDOMIZE) && (randomize_va_space > 1)) {
 		current->mm->brk = current->mm->start_brk =
 			arch_randomize_brk(current->mm);
+#ifdef CONFIG_COMPAT_BRK
+		current->brk_randomized = 1;
+#endif
+	}
 #endif
 
 	if (current->personality & MMAP_PAGE_ZERO) {

@@ -1254,6 +1254,9 @@ struct task_struct {
 #endif
 
 	struct mm_struct *mm, *active_mm;
+#ifdef CONFIG_COMPAT_BRK
+	unsigned brk_randomized:1;
+#endif
 #if defined(SPLIT_RSS_COUNTING)
 	struct task_rss_stat	rss_stat;
 #endif

@@ -259,7 +259,7 @@ SYSCALL_DEFINE1(brk, unsigned long, brk)
 	 * randomize_va_space to 2, which will still cause mm->start_brk
 	 * to be arbitrarily shifted
 	 */
-	if (mm->start_brk > PAGE_ALIGN(mm->end_data))
+	if (current->brk_randomized)
 		min_brk = mm->start_brk;
 	else
 		min_brk = mm->end_data;

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Felipe Balbi <[email protected]>

commit 7a180e70cfc56e131bfe4796773df2acfc7d4180 upstream.

Due to the recent changes to musb's glue layers, we can't compile musb-hdrc as a module - compilation will break due to undefined symbol musb_debug. In order to fix that, we need a big re-work of the debug support on the MUSB driver.

Because that would mean a lot of new code coming into the -rc series, it's best to defer that to next merge window and for now just disable module support for MUSB.

Once we get the refactor of the debugging support done, we can simply revert this patch and things will go back to normal again.

---
 drivers/usb/musb/Kconfig |    6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

@@ -14,7 +14,7 @@ config USB_MUSB_HDRC
 	select TWL4030_USB if MACH_OMAP_3430SDP
 	select TWL6030_USB if MACH_OMAP_4430SDP || MACH_OMAP4_PANDA
 	select USB_OTG_UTILS
-	tristate 'Inventra Highspeed Dual Role Controller (TI, ADI, ...)'
+	bool 'Inventra Highspeed Dual Role Controller (TI, ADI, ...)'
 	help
 	  Say Y here if your system has a dual role high speed USB
 	  controller based on the Mentor Graphics silicon IP.  Then

@@ -30,8 +30,8 @@ config USB_MUSB_HDRC
 
 	  If you do not know what this is, please say N.
 
-	  To compile this driver as a module, choose M here; the
-	  module will be called "musb-hdrc".
+#	  To compile this driver as a module, choose M here; the
+#	  module will be called "musb-hdrc".
 
 choice
 	prompt "Platform Glue Layer"

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Christian Simon <[email protected]>

commit 5a9443f08c83c294c5c806a689c1184b27cb26b3 upstream.

I added new ProdutIds for two devices from CTI GmbH Leipzig.

---
 drivers/usb/serial/ftdi_sio.c     |    2 ++
 drivers/usb/serial/ftdi_sio_ids.h |    9 +++++++++
 2 files changed, 11 insertions(+)

@@ -151,6 +151,8 @@ static struct ftdi_sio_quirk ftdi_stmcli
  * /sys/bus/usb/ftdi_sio/new_id, then send patch/report!
  */
 static struct usb_device_id id_table_combined [] = {
+	{ USB_DEVICE(FTDI_VID, FTDI_CTI_MINI_PID) },
+	{ USB_DEVICE(FTDI_VID, FTDI_CTI_NANO_PID) },
 	{ USB_DEVICE(FTDI_VID, FTDI_AMC232_PID) },
 	{ USB_DEVICE(FTDI_VID, FTDI_CANUSB_PID) },
 	{ USB_DEVICE(FTDI_VID, FTDI_CANDAPTER_PID) },

@@ -1140,3 +1140,12 @@
 #define QIHARDWARE_VID			0x20B7
 #define MILKYMISTONE_JTAGSERIAL_PID	0x0713
 
+/*
+ * CTI GmbH RS485 Converter http://www.cti-lean.com/
+ */
+/* USB-485-Mini*/
+#define FTDI_CTI_MINI_PID	0xF608
+/* USB-Nano-485*/
+#define FTDI_CTI_NANO_PID	0xF60B
+
+

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Johan Hovold <[email protected]>

commit 11a31d84129dc3133417d626643d714c9df5317e upstream.

Add PID 0x0103 for serial port of the OCT DK201 docking station.

---
 drivers/usb/serial/ftdi_sio.c     |    1 +
 drivers/usb/serial/ftdi_sio_ids.h |    1 +
 2 files changed, 2 insertions(+)

@@ -527,6 +527,7 @@ static struct usb_device_id id_table_com
 	{ USB_DEVICE(SEALEVEL_VID, SEALEVEL_2803_8_PID) },
 	{ USB_DEVICE(IDTECH_VID, IDTECH_IDT1221U_PID) },
 	{ USB_DEVICE(OCT_VID, OCT_US101_PID) },
+	{ USB_DEVICE(OCT_VID, OCT_DK201_PID) },
 	{ USB_DEVICE(FTDI_VID, FTDI_HE_TIRA1_PID),
 		.driver_info = (kernel_ulong_t)&ftdi_HE_TIRA1_quirk },
 	{ USB_DEVICE(FTDI_VID, FTDI_USB_UIRT_PID),

@@ -572,6 +572,7 @@
 /* Note: OCT US101 is also rebadged as Dick Smith Electronics (NZ) XH6381 */
 /* Also rebadged as Dick Smith Electronics (Aus) XH6451 */
 /* Also rebadged as SIIG Inc. model US2308 hardware version 1 */
+#define OCT_DK201_PID		0x0103	/* OCT DK201 USB docking station */
 #define OCT_US101_PID		0x0421	/* OCT US101 USB to RS-232 */
 
 /*

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Paul Friedrich <[email protected]>

commit c53c2fab40cf16e13af66f40bfd27200cda98d2f upstream.

usb serial: ftdi_sio: add two missing USB ID's for Hameg interfaces HO720 and HO730

---
 drivers/usb/serial/ftdi_sio.c     |    2 ++
 drivers/usb/serial/ftdi_sio_ids.h |    2 ++
 2 files changed, 4 insertions(+)

@@ -788,6 +788,8 @@ static struct usb_device_id id_table_com
 	{ USB_DEVICE(FTDI_VID, MARVELL_OPENRD_PID),
 		.driver_info = (kernel_ulong_t)&ftdi_jtag_quirk },
 	{ USB_DEVICE(FTDI_VID, HAMEG_HO820_PID) },
+	{ USB_DEVICE(FTDI_VID, HAMEG_HO720_PID) },
+	{ USB_DEVICE(FTDI_VID, HAMEG_HO730_PID) },
 	{ USB_DEVICE(FTDI_VID, HAMEG_HO870_PID) },
 	{ USB_DEVICE(FTDI_VID, MJSG_GENERIC_PID) },
 	{ USB_DEVICE(FTDI_VID, MJSG_SR_RADIO_PID) },

@@ -300,6 +300,8 @@
  * Hameg HO820 and HO870 interface (using VID 0x0403)
  */
 #define HAMEG_HO820_PID			0xed74
+#define HAMEG_HO730_PID			0xed73
+#define HAMEG_HO720_PID			0xed72
 #define HAMEG_HO870_PID			0xed71
 
 /*

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Marius B. Kotsbak <[email protected]>

commit 80f9df3e0093ad9f1eeefd2ff7fd27daaa518d25 upstream.

Bind only modem AT command endpoint to option.

---
 drivers/usb/serial/option.c |    5 +++++
 1 file changed, 5 insertions(+)

@@ -407,6 +407,10 @@ static void option_instat_callback(struc
 /* ONDA MT825UP HSDPA 14.2 modem */
 #define ONDA_MT825UP         0x000b
 
+/* Samsung products */
+#define SAMSUNG_VENDOR_ID                       0x04e8
+#define SAMSUNG_PRODUCT_GT_B3730                0x6889
+
 /* some devices interfaces need special handling due to a number of reasons */
 enum option_blacklist_reason {
 		OPTION_BLACKLIST_NONE = 0,

@@ -968,6 +972,7 @@ static const struct usb_device_id option
 	{ USB_DEVICE(OLIVETTI_VENDOR_ID, OLIVETTI_PRODUCT_OLICARD100) },
 	{ USB_DEVICE(CELOT_VENDOR_ID, CELOT_PRODUCT_CT680M) }, /* CT-650 CDMA 450 1xEVDO modem */
 	{ USB_DEVICE(ONDA_VENDOR_ID, ONDA_MT825UP) }, /* ONDA MT825UP modem */
+	{ USB_DEVICE_AND_INTERFACE_INFO(SAMSUNG_VENDOR_ID, SAMSUNG_PRODUCT_GT_B3730, USB_CLASS_CDC_DATA, 0x00, 0x00) }, /* Samsung GT-B3730/GT-B3710 LTE USB modem.*/
 	{ } /* Terminating entry */
 };
 MODULE_DEVICE_TABLE(usb, option_ids);

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Linus Torvalds <[email protected]>

commit c78193e9c7bcbf25b8237ad0dec82f805c4ea69b upstream.

next_pidmap() just quietly accepted whatever 'last' pid that was passed in, which is not all that safe when one of the users is /proc.

Admittedly the proc code should do some sanity checking on the range (and that will be the next commit), but that doesn't mean that the helper functions should just do that pidmap pointer arithmetic without checking the range of its arguments.

So clamp 'last' to PID_MAX_LIMIT. The fact that we then do "last+1" doesn't really matter, the for-loop does check against the end of the pidmap array properly (it's only the actual pointer arithmetic overflow case we need to worry about, and going one bit beyond isn't going to overflow).

[ Use PID_MAX_LIMIT rather than pid_max as per Eric Biederman ]

---
 include/linux/pid.h |    2 +-
 kernel/pid.c        |    5 ++++-
 2 files changed, 5 insertions(+), 2 deletions(-)

@@ -117,7 +117,7 @@ extern struct pid *find_vpid(int nr);
  */
 extern struct pid *find_get_pid(int nr);
 extern struct pid *find_ge_pid(int nr, struct pid_namespace *);
-int next_pidmap(struct pid_namespace *pid_ns, int last);
+int next_pidmap(struct pid_namespace *pid_ns, unsigned int last);
 
 extern struct pid *alloc_pid(struct pid_namespace *ns);
 extern void free_pid(struct pid *pid);

@@ -217,11 +217,14 @@ static int alloc_pidmap(struct pid_names
 	return -1;
 }
 
-int next_pidmap(struct pid_namespace *pid_ns, int last)
+int next_pidmap(struct pid_namespace *pid_ns, unsigned int last)
 {
 	int offset;
 	struct pidmap *map, *end;
 
+	if (last >= PID_MAX_LIMIT)
+		return -1;
+
 	offset = (last + 1) & BITS_PER_PAGE_MASK;
 	map = &pid_ns->pidmap[(last + 1)/BITS_PER_PAGE];
 	end = &pid_ns->pidmap[PIDMAP_ENTRIES];

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Linus Torvalds <[email protected]>

commit d8bdc59f215e62098bc5b4256fd9928bf27053a1 upstream.

Rather than pass in some random truncated offset to the pid-related functions, check that the offset is in range up-front.

This is just cleanup, the previous commit fixed the real problem.

---
 fs/proc/base.c |    9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

@@ -3066,11 +3066,16 @@ static int proc_pid_fill_cache(struct fi
 /* for the /proc/ directory itself, after non-process stuff has been done */
 int proc_pid_readdir(struct file * filp, void * dirent, filldir_t filldir)
 {
-	unsigned int nr = filp->f_pos - FIRST_PROCESS_ENTRY;
-	struct task_struct *reaper = get_proc_task(filp->f_path.dentry->d_inode);
+	unsigned int nr;
+	struct task_struct *reaper;
 	struct tgid_iter iter;
 	struct pid_namespace *ns;
 
+	if (filp->f_pos >= PID_MAX_LIMIT + TGID_OFFSET)
+		goto out_no_task;
+	nr = filp->f_pos - FIRST_PROCESS_ENTRY;
+
+	reaper = get_proc_task(filp->f_path.dentry->d_inode);
 	if (!reaper)
 		goto out_no_task;
 

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Anton Blanchard <[email protected]>

commit 84ffae55af79d7b8834fd0c08d0d1ebf2c77f91e upstream.

We currently enable interrupts before the dispatch log for the boot cpu is setup. If a timer interrupt comes in early enough we oops in scan_dispatch_log:

Unable to handle kernel paging request for data at address 0x00000010

...

.scan_dispatch_log+0xb0/0x170
.account_system_vtime+0xa0/0x220
.irq_enter+0x88/0xc0
.do_IRQ+0x48/0x230

The patch below adds a check to scan_dispatch_log to ensure the dispatch log has been allocated.

---
 arch/powerpc/kernel/time.c |    3 +++
 1 file changed, 3 insertions(+)

@@ -229,6 +229,9 @@ static u64 scan_dispatch_log(u64 stop_tb
 	u64 stolen = 0;
 	u64 dtb;
 
+	if (!dtl)
+		return 0;
+
 	if (i == vpa->dtl_idx)
 		return 0;
 	while (i < vpa->dtl_idx) {

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Eric B Munson <[email protected]>

commit 86c74ab317c1ef4d37325e0d7ca8a01a796b0bd7 upstream.

Because of speculative event roll back, it is possible for some event coutners to decrease between reads on POWER7. This causes a problem with the way that counters are updated. Delta calues are calculated in a 64 bit value and the top 32 bits are masked. If the register value has decreased, this leaves us with a very large positive value added to the kernel counters. This patch protects against this by skipping the update if the delta would be negative. This can lead to a lack of precision in the coutner values, but from my testing the value is typcially fewer than 10 samples at a time.

---
 arch/powerpc/kernel/perf_event.c |   37 ++++++++++++++++++++++++++++++-------
 1 file changed, 30 insertions(+), 7 deletions(-)

@@ -398,6 +398,25 @@ static int check_excludes(struct perf_ev
 	return 0;
 }
 
+static u64 check_and_compute_delta(u64 prev, u64 val)
+{
+	u64 delta = (val - prev) & 0xfffffffful;
+
+	/*
+	 * POWER7 can roll back counter values, if the new value is smaller
+	 * than the previous value it will cause the delta and the counter to
+	 * have bogus values unless we rolled a counter over.  If a coutner is
+	 * rolled back, it will be smaller, but within 256, which is the maximum
+	 * number of events to rollback at once.  If we dectect a rollback
+	 * return 0.  This can lead to a small lack of precision in the
+	 * counters.
+	 */
+	if (prev > val && (prev - val) < 256)
+		delta = 0;
+
+	return delta;
+}
+
 static void power_pmu_read(struct perf_event *event)
 {
 	s64 val, delta, prev;

@@ -416,10 +435,11 @@ static void power_pmu_read(struct perf_e
 		prev = local64_read(&event->hw.prev_count);
 		barrier();
 		val = read_pmc(event->hw.idx);
+		delta = check_and_compute_delta(prev, val);
+		if (!delta)
+			return;
 	} while (local64_cmpxchg(&event->hw.prev_count, prev, val) != prev);
 
-	/* The counters are only 32 bits wide */
-	delta = (val - prev) & 0xfffffffful;
 	local64_add(delta, &event->count);
 	local64_sub(delta, &event->hw.period_left);
 }

@@ -449,8 +469,9 @@ static void freeze_limited_counters(stru
 		val = (event->hw.idx == 5) ? pmc5 : pmc6;
 		prev = local64_read(&event->hw.prev_count);
 		event->hw.idx = 0;
-		delta = (val - prev) & 0xfffffffful;
-		local64_add(delta, &event->count);
+		delta = check_and_compute_delta(prev, val);
+		if (delta)
+			local64_add(delta, &event->count);
 	}
 }
 

@@ -458,14 +479,16 @@ static void thaw_limited_counters(struct
 				  unsigned long pmc5, unsigned long pmc6)
 {
 	struct perf_event *event;
-	u64 val;
+	u64 val, prev;
 	int i;
 
 	for (i = 0; i < cpuhw->n_limited; ++i) {
 		event = cpuhw->limited_counter[i];
 		event->hw.idx = cpuhw->limited_hwidx[i];
 		val = (event->hw.idx == 5) ? pmc5 : pmc6;
-		local64_set(&event->hw.prev_count, val);
+		prev = local64_read(&event->hw.prev_count);
+		if (check_and_compute_delta(prev, val))
+			local64_set(&event->hw.prev_count, val);
 		perf_event_update_userpage(event);
 	}
 }

@@ -1197,7 +1220,7 @@ static void record_and_restart(struct pe
 
 	/* we don't have to worry about interrupts here */
 	prev = local64_read(&event->hw.prev_count);
-	delta = (val - prev) & 0xfffffffful;
+	delta = check_and_compute_delta(prev, val);
 	local64_add(delta, &event->count);
 
 	/*

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Steven Hardy <[email protected]>

commit 10c9ab15d6aee153968d150c05b3ee3df89673de upstream.

qcprobe function allocates serial->private but this is never freed, this patch adds a new function qc_release() which frees serial->private, after calling usb_wwan_release

---
 drivers/usb/serial/qcserial.c |   14 +++++++++++++-
 1 file changed, 13 insertions(+), 1 deletion(-)

@@ -205,6 +205,18 @@ static int qcprobe(struct usb_serial *se
 	return retval;
 }
 
+static void qc_release(struct usb_serial *serial)
+{
+	struct usb_wwan_intf_private *priv = usb_get_serial_data(serial);
+
+	dbg("%s", __func__);
+
+	/* Call usb_wwan release & free the private data allocated in qcprobe */
+	usb_wwan_release(serial);
+	usb_set_serial_data(serial, NULL);
+	kfree(priv);
+}
+
 static struct usb_serial_driver qcdevice = {
 	.driver = {
 		.owner     = THIS_MODULE,

@@ -222,7 +234,7 @@ static struct usb_serial_driver qcdevice
 	.chars_in_buffer     = usb_wwan_chars_in_buffer,
 	.attach		     = usb_wwan_startup,
 	.disconnect	     = usb_wwan_disconnect,
-	.release	     = usb_wwan_release,
+	.release	     = qc_release,
 #ifdef CONFIG_PM
 	.suspend	     = usb_wwan_suspend,
 	.resume		     = usb_wwan_resume,

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Steven Hardy <[email protected]>

commit 99ab3f9e4eaec35fd2d7159c31b71f17f7e613e3 upstream.

Rework the qcprobe logic such that serial->private is not set when qcprobe exits with -ENODEV, otherwise serial->private will point to freed memory on -ENODEV

---
 drivers/usb/serial/qcserial.c |   15 +++++++++------
 1 file changed, 9 insertions(+), 6 deletions(-)

@@ -111,7 +111,7 @@ static int qcprobe(struct usb_serial *se
 	ifnum = intf->desc.bInterfaceNumber;
 	dbg("This Interface = %d", ifnum);
 
-	data = serial->private = kzalloc(sizeof(struct usb_wwan_intf_private),
+	data = kzalloc(sizeof(struct usb_wwan_intf_private),
 					 GFP_KERNEL);
 	if (!data)
 		return -ENOMEM;

@@ -134,8 +134,10 @@ static int qcprobe(struct usb_serial *se
 		    usb_endpoint_is_bulk_out(&intf->endpoint[1].desc)) {
 			dbg("QDL port found");
 
-			if (serial->interface->num_altsetting == 1)
-				return 0;
+			if (serial->interface->num_altsetting == 1) {
+				retval = 0; /* Success */
+				break;
+			}
 
 			retval = usb_set_interface(serial->dev, ifnum, 1);
 			if (retval < 0) {

@@ -145,7 +147,6 @@ static int qcprobe(struct usb_serial *se
 				retval = -ENODEV;
 				kfree(data);
 			}
-			return retval;
 		}
 		break;
 

@@ -177,7 +178,6 @@ static int qcprobe(struct usb_serial *se
 				retval = -ENODEV;
 				kfree(data);
 			}
-			return retval;
 		} else if (ifnum==3) {
 			/*
 			 * NMEA (serial line 9600 8N1)

@@ -199,9 +199,12 @@ static int qcprobe(struct usb_serial *se
 		dev_err(&serial->dev->dev,
 			"unknown number of interfaces: %d\n", nintf);
 		kfree(data);
-		return -ENODEV;
+		retval = -ENODEV;
 	}
 
+	/* Set serial->private if not returning -ENODEV */
+	if (retval != -ENODEV)
+		usb_set_serial_data(serial, data);
 	return retval;
 }
 

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Steven Hardy <[email protected]>

commit cb62d65f966146a39fdde548cb474dacf1d00fa5 upstream.

There are two -ENODEV error paths in qcprobe where the allocated private data is not freed, this patch adds the two missing kfrees to avoid leaking memory on the error path

---
 drivers/usb/serial/qcserial.c |    2 ++
 1 file changed, 2 insertions(+)

@@ -167,6 +167,7 @@ static int qcprobe(struct usb_serial *se
 					"Could not set interface, error %d\n",
 					retval);
 				retval = -ENODEV;
+				kfree(data);
 			}
 		} else if (ifnum == 2) {
 			dbg("Modem port found");

@@ -191,6 +192,7 @@ static int qcprobe(struct usb_serial *se
 					"Could not set interface, error %d\n",
 					retval);
 				retval = -ENODEV;
+				kfree(data);
 			}
 		}
 		break;

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Alan Stern <[email protected]>

commit 94ae4976e253757e9b03a44d27d41b20f1829d80 upstream.

This patch (as1458) fixes a problem affecting ultra-reliable systems:
When hardware failover of an EHCI controller occurs, the data structures do not get released correctly. This is because the routine responsible for removing unused QHs from the async schedule assumes the controller is running properly (the frame counter is used in determining how long the QH has been idle) -- but when a failover causes the controller to be electronically disconnected from the PCI bus, obviously it stops running.

The solution is simple: Allow scan_async() to remove a QH from the async schedule if it has been idle for long enough _or_ if the controller is stopped.

---
 drivers/usb/host/ehci-q.c |   15 +++++++++------
 1 file changed, 9 insertions(+), 6 deletions(-)

@@ -1245,24 +1245,27 @@ static void start_unlink_async (struct e
 
 static void scan_async (struct ehci_hcd *ehci)
 {
+	bool			stopped;
 	struct ehci_qh		*qh;
 	enum ehci_timer_action	action = TIMER_IO_WATCHDOG;
 
 	ehci->stamp = ehci_readl(ehci, &ehci->regs->frame_index);
 	timer_action_done (ehci, TIMER_ASYNC_SHRINK);
 rescan:
+	stopped = !HC_IS_RUNNING(ehci_to_hcd(ehci)->state);
 	qh = ehci->async->qh_next.qh;
 	if (likely (qh != NULL)) {
 		do {
 			/* clean any finished work for this qh */
-			if (!list_empty (&qh->qtd_list)
-					&& qh->stamp != ehci->stamp) {
+			if (!list_empty(&qh->qtd_list) && (stopped ||
+					qh->stamp != ehci->stamp)) {
 				int temp;
 
 				/* unlinks could happen here; completion
 				 * reporting drops the lock.  rescan using
 				 * the latest schedule, but don't rescan
-				 * qhs we already finished (no looping).
+				 * qhs we already finished (no looping)
+				 * unless the controller is stopped.
 				 */
 				qh = qh_get (qh);
 				qh->stamp = ehci->stamp;

@@ -1283,9 +1286,9 @@ rescan:
 			 */
 			if (list_empty(&qh->qtd_list)
 					&& qh->qh_state == QH_STATE_LINKED) {
-				if (!ehci->reclaim
-					&& ((ehci->stamp - qh->stamp) & 0x1fff)
-						>= (EHCI_SHRINK_FRAMES * 8))
+				if (!ehci->reclaim && (stopped ||
+					((ehci->stamp - qh->stamp) & 0x1fff)
+						>= EHCI_SHRINK_FRAMES * 8))
 					start_unlink_async(ehci, qh);
 				else
 					action = TIMER_ASYNC_SHRINK;

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Dmitry Torokhov <[email protected]>

commit 2868a2b1ba8f9c7f6c4170519ebb6c62934df70e upstream.

Isochronous and interrupt SuperSpeed endpoints use the same mechanisms for decoding bInterval values as HighSpeed ones so adjust the code accordingly.

Also bandwidth reservation for SuperSpeed matches highspeed, not low/full speed.

---
 drivers/usb/core/devices.c |   10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

@@ -221,7 +221,7 @@ static char *usb_dump_endpoint_descripto
 		break;
 	case USB_ENDPOINT_XFER_INT:
 		type = "Int.";
-		if (speed == USB_SPEED_HIGH)
+		if (speed == USB_SPEED_HIGH || speed == USB_SPEED_SUPER)
 			interval = 1 << (desc->bInterval - 1);
 		else
 			interval = desc->bInterval;

@@ -229,7 +229,8 @@ static char *usb_dump_endpoint_descripto
 	default:	/* "can't happen" */
 		return start;
 	}
-	interval *= (speed == USB_SPEED_HIGH) ? 125 : 1000;
+	interval *= (speed == USB_SPEED_HIGH ||
+		     speed == USB_SPEED_SUPER) ? 125 : 1000;
 	if (interval % 1000)
 		unit = 'u';
 	else {

@@ -542,8 +543,9 @@ static ssize_t usb_device_dump(char __us
 	if (level == 0) {
 		int	max;
 
-		/* high speed reserves 80%, full/low reserves 90% */
-		if (usbdev->speed == USB_SPEED_HIGH)
+		/* super/high speed reserves 80%, full/low reserves 90% */
+		if (usbdev->speed == USB_SPEED_HIGH ||
+		    usbdev->speed == USB_SPEED_SUPER)
 			max = 800;
 		else
 			max = FRAME_TIME_MAX_USECS_ALLOC;

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Dmitry Torokhov <[email protected]>

commit 5a6c2f3ff039154872ce597952f8b8900ea0d732 upstream.

Macro arguments used in expressions need to be enclosed in parenthesis to avoid unpleasant surprises.

This should be queued for kernels back to 2.6.31

---
 drivers/usb/host/xhci.h |    6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

@@ -232,7 +232,7 @@ struct xhci_op_regs {
  * notification type that matches a bit set in this bit field.
  */
 #define	DEV_NOTE_MASK		(0xffff)
-#define ENABLE_DEV_NOTE(x)	(1 << x)
+#define ENABLE_DEV_NOTE(x)	(1 << (x))
 /* Most of the device notification types should only be used for debug.
  * SW does need to pay attention to function wake notifications.
  */

@@ -601,11 +601,11 @@ struct xhci_ep_ctx {
 #define EP_STATE_STOPPED	3
 #define EP_STATE_ERROR		4
 /* Mult - Max number of burtst within an interval, in EP companion desc. */
-#define EP_MULT(p)		((p & 0x3) << 8)
+#define EP_MULT(p)		(((p) & 0x3) << 8)
 /* bits 10:14 are Max Primary Streams */
 /* bit 15 is Linear Stream Array */
 /* Interval - period between requests to an endpoint - 125u increments. */
-#define EP_INTERVAL(p)		((p & 0xff) << 16)
+#define EP_INTERVAL(p)		(((p) & 0xff) << 16)
 #define EP_INTERVAL_TO_UFRAMES(p)		(1 << (((p) >> 16) & 0xff))
 #define EP_MAXPSTREAMS_MASK	(0x1f << 10)
 #define EP_MAXPSTREAMS(p)	(((p) << 10) & EP_MAXPSTREAMS_MASK)

2.6.38-stable review patch. If anyone has any objections, please let us know.

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

From: Dmitry Torokhov <[email protected]>

commit dfa49c4ad120a784ef1ff0717168aa79f55a483a upstream.

When parsing exponent-expressed intervals we subtract 1 from the value and then expect it to match with original + 1, which is highly unlikely, and we end with frequent spew:

	usb 3-4: ep 0x83 - rounding interval to 512 microframes

Also, parsing interval for fullspeed isochronous endpoints was incorrect - according to USB spec they use exponent-based intervals (but xHCI spec claims frame-based intervals). I trust USB spec more, especially since USB core agrees with it.

This should be queued for stable kernels back to 2.6.31.

---
 drivers/usb/host/xhci-mem.c |   90 ++++++++++++++++++++++++++++++--------------
 1 file changed, 62 insertions(+), 28 deletions(-)

@@ -920,6 +920,47 @@ int xhci_setup_addressable_virt_dev(stru
 	return 0;
 }
 
+/*
+ * Convert interval expressed as 2^(bInterval - 1) == interval into
+ * straight exponent value 2^n == interval.
+ *
+ */
+static unsigned int xhci_parse_exponent_interval(struct usb_device *udev,
+		struct usb_host_endpoint *ep)
+{
+	unsigned int interval;
+
+	interval = clamp_val(ep->desc.bInterval, 1, 16) - 1;
+	if (interval != ep->desc.bInterval - 1)
+		dev_warn(&udev->dev,
+			 "ep %#x - rounding interval to %d microframes\n",
+			 ep->desc.bEndpointAddress,
+			 1 << interval);
+
+	return interval;
+}
+
+/*
+ * Convert bInterval expressed in frames (in 1-255 range) to exponent of
+ * microframes, rounded down to nearest power of 2.
+ */
+static unsigned int xhci_parse_frame_interval(struct usb_device *udev,
+		struct usb_host_endpoint *ep)
+{
+	unsigned int interval;
+
+	interval = fls(8 * ep->desc.bInterval) - 1;
+	interval = clamp_val(interval, 3, 10);
+	if ((1 << interval) != 8 * ep->desc.bInterval)
+		dev_warn(&udev->dev,
+			 "ep %#x - rounding interval to %d microframes, ep desc says %d microframes\n",
+			 ep->desc.bEndpointAddress,
+			 1 << interval,
+			 8 * ep->desc.bInterval);
+
+	return interval;
+}
+
 /* Return the polling or NAK interval.
  *
  * The polling interval is expressed in "microframes".  If xHCI's Interval field