Skip to contentSkip to conversation

4.19.16-stable review

PATCH In review Greg Kroah-Hartman 4 participants 54 messages Jan 15, 2019 – Jan 16, 2019

Subsystems (2)
lore
Share & tools
lore

Greg Kroah-Hartman

author maintainer
#0 lore raw

cover 4.19.16-stable review

This is the start of the stable review cycle for the 4.19.16 release. There are 50 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 me know.

Responses should be made by Thu Jan 17 15:48:36 UTC 2019. Anything received after that time might be too late.

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

	https://www.kernel.org/pub/linux/kernel/v4.x/stable-review/patch-4.19.16-rc1.gz

or in the git tree and branch at:
git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git linux-4.19.y
and the diffstat can be found below.

thanks,

greg k-h

-------------
Pseudo-Shortlog of commits:

Greg Kroah-Hartman <[email protected]>

    Linux 4.19.16-rc1

Filipe Manana <[email protected]>
Btrfs: use nofs context when initializing security xattrs to avoid deadlock

Filipe Manana <[email protected]>
Btrfs: fix deadlock when enabling quotas due to concurrent snapshot creation

Filipe Manana <[email protected]>
Btrfs: fix access to available allocation bits when starting balance

Will Deacon <[email protected]>
arm64: compat: Don't pull syscall number from regs in arm_compat_syscall

Christoffer Dall <[email protected]>
KVM: arm/arm64: Fix VMID alloc race by reverting to lock-less

Vasily Averin <[email protected]>
sunrpc: use-after-free in svc_process_common()

Jan Stancek <[email protected]>
mm: page_mapped: don't assume compound page is huge or THP

Theodore Ts'o <[email protected]>
ext4: fix special inode number checks in __ext4_iget()

Theodore Ts'o <[email protected]>
ext4: track writeback errors using the generic tracking infrastructure

Theodore Ts'o <[email protected]>
ext4: use ext4_write_inode() when fsyncing w/o a journal

Theodore Ts'o <[email protected]>
ext4: avoid kernel warning when writing the superblock to a dead device

Theodore Ts'o <[email protected]>
ext4: fix a potential fiemap/page fault deadlock w/ inline_data

Theodore Ts'o <[email protected]>
ext4: make sure enough credits are reserved for dioread_nolock writes

Ilya Dryomov <[email protected]>
rbd: don't return 0 on unmap if RBD_DEV_FLAG_REMOVING is set

Lyude Paul <[email protected]>
drm/amdgpu: Don't fail resume process if resuming atomic state fails

Lyude Paul <[email protected]>
drm/amdgpu: Don't ignore rc from drm_dp_mst_topology_mgr_resume()

Chris Wilson <[email protected]>
drm/i915: Unwind failure on pinning the gen7 ppgtt

Ivan Mironov <[email protected]>
drm/fb-helper: Partially bring back workaround for bugs of SDL 1.2

Neil Armstrong <[email protected]>
drm/fb_helper: Allow leaking fbdev smem_start

Jerry (Fangzhi) Zuo <[email protected]>
drm/amd/display: Fix MST dp_blank REG_WAIT timeout

Marc Zyngier <[email protected]>
PCI: dwc: Move interrupt acking into the proper callback

Marc Zyngier <[email protected]>
PCI: dwc: Take lock when ACKing an interrupt

Marc Zyngier <[email protected]>
PCI: dwc: Use interrupt masking instead of disabling

Alex Deucher <[email protected]>
drm/amdgpu: Add new VegaM pci id

Alex Williamson <[email protected]>
vfio/type1: Fix unmap overflow off-by-one

Christian Lamparter <[email protected]>
mtd: rawnand: qcom: fix memory corruption that causes panic

Yi Zeng <[email protected]>
i2c: dev: prevent adapter retries and timeout being set as minus value

Jean-Philippe Brucker <[email protected]>
ACPI/IORT: Fix rc_dma_get_range()

Hans de Goede <[email protected]>
ACPI / PMIC: xpower: Fix TS-pin current-source handling

Hans de Goede <[email protected]>
ACPI: power: Skip duplicate power resource references in _PRx

Michal Hocko <[email protected]>
mm, memcg: fix reclaim deadlock with writeback

Qian Cai <[email protected]>
mm/usercopy.c: no check page span for stack objects

Christoph Lameter <[email protected]>
slab: alien caches must not be initialized if the allocation of the alien cache failed

Jack Stocker <[email protected]>
USB: Add USB_QUIRK_DELAY_CTRL_MSG quirk for Corsair K70 RGB

Icenowy Zheng <[email protected]>
USB: storage: add quirk for SMI SM3350

Icenowy Zheng <[email protected]>
USB: storage: don't insert sane sense for SPC3+ when bad sense specified

Daniele Palmas <[email protected]>
usb: cdc-acm: send ZLP for Telit 3G Intel based modems

Ross Lagerwall <[email protected]>
cifs: Fix potential OOB access of lock element array

Pavel Shilovsky <[email protected]>
CIFS: Fix credit computation for compounded requests

Pavel Shilovsky <[email protected]>
CIFS: Do not hide EINTR after sending network packets

Pavel Shilovsky <[email protected]>
CIFS: Do not set credits to 1 if the server didn't grant anything

Pavel Shilovsky <[email protected]>
CIFS: Fix adjustment of credits for MTU requests

Kailang Yang <[email protected]>
ALSA: hda/realtek - Disable headset Mic VREF for headset mode of ALC225

Kailang Yang <[email protected]>
ALSA: hda/realtek - Add unplug function into unplug state of Headset Mode for ALC225

Kailang Yang <[email protected]>
ALSA: hda/realtek - Support Dell headset mode for New AIO platform

WANG Chao <[email protected]>
x86, modpost: Replace last remnants of RETPOLINE with CONFIG_RETPOLINE

Quentin Perret <[email protected]>
cpufreq: scmi: Fix frequency invariance in slow path

Larry Finger <[email protected]>
staging: rtl8188eu: Fix module loading from tasklet for WEP encryption

Larry Finger <[email protected]>
staging: rtl8188eu: Fix module loading from tasklet for CCMP encryption

Filipe Manana <[email protected]>
Btrfs: fix deadlock when using free space tree due to block group creation

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

Diffstat:

 Makefile                                          |   4 +-
 arch/arm64/kernel/sys_compat.c                    |   9 +-
 arch/arm64/kernel/syscall.c                       |   9 +-
 arch/x86/kernel/cpu/bugs.c                        |   2 +-
 drivers/acpi/arm64/iort.c                         |   3 +-
 drivers/acpi/pmic/intel_pmic_xpower.c             |  41 ++++--
 drivers/acpi/power.c                              |  22 ++++
 drivers/block/rbd.c                               |   9 +-
 drivers/cpufreq/scmi-cpufreq.c                    |   4 +-
 drivers/gpu/drm/Kconfig                           |  20 +++
 drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c           |   1 +
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c |  37 ++++--
 drivers/gpu/drm/amd/display/dc/core/dc_link.c     |   4 +-
 drivers/gpu/drm/drm_fb_helper.c                   | 151 ++++++++++++++--------
 drivers/gpu/drm/i915/i915_gem_gtt.c               |  15 ++-
 drivers/i2c/i2c-dev.c                             |   6 +
 drivers/mtd/nand/raw/qcom_nandc.c                 |  20 +--
 drivers/pci/controller/dwc/pcie-designware-host.c |  37 ++++--
 drivers/staging/rtl8188eu/core/rtw_security.c     |   6 +-
 drivers/usb/class/cdc-acm.c                       |   7 +
 drivers/usb/core/quirks.c                         |   3 +-
 drivers/usb/storage/scsiglue.c                    |   8 +-
 drivers/usb/storage/unusual_devs.h                |  12 ++
 drivers/vfio/vfio_iommu_type1.c                   |   2 +-
 fs/btrfs/ctree.c                                  |  16 ++-
 fs/btrfs/qgroup.c                                 |  14 +-
 fs/btrfs/volumes.c                                |  39 +++---
 fs/btrfs/xattr.c                                  |   8 ++
 fs/cifs/file.c                                    |   8 +-
 fs/cifs/smb2file.c                                |   4 +-
 fs/cifs/smb2pdu.c                                 |   8 +-
 fs/cifs/transport.c                               |  63 ++++++---
 fs/ext4/fsync.c                                   |  16 ++-
 fs/ext4/inline.c                                  |   6 +-
 fs/ext4/inode.c                                   |   5 +-
 fs/ext4/super.c                                   |   2 +-
 include/linux/compiler-gcc.h                      |   2 +-
 include/linux/module.h                            |   2 +-
 include/linux/sunrpc/svc.h                        |   5 +-
 include/trace/events/sunrpc.h                     |   6 +-
 mm/memory.c                                       |  23 ++++
 mm/slab.c                                         |   6 +-
 mm/usercopy.c                                     |   9 +-
 mm/util.c                                         |   2 +-
 net/sunrpc/svc.c                                  |  11 +-
 net/sunrpc/svc_xprt.c                             |   5 +-
 net/sunrpc/svcsock.c                              |   2 +-
 scripts/mod/modpost.c                             |   2 +-
 sound/pci/hda/patch_realtek.c                     |  18 ++-
 virt/kvm/arm/arm.c                                |  23 ++--
 50 files changed, 509 insertions(+), 228 deletions(-)
author maintainer
#1 lore raw

PATCH 1/50 Btrfs: fix deadlock when using free space tree due to block group creation

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Filipe Manana <[email protected]>

commit a6d8654d885d7d79a3fb82da64eaa489ca332a82 upstream.

When modifying the free space tree we can end up COWing one of its extent buffers which in turn might result in allocating a new chunk, which in turn can result in flushing (finish creation) of pending block groups. If that happens we can deadlock because creating a pending block group needs to update the free space tree, and if any of the updates tries to modify the same extent buffer that we are COWing, we end up in a deadlock since we try to write lock twice the same extent buffer.

So fix this by skipping pending block group creation if we are COWing an extent buffer from the free space tree. This is a case missed by commit 5ce555578e091 ("Btrfs: fix deadlock when writing out free space caches").

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=202173

Fixes 5ce555578e091 ("Btrfs: fix deadlock when writing out free space caches")
---
 fs/btrfs/ctree.c |   16 +++++++++-------
 1 file changed, 9 insertions(+), 7 deletions(-)
fs/btrfs/ctree.c +9 -7
@@ -1051,19 +1051,21 @@ static noinline int __btrfs_cow_block(st
 		parent_start = parent->start;
 
 	/*
-	 * If we are COWing a node/leaf from the extent, chunk or device trees,
-	 * make sure that we do not finish block group creation of pending block
-	 * groups. We do this to avoid a deadlock.
+	 * If we are COWing a node/leaf from the extent, chunk, device or free
+	 * space trees, make sure that we do not finish block group creation of
+	 * pending block groups. We do this to avoid a deadlock.
 	 * COWing can result in allocation of a new chunk, and flushing pending
 	 * block groups (btrfs_create_pending_block_groups()) can be triggered
 	 * when finishing allocation of a new chunk. Creation of a pending block
-	 * group modifies the extent, chunk and device trees, therefore we could
-	 * deadlock with ourselves since we are holding a lock on an extent
-	 * buffer that btrfs_create_pending_block_groups() may try to COW later.
+	 * group modifies the extent, chunk, device and free space trees,
+	 * therefore we could deadlock with ourselves since we are holding a
+	 * lock on an extent buffer that btrfs_create_pending_block_groups() may
+	 * try to COW later.
 	 */
 	if (root == fs_info->extent_root ||
 	    root == fs_info->chunk_root ||
-	    root == fs_info->dev_root)
+	    root == fs_info->dev_root ||
+	    root == fs_info->free_space_root)
 		trans->can_flush_pending_bgs = false;
 
 	cow = btrfs_alloc_tree_block(trans, root, parent_start,
author maintainer
#2 lore raw

PATCH 2/50 staging: rtl8188eu: Fix module loading from tasklet for CCMP encryption

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Larry Finger <[email protected]>

commit 84cad97a717f5749a0236abd5ce68da582ea074f upstream.

Commit 6bd082af7e36 ("staging:r8188eu: use lib80211 CCMP decrypt") causes scheduling while atomic bugs followed by a hard freeze whenever the driver tries to connect to a CCMP-encrypted network. Experimentation showed that the freezes were eliminated when module lib80211 was preloaded, which can be forced by calling lib80211_get_crypto_ops() directly rather than indirectly through try_then_request_module(). With this change, no BUG messages are logged.

Fixes 6bd082af7e36 ("staging:r8188eu: use lib80211 CCMP decrypt")
Cc Stable <[email protected]> # v4.17+
Cc Ivan Safonov <[email protected]>
---
 drivers/staging/rtl8188eu/core/rtw_security.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/staging/rtl8188eu/core/rtw_security.c +1 -1
@@ -1292,7 +1292,7 @@ u32 rtw_aes_decrypt(struct adapter *pada
 			struct sk_buff *skb = ((struct recv_frame *)precvframe)->pkt;
 			void *crypto_private = NULL;
 			u8 *key, *pframe = skb->data;
-			struct lib80211_crypto_ops *crypto_ops = try_then_request_module(lib80211_get_crypto_ops("CCMP"), "lib80211_crypt_ccmp");
+			struct lib80211_crypto_ops *crypto_ops = lib80211_get_crypto_ops("CCMP");
 			struct security_priv *psecuritypriv = &padapter->securitypriv;
 			char iv[8], icv[8];
 
author maintainer
#3 lore raw

PATCH 3/50 staging: rtl8188eu: Fix module loading from tasklet for WEP encryption

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Larry Finger <[email protected]>

commit 7775665aadc48a562051834a73519129bf717d73 upstream.

Commit 2b2ea09e74a5 ("staging:r8188eu: Use lib80211 to decrypt WEP-frames") causes scheduling while atomic bugs followed by a hard freeze whenever the driver tries to connect to a WEP-encrypted network. Experimentation showed that the freezes were eliminated when module lib80211 was preloaded, which can be forced by calling lib80211_get_crypto_ops() directly rather than indirectly through try_then_request_module(). With this change, no BUG messages are logged.

Fixes 2b2ea09e74a5 ("staging:r8188eu: Use lib80211 to decrypt WEP-frames")
Cc Stable <[email protected]> # v4.17+
Cc Michael Straube <[email protected]>
Cc Ivan Safonov <[email protected]>
---
 drivers/staging/rtl8188eu/core/rtw_security.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/staging/rtl8188eu/core/rtw_security.c +2 -2
@@ -154,7 +154,7 @@ void rtw_wep_encrypt(struct adapter *pad
 
 	pframe = ((struct xmit_frame *)pxmitframe)->buf_addr + hw_hdr_offset;
 
-	crypto_ops = try_then_request_module(lib80211_get_crypto_ops("WEP"), "lib80211_crypt_wep");
+	crypto_ops = lib80211_get_crypto_ops("WEP");
 
 	if (!crypto_ops)
 		return;
@@ -210,7 +210,7 @@ int rtw_wep_decrypt(struct adapter  *pad
 		void *crypto_private = NULL;
 		int status = _SUCCESS;
 		const int keyindex = prxattrib->key_index;
-		struct lib80211_crypto_ops *crypto_ops = try_then_request_module(lib80211_get_crypto_ops("WEP"), "lib80211_crypt_wep");
+		struct lib80211_crypto_ops *crypto_ops = lib80211_get_crypto_ops("WEP");
 		char iv[4], icv[4];
 
 		if (!crypto_ops) {
author maintainer
#4 lore raw

PATCH 4/50 cpufreq: scmi: Fix frequency invariance in slow path

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Quentin Perret <[email protected]>

commit 0e141d1c65c1dd31c914eb2e11651adcc1a15912 upstream.

The scmi-cpufreq driver calls the arch_set_freq_scale() callback on frequency changes to provide scale-invariant load-tracking signals to the scheduler. However, in the slow path, it does so while specifying the current and max frequencies in different units, hence resulting in a broken freq_scale factor.

Fix this by passing all frequencies in KHz, as stored in the CPUFreq frequency table.

Fixes 99d6bdf33877 (cpufreq: add support for CPU DVFS based on SCMI message protocol)
Cc 4.17+ <[email protected]> # v4.17+
---
 drivers/cpufreq/scmi-cpufreq.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/cpufreq/scmi-cpufreq.c +2 -2
@@ -52,9 +52,9 @@ scmi_cpufreq_set_target(struct cpufreq_p
 	int ret;
 	struct scmi_data *priv = policy->driver_data;
 	struct scmi_perf_ops *perf_ops = handle->perf_ops;
-	u64 freq = policy->freq_table[index].frequency * 1000;
+	u64 freq = policy->freq_table[index].frequency;
 
-	ret = perf_ops->freq_set(handle, priv->domain_id, freq, false);
+	ret = perf_ops->freq_set(handle, priv->domain_id, freq * 1000, false);
 	if (!ret)
 		arch_set_freq_scale(policy->related_cpus, freq,
 				    policy->cpuinfo.max_freq);
author maintainer
#5 lore raw

PATCH 5/50 x86, modpost: Replace last remnants of RETPOLINE with CONFIG_RETPOLINE

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: WANG Chao <[email protected]>

commit e4f358916d528d479c3c12bd2fd03f2d5a576380 upstream.

Commit

4cd24de3a098 ("x86/retpoline: Make CONFIG_RETPOLINE depend on compiler support")

replaced the RETPOLINE define with CONFIG_RETPOLINE checks. Remove the remaining pieces.

[ bp: Massage commit message. ]

Fixes 4cd24de3a098 ("x86/retpoline: Make CONFIG_RETPOLINE depend on compiler support")
Cc "H. Peter Anvin" <[email protected]>
Cc Andi Kleen <[email protected]>
Cc Andrew Morton <[email protected]>
Cc Andy Lutomirski <[email protected]>
Cc Arnd Bergmann <[email protected]>
Cc Daniel Borkmann <[email protected]>
Cc David Woodhouse <[email protected]>
Cc Geert Uytterhoeven <[email protected]>
Cc Jessica Yu <[email protected]>
Cc Jiri Kosina <[email protected]>
Cc Kees Cook <[email protected]>
Cc Konrad Rzeszutek Wilk <[email protected]>
Cc Luc Van Oostenryck <[email protected]>
Cc Michal Marek <[email protected]>
Cc Miguel Ojeda <[email protected]>
Cc Peter Zijlstra <[email protected]>
Cc Tim Chen <[email protected]>
Cc Vasily Gorbik <[email protected]>
Cc stable <[email protected]>
Cc x86-ml <[email protected]>
---
 arch/x86/kernel/cpu/bugs.c   |    2 +-
 include/linux/compiler-gcc.h |    2 +-
 include/linux/module.h       |    2 +-
 scripts/mod/modpost.c        |    2 +-
 4 files changed, 4 insertions(+), 4 deletions(-)
arch/x86/kernel/cpu/bugs.c +1 -1
@@ -213,7 +213,7 @@ static enum spectre_v2_mitigation spectr
 static enum spectre_v2_user_mitigation spectre_v2_user __ro_after_init =
 	SPECTRE_V2_USER_NONE;
 
-#ifdef RETPOLINE
+#ifdef CONFIG_RETPOLINE
 static bool spectre_v2_bad_module;
 
 bool retpoline_module_ok(bool has_retpoline)
include/linux/compiler-gcc.h +1 -1
@@ -75,7 +75,7 @@
 #define __must_be_array(a)	BUILD_BUG_ON_ZERO(__same_type((a), &(a)[0]))
 #endif
 
-#ifdef RETPOLINE
+#ifdef CONFIG_RETPOLINE
 #define __noretpoline __attribute__((indirect_branch("keep")))
 #endif
 
include/linux/module.h +1 -1
@@ -818,7 +818,7 @@ static inline void module_bug_finalize(c
 static inline void module_bug_cleanup(struct module *mod) {}
 #endif	/* CONFIG_GENERIC_BUG */
 
-#ifdef RETPOLINE
+#ifdef CONFIG_RETPOLINE
 extern bool retpoline_module_ok(bool has_retpoline);
 #else
 static inline bool retpoline_module_ok(bool has_retpoline)
scripts/mod/modpost.c +1 -1
@@ -2157,7 +2157,7 @@ static void add_intree_flag(struct buffe
 /* Cannot check for assembler */
 static void add_retpoline(struct buffer *b)
 {
-	buf_printf(b, "\n#ifdef RETPOLINE\n");
+	buf_printf(b, "\n#ifdef CONFIG_RETPOLINE\n");
 	buf_printf(b, "MODULE_INFO(retpoline, \"Y\");\n");
 	buf_printf(b, "#endif\n");
 }
author maintainer
#6 lore raw

PATCH 6/50 ALSA: hda/realtek - Support Dell headset mode for New AIO platform

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Kailang Yang <[email protected]>

commit c2a7c55a04065c3b0c32d23b099db7ea1dbf6250 upstream.

Dell has new platform for ALC274.
This will support to enable headset mode.

---
 sound/pci/hda/patch_realtek.c |    1 +
 1 file changed, 1 insertion(+)
sound/pci/hda/patch_realtek.c +1 -0
@@ -6503,6 +6503,7 @@ static const struct snd_pci_quirk alc269
 	SND_PCI_QUIRK(0x1028, 0x0871, "Dell Precision 3630", ALC255_FIXUP_DELL_HEADSET_MIC),
 	SND_PCI_QUIRK(0x1028, 0x0872, "Dell Precision 3630", ALC255_FIXUP_DELL_HEADSET_MIC),
 	SND_PCI_QUIRK(0x1028, 0x0873, "Dell Precision 3930", ALC255_FIXUP_DUMMY_LINEOUT_VERB),
+	SND_PCI_QUIRK(0x1028, 0x0935, "Dell", ALC274_FIXUP_DELL_AIO_LINEOUT_VERB),
 	SND_PCI_QUIRK(0x1028, 0x164a, "Dell", ALC293_FIXUP_DELL1_MIC_NO_PRESENCE),
 	SND_PCI_QUIRK(0x1028, 0x164b, "Dell", ALC293_FIXUP_DELL1_MIC_NO_PRESENCE),
 	SND_PCI_QUIRK(0x103c, 0x1586, "HP", ALC269_FIXUP_HP_MUTE_LED_MIC2),
author maintainer
#7 lore raw

PATCH 7/50 ALSA: hda/realtek - Add unplug function into unplug state of Headset Mode for ALC225

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Kailang Yang <[email protected]>

commit 4d4b0c52bde470c379f5d168d5c139ad866cb808 upstream.

Forgot to add unplug function to unplug state of headset mode for ALC225.

---
 sound/pci/hda/patch_realtek.c |    1 +
 1 file changed, 1 insertion(+)
sound/pci/hda/patch_realtek.c +1 -0
@@ -4102,6 +4102,7 @@ static void alc_headset_mode_unplugged(s
 	case 0x10ec0295:
 	case 0x10ec0289:
 	case 0x10ec0299:
+		alc_process_coef_fw(codec, alc225_pre_hsmode);
 		alc_process_coef_fw(codec, coef0225);
 		break;
 	case 0x10ec0867:
author maintainer
#8 lore raw

PATCH 8/50 ALSA: hda/realtek - Disable headset Mic VREF for headset mode of ALC225

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Kailang Yang <[email protected]>

commit d1dd42110d2727e81b9265841a62bc84c454c3a2 upstream.

Disable Headset Mic VREF for headset mode of ALC225. This will be controlled by coef bits of headset mode functions.

[ Fixed a compile warning and code simplification -- tiwai ]

---
 sound/pci/hda/patch_realtek.c |   16 +++++++++++++++-
 1 file changed, 15 insertions(+), 1 deletion(-)
sound/pci/hda/patch_realtek.c +15 -1
@@ -5381,6 +5381,13 @@ static void alc285_fixup_invalidate_dacs
 	snd_hda_override_wcaps(codec, 0x03, 0);
 }
 
+static void alc_fixup_disable_mic_vref(struct hda_codec *codec,
+				  const struct hda_fixup *fix, int action)
+{
+	if (action == HDA_FIXUP_ACT_PRE_PROBE)
+		snd_hda_codec_set_pin_target(codec, 0x19, PIN_VREFHIZ);
+}
+
 /* for hda_fixup_thinkpad_acpi() */
 #include "thinkpad_helper.c"
 
@@ -5493,6 +5500,7 @@ enum {
 	ALC293_FIXUP_LENOVO_SPK_NOISE,
 	ALC233_FIXUP_LENOVO_LINE2_MIC_HOTKEY,
 	ALC255_FIXUP_DELL_SPK_NOISE,
+	ALC225_FIXUP_DISABLE_MIC_VREF,
 	ALC225_FIXUP_DELL1_MIC_NO_PRESENCE,
 	ALC295_FIXUP_DISABLE_DAC3,
 	ALC280_FIXUP_HP_HEADSET_MIC,
@@ -6192,6 +6200,12 @@ static const struct hda_fixup alc269_fix
 		.chained = true,
 		.chain_id = ALC255_FIXUP_DELL1_MIC_NO_PRESENCE
 	},
+	[ALC225_FIXUP_DISABLE_MIC_VREF] = {
+		.type = HDA_FIXUP_FUNC,
+		.v.func = alc_fixup_disable_mic_vref,
+		.chained = true,
+		.chain_id = ALC269_FIXUP_DELL1_MIC_NO_PRESENCE
+	},
 	[ALC225_FIXUP_DELL1_MIC_NO_PRESENCE] = {
 		.type = HDA_FIXUP_VERBS,
 		.v.verbs = (const struct hda_verb[]) {
@@ -6201,7 +6215,7 @@ static const struct hda_fixup alc269_fix
 			{}
 		},
 		.chained = true,
-		.chain_id = ALC269_FIXUP_DELL1_MIC_NO_PRESENCE
+		.chain_id = ALC225_FIXUP_DISABLE_MIC_VREF
 	},
 	[ALC280_FIXUP_HP_HEADSET_MIC] = {
 		.type = HDA_FIXUP_FUNC,
author maintainer
#9 lore raw

PATCH 9/50 CIFS: Fix adjustment of credits for MTU requests

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Pavel Shilovsky <[email protected]>

commit b983f7e92348d7e7d091db1b78b7915e9dd3d63a upstream.

Currently for MTU requests we allocate maximum possible credits in advance and then adjust them according to the request size. While we were adjusting the number of credits belonging to the server, we were skipping adjustment of credits belonging to the request. This patch fixes it by setting request credits to CreditCharge field value of SMB2 packet header.

Also ask 1 credit more for async read and write operations to increase parallelism and match the behavior of other operations.

---
 fs/cifs/smb2pdu.c |    8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)
fs/cifs/smb2pdu.c +6 -2
@@ -3185,12 +3185,14 @@ smb2_async_readv(struct cifs_readdata *r
 	if (rdata->credits) {
 		shdr->CreditCharge = cpu_to_le16(DIV_ROUND_UP(rdata->bytes,
 						SMB2_MAX_BUFFER_SIZE));
-		shdr->CreditRequest = shdr->CreditCharge;
+		shdr->CreditRequest =
+			cpu_to_le16(le16_to_cpu(shdr->CreditCharge) + 1);
 		spin_lock(&server->req_lock);
 		server->credits += rdata->credits -
 						le16_to_cpu(shdr->CreditCharge);
 		spin_unlock(&server->req_lock);
 		wake_up(&server->request_q);
+		rdata->credits = le16_to_cpu(shdr->CreditCharge);
 		flags |= CIFS_HAS_CREDITS;
 	}
 
@@ -3462,12 +3464,14 @@ smb2_async_writev(struct cifs_writedata
 	if (wdata->credits) {
 		shdr->CreditCharge = cpu_to_le16(DIV_ROUND_UP(wdata->bytes,
 						    SMB2_MAX_BUFFER_SIZE));
-		shdr->CreditRequest = shdr->CreditCharge;
+		shdr->CreditRequest =
+			cpu_to_le16(le16_to_cpu(shdr->CreditCharge) + 1);
 		spin_lock(&server->req_lock);
 		server->credits += wdata->credits -
 						le16_to_cpu(shdr->CreditCharge);
 		spin_unlock(&server->req_lock);
 		wake_up(&server->request_q);
+		wdata->credits = le16_to_cpu(shdr->CreditCharge);
 		flags |= CIFS_HAS_CREDITS;
 	}
 
author maintainer
#10 lore raw

PATCH 10/50 CIFS: Do not set credits to 1 if the server didnt grant anything

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Pavel Shilovsky <[email protected]>

commit 33fa5c8b8a7dbe6353a56eaa654b790348890d42 upstream.

Currently we reset the number of total credits granted by the server to 1 if the server didn't grant us anything int the response. This violates the SMB3 protocol - we need to trust the server and use the credit values from the response. Fix this by removing the corresponding code.

---
 fs/cifs/transport.c |    2 --
 1 file changed, 2 deletions(-)
fs/cifs/transport.c +0 -2
@@ -885,8 +885,6 @@ compound_send_recv(const unsigned int xi
 	for (i = 0; i < num_rqst; i++)
 		if (midQ[i]->resp_buf)
 			credits += ses->server->ops->get_credits(midQ[i]);
-	if (!credits)
-		credits = 1;
 
 	for (i = 0; i < num_rqst; i++) {
 		if (rc < 0)
author maintainer
#11 lore raw

PATCH 11/50 CIFS: Do not hide EINTR after sending network packets

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Pavel Shilovsky <[email protected]>

commit ee13919c2e8d1f904e035ad4b4239029a8994131 upstream.

Currently we hide EINTR code returned from sock_sendmsg() and return 0 instead. This makes a caller think that we successfully completed the network operation which is not true. Fix this by properly returning EINTR to callers.

---
 fs/cifs/transport.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
fs/cifs/transport.c +1 -1
@@ -378,7 +378,7 @@ smbd_done:
 	if (rc < 0 && rc != -EINTR)
 		cifs_dbg(VFS, "Error %d sending data on socket to server\n",
 			 rc);
-	else
+	else if (rc > 0)
 		rc = 0;
 
 	return rc;
author maintainer
#12 lore raw

PATCH 12/50 CIFS: Fix credit computation for compounded requests

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Pavel Shilovsky <[email protected]>

commit 8544f4aa9dd19a04d1244dae10feecc813ccf175 upstream.

In SMB3 protocol every part of the compound chain consumes credits individually, so we need to call wait_for_free_credits() for each of the PDUs in the chain. If an operation is interrupted, we must ensure we return all credits taken from the server structure back.

Without this patch server can sometimes disconnect the session due to credit mismatches, especially when first operation(s) are large writes.

---
 fs/cifs/transport.c |   59 ++++++++++++++++++++++++++++++++++++----------------
 1 file changed, 41 insertions(+), 18 deletions(-)
fs/cifs/transport.c +41 -18
@@ -786,7 +786,8 @@ compound_send_recv(const unsigned int xi
 	int i, j, rc = 0;
 	int timeout, optype;
 	struct mid_q_entry *midQ[MAX_COMPOUND];
-	unsigned int credits = 0;
+	bool cancelled_mid[MAX_COMPOUND] = {false};
+	unsigned int credits[MAX_COMPOUND] = {0};
 	char *buf;
 
 	timeout = flags & CIFS_TIMEOUT_MASK;
@@ -804,13 +805,31 @@ compound_send_recv(const unsigned int xi
 		return -ENOENT;
 
 	/*
-	 * Ensure that we do not send more than 50 overlapping requests
-	 * to the same server. We may make this configurable later or
-	 * use ses->maxReq.
+	 * Ensure we obtain 1 credit per request in the compound chain.
+	 * It can be optimized further by waiting for all the credits
+	 * at once but this can wait long enough if we don't have enough
+	 * credits due to some heavy operations in progress or the server
+	 * not granting us much, so a fallback to the current approach is
+	 * needed anyway.
 	 */
-	rc = wait_for_free_request(ses->server, timeout, optype);
-	if (rc)
-		return rc;
+	for (i = 0; i < num_rqst; i++) {
+		rc = wait_for_free_request(ses->server, timeout, optype);
+		if (rc) {
+			/*
+			 * We haven't sent an SMB packet to the server yet but
+			 * we already obtained credits for i requests in the
+			 * compound chain - need to return those credits back
+			 * for future use. Note that we need to call add_credits
+			 * multiple times to match the way we obtained credits
+			 * in the first place and to account for in flight
+			 * requests correctly.
+			 */
+			for (j = 0; j < i; j++)
+				add_credits(ses->server, 1, optype);
+			return rc;
+		}
+		credits[i] = 1;
+	}
 
 	/*
 	 * Make sure that we sign in the same order that we send on this socket
@@ -826,8 +845,10 @@ compound_send_recv(const unsigned int xi
 			for (j = 0; j < i; j++)
 				cifs_delete_mid(midQ[j]);
 			mutex_unlock(&ses->server->srv_mutex);
+
 			/* Update # of requests on wire to server */
-			add_credits(ses->server, 1, optype);
+			for (j = 0; j < num_rqst; j++)
+				add_credits(ses->server, credits[j], optype);
 			return PTR_ERR(midQ[i]);
 		}
 
@@ -874,17 +895,16 @@ compound_send_recv(const unsigned int xi
 			if (midQ[i]->mid_state == MID_REQUEST_SUBMITTED) {
 				midQ[i]->mid_flags |= MID_WAIT_CANCELLED;
 				midQ[i]->callback = DeleteMidQEntry;
-				spin_unlock(&GlobalMid_Lock);
-				add_credits(ses->server, 1, optype);
-				return rc;
+				cancelled_mid[i] = true;
 			}
 			spin_unlock(&GlobalMid_Lock);
 		}
 	}
 
 	for (i = 0; i < num_rqst; i++)
-		if (midQ[i]->resp_buf)
-			credits += ses->server->ops->get_credits(midQ[i]);
+		if (!cancelled_mid[i] && midQ[i]->resp_buf
+		    && (midQ[i]->mid_state == MID_RESPONSE_RECEIVED))
+			credits[i] = ses->server->ops->get_credits(midQ[i]);
 
 	for (i = 0; i < num_rqst; i++) {
 		if (rc < 0)
@@ -892,8 +912,9 @@ compound_send_recv(const unsigned int xi
 
 		rc = cifs_sync_mid_result(midQ[i], ses->server);
 		if (rc != 0) {
-			add_credits(ses->server, credits, optype);
-			return rc;
+			/* mark this mid as cancelled to not free it below */
+			cancelled_mid[i] = true;
+			goto out;
 		}
 
 		if (!midQ[i]->resp_buf ||
@@ -940,9 +961,11 @@ out:
 	 * This is prevented above by using a noop callback that will not
 	 * wake this thread except for the very last PDU.
 	 */
-	for (i = 0; i < num_rqst; i++)
-		cifs_delete_mid(midQ[i]);
-	add_credits(ses->server, credits, optype);
+	for (i = 0; i < num_rqst; i++) {
+		if (!cancelled_mid[i])
+			cifs_delete_mid(midQ[i]);
+		add_credits(ses->server, credits[i], optype);
+	}
 
 	return rc;
 }
author maintainer
#13 lore raw

PATCH 13/50 cifs: Fix potential OOB access of lock element array

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Ross Lagerwall <[email protected]>

commit b9a74cde94957d82003fb9f7ab4777938ca851cd upstream.

If maxBuf is small but non-zero, it could result in a zero sized lock element array which we would then try and access OOB.

---
 fs/cifs/file.c     |    8 ++++----
 fs/cifs/smb2file.c |    4 ++--
 2 files changed, 6 insertions(+), 6 deletions(-)
fs/cifs/file.c +4 -4
@@ -1120,10 +1120,10 @@ cifs_push_mandatory_locks(struct cifsFil
 
 	/*
 	 * Accessing maxBuf is racy with cifs_reconnect - need to store value
-	 * and check it for zero before using.
+	 * and check it before using.
 	 */
 	max_buf = tcon->ses->server->maxBuf;
-	if (!max_buf) {
+	if (max_buf < (sizeof(struct smb_hdr) + sizeof(LOCKING_ANDX_RANGE))) {
 		free_xid(xid);
 		return -EINVAL;
 	}
@@ -1460,10 +1460,10 @@ cifs_unlock_range(struct cifsFileInfo *c
 
 	/*
 	 * Accessing maxBuf is racy with cifs_reconnect - need to store value
-	 * and check it for zero before using.
+	 * and check it before using.
 	 */
 	max_buf = tcon->ses->server->maxBuf;
-	if (!max_buf)
+	if (max_buf < (sizeof(struct smb_hdr) + sizeof(LOCKING_ANDX_RANGE)))
 		return -EINVAL;
 
 	max_num = (max_buf - sizeof(struct smb_hdr)) /
fs/cifs/smb2file.c +2 -2
@@ -122,10 +122,10 @@ smb2_unlock_range(struct cifsFileInfo *c
 
 	/*
 	 * Accessing maxBuf is racy with cifs_reconnect - need to store value
-	 * and check it for zero before using.
+	 * and check it before using.
 	 */
 	max_buf = tcon->ses->server->maxBuf;
-	if (!max_buf)
+	if (max_buf < sizeof(struct smb2_lock_element))
 		return -EINVAL;
 
 	max_num = max_buf / sizeof(struct smb2_lock_element);
author maintainer
#14 lore raw

PATCH 14/50 usb: cdc-acm: send ZLP for Telit 3G Intel based modems

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Daniele Palmas <[email protected]>

commit 34aabf918717dd14e05051896aaecd3b16b53d95 upstream.

Telit 3G Intel based modems require zero packet to be sent if out data size is equal to the endpoint max packet size.

---
 drivers/usb/class/cdc-acm.c |    7 +++++++
 1 file changed, 7 insertions(+)
drivers/usb/class/cdc-acm.c +7 -0
@@ -1880,6 +1880,13 @@ static const struct usb_device_id acm_id
 	.driver_info = IGNORE_DEVICE,
 	},
 
+	{ USB_DEVICE(0x1bc7, 0x0021), /* Telit 3G ACM only composition */
+	.driver_info = SEND_ZERO_PACKET,
+	},
+	{ USB_DEVICE(0x1bc7, 0x0023), /* Telit 3G ACM + ECM composition */
+	.driver_info = SEND_ZERO_PACKET,
+	},
+
 	/* control interfaces without any protocol set */
 	{ USB_INTERFACE_INFO(USB_CLASS_COMM, USB_CDC_SUBCLASS_ACM,
 		USB_CDC_PROTO_NONE) },
author maintainer
#15 lore raw

PATCH 15/50 USB: storage: dont insert sane sense for SPC3+ when bad sense specified

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Icenowy Zheng <[email protected]>

commit c5603d2fdb424849360fe7e3f8c1befc97571b8c upstream.

Currently the code will set US_FL_SANE_SENSE flag unconditionally if device claims SPC3+, however we should allow US_FL_BAD_SENSE flag to prevent this behavior, because SMI SM3350 UFS-USB bridge controller, which claims SPC4, will show strange behavior with 96-byte sense (put the chip into a wrong state that cannot read/write anything).

Check the presence of US_FL_BAD_SENSE when assuming US_FL_SANE_SENSE on SPC4+ devices.

---
 drivers/usb/storage/scsiglue.c |    8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)
drivers/usb/storage/scsiglue.c +6 -2
@@ -235,8 +235,12 @@ static int slave_configure(struct scsi_d
 		if (!(us->fflags & US_FL_NEEDS_CAP16))
 			sdev->try_rc_10_first = 1;
 
-		/* assume SPC3 or latter devices support sense size > 18 */
-		if (sdev->scsi_level > SCSI_SPC_2)
+		/*
+		 * assume SPC3 or latter devices support sense size > 18
+		 * unless US_FL_BAD_SENSE quirk is specified.
+		 */
+		if (sdev->scsi_level > SCSI_SPC_2 &&
+		    !(us->fflags & US_FL_BAD_SENSE))
 			us->fflags |= US_FL_SANE_SENSE;
 
 		/*
author maintainer
#16 lore raw

PATCH 16/50 USB: storage: add quirk for SMI SM3350

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Icenowy Zheng <[email protected]>

commit 0a99cc4b8ee83885ab9f097a3737d1ab28455ac0 upstream.

The SMI SM3350 USB-UFS bridge controller cannot handle long sense request correctly and will make the chip refuse to do read/write when requested long sense.

Add a bad sense quirk for it.

---
 drivers/usb/storage/unusual_devs.h |   12 ++++++++++++
 1 file changed, 12 insertions(+)
drivers/usb/storage/unusual_devs.h +12 -0
@@ -1266,6 +1266,18 @@ UNUSUAL_DEV( 0x090c, 0x1132, 0x0000, 0xf
 		US_FL_FIX_CAPACITY ),
 
 /*
+ * Reported by Icenowy Zheng <[email protected]>
+ * The SMI SM3350 USB-UFS bridge controller will enter a wrong state
+ * that do not process read/write command if a long sense is requested,
+ * so force to use 18-byte sense.
+ */
+UNUSUAL_DEV(  0x090c, 0x3350, 0x0000, 0xffff,
+		"SMI",
+		"SM3350 UFS-to-USB-Mass-Storage bridge",
+		USB_SC_DEVICE, USB_PR_DEVICE, NULL,
+		US_FL_BAD_SENSE ),
+
+/*
  * Reported by Paul Hartman <[email protected]>
  * This card reader returns "Illegal Request, Logical Block Address
  * Out of Range" for the first READ(10) after a new card is inserted.
author maintainer
#17 lore raw

PATCH 17/50 USB: Add USB_QUIRK_DELAY_CTRL_MSG quirk for Corsair K70 RGB

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Jack Stocker <[email protected]>

commit 3483254b89438e60f719937376c5e0ce2bc46761 upstream.

To match the Corsair Strafe RGB, the Corsair K70 RGB also requires USB_QUIRK_DELAY_CTRL_MSG to completely resolve boot connection issues discussed here: https://github.com/ckb-next/ckb-next/issues/42.
Otherwise roughly 1 in 10 boots the keyboard will fail to be detected.

Patch that applied delay control quirk for Corsair Strafe RGB:
cb88a0588717 ("usb: quirks: add control message delay for 1b1c:1b20")

Previous K70 RGB patch to add delay-init quirk:
7a1646d92257 ("Add delay-init quirk for Corsair K70 RGB keyboards")

---
 drivers/usb/core/quirks.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/usb/core/quirks.c +2 -1
@@ -394,7 +394,8 @@ static const struct usb_device_id usb_qu
 	{ USB_DEVICE(0x1a40, 0x0101), .driver_info = USB_QUIRK_HUB_SLOW_RESET },
 
 	/* Corsair K70 RGB */
-	{ USB_DEVICE(0x1b1c, 0x1b13), .driver_info = USB_QUIRK_DELAY_INIT },
+	{ USB_DEVICE(0x1b1c, 0x1b13), .driver_info = USB_QUIRK_DELAY_INIT |
+	  USB_QUIRK_DELAY_CTRL_MSG },
 
 	/* Corsair Strafe */
 	{ USB_DEVICE(0x1b1c, 0x1b15), .driver_info = USB_QUIRK_DELAY_INIT |
author maintainer
#18 lore raw

PATCH 18/50 slab: alien caches must not be initialized if the allocation of the alien cache failed

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Christoph Lameter <[email protected]>

commit 09c2e76ed734a1d36470d257a778aaba28e86531 upstream.

Callers of __alloc_alien() check for NULL. We must do the same check in __alloc_alien_cache to avoid NULL pointer dereferences on allocation failures.

Fixes 49dfc304ba241 ("slab: use the lock on alien_cache, instead of the lock on array_cache")
Fixes c8522a3a5832b ("Slab: introduce alloc_alien")
Reported-by [email protected]
Cc Pekka Enberg <[email protected]>
Cc David Rientjes <[email protected]>
Cc Joonsoo Kim <[email protected]>
---
 mm/slab.c |    6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
mm/slab.c +4 -2
@@ -679,8 +679,10 @@ static struct alien_cache *__alloc_alien
 	struct alien_cache *alc = NULL;
 
 	alc = kmalloc_node(memsize, gfp, node);
-	init_arraycache(&alc->ac, entries, batch);
-	spin_lock_init(&alc->lock);
+	if (alc) {
+		init_arraycache(&alc->ac, entries, batch);
+		spin_lock_init(&alc->lock);
+	}
 	return alc;
 }
 
author maintainer
#19 lore raw

PATCH 19/50 mm/usercopy.c: no check page span for stack objects

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Qian Cai <[email protected]>

commit 7bff3c06997374fb9b9991536a547b840549a813 upstream.

It is easy to trigger this with CONFIG_HARDENED_USERCOPY_PAGESPAN=y,

usercopy: Kernel memory overwrite attempt detected to spans multiple pages (offset 0, size 23)! kernel BUG at mm/usercopy.c:102!

For example,

print_worker_info
char name[WQ_NAME_LEN] = { };
char desc[WORKER_DESC_LEN] = { };
probe_kernel_read(name, wq->name, sizeof(name) - 1); probe_kernel_read(desc, worker->desc, sizeof(desc) - 1);
__copy_from_user_inatomic
check_object_size
check_heap_object
check_page_span

This is because on-stack variables could cross PAGE_SIZE boundary, and failed this check,

if (likely(((unsigned long)ptr & (unsigned long)PAGE_MASK) ==
((unsigned long)end & (unsigned long)PAGE_MASK)))

ptr = FFFF889007D7EFF8
end = FFFF889007D7F00E

Hence, fix it by checking if it is a stack object first.

---
 mm/usercopy.c |    9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)
mm/usercopy.c +5 -4
@@ -247,7 +247,8 @@ static DEFINE_STATIC_KEY_FALSE_RO(bypass
 /*
  * Validates that the given object is:
  * - not bogus address
- * - known-safe heap or stack object
+ * - fully contained by stack (or stack frame, when available)
+ * - fully within SLAB object (or object whitelist area, when available)
  * - not in kernel text
  */
 void __check_object_size(const void *ptr, unsigned long n, bool to_user)
@@ -262,9 +263,6 @@ void __check_object_size(const void *ptr
 	/* Check for invalid addresses. */
 	check_bogus_address((const unsigned long)ptr, n, to_user);
 
-	/* Check for bad heap object. */
-	check_heap_object(ptr, n, to_user);
-
 	/* Check for bad stack object. */
 	switch (check_stack_object(ptr, n)) {
 	case NOT_STACK:
@@ -282,6 +280,9 @@ void __check_object_size(const void *ptr
 		usercopy_abort("process stack", NULL, to_user, 0, n);
 	}
 
+	/* Check for bad heap object. */
+	check_heap_object(ptr, n, to_user);
+
 	/* Check for object in kernel to avoid text exposure. */
 	check_kernel_text_object((const unsigned long)ptr, n, to_user);
 }
author maintainer
#20 lore raw

PATCH 20/50 mm, memcg: fix reclaim deadlock with writeback

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Michal Hocko <[email protected]>

commit 63f3655f950186752236bb88a22f8252c11ce394 upstream.

Liu Bo has experienced a deadlock between memcg (legacy) reclaim and the ext4 writeback

task1:

    wait_on_page_bit+0x82/0xa0
    shrink_page_list+0x907/0x960
    shrink_inactive_list+0x2c7/0x680
    shrink_node_memcg+0x404/0x830
    shrink_node+0xd8/0x300
    do_try_to_free_pages+0x10d/0x330
    try_to_free_mem_cgroup_pages+0xd5/0x1b0
    try_charge+0x14d/0x720
    memcg_kmem_charge_memcg+0x3c/0xa0
    memcg_kmem_charge+0x7e/0xd0
    __alloc_pages_nodemask+0x178/0x260
    alloc_pages_current+0x95/0x140
    pte_alloc_one+0x17/0x40
    __pte_alloc+0x1e/0x110
    alloc_set_pte+0x5fe/0xc20
    do_fault+0x103/0x970
    handle_mm_fault+0x61e/0xd10
    __do_page_fault+0x252/0x4d0
    do_page_fault+0x30/0x80
    page_fault+0x28/0x30

task2:

    __lock_page+0x86/0xa0
    mpage_prepare_extent_to_map+0x2e7/0x310 [ext4]
    ext4_writepages+0x479/0xd60
    do_writepages+0x1e/0x30
    __writeback_single_inode+0x45/0x320
    writeback_sb_inodes+0x272/0x600
    __writeback_inodes_wb+0x92/0xc0
    wb_writeback+0x268/0x300
    wb_workfn+0xb4/0x390
    process_one_work+0x189/0x420
    worker_thread+0x4e/0x4b0
    kthread+0xe6/0x100
    ret_from_fork+0x41/0x50

He adds
"task1 is waiting for the PageWriteback bit of the page that task2 has
collected in mpd->io_submit->io_bio, and tasks2 is waiting for the LOCKED bit the page which tasks1 has locked"

More precisely task1 is handling a page fault and it has a page locked while it charges a new page table to a memcg. That in turn hits a memory limit reclaim and the memcg reclaim for legacy controller is waiting on the writeback but that is never going to finish because the writeback itself is waiting for the page locked in the #PF path. So this is essentially ABBA deadlock:

lock_page(A) SetPageWriteback(A) unlock_page(A)
lock_page(B)
lock_page(B)
pte_alloc_pne
shrink_page_list
wait_on_page_writeback(A)
SetPageWriteback(B) unlock_page(B)

# flush A, B to clear the writeback

This accumulating of more pages to flush is used by several filesystems to generate a more optimal IO patterns.

Waiting for the writeback in legacy memcg controller is a workaround for pre-mature OOM killer invocations because there is no dirty IO throttling available for the controller. There is no easy way around that unfortunately. Therefore fix this specific issue by pre-allocating the page table outside of the page lock. We have that handy infrastructure for that already so simply reuse the fault-around pattern which already does this.

There are probably other hidden __GFP_ACCOUNT | GFP_KERNEL allocations from under a fs page locked but they should be really rare. I am not aware of a better solution unfortunately.

[[email protected]: fix mm/memory.c:__do_fault()]
[[email protected]: coding-style fixes]
[[email protected]: enhance comment, per Johannes]
Fixes c3b94f44fcb0 ("memcg: further prevent OOM with too many dirty pages")
Cc Jan Kara <[email protected]>
Cc Dave Chinner <[email protected]>
Cc Theodore Ts'o <[email protected]>
Cc Vladimir Davydov <[email protected]>
Cc Shakeel Butt <[email protected]>
---
 mm/memory.c |   23 +++++++++++++++++++++++
 1 file changed, 23 insertions(+)
mm/memory.c +23 -0
@@ -3237,6 +3237,29 @@ static vm_fault_t __do_fault(struct vm_f
 	struct vm_area_struct *vma = vmf->vma;
 	vm_fault_t ret;
 
+	/*
+	 * Preallocate pte before we take page_lock because this might lead to
+	 * deadlocks for memcg reclaim which waits for pages under writeback:
+	 *				lock_page(A)
+	 *				SetPageWriteback(A)
+	 *				unlock_page(A)
+	 * lock_page(B)
+	 *				lock_page(B)
+	 * pte_alloc_pne
+	 *   shrink_page_list
+	 *     wait_on_page_writeback(A)
+	 *				SetPageWriteback(B)
+	 *				unlock_page(B)
+	 *				# flush A, B to clear the writeback
+	 */
+	if (pmd_none(*vmf->pmd) && !vmf->prealloc_pte) {
+		vmf->prealloc_pte = pte_alloc_one(vmf->vma->vm_mm,
+						  vmf->address);
+		if (!vmf->prealloc_pte)
+			return VM_FAULT_OOM;
+		smp_wmb(); /* See comment in __pte_alloc() */
+	}
+
 	ret = vma->vm_ops->fault(vmf);
 	if (unlikely(ret & (VM_FAULT_ERROR | VM_FAULT_NOPAGE | VM_FAULT_RETRY |
 			    VM_FAULT_DONE_COW)))
author maintainer
#21 lore raw

PATCH 21/50 ACPI: power: Skip duplicate power resource references in _PRx

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Hans de Goede <[email protected]>

commit 7d7b467cb95bf29597b417d4990160d4ea6d69b9 upstream.

Some ACPI tables contain duplicate power resource references like this:

        Name (_PR0, Package (0x04)  // _PR0: Power Resources for D0
        {
            P28P,
            P18P,
            P18P,
            CLK4
        })

This causes a WARN_ON in sysfs_add_link_to_group() because we end up adding a link to the same acpi_device twice:

sysfs: cannot create duplicate filename '/devices/LNXSYSTM:00/LNXSYBUS:00/PNP0A08:00/808622C1:00/OVTI2680:00/power_resources_D0/LNXPOWER:0a'

CPU: 0 PID: 1 Comm: swapper/0 Not tainted 4.19.12-301.fc29.x86_64 #1
Hardware name: Insyde CherryTrail/Type2 - Board Product Name, BIOS jumperx.T87.KFBNEEA02 04/13/2016
Call Trace:
 dump_stack+0x5c/0x80
 sysfs_warn_dup.cold.3+0x17/0x2a
 sysfs_do_create_link_sd.isra.2+0xa9/0xb0
 sysfs_add_link_to_group+0x30/0x50
 acpi_power_expose_list+0x74/0xa0
 acpi_power_add_remove_device+0x50/0xa0
 acpi_add_single_object+0x26b/0x5f0
 acpi_bus_check_add+0xc4/0x250
 ...

To address this issue, make acpi_extract_power_resources() check for duplicates and simply skip them when found.

Cc All applicable <[email protected]>
[ rjw: Subject & changelog, comments ]
---
 drivers/acpi/power.c |   22 ++++++++++++++++++++++
 1 file changed, 22 insertions(+)
drivers/acpi/power.c +22 -0
@@ -131,6 +131,23 @@ void acpi_power_resources_list_free(stru
 	}
 }
 
+static bool acpi_power_resource_is_dup(union acpi_object *package,
+				       unsigned int start, unsigned int i)
+{
+	acpi_handle rhandle, dup;
+	unsigned int j;
+
+	/* The caller is expected to check the package element types */
+	rhandle = package->package.elements[i].reference.handle;
+	for (j = start; j < i; j++) {
+		dup = package->package.elements[j].reference.handle;
+		if (dup == rhandle)
+			return true;
+	}
+
+	return false;
+}
+
 int acpi_extract_power_resources(union acpi_object *package, unsigned int start,
 				 struct list_head *list)
 {
@@ -150,6 +167,11 @@ int acpi_extract_power_resources(union a
 			err = -ENODEV;
 			break;
 		}
+
+		/* Some ACPI tables contain duplicate power resource references */
+		if (acpi_power_resource_is_dup(package, start, i))
+			continue;
+
 		err = acpi_add_power_resource(rhandle);
 		if (err)
 			break;
author maintainer
#22 lore raw

PATCH 22/50 ACPI / PMIC: xpower: Fix TS-pin current-source handling

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Hans de Goede <[email protected]>

commit 2b531d71595d2b5b12782a49b23c335869e2621e upstream.

The current-source used for the battery temp-sensor (TS) is shared with the GPADC. For proper fuel-gauge and charger operation the TS current-source needs to be permanently on. But to read the GPADC we need to temporary switch the TS current-source to ondemand, so that the GPADC can use it, otherwise we will always read an all 0 value.

The switching from on to on-ondemand is not necessary when the TS current-source is off (this happens on devices which do not have a TS).

Prior to this commit there were 2 issues with our handling of the TS current-source switching:

1) We were writing hardcoded values to the ADC TS pin-ctrl register, overwriting various other unrelated bits. Specifically we were overwriting the current-source setting for the TS and GPIO0 pins, forcing it to 80ųA independent of its original setting. On a Chuwi Vi10 tablet this was causing us to get a too high adc value (due to a too high current-source) resulting in acpi_lpat_raw_to_temp() returning -ENOENT, resulting in:

ACPI Error: AE_ERROR, Returned by Handler for [UserDefinedRegion] ACPI Error: Method parse/execution failed \_SB.SXP1._TMP, AE_ERROR

This commit fixes this by using regmap_update_bits to change only the relevant bits.

2) At the end of intel_xpower_pmic_get_raw_temp() we were unconditionally enabling the TS current-source even on devices where the TS-pin is not used and the current-source thus was off on entry of the function.

This commit fixes this by checking if the TS current-source is off when entering intel_xpower_pmic_get_raw_temp() and if so it is left as is.

Fixes 58eefe2f3f53 (ACPI / PMIC: xpower: Do pinswitch ... reading GPADC)
Cc 4.14+ <[email protected]> # 4.14+
---
 drivers/acpi/pmic/intel_pmic_xpower.c |   41 +++++++++++++++++++++++++++-------
 1 file changed, 33 insertions(+), 8 deletions(-)
drivers/acpi/pmic/intel_pmic_xpower.c +33 -8
@@ -27,8 +27,11 @@
 #define GPI1_LDO_ON		(3 << 0)
 #define GPI1_LDO_OFF		(4 << 0)
 
-#define AXP288_ADC_TS_PIN_GPADC	0xf2
-#define AXP288_ADC_TS_PIN_ON	0xf3
+#define AXP288_ADC_TS_CURRENT_ON_OFF_MASK		GENMASK(1, 0)
+#define AXP288_ADC_TS_CURRENT_OFF			(0 << 0)
+#define AXP288_ADC_TS_CURRENT_ON_WHEN_CHARGING		(1 << 0)
+#define AXP288_ADC_TS_CURRENT_ON_ONDEMAND		(2 << 0)
+#define AXP288_ADC_TS_CURRENT_ON			(3 << 0)
 
 static struct pmic_table power_table[] = {
 	{
@@ -211,22 +214,44 @@ static int intel_xpower_pmic_update_powe
  */
 static int intel_xpower_pmic_get_raw_temp(struct regmap *regmap, int reg)
 {
+	int ret, adc_ts_pin_ctrl;
 	u8 buf[2];
-	int ret;
 
-	ret = regmap_write(regmap, AXP288_ADC_TS_PIN_CTRL,
-			   AXP288_ADC_TS_PIN_GPADC);
+	/*
+	 * The current-source used for the battery temp-sensor (TS) is shared
+	 * with the GPADC. For proper fuel-gauge and charger operation the TS
+	 * current-source needs to be permanently on. But to read the GPADC we
+	 * need to temporary switch the TS current-source to ondemand, so that
+	 * the GPADC can use it, otherwise we will always read an all 0 value.
+	 *
+	 * Note that the switching from on to on-ondemand is not necessary
+	 * when the TS current-source is off (this happens on devices which
+	 * do not use the TS-pin).
+	 */
+	ret = regmap_read(regmap, AXP288_ADC_TS_PIN_CTRL, &adc_ts_pin_ctrl);
 	if (ret)
 		return ret;
 
-	/* After switching to the GPADC pin give things some time to settle */
-	usleep_range(6000, 10000);
+	if (adc_ts_pin_ctrl & AXP288_ADC_TS_CURRENT_ON_OFF_MASK) {
+		ret = regmap_update_bits(regmap, AXP288_ADC_TS_PIN_CTRL,
+					 AXP288_ADC_TS_CURRENT_ON_OFF_MASK,
+					 AXP288_ADC_TS_CURRENT_ON_ONDEMAND);
+		if (ret)
+			return ret;
+
+		/* Wait a bit after switching the current-source */
+		usleep_range(6000, 10000);
+	}
 
 	ret = regmap_bulk_read(regmap, AXP288_GP_ADC_H, buf, 2);
 	if (ret == 0)
 		ret = (buf[0] << 4) + ((buf[1] >> 4) & 0x0f);
 
-	regmap_write(regmap, AXP288_ADC_TS_PIN_CTRL, AXP288_ADC_TS_PIN_ON);
+	if (adc_ts_pin_ctrl & AXP288_ADC_TS_CURRENT_ON_OFF_MASK) {
+		regmap_update_bits(regmap, AXP288_ADC_TS_PIN_CTRL,
+				   AXP288_ADC_TS_CURRENT_ON_OFF_MASK,
+				   AXP288_ADC_TS_CURRENT_ON);
+	}
 
 	return ret;
 }
author maintainer
#23 lore raw

PATCH 23/50 ACPI/IORT: Fix rc_dma_get_range()

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Jean-Philippe Brucker <[email protected]>

commit c7777236dd8f587f6a8d6800c03df318fd4d2627 upstream.

When executed for a PCI_ROOT_COMPLEX type, iort_match_node_callback() expects the opaque pointer argument to be a PCI bus device. At the moment rc_dma_get_range() passes the PCI endpoint instead of the bus, and we've been lucky to have pci_domain_nr(ptr) return 0 instead of crashing. Pass the bus device to iort_scan_node().

Fixes 5ac65e8c8941 ("ACPI/IORT: Support address size limit for root complexes")
Cc Will Deacon <[email protected]>
Cc Hanjun Guo <[email protected]>
Cc Sudeep Holla <[email protected]>
Cc Catalin Marinas <[email protected]>
Cc "Rafael J. Wysocki" <[email protected]>
---
 drivers/acpi/arm64/iort.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/acpi/arm64/iort.c +2 -1
@@ -951,9 +951,10 @@ static int rc_dma_get_range(struct devic
 {
 	struct acpi_iort_node *node;
 	struct acpi_iort_root_complex *rc;
+	struct pci_bus *pbus = to_pci_dev(dev)->bus;
 
 	node = iort_scan_node(ACPI_IORT_NODE_PCI_ROOT_COMPLEX,
-			      iort_match_node_callback, dev);
+			      iort_match_node_callback, &pbus->dev);
 	if (!node || node->revision < 1)
 		return -ENODEV;
 
author maintainer
#24 lore raw

PATCH 24/50 i2c: dev: prevent adapter retries and timeout being set as minus value

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Yi Zeng <[email protected]>

commit 6ebec961d59bccf65d08b13fc1ad4e6272a89338 upstream.

If adapter->retries is set to a minus value from user space via ioctl, it will make __i2c_transfer and __i2c_smbus_xfer skip the calling to adapter->algo->master_xfer and adapter->algo->smbus_xfer that is registered by the underlying bus drivers, and return value 0 to all the callers. The bus driver will never be accessed anymore by all users, besides, the users may still get successful return value without any error or information log print out.

If adapter->timeout is set to minus value from user space via ioctl, it will make the retrying loop in __i2c_transfer and __i2c_smbus_xfer always break after the the first try, due to the time_after always returns true.

[wsa: minor grammar updates to commit message]
---
 drivers/i2c/i2c-dev.c |    6 ++++++
 1 file changed, 6 insertions(+)
drivers/i2c/i2c-dev.c +6 -0
@@ -470,9 +470,15 @@ static long i2cdev_ioctl(struct file *fi
 					  data_arg.data);
 	}
 	case I2C_RETRIES:
+		if (arg > INT_MAX)
+			return -EINVAL;
+
 		client->adapter->retries = arg;
 		break;
 	case I2C_TIMEOUT:
+		if (arg > INT_MAX)
+			return -EINVAL;
+
 		/* For historical reasons, user-space sets the timeout
 		 * value in units of 10 ms.
 		 */
author maintainer
#25 lore raw

PATCH 25/50 mtd: rawnand: qcom: fix memory corruption that causes panic

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Christian Lamparter <[email protected]>

commit 81d9bdf59092e4755fc4307c93c4589ef0fe2e0f upstream.

This patch fixes a memory corruption that occurred in the qcom-nandc driver since it was converted to nand_scan().

On boot, an affected device will panic from a NPE at a weird place:

| Unable to handle kernel NULL pointer dereference at virtual address 0
| pgd = (ptrval)
| [00000000] *pgd=00000000
| Internal error: Oops: 80000005 [#1] SMP ARM
| CPU: 0 PID: 1 Comm: swapper/0 Not tainted 4.19.9 #0
| Hardware name: Generic DT based system
| PC is at   (null)
| LR is at nand_block_isbad+0x90/0xa4
| pc : [<00000000>]    lr : [<c0592240>]    psr: 80000013
| sp : cf839d40  ip : 00000000  fp : cfae9e20
| r10: cf815810  r9 : 00000000  r8 : 00000000
| r7 : 00000000  r6 : 00000000  r5 : 00000001  r4 : cf815810
| r3 : 00000000  r2 : cfae9810  r1 : ffffffff  r0 : cf815810
| Flags: Nzcv  IRQs on  FIQs on  Mode SVC_32  ISA ARM  Segment none
| Control: 10c5387d  Table: 8020406a  DAC: 00000051
| Process swapper/0 (pid: 1, stack limit = 0x(ptrval))
| [<c0592240>] (nand_block_isbad) from [<c0580a94>]
| [<c0580a94>] (allocate_partition) from [<c05811e4>]
| [<c05811e4>] (add_mtd_partitions) from [<c0581164>]
| [<c0581164>] (parse_mtd_partitions) from [<c057def4>]
| [<c057def4>] (mtd_device_parse_register) from [<c059d274>]
| [<c059d274>] (qcom_nandc_probe) from [<c0567f00>]

The problem is that the nand_scan()'s qcom_nand_attach_chip callback is updating the nandc->max_cwperpage from 1 to 4. This causes the sg_init_table of clear_bam_transaction() in the driver's qcom_nandc_block_bad() to memset much more than what was initially allocated by alloc_bam_transaction().

This patch restores the old behavior by reallocating the shared bam transaction alloc_bam_transaction() after the chip was identified, but before mtd_device_parse_register() (which is an alias for mtd_device_register() - see panic) gets called. This fixes the corruption and the driver is working again.

Fixes 6a3cec64f18c ("mtd: rawnand: qcom: convert driver to nand_scan()")
---
 drivers/mtd/nand/raw/qcom_nandc.c |   20 ++++++++++----------
 1 file changed, 10 insertions(+), 10 deletions(-)
drivers/mtd/nand/raw/qcom_nandc.c +10 -10
@@ -2839,6 +2839,16 @@ static int qcom_nand_host_init_and_regis
 	if (ret)
 		return ret;
 
+	if (nandc->props->is_bam) {
+		free_bam_transaction(nandc);
+		nandc->bam_txn = alloc_bam_transaction(nandc);
+		if (!nandc->bam_txn) {
+			dev_err(nandc->dev,
+				"failed to allocate bam transaction\n");
+			return -ENOMEM;
+		}
+	}
+
 	ret = mtd_device_register(mtd, NULL, 0);
 	if (ret)
 		nand_cleanup(chip);
@@ -2853,16 +2863,6 @@ static int qcom_probe_nand_devices(struc
 	struct qcom_nand_host *host;
 	int ret;
 
-	if (nandc->props->is_bam) {
-		free_bam_transaction(nandc);
-		nandc->bam_txn = alloc_bam_transaction(nandc);
-		if (!nandc->bam_txn) {
-			dev_err(nandc->dev,
-				"failed to allocate bam transaction\n");
-			return -ENOMEM;
-		}
-	}
-
 	for_each_available_child_of_node(dn, child) {
 		host = devm_kzalloc(dev, sizeof(*host), GFP_KERNEL);
 		if (!host) {
author maintainer
#26 lore raw

PATCH 26/50 vfio/type1: Fix unmap overflow off-by-one

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Alex Williamson <[email protected]>

commit 58fec830fc19208354895d9832785505046d6c01 upstream.

The below referenced commit adds a test for integer overflow, but in doing so prevents the unmap ioctl from ever including the last page of the address space. Subtract one to compare to the last address of the unmap to avoid the overflow and wrap-around.

---
 drivers/vfio/vfio_iommu_type1.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/vfio/vfio_iommu_type1.c +1 -1
@@ -878,7 +878,7 @@ static int vfio_dma_do_unmap(struct vfio
 		return -EINVAL;
 	if (!unmap->size || unmap->size & mask)
 		return -EINVAL;
-	if (unmap->iova + unmap->size < unmap->iova ||
+	if (unmap->iova + unmap->size - 1 < unmap->iova ||
 	    unmap->size > SIZE_MAX)
 		return -EINVAL;
 
author maintainer
#27 lore raw

PATCH 27/50 drm/amdgpu: Add new VegaM pci id

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Alex Deucher <[email protected]>

commit f6653a0e0877572c87f6dab5351e7bd6b6b7100c upstream.

Add a new pci id.

---
 drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c |    1 +
 1 file changed, 1 insertion(+)
drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +1 -0
@@ -753,6 +753,7 @@ static const struct pci_device_id pciidl
 	/* VEGAM */
 	{0x1002, 0x694C, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGAM},
 	{0x1002, 0x694E, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGAM},
+	{0x1002, 0x694F, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGAM},
 	/* Vega 10 */
 	{0x1002, 0x6860, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA10},
 	{0x1002, 0x6861, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA10},
author maintainer
#28 lore raw

PATCH 28/50 PCI: dwc: Use interrupt masking instead of disabling

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Marc Zyngier <[email protected]>

commit 830920e065e90db318a0da98bf13a02b641eae7f upstream.

The dwc driver is showing an interesting level of brokeness, as it insists on using the enable/disable set of registers to mask/unmask MSIs, meaning that an MSIs being generated while the interrupt is in that "disabled" state will simply be lost.

Let's move to the mask/unmask set of registers, which offers the expected semantics.

Fixes 7c5925afbc58 ("PCI: dwc: Move MSI IRQs allocation to IRQ domains

hierarchical API")

---
 drivers/pci/controller/dwc/pcie-designware-host.c |   19 ++++++++++++-------
 1 file changed, 12 insertions(+), 7 deletions(-)
drivers/pci/controller/dwc/pcie-designware-host.c +12 -7
@@ -168,8 +168,8 @@ static void dw_pci_bottom_mask(struct ir
 		bit = data->hwirq % MAX_MSI_IRQS_PER_CTRL;
 
 		pp->irq_status[ctrl] &= ~(1 << bit);
-		dw_pcie_wr_own_conf(pp, PCIE_MSI_INTR0_ENABLE + res, 4,
-				    pp->irq_status[ctrl]);
+		dw_pcie_wr_own_conf(pp, PCIE_MSI_INTR0_MASK + res, 4,
+				    ~pp->irq_status[ctrl]);
 	}
 
 	raw_spin_unlock_irqrestore(&pp->lock, flags);
@@ -191,8 +191,8 @@ static void dw_pci_bottom_unmask(struct
 		bit = data->hwirq % MAX_MSI_IRQS_PER_CTRL;
 
 		pp->irq_status[ctrl] |= 1 << bit;
-		dw_pcie_wr_own_conf(pp, PCIE_MSI_INTR0_ENABLE + res, 4,
-				    pp->irq_status[ctrl]);
+		dw_pcie_wr_own_conf(pp, PCIE_MSI_INTR0_MASK + res, 4,
+				    ~pp->irq_status[ctrl]);
 	}
 
 	raw_spin_unlock_irqrestore(&pp->lock, flags);
@@ -658,10 +658,15 @@ void dw_pcie_setup_rc(struct pcie_port *
 	num_ctrls = pp->num_vectors / MAX_MSI_IRQS_PER_CTRL;
 
 	/* Initialize IRQ Status array */
-	for (ctrl = 0; ctrl < num_ctrls; ctrl++)
-		dw_pcie_rd_own_conf(pp, PCIE_MSI_INTR0_ENABLE +
+	for (ctrl = 0; ctrl < num_ctrls; ctrl++) {
+		dw_pcie_wr_own_conf(pp, PCIE_MSI_INTR0_MASK +
 					(ctrl * MSI_REG_CTRL_BLOCK_SIZE),
-				    4, &pp->irq_status[ctrl]);
+				    4, ~0);
+		dw_pcie_wr_own_conf(pp, PCIE_MSI_INTR0_ENABLE +
+					(ctrl * MSI_REG_CTRL_BLOCK_SIZE),
+				    4, ~0);
+		pp->irq_status[ctrl] = 0;
+	}
 
 	/* Setup RC BARs */
 	dw_pcie_writel_dbi(pci, PCI_BASE_ADDRESS_0, 0x00000004);
author maintainer
#29 lore raw

PATCH 29/50 PCI: dwc: Take lock when ACKing an interrupt

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Marc Zyngier <[email protected]>

commit fce5423e4f431c71933d6c1f850b540a314aa6ee upstream.

Bizarrely, there is no lock taken in the irq_ack() helper. This puts the ACK callback provided by a specific platform in a awkward situation where there is no synchronization that would be expected on other callback.

Introduce the required lock, giving some level of uniformity among callbacks.

Fixes 7c5925afbc58 ("PCI: dwc: Move MSI IRQs allocation to IRQ domains

hierarchical API")

---
 drivers/pci/controller/dwc/pcie-designware-host.c |    5 +++++
 1 file changed, 5 insertions(+)
drivers/pci/controller/dwc/pcie-designware-host.c +5 -0
@@ -202,11 +202,16 @@ static void dw_pci_bottom_ack(struct irq
 {
 	struct msi_desc *msi = irq_data_get_msi_desc(d);
 	struct pcie_port *pp;
+	unsigned long flags;
 
 	pp = msi_desc_to_pci_sysdata(msi);
 
+	raw_spin_lock_irqsave(&pp->lock, flags);
+
 	if (pp->ops->msi_irq_ack)
 		pp->ops->msi_irq_ack(d->hwirq, pp);
+
+	raw_spin_unlock_irqrestore(&pp->lock, flags);
 }
 
 static struct irq_chip dw_pci_msi_bottom_irq_chip = {
author maintainer
#30 lore raw

PATCH 30/50 PCI: dwc: Move interrupt acking into the proper callback

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Marc Zyngier <[email protected]>

commit 3f7bb2ec20ce07c02b2002349d256c91a463fcc5 upstream.

The write to the status register is really an ACK for the HW, and should be treated as such by the driver. Let's move it to the irq_ack() callback, which will prevent people from moving it around in order to paper over other bugs.

Fixes 8c934095fa2f ("PCI: dwc: Clear MSI interrupt status after it is handled,

not before")

Fixes 7c5925afbc58 ("PCI: dwc: Move MSI IRQs allocation to IRQ domains

hierarchical API")

---
 drivers/pci/controller/dwc/pcie-designware-host.c |   13 +++++++------
 1 file changed, 7 insertions(+), 6 deletions(-)
drivers/pci/controller/dwc/pcie-designware-host.c +7 -6
@@ -99,9 +99,6 @@ irqreturn_t dw_handle_msi_irq(struct pci
 					       (i * MAX_MSI_IRQS_PER_CTRL) +
 					       pos);
 			generic_handle_irq(irq);
-			dw_pcie_wr_own_conf(pp, PCIE_MSI_INTR0_STATUS +
-						(i * MSI_REG_CTRL_BLOCK_SIZE),
-					    4, 1 << pos);
 			pos++;
 		}
 	}
@@ -200,14 +197,18 @@ static void dw_pci_bottom_unmask(struct
 
 static void dw_pci_bottom_ack(struct irq_data *d)
 {
-	struct msi_desc *msi = irq_data_get_msi_desc(d);
-	struct pcie_port *pp;
+	struct pcie_port *pp  = irq_data_get_irq_chip_data(d);
+	unsigned int res, bit, ctrl;
 	unsigned long flags;
 
-	pp = msi_desc_to_pci_sysdata(msi);
+	ctrl = d->hwirq / MAX_MSI_IRQS_PER_CTRL;
+	res = ctrl * MSI_REG_CTRL_BLOCK_SIZE;
+	bit = d->hwirq % MAX_MSI_IRQS_PER_CTRL;
 
 	raw_spin_lock_irqsave(&pp->lock, flags);
 
+	dw_pcie_wr_own_conf(pp, PCIE_MSI_INTR0_STATUS + res, 4, 1 << bit);
+
 	if (pp->ops->msi_irq_ack)
 		pp->ops->msi_irq_ack(d->hwirq, pp);
 
author maintainer
#31 lore raw

PATCH 31/50 drm/amd/display: Fix MST dp_blank REG_WAIT timeout

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Jerry (Fangzhi) Zuo <[email protected]>

commit 8c9d90eebd23b6d40ddf4ce5df5ca2b932336a06 upstream.

Need to blank stream before deallocate MST payload.

[drm:generic_reg_wait [amdgpu]] *ERROR* REG_WAIT timeout 10us * 3000 tries - dce110_stream_encoder_dp_blank line:944

------------[ cut here ]------------
WARNING: CPU: 0 PID: 2201 at /var/lib/dkms/amdgpu/18.50-690240/build/amd/amdgpu/../display/dc/dc_helper.c:249 generic_reg_wait+0xe7/0x160 [amdgpu]
Call Trace:
 dce110_stream_encoder_dp_blank+0x11c/0x180 [amdgpu]
 core_link_disable_stream+0x40/0x230 [amdgpu]
 ? generic_reg_update_ex+0xdb/0x130 [amdgpu]
 dce110_reset_hw_ctx_wrap+0xb7/0x1f0 [amdgpu]
 dce110_apply_ctx_to_hw+0x30/0x430 [amdgpu]
 ? dce110_apply_ctx_for_surface+0x206/0x260 [amdgpu]
 dc_commit_state+0x2ba/0x4d0 [amdgpu]
 amdgpu_dm_atomic_commit_tail+0x297/0xd70 [amdgpu]
 ? amdgpu_bo_pin_restricted+0x58/0x260 [amdgpu]
 ? wait_for_completion_timeout+0x1f/0x120
 ? wait_for_completion_interruptible+0x1c/0x160
 commit_tail+0x3d/0x60 [drm_kms_helper]
 drm_atomic_helper_commit+0xf6/0x100 [drm_kms_helper]
 drm_atomic_connector_commit_dpms+0xe5/0xf0 [drm]
 drm_mode_obj_set_property_ioctl+0x14f/0x250 [drm]
 drm_mode_connector_property_set_ioctl+0x2e/0x40 [drm]
 drm_ioctl+0x1e0/0x430 [drm]
 ? drm_mode_connector_set_obj_prop+0x70/0x70 [drm]
 ? ep_read_events_proc+0xb0/0xb0
 ? ep_scan_ready_list.constprop.18+0x1e6/0x1f0
 ? timerqueue_add+0x52/0x80
 amdgpu_drm_ioctl+0x49/0x80 [amdgpu]
 do_vfs_ioctl+0x90/0x5f0
 SyS_ioctl+0x74/0x80
 do_syscall_64+0x74/0x140
 entry_SYSCALL_64_after_hwframe+0x3d/0xa2
---[ end trace 3ed7b77a97d60f72 ]---
---
 drivers/gpu/drm/amd/display/dc/core/dc_link.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/gpu/drm/amd/display/dc/core/dc_link.c +2 -2
@@ -2457,11 +2457,11 @@ void core_link_disable_stream(struct pip
 {
 	struct dc  *core_dc = pipe_ctx->stream->ctx->dc;
 
+	core_dc->hwss.blank_stream(pipe_ctx);
+
 	if (pipe_ctx->stream->signal == SIGNAL_TYPE_DISPLAY_PORT_MST)
 		deallocate_mst_payload(pipe_ctx);
 
-	core_dc->hwss.blank_stream(pipe_ctx);
-
 	core_dc->hwss.disable_stream(pipe_ctx, option);
 
 	disable_link(pipe_ctx->stream->sink->link, pipe_ctx->stream->signal);
author maintainer
#32 lore raw

PATCH 32/50 drm/fb_helper: Allow leaking fbdev smem_start

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Neil Armstrong <[email protected]>

commit 4be9bd10e22dfc7fc101c5cf5969ef2d3a042d8a upstream.

Since "drm/fb: Stop leaking physical address", the default behaviour of the DRM fbdev emulation is to set the smem_base to 0 and pass the new FBINFO_HIDE_SMEM_START flag.

The main reason is to avoid leaking physical addresse to user-space, and it follows a general move over the kernel code to avoid user-space to manipulate physical addresses and then use some other mechanisms like dma-buf to transfer physical buffer handles over multiple subsystems.

But, a lot of devices depends on closed sources binaries to enable OpenGL hardware acceleration that uses this smem_start value to pass physical addresses to out-of-tree modules in order to render into these physical adresses. These should use dma-buf buffers allocated from the DRM display device instead and stop relying on fbdev overallocation to gather DMA memory (some HW vendors delivers GBM and Wayland capable binaries, but older unsupported devices won't have these new binaries and are doomed until an Open Source solution like Lima finalizes).

Since these devices heavily depends on this kind of software and because the smem_start population was available for years, it's a breakage to stop leaking smem_start without any alternative solutions.

This patch adds a Kconfig depending on the EXPERT config and an unsafe kernel module parameter tainting the kernel when enabled.

A clear comment and Kconfig help text was added to clarify why and when this patch should be reverted, but in the meantime it's a necessary feature to keep.

---
 drivers/gpu/drm/Kconfig         |   20 ++++++++++++++++++++
 drivers/gpu/drm/drm_fb_helper.c |   25 +++++++++++++++++++++++++
 2 files changed, 45 insertions(+)
drivers/gpu/drm/Kconfig +20 -0
@@ -110,6 +110,26 @@ config DRM_FBDEV_OVERALLOC
 	  is 100. Typical values for double buffering will be 200,
 	  triple buffering 300.
 
+config DRM_FBDEV_LEAK_PHYS_SMEM
+	bool "Shamelessly allow leaking of fbdev physical address (DANGEROUS)"
+	depends on DRM_FBDEV_EMULATION && EXPERT
+	default n
+	help
+	  In order to keep user-space compatibility, we want in certain
+	  use-cases to keep leaking the fbdev physical address to the
+	  user-space program handling the fbdev buffer.
+	  This affects, not only, Amlogic, Allwinner or Rockchip devices
+	  with ARM Mali GPUs using an userspace Blob.
+	  This option is not supported by upstream developers and should be
+	  removed as soon as possible and be considered as a broken and
+	  legacy behaviour from a modern fbdev device driver.
+
+	  Please send any bug reports when using this to your proprietary
+	  software vendor that requires this.
+
+	  If in doubt, say "N" or spread the word to your closed source
+	  library vendor.
+
 config DRM_LOAD_EDID_FIRMWARE
 	bool "Allow to specify an EDID data set instead of probing for it"
 	depends on DRM
drivers/gpu/drm/drm_fb_helper.c +25 -0
@@ -56,6 +56,25 @@ MODULE_PARM_DESC(drm_fbdev_overalloc,
 		 "Overallocation of the fbdev buffer (%) [default="
 		 __MODULE_STRING(CONFIG_DRM_FBDEV_OVERALLOC) "]");
 
+/*
+ * In order to keep user-space compatibility, we want in certain use-cases
+ * to keep leaking the fbdev physical address to the user-space program
+ * handling the fbdev buffer.
+ * This is a bad habit essentially kept into closed source opengl driver
+ * that should really be moved into open-source upstream projects instead
+ * of using legacy physical addresses in user space to communicate with
+ * other out-of-tree kernel modules.
+ *
+ * This module_param *should* be removed as soon as possible and be
+ * considered as a broken and legacy behaviour from a modern fbdev device.
+ */
+#if IS_ENABLED(CONFIG_DRM_FBDEV_LEAK_PHYS_SMEM)
+static bool drm_leak_fbdev_smem = false;
+module_param_unsafe(drm_leak_fbdev_smem, bool, 0600);
+MODULE_PARM_DESC(fbdev_emulation,
+		 "Allow unsafe leaking fbdev physical smem address [default=false]");
+#endif
+
 static LIST_HEAD(kernel_fb_helper_list);
 static DEFINE_MUTEX(kernel_fb_helper_lock);
 
@@ -3041,6 +3060,12 @@ int drm_fb_helper_generic_probe(struct d
 	fbi->screen_size = fb->height * fb->pitches[0];
 	fbi->fix.smem_len = fbi->screen_size;
 	fbi->screen_buffer = buffer->vaddr;
+	/* Shamelessly leak the physical address to user-space */
+#if IS_ENABLED(CONFIG_DRM_FBDEV_LEAK_PHYS_SMEM)
+	if (drm_leak_fbdev_smem && fbi->fix.smem_start == 0)
+		fbi->fix.smem_start =
+			page_to_phys(virt_to_page(fbi->screen_buffer));
+#endif
 	strcpy(fbi->fix.id, "DRM emulated");
 
 	drm_fb_helper_fill_fix(fbi, fb->pitches[0], fb->format->depth);
author maintainer
#33 lore raw

PATCH 33/50 drm/fb-helper: Partially bring back workaround for bugs of SDL 1.2

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Ivan Mironov <[email protected]>

commit 62d85b3bf9d978ed4b6b2aeef5cf0ccf1423906e upstream.

SDL 1.2 sets all fields related to the pixel format to zero in some cases[1]. Prior to commit db05c48197759 ("drm: fb-helper: Reject all pixel format changing requests"), there was an unintentional workaround for this that existed for more than a decade. First in device-specific DRM drivers, then here in drm_fb_helper.c.

Previous code containing this workaround just ignores pixel format fields from userspace code. Not a good thing either, as this way, driver may silently use pixel format different from what client actually requested, and this in turn will lead to displaying garbage on the screen. I think that returning EINVAL to userspace in this particular case is the right option, so I decided to left code from problematic commit untouched instead of just reverting it entirely.

Here is the steps required to reproduce this problem exactly:
1) Compile fceux[2] with SDL 1.2.15 and without GTK or OpenGL support. SDL should be compiled with fbdev support (which is on by default).
2) Create /etc/fb.modes with following contents (values seems not used, and just required to trigger problematic code in SDL):

mode "test"
geometry 1 1 1 1 1
timings 1 1 1 1 1 1 1
endmode

3) Create ~/.fceux/fceux.cfg with following contents:

SDL.Hotkeys.Quit = 27
SDL.DoubleBuffering = 1

4) Ensure that screen resolution is at least 1280x960 (e.g. append "video=Virtual-1:1280x960-32" to the kernel cmdline for qemu/QXL).

5) Try to run fceux on VT with some ROM file[3]:

# ./fceux color_test.nes

[1] SDL 1.2.15 source code, src/video/fbcon/SDL_fbvideo.c, FB_SetVideoMode()
[2] http://www.fceux.com
[3] Example ROM: https://github.com/bokuweb/rustynes/blob/master/roms/color_test.nes

---
 drivers/gpu/drm/drm_fb_helper.c |  126 +++++++++++++++++++++++-----------------
 1 file changed, 73 insertions(+), 53 deletions(-)
drivers/gpu/drm/drm_fb_helper.c +73 -53
@@ -1621,6 +1621,64 @@ static bool drm_fb_pixel_format_equal(co
 	       var_1->transp.msb_right == var_2->transp.msb_right;
 }
 
+static void drm_fb_helper_fill_pixel_fmt(struct fb_var_screeninfo *var,
+					 u8 depth)
+{
+	switch (depth) {
+	case 8:
+		var->red.offset = 0;
+		var->green.offset = 0;
+		var->blue.offset = 0;
+		var->red.length = 8; /* 8bit DAC */
+		var->green.length = 8;
+		var->blue.length = 8;
+		var->transp.offset = 0;
+		var->transp.length = 0;
+		break;
+	case 15:
+		var->red.offset = 10;
+		var->green.offset = 5;
+		var->blue.offset = 0;
+		var->red.length = 5;
+		var->green.length = 5;
+		var->blue.length = 5;
+		var->transp.offset = 15;
+		var->transp.length = 1;
+		break;
+	case 16:
+		var->red.offset = 11;
+		var->green.offset = 5;
+		var->blue.offset = 0;
+		var->red.length = 5;
+		var->green.length = 6;
+		var->blue.length = 5;
+		var->transp.offset = 0;
+		break;
+	case 24:
+		var->red.offset = 16;
+		var->green.offset = 8;
+		var->blue.offset = 0;
+		var->red.length = 8;
+		var->green.length = 8;
+		var->blue.length = 8;
+		var->transp.offset = 0;
+		var->transp.length = 0;
+		break;
+	case 32:
+		var->red.offset = 16;
+		var->green.offset = 8;
+		var->blue.offset = 0;
+		var->red.length = 8;
+		var->green.length = 8;
+		var->blue.length = 8;
+		var->transp.offset = 24;
+		var->transp.length = 8;
+		break;
+	default:
+		break;
+	}
+}
+
 /**
  * drm_fb_helper_check_var - implementation for &fb_ops.fb_check_var
  * @var: screeninfo to check
@@ -1651,6 +1709,20 @@ int drm_fb_helper_check_var(struct fb_va
 	}
 
 	/*
+	 * Workaround for SDL 1.2, which is known to be setting all pixel format
+	 * fields values to zero in some cases. We treat this situation as a
+	 * kind of "use some reasonable autodetected values".
+	 */
+	if (!var->red.offset     && !var->green.offset    &&
+	    !var->blue.offset    && !var->transp.offset   &&
+	    !var->red.length     && !var->green.length    &&
+	    !var->blue.length    && !var->transp.length   &&
+	    !var->red.msb_right  && !var->green.msb_right &&
+	    !var->blue.msb_right && !var->transp.msb_right) {
+		drm_fb_helper_fill_pixel_fmt(var, fb->format->depth);
+	}
+
+	/*
 	 * drm fbdev emulation doesn't support changing the pixel format at all,
 	 * so reject all pixel format changing requests.
 	 */
@@ -1961,59 +2033,7 @@ void drm_fb_helper_fill_var(struct fb_in
 	info->var.yoffset = 0;
 	info->var.activate = FB_ACTIVATE_NOW;
 
-	switch (fb->format->depth) {
-	case 8:
-		info->var.red.offset = 0;
-		info->var.green.offset = 0;
-		info->var.blue.offset = 0;
-		info->var.red.length = 8; /* 8bit DAC */
-		info->var.green.length = 8;
-		info->var.blue.length = 8;
-		info->var.transp.offset = 0;
-		info->var.transp.length = 0;
-		break;
-	case 15:
-		info->var.red.offset = 10;
-		info->var.green.offset = 5;
-		info->var.blue.offset = 0;
-		info->var.red.length = 5;
-		info->var.green.length = 5;
-		info->var.blue.length = 5;
-		info->var.transp.offset = 15;
-		info->var.transp.length = 1;
-		break;
-	case 16:
-		info->var.red.offset = 11;
-		info->var.green.offset = 5;
-		info->var.blue.offset = 0;
-		info->var.red.length = 5;
-		info->var.green.length = 6;
-		info->var.blue.length = 5;
-		info->var.transp.offset = 0;
-		break;
-	case 24:
-		info->var.red.offset = 16;
-		info->var.green.offset = 8;
-		info->var.blue.offset = 0;
-		info->var.red.length = 8;
-		info->var.green.length = 8;
-		info->var.blue.length = 8;
-		info->var.transp.offset = 0;
-		info->var.transp.length = 0;
-		break;
-	case 32:
-		info->var.red.offset = 16;
-		info->var.green.offset = 8;
-		info->var.blue.offset = 0;
-		info->var.red.length = 8;
-		info->var.green.length = 8;
-		info->var.blue.length = 8;
-		info->var.transp.offset = 24;
-		info->var.transp.length = 8;
-		break;
-	default:
-		break;
-	}
+	drm_fb_helper_fill_pixel_fmt(&info->var, fb->format->depth);
 
 	info->var.xres = fb_width;
 	info->var.yres = fb_height;
author maintainer
#34 lore raw

PATCH 34/50 drm/i915: Unwind failure on pinning the gen7 ppgtt

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Chris Wilson <[email protected]>

commit 280d479b310298dfeb1d6f9a1617eca37beb6ce4 upstream.

If we fail to pin the ggtt vma slot for the ppgtt page tables, we need to unwind the locals before reporting the error. Or else on subsequent attempts to bind the page tables into the ggtt, we will already believe that the vma has been pinned and continue on blithely. If something else should happen to be at that location, choas ensues.

Fixes a2bbf7148342 ("drm/i915/gtt: Only keep gen6 page directories pinned while active")
Cc Joonas Lahtinen <[email protected]>
Cc Mika Kuoppala <[email protected]>
Cc Matthew Auld <[email protected]>
Cc <[email protected]> # v4.19+
(cherry picked from commit d4de753526f4d99f541f1b6ed1d963005c09700c)
---
 drivers/gpu/drm/i915/i915_gem_gtt.c |   15 ++++++++++++---
 1 file changed, 12 insertions(+), 3 deletions(-)
drivers/gpu/drm/i915/i915_gem_gtt.c +12 -3
@@ -2128,6 +2128,7 @@ static struct i915_vma *pd_vma_create(st
 int gen6_ppgtt_pin(struct i915_hw_ppgtt *base)
 {
 	struct gen6_hw_ppgtt *ppgtt = to_gen6_ppgtt(base);
+	int err;
 
 	/*
 	 * Workaround the limited maximum vma->pin_count and the aliasing_ppgtt
@@ -2143,9 +2144,17 @@ int gen6_ppgtt_pin(struct i915_hw_ppgtt
 	 * allocator works in address space sizes, so it's multiplied by page
 	 * size. We allocate at the top of the GTT to avoid fragmentation.
 	 */
-	return i915_vma_pin(ppgtt->vma,
-			    0, GEN6_PD_ALIGN,
-			    PIN_GLOBAL | PIN_HIGH);
+	err = i915_vma_pin(ppgtt->vma,
+			   0, GEN6_PD_ALIGN,
+			   PIN_GLOBAL | PIN_HIGH);
+	if (err)
+		goto unpin;
+
+	return 0;
+
+unpin:
+	ppgtt->pin_count = 0;
+	return err;
 }
 
 void gen6_ppgtt_unpin(struct i915_hw_ppgtt *base)
author maintainer
#35 lore raw

PATCH 35/50 drm/amdgpu: Dont ignore rc from drm_dp_mst_topology_mgr_resume()

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Lyude Paul <[email protected]>

commit fe7553bef8d676d1d8b40666868b33ec39b9df5d upstream.

drm_dp_mst_topology_mgr_resume() returns whether or not it managed to find the topology in question after a suspend resume cycle, and the driver is supposed to check this value and disable MST accordingly if it's gone-in addition to sending a hotplug in order to notify userspace that something changed during suspend.

Currently, amdgpu just makes the mistake of ignoring the return code from drm_dp_mst_topology_mgr_resume() which means that if a topology was removed in suspend, amdgpu never notices and assumes it's still connected which leads to all sorts of problems.

So, fix this by actually checking the rc from drm_dp_mst_topology_mgr_resume(). Also, reformat the rest of the function while we're at it to fix the over-indenting.

---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c |   34 +++++++++++++++-------
 1 file changed, 24 insertions(+), 10 deletions(-)
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +24 -10
@@ -565,22 +565,36 @@ static void s3_handle_mst(struct drm_dev
 {
 	struct amdgpu_dm_connector *aconnector;
 	struct drm_connector *connector;
+	struct drm_dp_mst_topology_mgr *mgr;
+	int ret;
+	bool need_hotplug = false;
 
 	drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
 
-	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
-		   aconnector = to_amdgpu_dm_connector(connector);
-		   if (aconnector->dc_link->type == dc_connection_mst_branch &&
-				   !aconnector->mst_port) {
-
-			   if (suspend)
-				   drm_dp_mst_topology_mgr_suspend(&aconnector->mst_mgr);
-			   else
-				   drm_dp_mst_topology_mgr_resume(&aconnector->mst_mgr);
-		   }
+	list_for_each_entry(connector, &dev->mode_config.connector_list,
+			    head) {
+		aconnector = to_amdgpu_dm_connector(connector);
+		if (aconnector->dc_link->type != dc_connection_mst_branch ||
+		    aconnector->mst_port)
+			continue;
+
+		mgr = &aconnector->mst_mgr;
+
+		if (suspend) {
+			drm_dp_mst_topology_mgr_suspend(mgr);
+		} else {
+			ret = drm_dp_mst_topology_mgr_resume(mgr);
+			if (ret < 0) {
+				drm_dp_mst_topology_mgr_set_mst(mgr, false);
+				need_hotplug = true;
+			}
+		}
 	}
 
 	drm_modeset_unlock(&dev->mode_config.connection_mutex);
+
+	if (need_hotplug)
+		drm_kms_helper_hotplug_event(dev);
 }
 
 static int dm_hw_init(void *handle)
author maintainer
#36 lore raw

PATCH 36/50 drm/amdgpu: Dont fail resume process if resuming atomic state fails

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Lyude Paul <[email protected]>

commit 2d1af6a11cb9d88e0e3dd10258904c437fe1b315 upstream.

This is an ugly one unfortunately. Currently, all DRM drivers supporting atomic modesetting will save the state that userspace had set before suspending, then attempt to restore that state on resume. This probably worked very well at one point, like many other things, until DP MST came into the picture. While it's easy to restore state on normal display connectors that were disconnected during suspend regardless of their state post-resume, this can't really be done with MST because of the fact that setting up a downstream sink requires performing sideband transactions between the source and the MST hub, sending out the ACT packets, etc.

Because of this, there isn't really a guarantee that we can restore the atomic state we had before suspend once we've resumed. This sucks pretty bad, but so far I haven't run into any compositors that this actually causes serious issues with. Most compositors will notice the hotplug we send afterwards, and then reprobe state.

Since nouveau and i915 also don't fail the suspend/resume process due to failing to restore the atomic state, let's make amdgpu match this behavior. Better to resume the GPU properly, then to stop the process half way because of a potentially unavoidable atomic commit failure.

Eventually, we'll have a real fix for this problem on the DRM level. But we've got some more important low-hanging fruit to deal with first.

---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c |    5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +2 -3
@@ -750,7 +750,6 @@ static int dm_resume(void *handle)
 	struct drm_plane_state *new_plane_state;
 	struct dm_plane_state *dm_new_plane_state;
 	enum dc_connection_type new_connection_type = dc_connection_none;
-	int ret;
 	int i;
 
 	/* power on hardware */
@@ -823,13 +822,13 @@ static int dm_resume(void *handle)
 		}
 	}
 
-	ret = drm_atomic_helper_resume(ddev, dm->cached_state);
+	drm_atomic_helper_resume(ddev, dm->cached_state);
 
 	dm->cached_state = NULL;
 
 	amdgpu_dm_irq_resume_late(adev);
 
-	return ret;
+	return 0;
 }
 
 static const struct amd_ip_funcs amdgpu_dm_funcs = {
author maintainer
#37 lore raw

PATCH 37/50 rbd: dont return 0 on unmap if RBD_DEV_FLAG_REMOVING is set

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Ilya Dryomov <[email protected]>

commit 85f5a4d666fd9be73856ed16bb36c5af5b406b29 upstream.

There is a window between when RBD_DEV_FLAG_REMOVING is set and when the device is removed from rbd_dev_list. During this window, we set "already" and return 0.

Returning 0 from write(2) can confuse userspace tools because 0 indicates that nothing was written. In particular, "rbd unmap" will retry the write multiple times a second:

  10:28:05.463299 write(4, "0", 1)        = 0
  10:28:05.463509 write(4, "0", 1)        = 0
  10:28:05.463720 write(4, "0", 1)        = 0
  10:28:05.463942 write(4, "0", 1)        = 0
  10:28:05.464155 write(4, "0", 1)        = 0
---
 drivers/block/rbd.c |    9 ++++-----
 1 file changed, 4 insertions(+), 5 deletions(-)
drivers/block/rbd.c +4 -5
@@ -5982,7 +5982,6 @@ static ssize_t do_rbd_remove(struct bus_
 	struct list_head *tmp;
 	int dev_id;
 	char opt_buf[6];
-	bool already = false;
 	bool force = false;
 	int ret;
 
@@ -6015,13 +6014,13 @@ static ssize_t do_rbd_remove(struct bus_
 		spin_lock_irq(&rbd_dev->lock);
 		if (rbd_dev->open_count && !force)
 			ret = -EBUSY;
-		else
-			already = test_and_set_bit(RBD_DEV_FLAG_REMOVING,
-							&rbd_dev->flags);
+		else if (test_and_set_bit(RBD_DEV_FLAG_REMOVING,
+					  &rbd_dev->flags))
+			ret = -EINPROGRESS;
 		spin_unlock_irq(&rbd_dev->lock);
 	}
 	spin_unlock(&rbd_dev_list_lock);
-	if (ret < 0 || already)
+	if (ret)
 		return ret;
 
 	if (force) {
author maintainer
#38 lore raw

PATCH 38/50 ext4: make sure enough credits are reserved for dioread_nolock writes

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Theodore Ts'o <[email protected]>

commit 812c0cab2c0dfad977605dbadf9148490ca5d93f upstream.

There are enough credits reserved for most dioread_nolock writes; however, if the extent tree is sufficiently deep, and/or quota is enabled, the code was not allowing for all eventualities when reserving journal credits for the unwritten extent conversion.

This problem can be seen using xfstests ext4/034:

   WARNING: CPU: 1 PID: 257 at fs/ext4/ext4_jbd2.c:271 __ext4_handle_dirty_metadata+0x10c/0x180
   Workqueue: ext4-rsv-conversion ext4_end_io_rsv_work
   RIP: 0010:__ext4_handle_dirty_metadata+0x10c/0x180
   	...
   EXT4-fs: ext4_free_blocks:4938: aborting transaction: error 28 in __ext4_handle_dirty_metadata
   EXT4: jbd2_journal_dirty_metadata failed: handle type 11 started at line 4921, credits 4/0, errcode -28
   EXT4-fs error (device dm-1) in ext4_free_blocks:4950: error 28
---
 fs/ext4/inode.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
fs/ext4/inode.c +2 -1
@@ -2748,7 +2748,8 @@ static int ext4_writepages(struct addres
 		 * We may need to convert up to one extent per block in
 		 * the page and we may dirty the inode.
 		 */
-		rsv_blocks = 1 + (PAGE_SIZE >> inode->i_blkbits);
+		rsv_blocks = 1 + ext4_chunk_trans_blocks(inode,
+						PAGE_SIZE >> inode->i_blkbits);
 	}
 
 	/*
author maintainer
#39 lore raw

PATCH 39/50 ext4: fix a potential fiemap/page fault deadlock w/ inline_data

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Theodore Ts'o <[email protected]>

commit 2b08b1f12cd664dc7d5c84ead9ff25ae97ad5491 upstream.

The ext4_inline_data_fiemap() function calls fiemap_fill_next_extent() while still holding the xattr semaphore. This is not necessary and it triggers a circular lockdep warning. This is because fiemap_fill_next_extent() could trigger a page fault when it writes into page which triggers a page fault. If that page is mmaped from the inline file in question, this could very well result in a deadlock.

This problem can be reproduced using generic/519 with a file system configuration which has the inline_data feature enabled.

---
 fs/ext4/inline.c |    6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
fs/ext4/inline.c +3 -3
@@ -1890,12 +1890,12 @@ int ext4_inline_data_fiemap(struct inode
 	physical += (char *)ext4_raw_inode(&iloc) - iloc.bh->b_data;
 	physical += offsetof(struct ext4_inode, i_block);
 
-	if (physical)
-		error = fiemap_fill_next_extent(fieinfo, start, physical,
-						inline_len, flags);
 	brelse(iloc.bh);
 out:
 	up_read(&EXT4_I(inode)->xattr_sem);
+	if (physical)
+		error = fiemap_fill_next_extent(fieinfo, start, physical,
+						inline_len, flags);
 	return (error < 0 ? error : 0);
 }
 
author maintainer
#40 lore raw

PATCH 40/50 ext4: avoid kernel warning when writing the superblock to a dead device

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Theodore Ts'o <[email protected]>

commit e86807862e6880809f191c4cea7f88a489f0ed34 upstream.

The xfstests generic/475 test switches the underlying device with dm-error while running a stress test. This results in a large number of file system errors, and since we can't lock the buffer head when marking the superblock dirty in the ext4_grp_locked_error() case, it's possible the superblock to be !buffer_uptodate() without buffer_write_io_error() being true.

We need to set buffer_uptodate() before we call mark_buffer_dirty() or this will trigger a WARN_ON. It's safe to do this since the superblock must have been properly read into memory or the mount would have been successful. So if buffer_uptodate() is not set, we can safely assume that this happened due to a failed attempt to write the superblock.

---
 fs/ext4/super.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
fs/ext4/super.c +1 -1
@@ -4904,7 +4904,7 @@ static int ext4_commit_super(struct supe
 	ext4_superblock_csum_set(sb);
 	if (sync)
 		lock_buffer(sbh);
-	if (buffer_write_io_error(sbh)) {
+	if (buffer_write_io_error(sbh) || !buffer_uptodate(sbh)) {
 		/*
 		 * Oh, dear.  A previous attempt to write the
 		 * superblock failed.  This could happen because the
author maintainer
#41 lore raw

PATCH 41/50 ext4: use ext4_write_inode() when fsyncing w/o a journal

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Theodore Ts'o <[email protected]>

commit ad211f3e94b314a910d4af03178a0b52a7d1ee0a upstream.

In no-journal mode, we previously used __generic_file_fsync() in no-journal mode. This triggers a lockdep warning, and in addition, it's not safe to depend on the inode writeback mechanism in the case ext4. We can solve both problems by calling ext4_write_inode() directly.

---
 fs/ext4/fsync.c |   13 +++++++++----
 1 file changed, 9 insertions(+), 4 deletions(-)
fs/ext4/fsync.c +9 -4
@@ -116,8 +116,16 @@ int ext4_sync_file(struct file *file, lo
 		goto out;
 	}
 
+	ret = file_write_and_wait_range(file, start, end);
+	if (ret)
+		return ret;
+
 	if (!journal) {
-		ret = __generic_file_fsync(file, start, end, datasync);
+		struct writeback_control wbc = {
+			.sync_mode = WB_SYNC_ALL
+		};
+
+		ret = ext4_write_inode(inode, &wbc);
 		if (!ret)
 			ret = ext4_sync_parent(inode);
 		if (test_opt(inode->i_sb, BARRIER))
@@ -125,9 +133,6 @@ int ext4_sync_file(struct file *file, lo
 		goto out;
 	}
 
-	ret = file_write_and_wait_range(file, start, end);
-	if (ret)
-		return ret;
 	/*
 	 * data=writeback,ordered:
 	 *  The caller's filemap_fdatawrite()/wait will sync the data.
author maintainer
#42 lore raw

PATCH 42/50 ext4: track writeback errors using the generic tracking infrastructure

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Theodore Ts'o <[email protected]>

commit 95cb67138746451cc84cf8e516e14989746e93b0 upstream.

We already using mapping_set_error() in fs/ext4/page_io.c, so all we need to do is to use file_check_and_advance_wb_err() when handling fsync() requests in ext4_sync_file().

---
 fs/ext4/fsync.c |    3 +++
 1 file changed, 3 insertions(+)
fs/ext4/fsync.c +3 -0
@@ -164,6 +164,9 @@ int ext4_sync_file(struct file *file, lo
 			ret = err;
 	}
 out:
+	err = file_check_and_advance_wb_err(file);
+	if (ret == 0)
+		ret = err;
 	trace_ext4_sync_file_exit(inode, ret);
 	return ret;
 }
author maintainer
#43 lore raw

PATCH 43/50 ext4: fix special inode number checks in __ext4_iget()

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Theodore Ts'o <[email protected]>

commit 191ce17876c9367819c4b0a25b503c0f6d9054d8 upstream.

The check for special (reserved) inode number checks in __ext4_iget() was broken by commit 8a363970d1dc: ("ext4: avoid declaring fs inconsistent due to invalid file handles"). This was caused by a botched reversal of the sense of the flag now known as EXT4_IGET_SPECIAL (when it was previously named EXT4_IGET_NORMAL). Fix the logic appropriately.

Fixes 8a363970d1dc ("ext4: avoid declaring fs inconsistent...")
---
 fs/ext4/inode.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
fs/ext4/inode.c +1 -1
@@ -4803,7 +4803,7 @@ struct inode *__ext4_iget(struct super_b
 	gid_t i_gid;
 	projid_t i_projid;
 
-	if (((flags & EXT4_IGET_NORMAL) &&
+	if ((!(flags & EXT4_IGET_SPECIAL) &&
 	     (ino < EXT4_FIRST_INO(sb) && ino != EXT4_ROOT_INO)) ||
 	    (ino < EXT4_ROOT_INO) ||
 	    (ino > le32_to_cpu(EXT4_SB(sb)->s_es->s_inodes_count))) {
author maintainer
#44 lore raw

PATCH 44/50 mm: page_mapped: dont assume compound page is huge or THP

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Jan Stancek <[email protected]>

commit 8ab88c7169b7fba98812ead6524b9d05bc76cf00 upstream.

LTP proc01 testcase has been observed to rarely trigger crashes on arm64:

    page_mapped+0x78/0xb4
    stable_page_flags+0x27c/0x338
    kpageflags_read+0xfc/0x164
    proc_reg_read+0x7c/0xb8
    __vfs_read+0x58/0x178
    vfs_read+0x90/0x14c
    SyS_read+0x60/0xc0

The issue is that page_mapped() assumes that if compound page is not huge, then it must be THP. But if this is 'normal' compound page (COMPOUND_PAGE_DTOR), then following loop can keep running (for HPAGE_PMD_NR iterations) until it tries to read from memory that isn't mapped and triggers a panic:

        for (i = 0; i < hpage_nr_pages(page); i++) {
                if (atomic_read(&page[i]._mapcount) >= 0)
                        return true;
	}

I could replicate this on x86 (v4.20-rc4-98-g60b548237fed) only with a custom kernel module [1] which:
- allocates compound page (PAGEC) of order 1
- allocates 2 normal pages (COPY), which are initialized to 0xff (to satisfy _mapcount >= 0)
- 2 PAGEC page structs are copied to address of first COPY page
- second page of COPY is marked as not present
- call to page_mapped(COPY) now triggers fault on access to 2nd COPY page at offset 0x30 (_mapcount)

[1] https://github.com/jstancek/reproducers/blob/master/kernel/page_mapped_crash/repro.c

Fix the loop to iterate for "1 << compound_order" pages.

Kirrill said "IIRC, sound subsystem can producuce custom mapped compound pages".

---
 mm/util.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
mm/util.c +1 -1
@@ -485,7 +485,7 @@ bool page_mapped(struct page *page)
 		return true;
 	if (PageHuge(page))
 		return false;
-	for (i = 0; i < hpage_nr_pages(page); i++) {
+	for (i = 0; i < (1 << compound_order(page)); i++) {
 		if (atomic_read(&page[i]._mapcount) >= 0)
 			return true;
 	}
author maintainer
#45 lore raw

PATCH 45/50 sunrpc: use-after-free in svc_process_common()

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Vasily Averin <[email protected]>

commit d4b09acf924b84bae77cad090a9d108e70b43643 upstream.

if node have NFSv41+ mounts inside several net namespaces it can lead to use-after-free in svc_process_common()

svc_process_common()

        /* Setup reply header */
        rqstp->rq_xprt->xpt_ops->xpo_prep_reply_hdr(rqstp); <<< HERE

svc_process_common() can use incorrect rqstp->rq_xprt, its caller function bc_svc_process() takes it from serv->sv_bc_xprt. The problem is that serv is global structure but sv_bc_xprt is assigned per-netnamespace.

According to Trond, the whole "let's set up rqstp->rq_xprt for the back channel" is nothing but a giant hack in order to work around the fact that svc_process_common() uses it to find the xpt_ops, and perform a couple of (meaningless for the back channel) tests of xpt_flags.

All we really need in svc_process_common() is to be able to run rqstp->rq_xprt->xpt_ops->xpo_prep_reply_hdr()

Bruce J Fields points that this xpo_prep_reply_hdr() call is an awfully roundabout way just to do "svc_putnl(resv, 0);" in the tcp case.

This patch does not initialiuze rqstp->rq_xprt in bc_svc_process(), now it calls svc_process_common() with rqstp->rq_xprt = NULL.

To adjust reply header svc_process_common() just check rqstp->rq_prot and calls svc_tcp_prep_reply_hdr() for tcp case.

To handle rqstp->rq_xprt = NULL case in functions called from svc_process_common() patch intruduces net namespace pointer svc_rqst->rq_bc_net and adjust SVC_NET() definition.
Some other function was also adopted to properly handle described case.

Fixes 23c20ecd4475 ("NFS: callback up - users counting cleanup")

v2: added lost extern svc_tcp_prep_reply_hdr()

---
 include/linux/sunrpc/svc.h    |    5 ++++-
 include/trace/events/sunrpc.h |    6 ++++--
 net/sunrpc/svc.c              |   11 +++++++----
 net/sunrpc/svc_xprt.c         |    5 +++--
 net/sunrpc/svcsock.c          |    2 +-
 5 files changed, 19 insertions(+), 10 deletions(-)
include/linux/sunrpc/svc.h +4 -1
@@ -295,9 +295,12 @@ struct svc_rqst {
 	struct svc_cacherep *	rq_cacherep;	/* cache info */
 	struct task_struct	*rq_task;	/* service thread */
 	spinlock_t		rq_lock;	/* per-request lock */
+	struct net		*rq_bc_net;	/* pointer to backchannel's
+						 * net namespace
+						 */
 };
 
-#define SVC_NET(svc_rqst)	(svc_rqst->rq_xprt->xpt_net)
+#define SVC_NET(rqst) (rqst->rq_xprt ? rqst->rq_xprt->xpt_net : rqst->rq_bc_net)
 
 /*
  * Rigorous type checking on sockaddr type conversions
include/trace/events/sunrpc.h +4 -2
@@ -582,7 +582,8 @@ TRACE_EVENT(svc_process,
 		__field(u32, vers)
 		__field(u32, proc)
 		__string(service, name)
-		__string(addr, rqst->rq_xprt->xpt_remotebuf)
+		__string(addr, rqst->rq_xprt ?
+			 rqst->rq_xprt->xpt_remotebuf : "(null)")
 	),
 
 	TP_fast_assign(
@@ -590,7 +591,8 @@ TRACE_EVENT(svc_process,
 		__entry->vers = rqst->rq_vers;
 		__entry->proc = rqst->rq_proc;
 		__assign_str(service, name);
-		__assign_str(addr, rqst->rq_xprt->xpt_remotebuf);
+		__assign_str(addr, rqst->rq_xprt ?
+			     rqst->rq_xprt->xpt_remotebuf : "(null)");
 	),
 
 	TP_printk("addr=%s xid=0x%08x service=%s vers=%u proc=%u",
net/sunrpc/svc.c +7 -4
@@ -1144,6 +1144,8 @@ void svc_printk(struct svc_rqst *rqstp,
 static __printf(2,3) void svc_printk(struct svc_rqst *rqstp, const char *fmt, ...) {}
 #endif
 
+extern void svc_tcp_prep_reply_hdr(struct svc_rqst *);
+
 /*
  * Common routine for processing the RPC request.
  */
@@ -1172,7 +1174,8 @@ svc_process_common(struct svc_rqst *rqst
 	clear_bit(RQ_DROPME, &rqstp->rq_flags);
 
 	/* Setup reply header */
-	rqstp->rq_xprt->xpt_ops->xpo_prep_reply_hdr(rqstp);
+	if (rqstp->rq_prot == IPPROTO_TCP)
+		svc_tcp_prep_reply_hdr(rqstp);
 
 	svc_putu32(resv, rqstp->rq_xid);
 
@@ -1244,7 +1247,7 @@ svc_process_common(struct svc_rqst *rqst
 	 * for lower versions. RPC_PROG_MISMATCH seems to be the closest
 	 * fit.
 	 */
-	if (versp->vs_need_cong_ctrl &&
+	if (versp->vs_need_cong_ctrl && rqstp->rq_xprt &&
 	    !test_bit(XPT_CONG_CTRL, &rqstp->rq_xprt->xpt_flags))
 		goto err_bad_vers;
 
@@ -1336,7 +1339,7 @@ svc_process_common(struct svc_rqst *rqst
 	return 0;
 
  close:
-	if (test_bit(XPT_TEMP, &rqstp->rq_xprt->xpt_flags))
+	if (rqstp->rq_xprt && test_bit(XPT_TEMP, &rqstp->rq_xprt->xpt_flags))
 		svc_close_xprt(rqstp->rq_xprt);
 	dprintk("svc: svc_process close\n");
 	return 0;
@@ -1459,10 +1462,10 @@ bc_svc_process(struct svc_serv *serv, st
 	dprintk("svc: %s(%p)\n", __func__, req);
 
 	/* Build the svc_rqst used by the common processing routine */
-	rqstp->rq_xprt = serv->sv_bc_xprt;
 	rqstp->rq_xid = req->rq_xid;
 	rqstp->rq_prot = req->rq_xprt->prot;
 	rqstp->rq_server = serv;
+	rqstp->rq_bc_net = req->rq_xprt->xprt_net;
 
 	rqstp->rq_addrlen = sizeof(req->rq_xprt->addr);
 	memcpy(&rqstp->rq_addr, &req->rq_xprt->addr, rqstp->rq_addrlen);
net/sunrpc/svc_xprt.c +3 -2
@@ -469,10 +469,11 @@ out:
  */
 void svc_reserve(struct svc_rqst *rqstp, int space)
 {
+	struct svc_xprt *xprt = rqstp->rq_xprt;
+
 	space += rqstp->rq_res.head[0].iov_len;
 
-	if (space < rqstp->rq_reserved) {
-		struct svc_xprt *xprt = rqstp->rq_xprt;
+	if (xprt && space < rqstp->rq_reserved) {
 		atomic_sub((rqstp->rq_reserved - space), &xprt->xpt_reserved);
 		rqstp->rq_reserved = space;
 
net/sunrpc/svcsock.c +1 -1
@@ -1198,7 +1198,7 @@ static int svc_tcp_sendto(struct svc_rqs
 /*
  * Setup response header. TCP has a 4B record length field.
  */
-static void svc_tcp_prep_reply_hdr(struct svc_rqst *rqstp)
+void svc_tcp_prep_reply_hdr(struct svc_rqst *rqstp)
 {
 	struct kvec *resv = &rqstp->rq_res.head[0];
 
author maintainer
#46 lore raw

PATCH 46/50 KVM: arm/arm64: Fix VMID alloc race by reverting to lock-less

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Christoffer Dall <[email protected]>

commit fb544d1ca65a89f7a3895f7531221ceeed74ada7 upstream.

We recently addressed a VMID generation race by introducing a read/write lock around accesses and updates to the vmid generation values.

However, kvm_arch_vcpu_ioctl_run() also calls need_new_vmid_gen() but does so without taking the read lock.

As far as I can tell, this can lead to the same kind of race:

VM 0, VCPU 0 VM 0, VCPU 1
------------ ------------
update_vttbr (vmid 254)
update_vttbr (vmid 1) // roll over

				read_lock(kvm_vmid_lock);
				force_vm_exit()

local_irq_disable
need_new_vmid_gen == false //because vmid gen matches

enter_guest (vmid 254)
kvm_arch.vttbr = <PGD>:<VMID 1>

				read_unlock(kvm_vmid_lock);

enter_guest (vmid 1)

Which results in running two VCPUs in the same VM with different VMIDs and (even worse) other VCPUs from other VMs could now allocate clashing VMID 254 from the new generation as long as VCPU 0 is not exiting.

Attempt to solve this by making sure vttbr is updated before another CPU can observe the updated VMID generation.

Fixes f0cf47d939d0 "KVM: arm/arm64: Close VMID generation race"
---
 virt/kvm/arm/arm.c |   23 +++++++++++------------
 1 file changed, 11 insertions(+), 12 deletions(-)
virt/kvm/arm/arm.c +11 -12
@@ -66,7 +66,7 @@ static DEFINE_PER_CPU(struct kvm_vcpu *,
 static atomic64_t kvm_vmid_gen = ATOMIC64_INIT(1);
 static u32 kvm_next_vmid;
 static unsigned int kvm_vmid_bits __read_mostly;
-static DEFINE_RWLOCK(kvm_vmid_lock);
+static DEFINE_SPINLOCK(kvm_vmid_lock);
 
 static bool vgic_present;
 
@@ -482,7 +482,9 @@ void force_vm_exit(const cpumask_t *mask
  */
 static bool need_new_vmid_gen(struct kvm *kvm)
 {
-	return unlikely(kvm->arch.vmid_gen != atomic64_read(&kvm_vmid_gen));
+	u64 current_vmid_gen = atomic64_read(&kvm_vmid_gen);
+	smp_rmb(); /* Orders read of kvm_vmid_gen and kvm->arch.vmid */
+	return unlikely(READ_ONCE(kvm->arch.vmid_gen) != current_vmid_gen);
 }
 
 /**
@@ -497,16 +499,11 @@ static void update_vttbr(struct kvm *kvm
 {
 	phys_addr_t pgd_phys;
 	u64 vmid;
-	bool new_gen;
 
-	read_lock(&kvm_vmid_lock);
-	new_gen = need_new_vmid_gen(kvm);
-	read_unlock(&kvm_vmid_lock);
-
-	if (!new_gen)
+	if (!need_new_vmid_gen(kvm))
 		return;
 
-	write_lock(&kvm_vmid_lock);
+	spin_lock(&kvm_vmid_lock);
 
 	/*
 	 * We need to re-check the vmid_gen here to ensure that if another vcpu
@@ -514,7 +511,7 @@ static void update_vttbr(struct kvm *kvm
 	 * use the same vmid.
 	 */
 	if (!need_new_vmid_gen(kvm)) {
-		write_unlock(&kvm_vmid_lock);
+		spin_unlock(&kvm_vmid_lock);
 		return;
 	}
 
@@ -537,7 +534,6 @@ static void update_vttbr(struct kvm *kvm
 		kvm_call_hyp(__kvm_flush_vm_context);
 	}
 
-	kvm->arch.vmid_gen = atomic64_read(&kvm_vmid_gen);
 	kvm->arch.vmid = kvm_next_vmid;
 	kvm_next_vmid++;
 	kvm_next_vmid &= (1 << kvm_vmid_bits) - 1;
@@ -548,7 +544,10 @@ static void update_vttbr(struct kvm *kvm
 	vmid = ((u64)(kvm->arch.vmid) << VTTBR_VMID_SHIFT) & VTTBR_VMID_MASK(kvm_vmid_bits);
 	kvm->arch.vttbr = kvm_phys_to_vttbr(pgd_phys) | vmid;
 
-	write_unlock(&kvm_vmid_lock);
+	smp_wmb();
+	WRITE_ONCE(kvm->arch.vmid_gen, atomic64_read(&kvm_vmid_gen));
+
+	spin_unlock(&kvm_vmid_lock);
 }
 
 static int kvm_vcpu_first_run_init(struct kvm_vcpu *vcpu)
author maintainer
#47 lore raw

PATCH 47/50 arm64: compat: Dont pull syscall number from regs in arm_compat_syscall

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Will Deacon <[email protected]>

commit 53290432145a8eb143fe29e06e9c1465d43dc723 upstream.

The syscall number may have been changed by a tracer, so we should pass the actual number in from the caller instead of pulling it from the saved r7 value directly.

---
 arch/arm64/kernel/sys_compat.c |    9 ++++-----
 arch/arm64/kernel/syscall.c    |    9 ++++-----
 2 files changed, 8 insertions(+), 10 deletions(-)
arch/arm64/kernel/sys_compat.c +4 -5
@@ -66,12 +66,11 @@ do_compat_cache_op(unsigned long start,
 /*
  * Handle all unrecognised system calls.
  */
-long compat_arm_syscall(struct pt_regs *regs)
+long compat_arm_syscall(struct pt_regs *regs, int scno)
 {
 	siginfo_t info;
-	unsigned int no = regs->regs[7];
 
-	switch (no) {
+	switch (scno) {
 	/*
 	 * Flush a region from virtual address 'r0' to virtual address 'r1'
 	 * _exclusive_.  There is no alignment requirement on either address;
@@ -107,7 +106,7 @@ long compat_arm_syscall(struct pt_regs *
 		 * way the calling program can gracefully determine whether
 		 * a feature is supported.
 		 */
-		if (no < __ARM_NR_COMPAT_END)
+		if (scno < __ARM_NR_COMPAT_END)
 			return -ENOSYS;
 		break;
 	}
@@ -119,6 +118,6 @@ long compat_arm_syscall(struct pt_regs *
 	info.si_addr  = (void __user *)instruction_pointer(regs) -
 			 (compat_thumb_mode(regs) ? 2 : 4);
 
-	arm64_notify_die("Oops - bad compat syscall(2)", regs, &info, no);
+	arm64_notify_die("Oops - bad compat syscall(2)", regs, &info, scno);
 	return 0;
 }
arch/arm64/kernel/syscall.c +4 -5
@@ -13,16 +13,15 @@
 #include <asm/thread_info.h>
 #include <asm/unistd.h>
 
-long compat_arm_syscall(struct pt_regs *regs);
-
+long compat_arm_syscall(struct pt_regs *regs, int scno);
 long sys_ni_syscall(void);
 
-asmlinkage long do_ni_syscall(struct pt_regs *regs)
+static long do_ni_syscall(struct pt_regs *regs, int scno)
 {
 #ifdef CONFIG_COMPAT
 	long ret;
 	if (is_compat_task()) {
-		ret = compat_arm_syscall(regs);
+		ret = compat_arm_syscall(regs, scno);
 		if (ret != -ENOSYS)
 			return ret;
 	}
@@ -47,7 +46,7 @@ static void invoke_syscall(struct pt_reg
 		syscall_fn = syscall_table[array_index_nospec(scno, sc_nr)];
 		ret = __invoke_syscall(regs, syscall_fn);
 	} else {
-		ret = do_ni_syscall(regs);
+		ret = do_ni_syscall(regs, scno);
 	}
 
 	regs->regs[0] = ret;
author maintainer
#48 lore raw

PATCH 48/50 Btrfs: fix access to available allocation bits when starting balance

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Filipe Manana <[email protected]>

commit 5a8067c0d17feb7579db0476191417b441a8996e upstream.

The available allocation bits members from struct btrfs_fs_info are protected by a sequence lock, and when starting balance we access them incorrectly in two different ways:

1) In the read sequence lock loop at btrfs_balance() we use the values we read from fs_info->avail_*_alloc_bits and we can immediately do actions that have side effects and can not be undone (printing a message and jumping to a label). This is wrong because a retry might be needed, so our actions must not have side effects and must be repeatable as long as read_seqretry() returns a non-zero value. In other words, we were essentially ignoring the sequence lock;

2) Right below the read sequence lock loop, we were reading the values from avail_metadata_alloc_bits and avail_data_alloc_bits without any protection from concurrent writers, that is, reading them outside of the read sequence lock critical section.

So fix this by making sure we only read the available allocation bits while in a read sequence lock critical section and that what we do in the critical section is repeatable (has nothing that can not be undone) so that any eventual retry that is needed is handled properly.

Fixes de98ced9e743 ("Btrfs: use seqlock to protect fs_info->avail_{data, metadata, system}_alloc_bits")
Fixes 14506127979a ("btrfs: fix a bogus warning when converting only data or metadata")
---
 fs/btrfs/volumes.c |   39 +++++++++++++++++++++++----------------
 1 file changed, 23 insertions(+), 16 deletions(-)
fs/btrfs/volumes.c +23 -16
@@ -3712,6 +3712,7 @@ int btrfs_balance(struct btrfs_fs_info *
 	int ret;
 	u64 num_devices;
 	unsigned seq;
+	bool reducing_integrity;
 
 	if (btrfs_fs_closing(fs_info) ||
 	    atomic_read(&fs_info->balance_pause_req) ||
@@ -3796,24 +3797,30 @@ int btrfs_balance(struct btrfs_fs_info *
 		     !(bctl->sys.target & allowed)) ||
 		    ((bctl->meta.flags & BTRFS_BALANCE_ARGS_CONVERT) &&
 		     (fs_info->avail_metadata_alloc_bits & allowed) &&
-		     !(bctl->meta.target & allowed))) {
-			if (bctl->flags & BTRFS_BALANCE_FORCE) {
-				btrfs_info(fs_info,
-				"balance: force reducing metadata integrity");
-			} else {
-				btrfs_err(fs_info,
-	"balance: reduces metadata integrity, use --force if you want this");
-				ret = -EINVAL;
-				goto out;
-			}
-		}
+		     !(bctl->meta.target & allowed)))
+			reducing_integrity = true;
+		else
+			reducing_integrity = false;
+
+		/* if we're not converting, the target field is uninitialized */
+		meta_target = (bctl->meta.flags & BTRFS_BALANCE_ARGS_CONVERT) ?
+			bctl->meta.target : fs_info->avail_metadata_alloc_bits;
+		data_target = (bctl->data.flags & BTRFS_BALANCE_ARGS_CONVERT) ?
+			bctl->data.target : fs_info->avail_data_alloc_bits;
 	} while (read_seqretry(&fs_info->profiles_lock, seq));
 
-	/* if we're not converting, the target field is uninitialized */
-	meta_target = (bctl->meta.flags & BTRFS_BALANCE_ARGS_CONVERT) ?
-		bctl->meta.target : fs_info->avail_metadata_alloc_bits;
-	data_target = (bctl->data.flags & BTRFS_BALANCE_ARGS_CONVERT) ?
-		bctl->data.target : fs_info->avail_data_alloc_bits;
+	if (reducing_integrity) {
+		if (bctl->flags & BTRFS_BALANCE_FORCE) {
+			btrfs_info(fs_info,
+				   "balance: force reducing metadata integrity");
+		} else {
+			btrfs_err(fs_info,
+	  "balance: reduces metadata integrity, use --force if you want this");
+			ret = -EINVAL;
+			goto out;
+		}
+	}
+
 	if (btrfs_get_num_tolerated_disk_barrier_failures(meta_target) <
 		btrfs_get_num_tolerated_disk_barrier_failures(data_target)) {
 		int meta_index = btrfs_bg_flags_to_raid_index(meta_target);
author maintainer
#49 lore raw

PATCH 49/50 Btrfs: fix deadlock when enabling quotas due to concurrent snapshot creation

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Filipe Manana <[email protected]>

commit 9a6f209e36500efac51528132a3e3083586eda5f upstream.

If the quota enable and snapshot creation ioctls are called concurrently we can get into a deadlock where the task enabling quotas will deadlock on the fs_info->qgroup_ioctl_lock mutex because it attempts to lock it twice, or the task creating a snapshot tries to commit the transaction while the task enabling quota waits for the former task to commit the transaction while holding the mutex. The following time diagrams show how both cases happen.

First scenario:

CPU 0 CPU 1

btrfs_ioctl()
btrfs_ioctl_quota_ctl()
btrfs_quota_enable()

    mutex_lock(fs_info->qgroup_ioctl_lock)
    btrfs_start_transaction()

                                             btrfs_ioctl()
                                              btrfs_ioctl_snap_create_v2
                                               create_snapshot()
                                                --> adds snapshot to the
                                                    list pending_snapshots
                                                    of the current
                                                    transaction

    btrfs_commit_transaction()
     create_pending_snapshots()
       create_pending_snapshot()
        qgroup_account_snapshot()
         btrfs_qgroup_inherit()
	   mutex_lock(fs_info->qgroup_ioctl_lock)
	    --> deadlock, mutex already locked
	        by this task at
		btrfs_quota_enable()

Second scenario:

CPU 0 CPU 1

btrfs_ioctl()
btrfs_ioctl_quota_ctl()
btrfs_quota_enable()
mutex_lock(fs_info->qgroup_ioctl_lock)
btrfs_start_transaction()

btrfs_ioctl()
btrfs_ioctl_snap_create_v2
create_snapshot()
--> adds snapshot to the
list pending_snapshots of the current transaction

btrfs_commit_transaction()
--> waits for task at
CPU 0 to release its transaction handle

btrfs_commit_transaction()
--> sees another task started
the transaction commit first
--> releases its transaction
handle
--> waits for the transaction
commit to be completed by
the task at CPU 1

create_pending_snapshot()
qgroup_account_snapshot()
btrfs_qgroup_inherit()
mutex_lock(fs_info->qgroup_ioctl_lock)
--> deadlock, task at CPU 0
has the mutex locked but it is waiting for us to finish the transaction commit

So fix this by setting the quota enabled flag in fs_info after committing the transaction at btrfs_quota_enable(). This ends up serializing quota enable and snapshot creation as if the snapshot creation happened just before the quota enable request. The quota rescan task, scheduled after committing the transaction in btrfs_quote_enable(), will do the accounting.

Fixes 6426c7ad697d ("btrfs: qgroup: Fix qgroup accounting when creating snapshot")
---
 fs/btrfs/qgroup.c |   14 ++++++++++----
 1 file changed, 10 insertions(+), 4 deletions(-)
fs/btrfs/qgroup.c +10 -4
@@ -1013,16 +1013,22 @@ out_add_root:
 		btrfs_abort_transaction(trans, ret);
 		goto out_free_path;
 	}
-	spin_lock(&fs_info->qgroup_lock);
-	fs_info->quota_root = quota_root;
-	set_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags);
-	spin_unlock(&fs_info->qgroup_lock);
 
 	ret = btrfs_commit_transaction(trans);
 	trans = NULL;
 	if (ret)
 		goto out_free_path;
 
+	/*
+	 * Set quota enabled flag after committing the transaction, to avoid
+	 * deadlocks on fs_info->qgroup_ioctl_lock with concurrent snapshot
+	 * creation.
+	 */
+	spin_lock(&fs_info->qgroup_lock);
+	fs_info->quota_root = quota_root;
+	set_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags);
+	spin_unlock(&fs_info->qgroup_lock);
+
 	ret = qgroup_rescan_init(fs_info, 0, 1);
 	if (!ret) {
 	        qgroup_rescan_zero_tracking(fs_info);
author maintainer
#50 lore raw

PATCH 50/50 Btrfs: use nofs context when initializing security xattrs to avoid deadlock

In reply to Greg Kroah-Hartman

4.19-stable review patch. If anyone has any objections, please let me know.

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

From: Filipe Manana <[email protected]>

commit 827aa18e7b903c5ff3b3cd8fec328a99b1dbd411 upstream.

When initializing the security xattrs, we are holding a transaction handle therefore we need to use a GFP_NOFS context in order to avoid a deadlock with reclaim in case it's triggered.

Fixes 39a27ec1004e8 ("btrfs: use GFP_KERNEL for xattr and acl allocations")
---
 fs/btrfs/xattr.c |    8 ++++++++
 1 file changed, 8 insertions(+)
fs/btrfs/xattr.c +8 -0
@@ -11,6 +11,7 @@
 #include <linux/security.h>
 #include <linux/posix_acl_xattr.h>
 #include <linux/iversion.h>
+#include <linux/sched/mm.h>
 #include "ctree.h"
 #include "btrfs_inode.h"
 #include "transaction.h"
@@ -422,9 +423,15 @@ static int btrfs_initxattrs(struct inode
 {
 	const struct xattr *xattr;
 	struct btrfs_trans_handle *trans = fs_info;
+	unsigned int nofs_flag;
 	char *name;
 	int err = 0;
 
+	/*
+	 * We're holding a transaction handle, so use a NOFS memory allocation
+	 * context to avoid deadlock if reclaim happens.
+	 */
+	nofs_flag = memalloc_nofs_save();
 	for (xattr = xattr_array; xattr->name != NULL; xattr++) {
 		name = kmalloc(XATTR_SECURITY_PREFIX_LEN +
 			       strlen(xattr->name) + 1, GFP_KERNEL);
@@ -440,6 +447,7 @@ static int btrfs_initxattrs(struct inode
 		if (err < 0)
 			break;
 	}
+	memalloc_nofs_restore(nofs_flag);
 	return err;
 }
 
#51 lore raw
In reply to Greg Kroah-Hartman
Quoted 9 lines · Greg Kroah-Hartman

This is the start of the stable review cycle for the 4.19.16 release. There are 50 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 me know.

Responses should be made by Thu Jan 17 15:48:36 UTC 2019. Anything received after that time might be too late.

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

Quoted 9 lines
	https://www.kernel.org/pub/linux/kernel/v4.x/stable-review/patch-4.19.16-rc1.gz

or in the git tree and branch at:
git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git linux-4.19.y
and the diffstat can be found below.

thanks,

greg k-h

Compiled and booted on my test system. No dmesg regressions.

thanks,
-- Shuah

#52 lore raw
In reply to Greg Kroah-Hartman
Quoted 10 lines · Greg Kroah-Hartman

This is the start of the stable review cycle for the 4.19.16 release. There are 50 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 me know.

Responses should be made by Thu Jan 17 15:48:36 UTC 2019. Anything received after that time might be too late.

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

Quoted 8 lines
        https://www.kernel.org/pub/linux/kernel/v4.x/stable-review/patch-4.19.16-rc1.gz

or in the git tree and branch at:
git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git linux-4.19.y
and the diffstat can be found below.

thanks,

greg k-h

Results from Linaro’s test farm.
No regressions on arm64, arm, x86_64, and i386.

Summary
------------------------------------------------------------------------

kernel: 4.19.16-rc1
git repo: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git
git branch: linux-4.19.y
git commit: 3fd0fa89eb9f95b7989c649564ee07b6f36559cc
git describe: v4.19.15-51-g3fd0fa89eb9f
Test details: https://qa-reports.linaro.org/lkft/linux-stable-rc-4.19-oe/build/v4.19.15-51-g3fd0fa89eb9f

No regressions (compared to build v4.19.15)

No fixes (compared to build v4.19.15)

Ran 20640 total tests in the following environments and test suites.

Environments
--------------
- dragonboard-410c - arm64
- hi6220-hikey - arm64
- i386
- juno-r2 - arm64
- qemu_arm
- qemu_arm64
- qemu_i386
- qemu_x86_64
- x15 - arm
- x86_64

Test Suites
-----------
* boot
* install-android-platform-tools-r2600
* kselftest
* libhugetlbfs
* ltp-cap_bounds-tests
* ltp-containers-tests
* ltp-cpuhotplug-tests
* ltp-cve-tests
* ltp-fcntl-locktests-tests
* ltp-filecaps-tests
* ltp-fs-tests
* ltp-fs_bind-tests
* ltp-fs_perms_simple-tests
* ltp-fsx-tests
* ltp-hugetlb-tests
* ltp-io-tests
* ltp-ipc-tests
* ltp-math-tests
* ltp-nptl-tests
* ltp-pty-tests
* ltp-sched-tests
* ltp-securebits-tests
* ltp-syscalls-tests
* ltp-timers-tests
* spectre-meltdown-checker-test
* ltp-open-posix-tests
* kselftest-vsyscall-mode-native
* kselftest-vsyscall-mode-none

Signature · 2 lines
Linaro LKFT
https://lkft.linaro.org
#53 lore raw
In reply to Greg Kroah-Hartman
Quoted 8 lines · Greg Kroah-Hartman

This is the start of the stable review cycle for the 4.19.16 release. There are 50 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 me know.

Responses should be made by Thu Jan 17 15:48:36 UTC 2019. Anything received after that time might be too late.

Build results:

	total: 156 pass: 156 fail: 0
Qemu test results:
	total: 343 pass: 343 fail: 0

Guenter