Skip to contentSkip to conversation

qed: Fix the VF msix vectors flow

PATCH New Sasha Levin 1 participant 5 messages Aug 30, 2021

Subsystems (5)
lore
Share & tools
lore

Sasha Levin

author maintainer
#0 lore raw

PATCH 1/5 qed: Fix the VF msix vectors flow

From: Shai Malin <[email protected]>

[ Upstream commit b0cd08537db8d2fbb227cdb2e5835209db295a24 ]

For VFs we should return with an error in case we didn't get the exact number of msix vectors as we requested.
Not doing that will lead to a crash when starting queues for this VF.

---
 drivers/net/ethernet/qlogic/qed/qed_main.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)
drivers/net/ethernet/qlogic/qed/qed_main.c +6 -1
index 708117fc6f73..7669d36151c6 100644
@@ -374,7 +374,12 @@ static int qed_enable_msix(struct qed_dev *cdev,
 			rc = cnt;
 	}
 
-	if (rc > 0) {
+	/* For VFs, we should return with an error in case we didn't get the
+	 * exact number of msix vectors as we requested.
+	 * Not doing that will lead to a crash when starting queues for
+	 * this VF.
+	 */
+	if ((IS_PF(cdev) && rc > 0) || (IS_VF(cdev) && rc == cnt)) {
 		/* MSI-x configuration was achieved */
 		int_params->out.int_mode = QED_INT_MODE_MSIX;
 		int_params->out.num_vectors = rc;
Signature · 1 line
2.30.2
author maintainer
#1 lore raw

PATCH 2/5 qede: Fix memset corruption

From: Shai Malin <[email protected]>

[ Upstream commit e543468869e2532f5d7926e8f417782b48eca3dc ]

Thanks to Kees Cook who detected the problem of memset that starting from not the first member, but sized for the whole struct. The better change will be to remove the redundant memset and to clear only the msix_cnt member.

---
 drivers/net/ethernet/qlogic/qede/qede_main.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/qlogic/qede/qede_main.c +1 -1
index 9b1920b58594..d21a73bc4cde 100644
@@ -3145,6 +3145,7 @@ static void qede_sync_free_irqs(struct qede_dev *edev)
 	}
 
 	edev->int_info.used_cnt = 0;
+	edev->int_info.msix_cnt = 0;
 }
 
 static int qede_req_msix_irqs(struct qede_dev *edev)
@@ -3644,7 +3645,6 @@ static int qede_load(struct qede_dev *edev, enum qede_load_mode mode)
 
 err4:
 	qede_sync_free_irqs(edev);
-	memset(&edev->int_info.msix_cnt, 0, sizeof(struct qed_int_info));
 err3:
 	qede_napi_disable_remove(edev);
 err2:
Signature · 1 line
2.30.2
author maintainer
#2 lore raw

PATCH 3/5 perf/x86/amd/ibs: Work around erratum #1197

In reply to Sasha Levin

From: Kim Phillips <[email protected]>

[ Upstream commit 26db2e0c51fe83e1dd852c1321407835b481806e ]

Erratum #1197 "IBS (Instruction Based Sampling) Register State May be Incorrect After Restore From CC6" is published in a document:

"Revision Guide for AMD Family 19h Models 00h-0Fh Processors" 56683 Rev. 1.04 July 2021

https://bugzilla.kernel.org/show_bug.cgi?id=206537

Implement the erratum's suggested workaround and ignore IBS samples if MSRC001_1031 == 0.

---
 arch/x86/events/amd/ibs.c | 8 ++++++++
 1 file changed, 8 insertions(+)
arch/x86/events/amd/ibs.c +8 -0
index f6e57bebbc6b..5c6aad14cdd9 100644
@@ -89,6 +89,7 @@ struct perf_ibs {
 	unsigned long			offset_mask[1];
 	int				offset_max;
 	unsigned int			fetch_count_reset_broken : 1;
+	unsigned int			fetch_ignore_if_zero_rip : 1;
 	struct cpu_perf_ibs __percpu	*pcpu;
 
 	struct attribute		**format_attrs;
@@ -673,6 +674,10 @@ static int perf_ibs_handle_irq(struct perf_ibs *perf_ibs, struct pt_regs *iregs)
 	if (check_rip && (ibs_data.regs[2] & IBS_RIP_INVALID)) {
 		regs.flags &= ~PERF_EFLAGS_EXACT;
 	} else {
+		/* Workaround for erratum #1197 */
+		if (perf_ibs->fetch_ignore_if_zero_rip && !(ibs_data.regs[1]))
+			goto out;
+
 		set_linear_ip(&regs, ibs_data.regs[1]);
 		regs.flags |= PERF_EFLAGS_EXACT;
 	}
@@ -766,6 +771,9 @@ static __init void perf_event_ibs_init(void)
 	if (boot_cpu_data.x86 >= 0x16 && boot_cpu_data.x86 <= 0x18)
 		perf_ibs_fetch.fetch_count_reset_broken = 1;
 
+	if (boot_cpu_data.x86 == 0x19 && boot_cpu_data.x86_model < 0x10)
+		perf_ibs_fetch.fetch_ignore_if_zero_rip = 1;
+
 	perf_ibs_pmu_init(&perf_ibs_fetch, "ibs_fetch");
 
 	if (ibs_caps & IBS_CAPS_OPCNT) {
Signature · 1 line
2.30.2
author maintainer
#3 lore raw

PATCH 4/5 net: fix NULL pointer reference in cipso_v4_doi_free

In reply to Sasha Levin

From: 王贇 <[email protected]>

[ Upstream commit 733c99ee8be9a1410287cdbb943887365e83b2d6 ]

In netlbl_cipsov4_add_std() when 'doi_def->map.std' alloc failed, we sometime observe panic:

  BUG: kernel NULL pointer dereference, address:
  ...
  RIP: 0010:cipso_v4_doi_free+0x3a/0x80
  ...
  Call Trace:
   netlbl_cipsov4_add_std+0xf4/0x8c0
   netlbl_cipsov4_add+0x13f/0x1b0
   genl_family_rcv_msg_doit.isra.15+0x132/0x170
   genl_rcv_msg+0x125/0x240

This is because in cipso_v4_doi_free() there is no check on 'doi_def->map.std' when 'doi_def->type' equal 1, which is possibe, since netlbl_cipsov4_add_std() haven't initialize it before alloc 'doi_def->map.std'.

This patch just add the check to prevent panic happen for similar cases.

---
 net/ipv4/cipso_ipv4.c | 18 ++++++++++--------
 1 file changed, 10 insertions(+), 8 deletions(-)
net/ipv4/cipso_ipv4.c +10 -8
index 553cda6f887a..160800d21f10 100644
@@ -480,14 +480,16 @@ void cipso_v4_doi_free(struct cipso_v4_doi *doi_def)
 	if (!doi_def)
 		return;
 
-	switch (doi_def->type) {
-	case CIPSO_V4_MAP_TRANS:
-		kfree(doi_def->map.std->lvl.cipso);
-		kfree(doi_def->map.std->lvl.local);
-		kfree(doi_def->map.std->cat.cipso);
-		kfree(doi_def->map.std->cat.local);
-		kfree(doi_def->map.std);
-		break;
+	if (doi_def->map.std) {
+		switch (doi_def->type) {
+		case CIPSO_V4_MAP_TRANS:
+			kfree(doi_def->map.std->lvl.cipso);
+			kfree(doi_def->map.std->lvl.local);
+			kfree(doi_def->map.std->cat.cipso);
+			kfree(doi_def->map.std->cat.local);
+			kfree(doi_def->map.std);
+			break;
+		}
 	}
 	kfree(doi_def);
 }
Signature · 1 line
2.30.2
author maintainer
#4 lore raw

PATCH 5/5 cryptoloop: add a deprecation warning

In reply to Sasha Levin

From: Christoph Hellwig <[email protected]>

[ Upstream commit 222013f9ac30b9cec44301daa8dbd0aae38abffb ]

Support for cryptoloop has been officially marked broken and deprecated in favor of dm-crypt (which supports the same broken algorithms if needed) in Linux 2.6.4 (released in March 2004), and support for it has been entirely removed from losetup in util-linux 2.23 (released in April 2013). Add a warning and a deprecation schedule.

---
 drivers/block/Kconfig      | 4 ++--
 drivers/block/cryptoloop.c | 2 ++
 2 files changed, 4 insertions(+), 2 deletions(-)
drivers/block/Kconfig +2 -2
index 894102fd5a06..b701c79f07e5 100644
@@ -257,7 +257,7 @@ config BLK_DEV_LOOP_MIN_COUNT
 	  dynamically allocated with the /dev/loop-control interface.
 
 config BLK_DEV_CRYPTOLOOP
-	tristate "Cryptoloop Support"
+	tristate "Cryptoloop Support (DEPRECATED)"
 	select CRYPTO
 	select CRYPTO_CBC
 	depends on BLK_DEV_LOOP
@@ -269,7 +269,7 @@ config BLK_DEV_CRYPTOLOOP
 	  WARNING: This device is not safe for journaled file systems like
 	  ext3 or Reiserfs. Please use the Device Mapper crypto module
 	  instead, which can be configured to be on-disk compatible with the
-	  cryptoloop device.
+	  cryptoloop device.  cryptoloop support will be removed in Linux 5.16.
 
 source "drivers/block/drbd/Kconfig"
 
drivers/block/cryptoloop.c +2 -0
index 3d31761c0ed0..adbfd3e2a60f 100644
@@ -203,6 +203,8 @@ init_cryptoloop(void)
 
 	if (rc)
 		printk(KERN_ERR "cryptoloop: loop_register_transfer failed\n");
+	else
+		pr_warn("the cryptoloop driver has been deprecated and will be removed in in Linux 5.16\n");
 	return rc;
 }
 
Signature · 1 line
2.30.2