Skip to contentSkip to conversation

irqchip/gic-v3: Memory-safety fixes on the probe paths

PATCH Applied Fuad Tabba 4 participants 13 messages Sep 25 – Oct 15

Subsystems (2)
lore
Share & tools
lore

Fuad Tabba

author
#0 lore raw

v2 cover irqchip/gic-v3: Memory-safety fixes on the probe paths

Hi folks,

Changes since v1 [1]:
- Patch 3 marks the skipped CPU in broken_rdists, so it's kept offline. (Marc, sashiko)

Three probe-time bugs in the GICv3 drivers. The kernel keeps running after each of them, so the memory involved is live rather than on the way to a reset.

The first two free memory an enabled ITS still points at. Boot carries on because its_init() discards a failed probe and gic_init_bases() discards what its_init() returns. The third is a write past the end of redist_regs[] for an enabled GICC with a zero GICR base, on a MADT with no GICR entries. fa2dabe57220e's commit message says such tables exist, and I haven't seen one.

Based on Linux 7.3-rc4 (93f51579e7df2).

Cheers,
/fuad

[1] https://lore.kernel.org/all/[email protected]/

Fuad Tabba (3):
irqchip/gic-v3-its: Don't free a vPE table shared with another ITS irqchip/gic-v3-its: Don't free the tables of an enabled ITS irqchip/gic-v3: Don't register a redistributor that was never counted

 drivers/irqchip/irq-gic-v3-its.c | 16 +++++++++++-----
 drivers/irqchip/irq-gic-v3.c     | 11 +++++++++++
 2 files changed, 22 insertions(+), 5 deletions(-)

base-commit: 93f51579e7df248780214094418f205253383cc5

Signature · 1 line
2.39.5
author
#1 lore raw

PATCH v2 1/3 irqchip/gic-v3-its: Don't free a vPE table shared with another ITS

On GICv4.1, its_alloc_tables() copies a sibling's vPE table baser rather than allocating one, but its_free_tables() frees it like the rest. A probe failure on the copying ITS then frees pages the enabled sibling still points GITS_BASER2 at, and the kernel comes up running on that sibling.

Mark a shared table so its_free_tables() skips it.

Fixes 5e5168461c22c ("irqchip/gic-v4.1: VPE table (aka GICR_VPROPBASER) allocation")
---
 drivers/irqchip/irq-gic-v3-its.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/irqchip/irq-gic-v3-its.c +5 -1
index e9807af235373..58e52095e6ea8 100644
@@ -87,6 +87,8 @@ struct its_baser {
 	u64		val;
 	u32		order;
 	u32		psz;
+	/* Inherited from a sibling ITS, not freed here */
+	bool		shared;
 };
 
 struct its_device;
@@ -2604,7 +2606,8 @@ static void its_free_tables(struct its_node *its)
 
 	for (i = 0; i < GITS_BASER_NR_REGS; i++) {
 		if (its->tables[i].base) {
-			its_free_pages(its->tables[i].base, its->tables[i].order);
+			if (!its->tables[i].shared)
+				its_free_pages(its->tables[i].base, its->tables[i].order);
 			its->tables[i].base = NULL;
 		}
 	}
@@ -2703,6 +2706,7 @@ static int its_alloc_tables(struct its_node *its)
 				WARN_ON(i != 2);
 				if ((sibling = find_sibling_its(its))) {
 					*baser = sibling->tables[2];
+					baser->shared = true;
 					its_write_baser(its, baser, baser->val);
 					continue;
 				}
Signature · 1 line
2.39.5
author
#4 lore raw

PATCH v2 2/3 irqchip/gic-v3-its: Don't free the tables of an enabled ITS

In reply to Fuad Tabba

its_probe_one() enables the ITS before its_init_domain(), the last step that can fail, and the unwind from there frees the tables and the command queue without disabling it. The kernel comes up with an enabled ITS pointing at freed pages.

Enable the ITS last. Nothing reaches the new domain before then: this runs from init_IRQ(), and the ITS joins its_nodes only after the enable.

Fixes 4c21f3c26ecc2 ("irqchip: GICv3: ITS: DT probing and initialization")
---
 drivers/irqchip/irq-gic-v3-its.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)
drivers/irqchip/irq-gic-v3-its.c +6 -4
index 58e52095e6ea8..a5690721f0feb 100644
@@ -5332,16 +5332,18 @@ static int __init its_probe_one(struct its_node *its)
 	}
 
 	gits_write_cwriter(0, its->base + GITS_CWRITER);
+
+	err = its_init_domain(its);
+	if (err)
+		goto out_free_collection;
+
+	/* Enable last: the unwind frees memory an enabled ITS may access */
 	ctlr = readl_relaxed(its->base + GITS_CTLR);
 	ctlr |= GITS_CTLR_ENABLE;
 	if (is_v4(its))
 		ctlr |= GITS_CTLR_ImDe;
 	writel_relaxed(ctlr, its->base + GITS_CTLR);
 
-	err = its_init_domain(its);
-	if (err)
-		goto out_free_collection;
-
 	raw_spin_lock(&its_lock);
 	list_add(&its->entry, &its_nodes);
 	raw_spin_unlock(&its_lock);
Signature · 1 line
2.39.5
author
#7 lore raw

PATCH v2 3/3 irqchip/gic-v3: Don't register a redistributor that was never counted

In reply to Fuad Tabba

gic_acpi_match_gicc() counts only the enabled GICCs with a non-zero gicr_base_address, and that count sizes redist_regs[], but gic_acpi_parse_madt_gicc() registers every enabled one. On a MADT with no GICR entries, an enabled GICC with a zero GICR base therefore makes gic_acpi_register_redist() write a struct redist_region past the end of the array.

Commit fa2dabe57220e ("irqchip/gic-v3: Don't return errors from gic_acpi_match_gicc()") removed the check that kept the two consistent. Its message says such entries are still caught by gic_populate_rdist(), but that runs from gic_cpu_init(), after the write.

Skip the entry instead, and mark the CPU's redistributor broken as the online-capable path does, so that gic_check_rdist() rejects the CPU. gic_starting_cpu() runs past cpuhp's last failure point, so without the mark the CPU would come online with its redistributor and CPU interface unconfigured. With an ITS, its_cpu_init() would then dereference a NULL rd_base.

Fixes fa2dabe57220e ("irqchip/gic-v3: Don't return errors from gic_acpi_match_gicc()")
---
 drivers/irqchip/irq-gic-v3.c | 11 +++++++++++
 1 file changed, 11 insertions(+)
drivers/irqchip/irq-gic-v3.c +11 -0
index 6e1fa5b247fc4..bfc10d657fd57 100644
@@ -2344,6 +2344,17 @@ gic_acpi_parse_madt_gicc(union acpi_subtable_headers *header,
 		return 0;
 	}
 
+	/* Not counted by gic_acpi_match_gicc(), so there is no slot for it */
+	if (!gicc->gicr_base_address) {
+		int cpu = get_cpu_for_acpi_id(gicc->uid);
+
+		pr_warn(FW_BUG "GICC entry with ACPI UID %u has no GICR base address, CPU kept offline\n",
+			gicc->uid);
+		if (cpu >= 0)
+			cpumask_set_cpu(cpu, &broken_rdists);
+		return 0;
+	}
+
 	redist_base = ioremap(gicc->gicr_base_address, size);
 	if (!redist_base)
 		return -ENOMEM;
Signature · 1 line
2.39.5
maintainer Reviewed-by on the v2 series
#10 lore raw
In reply to Fuad Tabba
Quoted 30 lines · Fuad Tabba

Hi folks,

Changes since v1 [1]:
- Patch 3 marks the skipped CPU in broken_rdists, so it's kept offline. (Marc, sashiko)

Three probe-time bugs in the GICv3 drivers. The kernel keeps running after each of them, so the memory involved is live rather than on the way to a reset.

The first two free memory an enabled ITS still points at. Boot carries on because its_init() discards a failed probe and gic_init_bases() discards what its_init() returns. The third is a write past the end of redist_regs[] for an enabled GICC with a zero GICR base, on a MADT with no GICR entries. fa2dabe57220e's commit message says such tables exist, and I haven't seen one.

Based on Linux 7.3-rc4 (93f51579e7df2).

Cheers,
/fuad

[1] https://lore.kernel.org/all/[email protected]/

Fuad Tabba (3):
irqchip/gic-v3-its: Don't free a vPE table shared with another ITS irqchip/gic-v3-its: Don't free the tables of an enabled ITS irqchip/gic-v3: Don't register a redistributor that was never counted

Quoted 6 lines
 drivers/irqchip/irq-gic-v3-its.c | 16 +++++++++++-----
 drivers/irqchip/irq-gic-v3.c     | 11 +++++++++++
 2 files changed, 22 insertions(+), 5 deletions(-)

base-commit: 93f51579e7df248780214094418f205253383cc5

	M.
Signature · 1 line
Jazz isn't dead. It just smells funny.
maintainer
#12 lore raw
Quoted 4 lines · Fuad Tabba

Hi Thomas,

Gentle ping on this series, which Marc has reviewed, and on the lifetime and range fixes [1], which he has also reviewed.

All of that is in my todo list which contains the backlog of vacation and conference travel. I'll get there in the foreseeable future...

merge bot Applied
#8 lore raw

[tip: irq/drivers] irqchip/gic-v3: Don't register a redistributor that was never counted

In reply to Fuad Tabba
The following commit has been merged into the irq/drivers branch of tip:

The following commit has been merged into the irq/drivers branch of tip:

Commit-ID:     7bcd2fa22186587e06c456d1c2e3d336c83136f6
Gitweb:        https://git.kernel.org/tip/7bcd2fa22186587e06c456d1c2e3d336c83136f6
Author:        Fuad Tabba <[email protected]>
AuthorDate:    Fri, 25 Sep 2026 09:50:00 +01:00
Committer:     Thomas Gleixner <[email protected]>
CommitterDate: Thu, 15 Oct 2026 11:50:48 +02:00

irqchip/gic-v3: Don't register a redistributor that was never counted

gic_acpi_match_gicc() counts only the enabled GICCs with a non-zero gicr_base_address, and that count sizes redist_regs[], but gic_acpi_parse_madt_gicc() registers every enabled one. On a MADT with no GICR entries, an enabled GICC with a zero GICR base therefore makes gic_acpi_register_redist() write a struct redist_region past the end of the array.

Commit fa2dabe57220e ("irqchip/gic-v3: Don't return errors from gic_acpi_match_gicc()") removed the check that kept the two consistent. Its message says such entries are still caught by gic_populate_rdist(), but that runs from gic_cpu_init(), after the write.

Skip the entry instead, and mark the CPU's redistributor broken as the online-capable path does, so that gic_check_rdist() rejects the CPU. gic_starting_cpu() runs past cpuhp's last failure point, so without the mark the CPU would come online with its redistributor and CPU interface unconfigured. With an ITS, its_cpu_init() would then dereference a NULL rd_base.

Fixes fa2dabe57220e ("irqchip/gic-v3: Don't return errors from gic_acpi_match_gicc()")
---
 drivers/irqchip/irq-gic-v3.c | 11 +++++++++++
 1 file changed, 11 insertions(+)
drivers/irqchip/irq-gic-v3.c +11 -0
index 9231688..32a8ea4 100644
@@ -2344,6 +2344,17 @@ gic_acpi_parse_madt_gicc(union acpi_subtable_headers *header,
 		return 0;
 	}
 
+	/* Not counted by gic_acpi_match_gicc(), so there is no slot for it */
+	if (!gicc->gicr_base_address) {
+		int cpu = get_cpu_for_acpi_id(gicc->uid);
+
+		pr_warn(FW_BUG "GICC entry with ACPI UID %u has no GICR base address, CPU kept offline\n",
+			gicc->uid);
+		if (cpu >= 0)
+			cpumask_set_cpu(cpu, &broken_rdists);
+		return 0;
+	}
+
 	redist_base = ioremap(gicc->gicr_base_address, size);
 	if (!redist_base)
 		return -ENOMEM;
merge bot Applied
#5 lore raw

[tip: irq/drivers] irqchip/gic-v3-its: Don't free the tables of an enabled ITS

In reply to Fuad Tabba
The following commit has been merged into the irq/drivers branch of tip:

The following commit has been merged into the irq/drivers branch of tip:

Commit-ID:     a5b8c23c5c93438d62f81a034fa4e39001d283fb
Gitweb:        https://git.kernel.org/tip/a5b8c23c5c93438d62f81a034fa4e39001d283fb
Author:        Fuad Tabba <[email protected]>
AuthorDate:    Fri, 25 Sep 2026 09:49:59 +01:00
Committer:     Thomas Gleixner <[email protected]>
CommitterDate: Thu, 15 Oct 2026 11:50:48 +02:00

irqchip/gic-v3-its: Don't free the tables of an enabled ITS

its_probe_one() enables the ITS before its_init_domain(), the last step that can fail, and the unwind from there frees the tables and the command queue without disabling it. The kernel comes up with an enabled ITS pointing at freed pages.

Enable the ITS last. Nothing reaches the new domain before then: this runs from init_IRQ(), and the ITS joins its_nodes only after the enable.

Fixes 4c21f3c26ecc2 ("irqchip: GICv3: ITS: DT probing and initialization")
---
 drivers/irqchip/irq-gic-v3-its.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)
drivers/irqchip/irq-gic-v3-its.c +6 -4
index 259ec0c..282e4ba 100644
@@ -5337,16 +5337,18 @@ static int __init its_probe_one(struct its_node *its)
 	}
 
 	gits_write_cwriter(0, its->base + GITS_CWRITER);
+
+	err = its_init_domain(its);
+	if (err)
+		goto out_free_collection;
+
+	/* Enable last: the unwind frees memory an enabled ITS may access */
 	ctlr = readl_relaxed(its->base + GITS_CTLR);
 	ctlr |= GITS_CTLR_ENABLE;
 	if (is_v4(its))
 		ctlr |= GITS_CTLR_ImDe;
 	writel_relaxed(ctlr, its->base + GITS_CTLR);
 
-	err = its_init_domain(its);
-	if (err)
-		goto out_free_collection;
-
 	raw_spin_lock(&its_lock);
 	list_add(&its->entry, &its_nodes);
 	raw_spin_unlock(&its_lock);
merge bot Applied
#2 lore raw

[tip: irq/drivers] irqchip/gic-v3-its: Don't free a vPE table shared with another ITS

In reply to Fuad Tabba
The following commit has been merged into the irq/drivers branch of tip:

The following commit has been merged into the irq/drivers branch of tip:

Commit-ID:     496375c937cc76d36bacaf62b4d599c3ee41115d
Gitweb:        https://git.kernel.org/tip/496375c937cc76d36bacaf62b4d599c3ee41115d
Author:        Fuad Tabba <[email protected]>
AuthorDate:    Fri, 25 Sep 2026 09:49:58 +01:00
Committer:     Thomas Gleixner <[email protected]>
CommitterDate: Thu, 15 Oct 2026 11:50:48 +02:00

irqchip/gic-v3-its: Don't free a vPE table shared with another ITS

On GICv4.1, its_alloc_tables() copies a sibling's vPE table baser rather than allocating one, but its_free_tables() frees it like the rest. A probe failure on the copying ITS then frees pages the enabled sibling still points GITS_BASER2 at, and the kernel comes up running on that sibling.

Mark a shared table so its_free_tables() skips it.

Fixes 5e5168461c22c ("irqchip/gic-v4.1: VPE table (aka GICR_VPROPBASER) allocation")
---
 drivers/irqchip/irq-gic-v3-its.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/irqchip/irq-gic-v3-its.c +5 -1
index 872d342..259ec0c 100644
@@ -87,6 +87,8 @@ struct its_baser {
 	u64		val;
 	u32		order;
 	u32		psz;
+	/* Inherited from a sibling ITS, not freed here */
+	bool		shared;
 };
 
 struct its_device;
@@ -2605,7 +2607,8 @@ static void its_free_tables(struct its_node *its)
 
 	for (i = 0; i < GITS_BASER_NR_REGS; i++) {
 		if (its->tables[i].base) {
-			its_free_pages(its->tables[i].base, its->tables[i].order);
+			if (!its->tables[i].shared)
+				its_free_pages(its->tables[i].base, its->tables[i].order);
 			its->tables[i].base = NULL;
 		}
 	}
@@ -2704,6 +2707,7 @@ static int its_alloc_tables(struct its_node *its)
 				WARN_ON(i != 2);
 				if ((sibling = find_sibling_its(its))) {
 					*baser = sibling->tables[2];
+					baser->shared = true;
 					its_write_baser(its, baser, baser->val);
 					continue;
 				}
merge bot Applied
#6 lore raw

[tip: irq/drivers] irqchip/gic-v3-its: Don't free the tables of an enabled ITS

In reply to Fuad Tabba
The following commit has been merged into the irq/drivers branch of tip:

The following commit has been merged into the irq/drivers branch of tip:

Commit-ID:     d09b36d8e0e60f35bd1cd7d63aadc6dfad61700c
Gitweb:        https://git.kernel.org/tip/d09b36d8e0e60f35bd1cd7d63aadc6dfad61700c
Author:        Fuad Tabba <[email protected]>
AuthorDate:    Fri, 25 Sep 2026 09:49:59 +01:00
Committer:     Thomas Gleixner <[email protected]>
CommitterDate: Thu, 15 Oct 2026 11:55:41 +02:00

irqchip/gic-v3-its: Don't free the tables of an enabled ITS

its_probe_one() enables the ITS before its_init_domain(), the last step that can fail, and the unwind from there frees the tables and the command queue without disabling it. The kernel comes up with an enabled ITS pointing at freed pages.

Enable the ITS last. Nothing reaches the new domain before then: this runs from init_IRQ(), and the ITS joins its_nodes only after the enable.

Fixes 4c21f3c26ecc2 ("irqchip: GICv3: ITS: DT probing and initialization")
---
 drivers/irqchip/irq-gic-v3-its.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)
drivers/irqchip/irq-gic-v3-its.c +6 -4
index 259ec0c..282e4ba 100644
@@ -5337,16 +5337,18 @@ static int __init its_probe_one(struct its_node *its)
 	}
 
 	gits_write_cwriter(0, its->base + GITS_CWRITER);
+
+	err = its_init_domain(its);
+	if (err)
+		goto out_free_collection;
+
+	/* Enable last: the unwind frees memory an enabled ITS may access */
 	ctlr = readl_relaxed(its->base + GITS_CTLR);
 	ctlr |= GITS_CTLR_ENABLE;
 	if (is_v4(its))
 		ctlr |= GITS_CTLR_ImDe;
 	writel_relaxed(ctlr, its->base + GITS_CTLR);
 
-	err = its_init_domain(its);
-	if (err)
-		goto out_free_collection;
-
 	raw_spin_lock(&its_lock);
 	list_add(&its->entry, &its_nodes);
 	raw_spin_unlock(&its_lock);
merge bot Applied
#9 lore raw

[tip: irq/drivers] irqchip/gic-v3: Don't register a redistributor that was never counted

In reply to Fuad Tabba
The following commit has been merged into the irq/drivers branch of tip:

The following commit has been merged into the irq/drivers branch of tip:

Commit-ID:     6a21693b97ee5cb0a21d04bfd051a05dde5aa783
Gitweb:        https://git.kernel.org/tip/6a21693b97ee5cb0a21d04bfd051a05dde5aa783
Author:        Fuad Tabba <[email protected]>
AuthorDate:    Fri, 25 Sep 2026 09:50:00 +01:00
Committer:     Thomas Gleixner <[email protected]>
CommitterDate: Thu, 15 Oct 2026 11:55:41 +02:00

irqchip/gic-v3: Don't register a redistributor that was never counted

gic_acpi_match_gicc() counts only the enabled GICCs with a non-zero gicr_base_address, and that count sizes redist_regs[], but gic_acpi_parse_madt_gicc() registers every enabled one. On a MADT with no GICR entries, an enabled GICC with a zero GICR base therefore makes gic_acpi_register_redist() write a struct redist_region past the end of the array.

Commit fa2dabe57220e ("irqchip/gic-v3: Don't return errors from gic_acpi_match_gicc()") removed the check that kept the two consistent. Its message says such entries are still caught by gic_populate_rdist(), but that runs from gic_cpu_init(), after the write.

Skip the entry instead, and mark the CPU's redistributor broken as the online-capable path does, so that gic_check_rdist() rejects the CPU. gic_starting_cpu() runs past cpuhp's last failure point, so without the mark the CPU would come online with its redistributor and CPU interface unconfigured. With an ITS, its_cpu_init() would then dereference a NULL rd_base.

Fixes fa2dabe57220e ("irqchip/gic-v3: Don't return errors from gic_acpi_match_gicc()")
---
 drivers/irqchip/irq-gic-v3.c | 11 +++++++++++
 1 file changed, 11 insertions(+)
drivers/irqchip/irq-gic-v3.c +11 -0
index 9231688..32a8ea4 100644
@@ -2344,6 +2344,17 @@ gic_acpi_parse_madt_gicc(union acpi_subtable_headers *header,
 		return 0;
 	}
 
+	/* Not counted by gic_acpi_match_gicc(), so there is no slot for it */
+	if (!gicc->gicr_base_address) {
+		int cpu = get_cpu_for_acpi_id(gicc->uid);
+
+		pr_warn(FW_BUG "GICC entry with ACPI UID %u has no GICR base address, CPU kept offline\n",
+			gicc->uid);
+		if (cpu >= 0)
+			cpumask_set_cpu(cpu, &broken_rdists);
+		return 0;
+	}
+
 	redist_base = ioremap(gicc->gicr_base_address, size);
 	if (!redist_base)
 		return -ENOMEM;
merge bot Applied
#3 lore raw

[tip: irq/drivers] irqchip/gic-v3-its: Don't free a vPE table shared with another ITS

In reply to Fuad Tabba
The following commit has been merged into the irq/drivers branch of tip:

The following commit has been merged into the irq/drivers branch of tip:

Commit-ID:     32e056cc1d1582951f981113ca8b98914f29c32a
Gitweb:        https://git.kernel.org/tip/32e056cc1d1582951f981113ca8b98914f29c32a
Author:        Fuad Tabba <[email protected]>
AuthorDate:    Fri, 25 Sep 2026 09:49:58 +01:00
Committer:     Thomas Gleixner <[email protected]>
CommitterDate: Thu, 15 Oct 2026 11:55:41 +02:00

irqchip/gic-v3-its: Don't free a vPE table shared with another ITS

On GICv4.1, its_alloc_tables() copies a sibling's vPE table baser rather than allocating one, but its_free_tables() frees it like the rest. A probe failure on the copying ITS then frees pages the enabled sibling still points GITS_BASER2 at, and the kernel comes up running on that sibling.

Mark a shared table so its_free_tables() skips it.

Fixes 5e5168461c22c ("irqchip/gic-v4.1: VPE table (aka GICR_VPROPBASER) allocation")
---
 drivers/irqchip/irq-gic-v3-its.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/irqchip/irq-gic-v3-its.c +5 -1
index 872d342..259ec0c 100644
@@ -87,6 +87,8 @@ struct its_baser {
 	u64		val;
 	u32		order;
 	u32		psz;
+	/* Inherited from a sibling ITS, not freed here */
+	bool		shared;
 };
 
 struct its_device;
@@ -2605,7 +2607,8 @@ static void its_free_tables(struct its_node *its)
 
 	for (i = 0; i < GITS_BASER_NR_REGS; i++) {
 		if (its->tables[i].base) {
-			its_free_pages(its->tables[i].base, its->tables[i].order);
+			if (!its->tables[i].shared)
+				its_free_pages(its->tables[i].base, its->tables[i].order);
 			its->tables[i].base = NULL;
 		}
 	}
@@ -2704,6 +2707,7 @@ static int its_alloc_tables(struct its_node *its)
 				WARN_ON(i != 2);
 				if ((sibling = find_sibling_its(its))) {
 					*baser = sibling->tables[2];
+					baser->shared = true;
 					its_write_baser(its, baser, baser->val);
 					continue;
 				}