From: Matt Redfearn <matt.redfearn@imgtec.com>
To: Thomas Gleixner <tglx@linutronix.de>,
Arvind Yadav <arvind.yadav.cs@gmail.com>
Cc: Jason Cooper <jason@lakedaemon.net>,
Marc Zyngier <marc.zyngier@arm.com>,
LKML <linux-kernel@vger.kernel.org>,
Paul Burton <paul.burton@imgtec.com>, <linux-mips@linux-mips.org>
Subject: Re: [PATCH v2] irqchip: irq-mips-gic:- Avoiding Kernel panics due to Error.
Date: Tue, 17 Jan 2017 09:45:48 +0000 [thread overview]
Message-ID: <f308f402-70d3-a7a1-9cc8-fcf8a14aa741@imgtec.com> (raw)
In-Reply-To: <alpine.DEB.2.20.1701170936410.3495@nanos>
On 17/01/17 08:38, Thomas Gleixner wrote:
> On Mon, 16 Jan 2017, Arvind Yadav wrote:
>
> Cc'ing the MIPS folks.
>
>> Eliminating kernel panic due to NULL pointer dereferencing and
>> other failure in __gic_init.
>> Here, __gic_init can fail. This error check will avoid NULL pointer
>> dereference and kernel panic.
Hi Aravind,
While panicing due to a null dereference is not ideal, I don't think
simply printing an error an attempting to continue on without the GIC is
the way forward. If the GIC fails to initialize, there will be no
interrupts, from devices, timers, IPI's etc, in other words the system
will be completely broken.
The driver should panic since the system will not be usable if the set
up fails.
Your v1 https://lkml.org/lkml/2017/1/9/106 would be fine.
>>
>> Signed-off-by: Arvind Yadav <arvind.yadav.cs@gmail.com>
>> ---
>> drivers/irqchip/irq-mips-gic.c | 40 +++++++++++++++++++++++++++++++---------
>> 1 file changed, 31 insertions(+), 9 deletions(-)
>>
>> diff --git a/drivers/irqchip/irq-mips-gic.c b/drivers/irqchip/irq-mips-gic.c
>> index c01c09e..bf0816f 100644
>> --- a/drivers/irqchip/irq-mips-gic.c
>> +++ b/drivers/irqchip/irq-mips-gic.c
>> @@ -968,7 +968,7 @@ int gic_ipi_domain_match(struct irq_domain *d, struct device_node *node,
>> .match = gic_ipi_domain_match,
>> };
>>
>> -static void __init __gic_init(unsigned long gic_base_addr,
>> +static int __init __gic_init(unsigned long gic_base_addr,
>> unsigned long gic_addrspace_size,
>> unsigned int cpu_vec, unsigned int irqbase,
>> struct device_node *node)
>> @@ -979,6 +979,10 @@ static void __init __gic_init(unsigned long gic_base_addr,
>> __gic_base_addr = gic_base_addr;
>>
>> gic_base = ioremap_nocache(gic_base_addr, gic_addrspace_size);
>> + if (!gic_base) {
>> + pr_err("Failed to map GIC memory");
Just panic here and drop the rest.
Thanks,
Matt
>> + return -ENOMEM;
>> + }
>>
>> gicconfig = gic_read(GIC_REG(SHARED, GIC_SH_CONFIG));
>> gic_shared_intrs = (gicconfig & GIC_SH_CONFIG_NUMINTRS_MSK) >>
>> @@ -1035,23 +1039,29 @@ static void __init __gic_init(unsigned long gic_base_addr,
>> gic_irq_domain = irq_domain_add_simple(node, GIC_NUM_LOCAL_INTRS +
>> gic_shared_intrs, irqbase,
>> &gic_irq_domain_ops, NULL);
>> - if (!gic_irq_domain)
>> - panic("Failed to add GIC IRQ domain");
>> + if (!gic_irq_domain) {
>> + pr_err("Failed to add GIC IRQ domain");
>> + goto iounmap;
>> + }
>> gic_irq_domain->name = "mips-gic-irq";
>>
>> gic_dev_domain = irq_domain_add_hierarchy(gic_irq_domain, 0,
>> GIC_NUM_LOCAL_INTRS + gic_shared_intrs,
>> node, &gic_dev_domain_ops, NULL);
>> - if (!gic_dev_domain)
>> - panic("Failed to add GIC DEV domain");
>> + if (!gic_dev_domain) {
>> + pr_err("Failed to add GIC DEV domain");
>> + goto iounmap;
>> + }
>> gic_dev_domain->name = "mips-gic-dev";
>>
>> gic_ipi_domain = irq_domain_add_hierarchy(gic_irq_domain,
>> IRQ_DOMAIN_FLAG_IPI_PER_CPU,
>> GIC_NUM_LOCAL_INTRS + gic_shared_intrs,
>> node, &gic_ipi_domain_ops, NULL);
>> - if (!gic_ipi_domain)
>> - panic("Failed to add GIC IPI domain");
>> + if (!gic_ipi_domain) {
>> + pr_err("Failed to add GIC IPI domain");
>> + goto iounmap;
>> + }
>>
>> gic_ipi_domain->name = "mips-gic-ipi";
>> gic_ipi_domain->bus_token = DOMAIN_BUS_IPI;
>> @@ -1067,13 +1077,22 @@ static void __init __gic_init(unsigned long gic_base_addr,
>> }
>>
>> gic_basic_init();
>> + return 0;
>> +iounmap:
>> + iounmap(gic_base);
>> + return -ENOMEM;
>> }
>>
>> void __init gic_init(unsigned long gic_base_addr,
>> unsigned long gic_addrspace_size,
>> unsigned int cpu_vec, unsigned int irqbase)
>> {
>> - __gic_init(gic_base_addr, gic_addrspace_size, cpu_vec, irqbase, NULL);
>> + int ret;
>> +
>> + ret = __gic_init(gic_base_addr, gic_addrspace_size, cpu_vec, irqbase,
>> + NULL);
>> + if (ret)
>> + pr_err("Failed to initialize GIC");
>> }
>>
>> static int __init gic_of_init(struct device_node *node,
>> @@ -1083,6 +1102,7 @@ static int __init gic_of_init(struct device_node *node,
>> unsigned int cpu_vec, i = 0, reserved = 0;
>> phys_addr_t gic_base;
>> size_t gic_len;
>> + int ret;
>>
>> /* Find the first available CPU vector. */
>> while (!of_property_read_u32_index(node, "mti,reserved-cpu-vectors",
>> @@ -1119,7 +1139,9 @@ static int __init gic_of_init(struct device_node *node,
>> write_gcr_gic_base(gic_base | CM_GCR_GIC_BASE_GICEN_MSK);
>> gic_present = true;
>>
>> - __gic_init(gic_base, gic_len, cpu_vec, 0, node);
>> + ret = __gic_init(gic_base, gic_len, cpu_vec, 0, node);
>> + if (ret)
>> + return ret;
>>
>> return 0;
>> }
>> --
>> 1.9.1
>>
>>
prev parent reply other threads:[~2017-01-17 10:34 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-01-16 6:07 Arvind Yadav
2017-01-17 8:38 ` Thomas Gleixner
2017-01-17 9:45 ` Matt Redfearn [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=f308f402-70d3-a7a1-9cc8-fcf8a14aa741@imgtec.com \
--to=matt.redfearn@imgtec.com \
--cc=arvind.yadav.cs@gmail.com \
--cc=jason@lakedaemon.net \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mips@linux-mips.org \
--cc=marc.zyngier@arm.com \
--cc=paul.burton@imgtec.com \
--cc=tglx@linutronix.de \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®