[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [PATCH] xen/arm: gic-v3-lpi: Allocate the pending table while preparing the CPU


  • To: Julien Grall <julien@xxxxxxx>
  • From: Bertrand Marquis <Bertrand.Marquis@xxxxxxx>
  • Date: Mon, 16 May 2022 09:24:02 +0000
  • Accept-language: en-GB, en-US
  • Arc-authentication-results: i=2; mx.microsoft.com 1; spf=pass (sender ip is 63.35.35.123) smtp.rcpttodomain=lists.xenproject.org smtp.mailfrom=arm.com; dmarc=pass (p=none sp=none pct=100) action=none header.from=arm.com; dkim=pass (signature was verified) header.d=armh.onmicrosoft.com; arc=pass (0 oda=1 ltdi=1 spf=[1,1,smtp.mailfrom=arm.com] dkim=[1,1,header.d=arm.com] dmarc=[1,1,header.from=arm.com])
  • Arc-authentication-results: i=1; mx.microsoft.com 1; spf=pass smtp.mailfrom=arm.com; dmarc=pass action=none header.from=arm.com; dkim=pass header.d=arm.com; arc=none
  • Arc-message-signature: i=2; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com; s=arcselector9901; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-AntiSpam-MessageData-ChunkCount:X-MS-Exchange-AntiSpam-MessageData-0:X-MS-Exchange-AntiSpam-MessageData-1; bh=sVbPCDFxyAP+r+z3b6GliUc8MmqU/UawYx8E3bV3zJ0=; b=X6pqZW8Vpeunmcpts1gCsRmlPDeCzkYOQX4CjJ8ELOpipCti9j78Lh2o5Q9MSaUfXSNe20JwI7oC5v0N/pgRP9ETQzLc7R09a/+QSQM/KYUpYWuPiq8yt92CWtZlkGsz76J4ellDjHT1Ias7IlVOiCLvnixfg30W07llnx5VHop5lf7hH5ECprv2ucnGRrTxH4l6zPF1wFbu+f0k3I+WHdGrrhKvHw1ixVNfypjT1XlyifsQje0vxS8R4PTOexjs+JkQNzLvIFd2xZHS/ft826oLpXTpYIwB3Bemf8Jg4Tu6Wvvv3+zHD6JMhKdzSw6oMZWkh44fsStiu8p2hrPI1w==
  • Arc-message-signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com; s=arcselector9901; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-AntiSpam-MessageData-ChunkCount:X-MS-Exchange-AntiSpam-MessageData-0:X-MS-Exchange-AntiSpam-MessageData-1; bh=sVbPCDFxyAP+r+z3b6GliUc8MmqU/UawYx8E3bV3zJ0=; b=Dyhf2TLGd6tziysi0vG86QNpyEUGtDTXZLn4sa846KEqd3ccwOgtd9AGHOUU0kShoS22ts2rHhQzByvNy744mYXHJflOMHXKZGhm1PNMPS97ld+2y1U0HjqKBpyWAKlopsb4kSXxTiU0vLkDYgdGQgUTApca393C1Grh5iC9Nr9Q+hRnOYpXx3P85rdxLH1SJFl37+KPE7FJAafgRMcRU/G6lXyhL21Cjv1M0YNi+sx5+yBcdX/oTTKWAuPLfhtas/kEAIR79dXhLtKO3VeLVTiivNIq+QDEM6Gmg13vBfBB9q/950kFXd03Eh2N9kqIhd+Vbx4t7O+3PtXXCAFQ5g==
  • Arc-seal: i=2; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=pass; b=iQ3rRUPezfVlEiNcnr3THAW9fmgpX+srOmwaN4SY/ZUc11yVGd0BxVo3s0gCxRtQaHLgsloY0xYCEgxfMBOYfKo1tFXN3Tae2Y01AM/yWTrPEefCdktdqPehNcIpGsf70pLjuaI46ZsTpBnid8yzTRbyocgLDpCgMxJPB8/AHpdfagfldYMU3Sz7EeoNPi6j5oIdLXoIwOklSNA1F35LuxEymzdkMNh6tvcHuBNg2SESNZ69cWMpozQY12cW4KhDy4nHfNbNfLZ+qClshGsHhqNkyizezbF+TqeJRoKwhjVAYGCFkta1sHVld78s5thmme6A20op6bWKVSP1vi1q+g==
  • Arc-seal: i=1; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=none; b=GNGMkSRgjljT0lFFqToGVGKaRiMcyLbCKkPhP1lwGV3eIvtnmp8ZBlL90er8eK7CTF6Znherwzl1XuXjbu8ohF2ysAY2X9PO1eht85R3Jc73ewzNuvOgCAoMJRCh+q4KF8AXqke1LHduhtEy0AzXPeN9AskuPn0JoaExe75tpv3PVjYcKFaKvmPzkKcX/Yz253Rz511p0VXvfmdGoU47y3WaVSBpKcpfvPjwptvOJuwVfT6wrNt4Cf+VG7LSBESjkDBO6rvS+eT3klY3TBt+C7AU4+4Q1+g/+tyD8h34Zom4LZwmufRi2XhHKo+P96Bvbh3WnepkgaNp2Mj52YeZXw==
  • Authentication-results-original: dkim=none (message not signed) header.d=none;dmarc=none action=none header.from=arm.com;
  • Cc: "xen-devel@xxxxxxxxxxxxxxxxxxxx" <xen-devel@xxxxxxxxxxxxxxxxxxxx>, Julien Grall <jgrall@xxxxxxxxxx>, Stefano Stabellini <sstabellini@xxxxxxxxxx>, Volodymyr Babchuk <Volodymyr_Babchuk@xxxxxxxx>
  • Delivery-date: Mon, 16 May 2022 09:24:34 +0000
  • List-id: Xen developer discussion <xen-devel.lists.xenproject.org>
  • Nodisclaimer: true
  • Original-authentication-results: dkim=none (message not signed) header.d=none;dmarc=none action=none header.from=arm.com;
  • Thread-index: AQHYaQFPaa+cmN6NyUGhGvdTzOfzhq0hO1kA
  • Thread-topic: [PATCH] xen/arm: gic-v3-lpi: Allocate the pending table while preparing the CPU

Hi Julien,

> On 16 May 2022, at 09:45, Julien Grall <julien@xxxxxxx> wrote:
> 
> From: Julien Grall <jgrall@xxxxxxxxxx>
> 
> Commit 88a037e2cfe1 "page_alloc: assert IRQs are enabled in heap
> alloc/free" extended the checks in the buddy allocator to catch any
> use of the helpers from context with interrupts disabled.
> 
> Unfortunately, the rule is not followed in the LPI code when allocating
> the pending table:
> 
> (XEN) Xen call trace:
> (XEN)    [<000000000022a678>] alloc_xenheap_pages+0x178/0x194 (PC)
> (XEN)    [<000000000022a670>] alloc_xenheap_pages+0x170/0x194 (LR)
> (XEN)    [<0000000000237770>] _xmalloc+0x144/0x294
> (XEN)    [<00000000002378d4>] _xzalloc+0x14/0x30
> (XEN)    [<000000000027b4e4>] gicv3_lpi_init_rdist+0x54/0x324
> (XEN)    [<0000000000279898>] arch/arm/gic-v3.c#gicv3_cpu_init+0x128/0x46
> (XEN)    [<0000000000279bfc>] 
> arch/arm/gic-v3.c#gicv3_secondary_cpu_init+0x20/0x50
> (XEN)    [<0000000000277054>] gic_init_secondary_cpu+0x18/0x30
> (XEN)    [<0000000000284518>] start_secondary+0x1a8/0x234
> (XEN)    [<0000010722aa4200>] 0000010722aa4200
> (XEN)
> (XEN)
> (XEN) ****************************************
> (XEN) Panic on CPU 2:
> (XEN) Assertion '!in_irq() && (local_irq_is_enabled() || num_online_cpus() <= 
> 1)' failed at common/page_alloc.c:2212
> (XEN) ****************************************
> 
> For now the patch extending the checks has been reverted, but it would
> be good to re-introduce it (allocation with interrupt is not desirable).
> 
> The logic is reworked to allocate the pending table when preparing the
> CPU.
> 
> Signed-off-by: Julien Grall <jgrall@xxxxxxxxxx>
> ---
> xen/arch/arm/gic-v3-lpi.c | 81 ++++++++++++++++++++++++++++-----------
> 1 file changed, 59 insertions(+), 22 deletions(-)
> 
> diff --git a/xen/arch/arm/gic-v3-lpi.c b/xen/arch/arm/gic-v3-lpi.c
> index e1594dd20e4c..77d9d05c35a6 100644
> --- a/xen/arch/arm/gic-v3-lpi.c
> +++ b/xen/arch/arm/gic-v3-lpi.c
> @@ -18,6 +18,7 @@
>  * along with this program; If not, see <http://www.gnu.org/licenses/>.
>  */
> 
> +#include <xen/cpu.h>
> #include <xen/lib.h>
> #include <xen/mm.h>
> #include <xen/param.h>
> @@ -234,18 +235,13 @@ void gicv3_lpi_update_host_entry(uint32_t host_lpi, int 
> domain_id,
>     write_u64_atomic(&hlpip->data, hlpi.data);
> }
> 
> -static int gicv3_lpi_allocate_pendtable(uint64_t *reg)
> +static int gicv3_lpi_allocate_pendtable(unsigned int cpu)
> {
> -    uint64_t val;
>     void *pendtable;
> 
> -    if ( this_cpu(lpi_redist).pending_table )
> +    if ( per_cpu(lpi_redist, cpu).pending_table )
>         return -EBUSY;
> 
> -    val  = GIC_BASER_CACHE_RaWaWb << GICR_PENDBASER_INNER_CACHEABILITY_SHIFT;
> -    val |= GIC_BASER_CACHE_SameAsInner << 
> GICR_PENDBASER_OUTER_CACHEABILITY_SHIFT;
> -    val |= GIC_BASER_InnerShareable << GICR_PENDBASER_SHAREABILITY_SHIFT;
> -
>     /*
>      * The pending table holds one bit per LPI and even covers bits for
>      * interrupt IDs below 8192, so we allocate the full range.
> @@ -265,13 +261,38 @@ static int gicv3_lpi_allocate_pendtable(uint64_t *reg)
>     clean_and_invalidate_dcache_va_range(pendtable,
>                                          lpi_data.max_host_lpi_ids / 8);
> 
> -    this_cpu(lpi_redist).pending_table = pendtable;
> +    per_cpu(lpi_redist, cpu).pending_table = pendtable;
> 
> -    val |= GICR_PENDBASER_PTZ;
> +    return 0;
> +}
> +
> +static int gicv3_lpi_set_pendtable(void __iomem *rdist_base)
> +{
> +    const void *pendtable = this_cpu(lpi_redist).pending_table;
> +    uint64_t val;
> +

Should we add an assert here to check if we are to early in boot ?
That would also implicitly explain that allocation is done during CPU_PREPARE 
so this should not be called before.

> +    if ( !pendtable )
> +        return -ENOMEM;
> 
> +    ASSERT(!(virt_to_maddr(pendtable) & ~GENMASK(51, 16)));

This ASSERT is already done in gicv3_lpi_allocate_pendtable but it makes sense 
to have it closer to the place where we actually set the register.
Otherwise this assert can never be triggered.
Can you remove the one in the allocation function and also copy the comment 
that was on top of it here ?

> +
> +    val  = GIC_BASER_CACHE_RaWaWb << GICR_PENDBASER_INNER_CACHEABILITY_SHIFT;
> +    val |= GIC_BASER_CACHE_SameAsInner << 
> GICR_PENDBASER_OUTER_CACHEABILITY_SHIFT;
> +    val |= GIC_BASER_InnerShareable << GICR_PENDBASER_SHAREABILITY_SHIFT;
> +    val |= GICR_PENDBASER_PTZ;
>     val |= virt_to_maddr(pendtable);
> 
> -    *reg = val;
> +    writeq_relaxed(val, rdist_base + GICR_PENDBASER);
> +    val = readq_relaxed(rdist_base + GICR_PENDBASER);
> +
> +    /* If the hardware reports non-shareable, drop cacheability as well. */
> +    if ( !(val & GICR_PENDBASER_SHAREABILITY_MASK) )
> +    {
> +        val &= ~GICR_PENDBASER_INNER_CACHEABILITY_MASK;
> +        val |= GIC_BASER_CACHE_nC << GICR_PENDBASER_INNER_CACHEABILITY_SHIFT;
> +
> +        writeq_relaxed(val, rdist_base + GICR_PENDBASER);
> +    }
> 
>     return 0;
> }
> @@ -340,7 +361,6 @@ static int gicv3_lpi_set_proptable(void __iomem * 
> rdist_base)
> int gicv3_lpi_init_rdist(void __iomem * rdist_base)
> {
>     uint32_t reg;
> -    uint64_t table_reg;
>     int ret;
> 
>     /* We don't support LPIs without an ITS. */
> @@ -352,24 +372,33 @@ int gicv3_lpi_init_rdist(void __iomem * rdist_base)
>     if ( reg & GICR_CTLR_ENABLE_LPIS )
>         return -EBUSY;
> 
> -    ret = gicv3_lpi_allocate_pendtable(&table_reg);
> +    ret = gicv3_lpi_set_pendtable(rdist_base);
>     if ( ret )
>         return ret;
> -    writeq_relaxed(table_reg, rdist_base + GICR_PENDBASER);
> -    table_reg = readq_relaxed(rdist_base + GICR_PENDBASER);
> 
> -    /* If the hardware reports non-shareable, drop cacheability as well. */
> -    if ( !(table_reg & GICR_PENDBASER_SHAREABILITY_MASK) )
> -    {
> -        table_reg &= ~GICR_PENDBASER_INNER_CACHEABILITY_MASK;
> -        table_reg |= GIC_BASER_CACHE_nC << 
> GICR_PENDBASER_INNER_CACHEABILITY_SHIFT;
> +    return gicv3_lpi_set_proptable(rdist_base);
> +}
> +
> +static int cpu_callback(struct notifier_block *nfb, unsigned long action,
> +                        void *hcpu)
> +{
> +    unsigned long cpu = (unsigned long)hcpu;
> +    int rc = 0;
> 
> -        writeq_relaxed(table_reg, rdist_base + GICR_PENDBASER);
> +    switch ( action )
> +    {
> +    case CPU_UP_PREPARE:
> +        rc = gicv3_lpi_allocate_pendtable(cpu);
> +        break;
>     }
> 
> -    return gicv3_lpi_set_proptable(rdist_base);
> +    return !rc ? NOTIFY_DONE : notifier_from_errno(rc);
> }
> 
> +static struct notifier_block cpu_nfb = {
> +    .notifier_call = cpu_callback,
> +};
> +
> static unsigned int max_lpi_bits = 20;
> integer_param("max_lpi_bits", max_lpi_bits);
> 
> @@ -381,6 +410,7 @@ integer_param("max_lpi_bits", max_lpi_bits);
> int gicv3_lpi_init_host_lpis(unsigned int host_lpi_bits)
> {
>     unsigned int nr_lpi_ptrs;
> +    int rc;
> 
>     /* We rely on the data structure being atomically accessible. */
>     BUILD_BUG_ON(sizeof(union host_lpi) > sizeof(unsigned long));
> @@ -413,7 +443,14 @@ int gicv3_lpi_init_host_lpis(unsigned int host_lpi_bits)
> 
>     printk("GICv3: using at most %lu LPIs on the host.\n", MAX_NR_HOST_LPIS);
> 
> -    return 0;
> +    /* Register the CPU notifier and allocate memory for the boot CPU */
> +    register_cpu_notifier(&cpu_nfb);
> +    rc = gicv3_lpi_allocate_pendtable(smp_processor_id());
> +    if ( rc )
> +        printk(XENLOG_ERR "Unable to allocate the pendtable for CPU%u\n",
> +               smp_processor_id());

On secondary cores nothing equivalent will be printed and in the cal path there
will be nothing printed at all which could make debugging complex.
Can you move this print into gicv3_lpi_allocate_pendtable ?


> +
> +    return rc;
> }
> 
> static int find_unused_host_lpi(uint32_t start, uint32_t *index)
> -- 
> 2.32.0
> 

Cheers
Bertrand




 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.