The patch below does not apply to the 4.19-stable tree.
If someone wants it applied there, or to any other stable or longterm
tree, then please email the backport, including the original git commit
id to <stable(a)vger.kernel.org>.
thanks,
greg k-h
------------------ original commit in Linus's tree ------------------
From 748bc4dd9e663f23448d8ad7e58c011a67ea1eca Mon Sep 17 00:00:00 2001
From: "Jason A. Donenfeld" <Jason(a)zx2c4.com>
Date: Thu, 22 Sep 2022 18:46:04 +0200
Subject: [PATCH] random: use expired timer rather than wq for mixing fast pool
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Previously, the fast pool was dumped into the main pool periodically in
the fast pool's hard IRQ handler. This worked fine and there weren't
problems with it, until RT came around. Since RT converts spinlocks into
sleeping locks, problems cropped up. Rather than switching to raw
spinlocks, the RT developers preferred we make the transformation from
originally doing:
do_some_stuff()
spin_lock()
do_some_other_stuff()
spin_unlock()
to doing:
do_some_stuff()
queue_work_on(some_other_stuff_worker)
This is an ordinary pattern done all over the kernel. However, Sherry
noticed a 10% performance regression in qperf TCP over a 40gbps
InfiniBand card. Quoting her message:
> MT27500 Family [ConnectX-3] cards:
> Infiniband device 'mlx4_0' port 1 status:
> default gid: fe80:0000:0000:0000:0010:e000:0178:9eb1
> base lid: 0x6
> sm lid: 0x1
> state: 4: ACTIVE
> phys state: 5: LinkUp
> rate: 40 Gb/sec (4X QDR)
> link_layer: InfiniBand
>
> Cards are configured with IP addresses on private subnet for IPoIB
> performance testing.
> Regression identified in this bug is in TCP latency in this stack as reported
> by qperf tcp_lat metric:
>
> We have one system listen as a qperf server:
> [root@yourQperfServer ~]# qperf
>
> Have the other system connect to qperf server as a client (in this
> case, it’s X7 server with Mellanox card):
> [root@yourQperfClient ~]# numactl -m0 -N0 qperf 20.20.20.101 -v -uu -ub --time 60 --wait_server 20 -oo msg_size:4K:1024K:*2 tcp_lat
Rather than incur the scheduling latency from queue_work_on, we can
instead switch to running on the next timer tick, on the same core. This
also batches things a bit more -- once per jiffy -- which is okay now
that mix_interrupt_randomness() can credit multiple bits at once.
Reported-by: Sherry Yang <sherry.yang(a)oracle.com>
Tested-by: Paul Webb <paul.x.webb(a)oracle.com>
Cc: Sherry Yang <sherry.yang(a)oracle.com>
Cc: Phillip Goerl <phillip.goerl(a)oracle.com>
Cc: Jack Vogel <jack.vogel(a)oracle.com>
Cc: Nicky Veitch <nicky.veitch(a)oracle.com>
Cc: Colm Harrington <colm.harrington(a)oracle.com>
Cc: Ramanan Govindarajan <ramanan.govindarajan(a)oracle.com>
Cc: Sebastian Andrzej Siewior <bigeasy(a)linutronix.de>
Cc: Dominik Brodowski <linux(a)dominikbrodowski.net>
Cc: Tejun Heo <tj(a)kernel.org>
Cc: Sultan Alsawaf <sultan(a)kerneltoast.com>
Cc: stable(a)vger.kernel.org
Fixes: 58340f8e952b ("random: defer fast pool mixing to worker")
Signed-off-by: Jason A. Donenfeld <Jason(a)zx2c4.com>
diff --git a/drivers/char/random.c b/drivers/char/random.c
index a90d96f4b3bb..e591c6aadca4 100644
--- a/drivers/char/random.c
+++ b/drivers/char/random.c
@@ -921,17 +921,20 @@ struct fast_pool {
unsigned long pool[4];
unsigned long last;
unsigned int count;
- struct work_struct mix;
+ struct timer_list mix;
};
+static void mix_interrupt_randomness(struct timer_list *work);
+
static DEFINE_PER_CPU(struct fast_pool, irq_randomness) = {
#ifdef CONFIG_64BIT
#define FASTMIX_PERM SIPHASH_PERMUTATION
- .pool = { SIPHASH_CONST_0, SIPHASH_CONST_1, SIPHASH_CONST_2, SIPHASH_CONST_3 }
+ .pool = { SIPHASH_CONST_0, SIPHASH_CONST_1, SIPHASH_CONST_2, SIPHASH_CONST_3 },
#else
#define FASTMIX_PERM HSIPHASH_PERMUTATION
- .pool = { HSIPHASH_CONST_0, HSIPHASH_CONST_1, HSIPHASH_CONST_2, HSIPHASH_CONST_3 }
+ .pool = { HSIPHASH_CONST_0, HSIPHASH_CONST_1, HSIPHASH_CONST_2, HSIPHASH_CONST_3 },
#endif
+ .mix = __TIMER_INITIALIZER(mix_interrupt_randomness, 0)
};
/*
@@ -973,7 +976,7 @@ int __cold random_online_cpu(unsigned int cpu)
}
#endif
-static void mix_interrupt_randomness(struct work_struct *work)
+static void mix_interrupt_randomness(struct timer_list *work)
{
struct fast_pool *fast_pool = container_of(work, struct fast_pool, mix);
/*
@@ -1027,10 +1030,11 @@ void add_interrupt_randomness(int irq)
if (new_count < 1024 && !time_is_before_jiffies(fast_pool->last + HZ))
return;
- if (unlikely(!fast_pool->mix.func))
- INIT_WORK(&fast_pool->mix, mix_interrupt_randomness);
fast_pool->count |= MIX_INFLIGHT;
- queue_work_on(raw_smp_processor_id(), system_highpri_wq, &fast_pool->mix);
+ if (!timer_pending(&fast_pool->mix)) {
+ fast_pool->mix.expires = jiffies;
+ add_timer_on(&fast_pool->mix, raw_smp_processor_id());
+ }
}
EXPORT_SYMBOL_GPL(add_interrupt_randomness);
The patch below does not apply to the 5.4-stable tree.
If someone wants it applied there, or to any other stable or longterm
tree, then please email the backport, including the original git commit
id to <stable(a)vger.kernel.org>.
thanks,
greg k-h
------------------ original commit in Linus's tree ------------------
From 748bc4dd9e663f23448d8ad7e58c011a67ea1eca Mon Sep 17 00:00:00 2001
From: "Jason A. Donenfeld" <Jason(a)zx2c4.com>
Date: Thu, 22 Sep 2022 18:46:04 +0200
Subject: [PATCH] random: use expired timer rather than wq for mixing fast pool
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Previously, the fast pool was dumped into the main pool periodically in
the fast pool's hard IRQ handler. This worked fine and there weren't
problems with it, until RT came around. Since RT converts spinlocks into
sleeping locks, problems cropped up. Rather than switching to raw
spinlocks, the RT developers preferred we make the transformation from
originally doing:
do_some_stuff()
spin_lock()
do_some_other_stuff()
spin_unlock()
to doing:
do_some_stuff()
queue_work_on(some_other_stuff_worker)
This is an ordinary pattern done all over the kernel. However, Sherry
noticed a 10% performance regression in qperf TCP over a 40gbps
InfiniBand card. Quoting her message:
> MT27500 Family [ConnectX-3] cards:
> Infiniband device 'mlx4_0' port 1 status:
> default gid: fe80:0000:0000:0000:0010:e000:0178:9eb1
> base lid: 0x6
> sm lid: 0x1
> state: 4: ACTIVE
> phys state: 5: LinkUp
> rate: 40 Gb/sec (4X QDR)
> link_layer: InfiniBand
>
> Cards are configured with IP addresses on private subnet for IPoIB
> performance testing.
> Regression identified in this bug is in TCP latency in this stack as reported
> by qperf tcp_lat metric:
>
> We have one system listen as a qperf server:
> [root@yourQperfServer ~]# qperf
>
> Have the other system connect to qperf server as a client (in this
> case, it’s X7 server with Mellanox card):
> [root@yourQperfClient ~]# numactl -m0 -N0 qperf 20.20.20.101 -v -uu -ub --time 60 --wait_server 20 -oo msg_size:4K:1024K:*2 tcp_lat
Rather than incur the scheduling latency from queue_work_on, we can
instead switch to running on the next timer tick, on the same core. This
also batches things a bit more -- once per jiffy -- which is okay now
that mix_interrupt_randomness() can credit multiple bits at once.
Reported-by: Sherry Yang <sherry.yang(a)oracle.com>
Tested-by: Paul Webb <paul.x.webb(a)oracle.com>
Cc: Sherry Yang <sherry.yang(a)oracle.com>
Cc: Phillip Goerl <phillip.goerl(a)oracle.com>
Cc: Jack Vogel <jack.vogel(a)oracle.com>
Cc: Nicky Veitch <nicky.veitch(a)oracle.com>
Cc: Colm Harrington <colm.harrington(a)oracle.com>
Cc: Ramanan Govindarajan <ramanan.govindarajan(a)oracle.com>
Cc: Sebastian Andrzej Siewior <bigeasy(a)linutronix.de>
Cc: Dominik Brodowski <linux(a)dominikbrodowski.net>
Cc: Tejun Heo <tj(a)kernel.org>
Cc: Sultan Alsawaf <sultan(a)kerneltoast.com>
Cc: stable(a)vger.kernel.org
Fixes: 58340f8e952b ("random: defer fast pool mixing to worker")
Signed-off-by: Jason A. Donenfeld <Jason(a)zx2c4.com>
diff --git a/drivers/char/random.c b/drivers/char/random.c
index a90d96f4b3bb..e591c6aadca4 100644
--- a/drivers/char/random.c
+++ b/drivers/char/random.c
@@ -921,17 +921,20 @@ struct fast_pool {
unsigned long pool[4];
unsigned long last;
unsigned int count;
- struct work_struct mix;
+ struct timer_list mix;
};
+static void mix_interrupt_randomness(struct timer_list *work);
+
static DEFINE_PER_CPU(struct fast_pool, irq_randomness) = {
#ifdef CONFIG_64BIT
#define FASTMIX_PERM SIPHASH_PERMUTATION
- .pool = { SIPHASH_CONST_0, SIPHASH_CONST_1, SIPHASH_CONST_2, SIPHASH_CONST_3 }
+ .pool = { SIPHASH_CONST_0, SIPHASH_CONST_1, SIPHASH_CONST_2, SIPHASH_CONST_3 },
#else
#define FASTMIX_PERM HSIPHASH_PERMUTATION
- .pool = { HSIPHASH_CONST_0, HSIPHASH_CONST_1, HSIPHASH_CONST_2, HSIPHASH_CONST_3 }
+ .pool = { HSIPHASH_CONST_0, HSIPHASH_CONST_1, HSIPHASH_CONST_2, HSIPHASH_CONST_3 },
#endif
+ .mix = __TIMER_INITIALIZER(mix_interrupt_randomness, 0)
};
/*
@@ -973,7 +976,7 @@ int __cold random_online_cpu(unsigned int cpu)
}
#endif
-static void mix_interrupt_randomness(struct work_struct *work)
+static void mix_interrupt_randomness(struct timer_list *work)
{
struct fast_pool *fast_pool = container_of(work, struct fast_pool, mix);
/*
@@ -1027,10 +1030,11 @@ void add_interrupt_randomness(int irq)
if (new_count < 1024 && !time_is_before_jiffies(fast_pool->last + HZ))
return;
- if (unlikely(!fast_pool->mix.func))
- INIT_WORK(&fast_pool->mix, mix_interrupt_randomness);
fast_pool->count |= MIX_INFLIGHT;
- queue_work_on(raw_smp_processor_id(), system_highpri_wq, &fast_pool->mix);
+ if (!timer_pending(&fast_pool->mix)) {
+ fast_pool->mix.expires = jiffies;
+ add_timer_on(&fast_pool->mix, raw_smp_processor_id());
+ }
}
EXPORT_SYMBOL_GPL(add_interrupt_randomness);
The patch below does not apply to the 5.10-stable tree.
If someone wants it applied there, or to any other stable or longterm
tree, then please email the backport, including the original git commit
id to <stable(a)vger.kernel.org>.
thanks,
greg k-h
------------------ original commit in Linus's tree ------------------
From 748bc4dd9e663f23448d8ad7e58c011a67ea1eca Mon Sep 17 00:00:00 2001
From: "Jason A. Donenfeld" <Jason(a)zx2c4.com>
Date: Thu, 22 Sep 2022 18:46:04 +0200
Subject: [PATCH] random: use expired timer rather than wq for mixing fast pool
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Previously, the fast pool was dumped into the main pool periodically in
the fast pool's hard IRQ handler. This worked fine and there weren't
problems with it, until RT came around. Since RT converts spinlocks into
sleeping locks, problems cropped up. Rather than switching to raw
spinlocks, the RT developers preferred we make the transformation from
originally doing:
do_some_stuff()
spin_lock()
do_some_other_stuff()
spin_unlock()
to doing:
do_some_stuff()
queue_work_on(some_other_stuff_worker)
This is an ordinary pattern done all over the kernel. However, Sherry
noticed a 10% performance regression in qperf TCP over a 40gbps
InfiniBand card. Quoting her message:
> MT27500 Family [ConnectX-3] cards:
> Infiniband device 'mlx4_0' port 1 status:
> default gid: fe80:0000:0000:0000:0010:e000:0178:9eb1
> base lid: 0x6
> sm lid: 0x1
> state: 4: ACTIVE
> phys state: 5: LinkUp
> rate: 40 Gb/sec (4X QDR)
> link_layer: InfiniBand
>
> Cards are configured with IP addresses on private subnet for IPoIB
> performance testing.
> Regression identified in this bug is in TCP latency in this stack as reported
> by qperf tcp_lat metric:
>
> We have one system listen as a qperf server:
> [root@yourQperfServer ~]# qperf
>
> Have the other system connect to qperf server as a client (in this
> case, it’s X7 server with Mellanox card):
> [root@yourQperfClient ~]# numactl -m0 -N0 qperf 20.20.20.101 -v -uu -ub --time 60 --wait_server 20 -oo msg_size:4K:1024K:*2 tcp_lat
Rather than incur the scheduling latency from queue_work_on, we can
instead switch to running on the next timer tick, on the same core. This
also batches things a bit more -- once per jiffy -- which is okay now
that mix_interrupt_randomness() can credit multiple bits at once.
Reported-by: Sherry Yang <sherry.yang(a)oracle.com>
Tested-by: Paul Webb <paul.x.webb(a)oracle.com>
Cc: Sherry Yang <sherry.yang(a)oracle.com>
Cc: Phillip Goerl <phillip.goerl(a)oracle.com>
Cc: Jack Vogel <jack.vogel(a)oracle.com>
Cc: Nicky Veitch <nicky.veitch(a)oracle.com>
Cc: Colm Harrington <colm.harrington(a)oracle.com>
Cc: Ramanan Govindarajan <ramanan.govindarajan(a)oracle.com>
Cc: Sebastian Andrzej Siewior <bigeasy(a)linutronix.de>
Cc: Dominik Brodowski <linux(a)dominikbrodowski.net>
Cc: Tejun Heo <tj(a)kernel.org>
Cc: Sultan Alsawaf <sultan(a)kerneltoast.com>
Cc: stable(a)vger.kernel.org
Fixes: 58340f8e952b ("random: defer fast pool mixing to worker")
Signed-off-by: Jason A. Donenfeld <Jason(a)zx2c4.com>
diff --git a/drivers/char/random.c b/drivers/char/random.c
index a90d96f4b3bb..e591c6aadca4 100644
--- a/drivers/char/random.c
+++ b/drivers/char/random.c
@@ -921,17 +921,20 @@ struct fast_pool {
unsigned long pool[4];
unsigned long last;
unsigned int count;
- struct work_struct mix;
+ struct timer_list mix;
};
+static void mix_interrupt_randomness(struct timer_list *work);
+
static DEFINE_PER_CPU(struct fast_pool, irq_randomness) = {
#ifdef CONFIG_64BIT
#define FASTMIX_PERM SIPHASH_PERMUTATION
- .pool = { SIPHASH_CONST_0, SIPHASH_CONST_1, SIPHASH_CONST_2, SIPHASH_CONST_3 }
+ .pool = { SIPHASH_CONST_0, SIPHASH_CONST_1, SIPHASH_CONST_2, SIPHASH_CONST_3 },
#else
#define FASTMIX_PERM HSIPHASH_PERMUTATION
- .pool = { HSIPHASH_CONST_0, HSIPHASH_CONST_1, HSIPHASH_CONST_2, HSIPHASH_CONST_3 }
+ .pool = { HSIPHASH_CONST_0, HSIPHASH_CONST_1, HSIPHASH_CONST_2, HSIPHASH_CONST_3 },
#endif
+ .mix = __TIMER_INITIALIZER(mix_interrupt_randomness, 0)
};
/*
@@ -973,7 +976,7 @@ int __cold random_online_cpu(unsigned int cpu)
}
#endif
-static void mix_interrupt_randomness(struct work_struct *work)
+static void mix_interrupt_randomness(struct timer_list *work)
{
struct fast_pool *fast_pool = container_of(work, struct fast_pool, mix);
/*
@@ -1027,10 +1030,11 @@ void add_interrupt_randomness(int irq)
if (new_count < 1024 && !time_is_before_jiffies(fast_pool->last + HZ))
return;
- if (unlikely(!fast_pool->mix.func))
- INIT_WORK(&fast_pool->mix, mix_interrupt_randomness);
fast_pool->count |= MIX_INFLIGHT;
- queue_work_on(raw_smp_processor_id(), system_highpri_wq, &fast_pool->mix);
+ if (!timer_pending(&fast_pool->mix)) {
+ fast_pool->mix.expires = jiffies;
+ add_timer_on(&fast_pool->mix, raw_smp_processor_id());
+ }
}
EXPORT_SYMBOL_GPL(add_interrupt_randomness);
The patch below does not apply to the 5.15-stable tree.
If someone wants it applied there, or to any other stable or longterm
tree, then please email the backport, including the original git commit
id to <stable(a)vger.kernel.org>.
thanks,
greg k-h
------------------ original commit in Linus's tree ------------------
From 748bc4dd9e663f23448d8ad7e58c011a67ea1eca Mon Sep 17 00:00:00 2001
From: "Jason A. Donenfeld" <Jason(a)zx2c4.com>
Date: Thu, 22 Sep 2022 18:46:04 +0200
Subject: [PATCH] random: use expired timer rather than wq for mixing fast pool
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Previously, the fast pool was dumped into the main pool periodically in
the fast pool's hard IRQ handler. This worked fine and there weren't
problems with it, until RT came around. Since RT converts spinlocks into
sleeping locks, problems cropped up. Rather than switching to raw
spinlocks, the RT developers preferred we make the transformation from
originally doing:
do_some_stuff()
spin_lock()
do_some_other_stuff()
spin_unlock()
to doing:
do_some_stuff()
queue_work_on(some_other_stuff_worker)
This is an ordinary pattern done all over the kernel. However, Sherry
noticed a 10% performance regression in qperf TCP over a 40gbps
InfiniBand card. Quoting her message:
> MT27500 Family [ConnectX-3] cards:
> Infiniband device 'mlx4_0' port 1 status:
> default gid: fe80:0000:0000:0000:0010:e000:0178:9eb1
> base lid: 0x6
> sm lid: 0x1
> state: 4: ACTIVE
> phys state: 5: LinkUp
> rate: 40 Gb/sec (4X QDR)
> link_layer: InfiniBand
>
> Cards are configured with IP addresses on private subnet for IPoIB
> performance testing.
> Regression identified in this bug is in TCP latency in this stack as reported
> by qperf tcp_lat metric:
>
> We have one system listen as a qperf server:
> [root@yourQperfServer ~]# qperf
>
> Have the other system connect to qperf server as a client (in this
> case, it’s X7 server with Mellanox card):
> [root@yourQperfClient ~]# numactl -m0 -N0 qperf 20.20.20.101 -v -uu -ub --time 60 --wait_server 20 -oo msg_size:4K:1024K:*2 tcp_lat
Rather than incur the scheduling latency from queue_work_on, we can
instead switch to running on the next timer tick, on the same core. This
also batches things a bit more -- once per jiffy -- which is okay now
that mix_interrupt_randomness() can credit multiple bits at once.
Reported-by: Sherry Yang <sherry.yang(a)oracle.com>
Tested-by: Paul Webb <paul.x.webb(a)oracle.com>
Cc: Sherry Yang <sherry.yang(a)oracle.com>
Cc: Phillip Goerl <phillip.goerl(a)oracle.com>
Cc: Jack Vogel <jack.vogel(a)oracle.com>
Cc: Nicky Veitch <nicky.veitch(a)oracle.com>
Cc: Colm Harrington <colm.harrington(a)oracle.com>
Cc: Ramanan Govindarajan <ramanan.govindarajan(a)oracle.com>
Cc: Sebastian Andrzej Siewior <bigeasy(a)linutronix.de>
Cc: Dominik Brodowski <linux(a)dominikbrodowski.net>
Cc: Tejun Heo <tj(a)kernel.org>
Cc: Sultan Alsawaf <sultan(a)kerneltoast.com>
Cc: stable(a)vger.kernel.org
Fixes: 58340f8e952b ("random: defer fast pool mixing to worker")
Signed-off-by: Jason A. Donenfeld <Jason(a)zx2c4.com>
diff --git a/drivers/char/random.c b/drivers/char/random.c
index a90d96f4b3bb..e591c6aadca4 100644
--- a/drivers/char/random.c
+++ b/drivers/char/random.c
@@ -921,17 +921,20 @@ struct fast_pool {
unsigned long pool[4];
unsigned long last;
unsigned int count;
- struct work_struct mix;
+ struct timer_list mix;
};
+static void mix_interrupt_randomness(struct timer_list *work);
+
static DEFINE_PER_CPU(struct fast_pool, irq_randomness) = {
#ifdef CONFIG_64BIT
#define FASTMIX_PERM SIPHASH_PERMUTATION
- .pool = { SIPHASH_CONST_0, SIPHASH_CONST_1, SIPHASH_CONST_2, SIPHASH_CONST_3 }
+ .pool = { SIPHASH_CONST_0, SIPHASH_CONST_1, SIPHASH_CONST_2, SIPHASH_CONST_3 },
#else
#define FASTMIX_PERM HSIPHASH_PERMUTATION
- .pool = { HSIPHASH_CONST_0, HSIPHASH_CONST_1, HSIPHASH_CONST_2, HSIPHASH_CONST_3 }
+ .pool = { HSIPHASH_CONST_0, HSIPHASH_CONST_1, HSIPHASH_CONST_2, HSIPHASH_CONST_3 },
#endif
+ .mix = __TIMER_INITIALIZER(mix_interrupt_randomness, 0)
};
/*
@@ -973,7 +976,7 @@ int __cold random_online_cpu(unsigned int cpu)
}
#endif
-static void mix_interrupt_randomness(struct work_struct *work)
+static void mix_interrupt_randomness(struct timer_list *work)
{
struct fast_pool *fast_pool = container_of(work, struct fast_pool, mix);
/*
@@ -1027,10 +1030,11 @@ void add_interrupt_randomness(int irq)
if (new_count < 1024 && !time_is_before_jiffies(fast_pool->last + HZ))
return;
- if (unlikely(!fast_pool->mix.func))
- INIT_WORK(&fast_pool->mix, mix_interrupt_randomness);
fast_pool->count |= MIX_INFLIGHT;
- queue_work_on(raw_smp_processor_id(), system_highpri_wq, &fast_pool->mix);
+ if (!timer_pending(&fast_pool->mix)) {
+ fast_pool->mix.expires = jiffies;
+ add_timer_on(&fast_pool->mix, raw_smp_processor_id());
+ }
}
EXPORT_SYMBOL_GPL(add_interrupt_randomness);
The patch below does not apply to the 5.19-stable tree.
If someone wants it applied there, or to any other stable or longterm
tree, then please email the backport, including the original git commit
id to <stable(a)vger.kernel.org>.
thanks,
greg k-h
------------------ original commit in Linus's tree ------------------
From 748bc4dd9e663f23448d8ad7e58c011a67ea1eca Mon Sep 17 00:00:00 2001
From: "Jason A. Donenfeld" <Jason(a)zx2c4.com>
Date: Thu, 22 Sep 2022 18:46:04 +0200
Subject: [PATCH] random: use expired timer rather than wq for mixing fast pool
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Previously, the fast pool was dumped into the main pool periodically in
the fast pool's hard IRQ handler. This worked fine and there weren't
problems with it, until RT came around. Since RT converts spinlocks into
sleeping locks, problems cropped up. Rather than switching to raw
spinlocks, the RT developers preferred we make the transformation from
originally doing:
do_some_stuff()
spin_lock()
do_some_other_stuff()
spin_unlock()
to doing:
do_some_stuff()
queue_work_on(some_other_stuff_worker)
This is an ordinary pattern done all over the kernel. However, Sherry
noticed a 10% performance regression in qperf TCP over a 40gbps
InfiniBand card. Quoting her message:
> MT27500 Family [ConnectX-3] cards:
> Infiniband device 'mlx4_0' port 1 status:
> default gid: fe80:0000:0000:0000:0010:e000:0178:9eb1
> base lid: 0x6
> sm lid: 0x1
> state: 4: ACTIVE
> phys state: 5: LinkUp
> rate: 40 Gb/sec (4X QDR)
> link_layer: InfiniBand
>
> Cards are configured with IP addresses on private subnet for IPoIB
> performance testing.
> Regression identified in this bug is in TCP latency in this stack as reported
> by qperf tcp_lat metric:
>
> We have one system listen as a qperf server:
> [root@yourQperfServer ~]# qperf
>
> Have the other system connect to qperf server as a client (in this
> case, it’s X7 server with Mellanox card):
> [root@yourQperfClient ~]# numactl -m0 -N0 qperf 20.20.20.101 -v -uu -ub --time 60 --wait_server 20 -oo msg_size:4K:1024K:*2 tcp_lat
Rather than incur the scheduling latency from queue_work_on, we can
instead switch to running on the next timer tick, on the same core. This
also batches things a bit more -- once per jiffy -- which is okay now
that mix_interrupt_randomness() can credit multiple bits at once.
Reported-by: Sherry Yang <sherry.yang(a)oracle.com>
Tested-by: Paul Webb <paul.x.webb(a)oracle.com>
Cc: Sherry Yang <sherry.yang(a)oracle.com>
Cc: Phillip Goerl <phillip.goerl(a)oracle.com>
Cc: Jack Vogel <jack.vogel(a)oracle.com>
Cc: Nicky Veitch <nicky.veitch(a)oracle.com>
Cc: Colm Harrington <colm.harrington(a)oracle.com>
Cc: Ramanan Govindarajan <ramanan.govindarajan(a)oracle.com>
Cc: Sebastian Andrzej Siewior <bigeasy(a)linutronix.de>
Cc: Dominik Brodowski <linux(a)dominikbrodowski.net>
Cc: Tejun Heo <tj(a)kernel.org>
Cc: Sultan Alsawaf <sultan(a)kerneltoast.com>
Cc: stable(a)vger.kernel.org
Fixes: 58340f8e952b ("random: defer fast pool mixing to worker")
Signed-off-by: Jason A. Donenfeld <Jason(a)zx2c4.com>
diff --git a/drivers/char/random.c b/drivers/char/random.c
index a90d96f4b3bb..e591c6aadca4 100644
--- a/drivers/char/random.c
+++ b/drivers/char/random.c
@@ -921,17 +921,20 @@ struct fast_pool {
unsigned long pool[4];
unsigned long last;
unsigned int count;
- struct work_struct mix;
+ struct timer_list mix;
};
+static void mix_interrupt_randomness(struct timer_list *work);
+
static DEFINE_PER_CPU(struct fast_pool, irq_randomness) = {
#ifdef CONFIG_64BIT
#define FASTMIX_PERM SIPHASH_PERMUTATION
- .pool = { SIPHASH_CONST_0, SIPHASH_CONST_1, SIPHASH_CONST_2, SIPHASH_CONST_3 }
+ .pool = { SIPHASH_CONST_0, SIPHASH_CONST_1, SIPHASH_CONST_2, SIPHASH_CONST_3 },
#else
#define FASTMIX_PERM HSIPHASH_PERMUTATION
- .pool = { HSIPHASH_CONST_0, HSIPHASH_CONST_1, HSIPHASH_CONST_2, HSIPHASH_CONST_3 }
+ .pool = { HSIPHASH_CONST_0, HSIPHASH_CONST_1, HSIPHASH_CONST_2, HSIPHASH_CONST_3 },
#endif
+ .mix = __TIMER_INITIALIZER(mix_interrupt_randomness, 0)
};
/*
@@ -973,7 +976,7 @@ int __cold random_online_cpu(unsigned int cpu)
}
#endif
-static void mix_interrupt_randomness(struct work_struct *work)
+static void mix_interrupt_randomness(struct timer_list *work)
{
struct fast_pool *fast_pool = container_of(work, struct fast_pool, mix);
/*
@@ -1027,10 +1030,11 @@ void add_interrupt_randomness(int irq)
if (new_count < 1024 && !time_is_before_jiffies(fast_pool->last + HZ))
return;
- if (unlikely(!fast_pool->mix.func))
- INIT_WORK(&fast_pool->mix, mix_interrupt_randomness);
fast_pool->count |= MIX_INFLIGHT;
- queue_work_on(raw_smp_processor_id(), system_highpri_wq, &fast_pool->mix);
+ if (!timer_pending(&fast_pool->mix)) {
+ fast_pool->mix.expires = jiffies;
+ add_timer_on(&fast_pool->mix, raw_smp_processor_id());
+ }
}
EXPORT_SYMBOL_GPL(add_interrupt_randomness);
The patch below does not apply to the 6.0-stable tree.
If someone wants it applied there, or to any other stable or longterm
tree, then please email the backport, including the original git commit
id to <stable(a)vger.kernel.org>.
thanks,
greg k-h
------------------ original commit in Linus's tree ------------------
From 748bc4dd9e663f23448d8ad7e58c011a67ea1eca Mon Sep 17 00:00:00 2001
From: "Jason A. Donenfeld" <Jason(a)zx2c4.com>
Date: Thu, 22 Sep 2022 18:46:04 +0200
Subject: [PATCH] random: use expired timer rather than wq for mixing fast pool
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Previously, the fast pool was dumped into the main pool periodically in
the fast pool's hard IRQ handler. This worked fine and there weren't
problems with it, until RT came around. Since RT converts spinlocks into
sleeping locks, problems cropped up. Rather than switching to raw
spinlocks, the RT developers preferred we make the transformation from
originally doing:
do_some_stuff()
spin_lock()
do_some_other_stuff()
spin_unlock()
to doing:
do_some_stuff()
queue_work_on(some_other_stuff_worker)
This is an ordinary pattern done all over the kernel. However, Sherry
noticed a 10% performance regression in qperf TCP over a 40gbps
InfiniBand card. Quoting her message:
> MT27500 Family [ConnectX-3] cards:
> Infiniband device 'mlx4_0' port 1 status:
> default gid: fe80:0000:0000:0000:0010:e000:0178:9eb1
> base lid: 0x6
> sm lid: 0x1
> state: 4: ACTIVE
> phys state: 5: LinkUp
> rate: 40 Gb/sec (4X QDR)
> link_layer: InfiniBand
>
> Cards are configured with IP addresses on private subnet for IPoIB
> performance testing.
> Regression identified in this bug is in TCP latency in this stack as reported
> by qperf tcp_lat metric:
>
> We have one system listen as a qperf server:
> [root@yourQperfServer ~]# qperf
>
> Have the other system connect to qperf server as a client (in this
> case, it’s X7 server with Mellanox card):
> [root@yourQperfClient ~]# numactl -m0 -N0 qperf 20.20.20.101 -v -uu -ub --time 60 --wait_server 20 -oo msg_size:4K:1024K:*2 tcp_lat
Rather than incur the scheduling latency from queue_work_on, we can
instead switch to running on the next timer tick, on the same core. This
also batches things a bit more -- once per jiffy -- which is okay now
that mix_interrupt_randomness() can credit multiple bits at once.
Reported-by: Sherry Yang <sherry.yang(a)oracle.com>
Tested-by: Paul Webb <paul.x.webb(a)oracle.com>
Cc: Sherry Yang <sherry.yang(a)oracle.com>
Cc: Phillip Goerl <phillip.goerl(a)oracle.com>
Cc: Jack Vogel <jack.vogel(a)oracle.com>
Cc: Nicky Veitch <nicky.veitch(a)oracle.com>
Cc: Colm Harrington <colm.harrington(a)oracle.com>
Cc: Ramanan Govindarajan <ramanan.govindarajan(a)oracle.com>
Cc: Sebastian Andrzej Siewior <bigeasy(a)linutronix.de>
Cc: Dominik Brodowski <linux(a)dominikbrodowski.net>
Cc: Tejun Heo <tj(a)kernel.org>
Cc: Sultan Alsawaf <sultan(a)kerneltoast.com>
Cc: stable(a)vger.kernel.org
Fixes: 58340f8e952b ("random: defer fast pool mixing to worker")
Signed-off-by: Jason A. Donenfeld <Jason(a)zx2c4.com>
diff --git a/drivers/char/random.c b/drivers/char/random.c
index a90d96f4b3bb..e591c6aadca4 100644
--- a/drivers/char/random.c
+++ b/drivers/char/random.c
@@ -921,17 +921,20 @@ struct fast_pool {
unsigned long pool[4];
unsigned long last;
unsigned int count;
- struct work_struct mix;
+ struct timer_list mix;
};
+static void mix_interrupt_randomness(struct timer_list *work);
+
static DEFINE_PER_CPU(struct fast_pool, irq_randomness) = {
#ifdef CONFIG_64BIT
#define FASTMIX_PERM SIPHASH_PERMUTATION
- .pool = { SIPHASH_CONST_0, SIPHASH_CONST_1, SIPHASH_CONST_2, SIPHASH_CONST_3 }
+ .pool = { SIPHASH_CONST_0, SIPHASH_CONST_1, SIPHASH_CONST_2, SIPHASH_CONST_3 },
#else
#define FASTMIX_PERM HSIPHASH_PERMUTATION
- .pool = { HSIPHASH_CONST_0, HSIPHASH_CONST_1, HSIPHASH_CONST_2, HSIPHASH_CONST_3 }
+ .pool = { HSIPHASH_CONST_0, HSIPHASH_CONST_1, HSIPHASH_CONST_2, HSIPHASH_CONST_3 },
#endif
+ .mix = __TIMER_INITIALIZER(mix_interrupt_randomness, 0)
};
/*
@@ -973,7 +976,7 @@ int __cold random_online_cpu(unsigned int cpu)
}
#endif
-static void mix_interrupt_randomness(struct work_struct *work)
+static void mix_interrupt_randomness(struct timer_list *work)
{
struct fast_pool *fast_pool = container_of(work, struct fast_pool, mix);
/*
@@ -1027,10 +1030,11 @@ void add_interrupt_randomness(int irq)
if (new_count < 1024 && !time_is_before_jiffies(fast_pool->last + HZ))
return;
- if (unlikely(!fast_pool->mix.func))
- INIT_WORK(&fast_pool->mix, mix_interrupt_randomness);
fast_pool->count |= MIX_INFLIGHT;
- queue_work_on(raw_smp_processor_id(), system_highpri_wq, &fast_pool->mix);
+ if (!timer_pending(&fast_pool->mix)) {
+ fast_pool->mix.expires = jiffies;
+ add_timer_on(&fast_pool->mix, raw_smp_processor_id());
+ }
}
EXPORT_SYMBOL_GPL(add_interrupt_randomness);