Change vring space from nomal memory to dma coherent memory
Subsystems (2)
v3 cover Change vring space from nomal memory to dma coherent memory
Changes in V3:
1. Change the device page allocation method of the Intel mic layer in Patch 1 to align with the vring allocation.
2. Move the vring physical address changes in mmap callback from Prach 3 to 1.
3. Use must_be_zero instead of directly deleting used_address_updated in Patch 2 to avoid the influence of ABI change.
(I tried to use dma_mmap_coherent api in Patch 4 as Christoph suggested, but
met some issues explained here https://lore.kernel.org/patchwork/patch/1313327/,
so there is no change to Patch 4, and still can't find a better way than patch 3)
The original vop driver only supports dma coherent device, as it allocates and maps vring by _get_free_pages and dma_map_single, but not use dma_sync_single_for_cpu/device to sync the updates of device_page/vring between EP and RC, which will cause memory synchronization problem for device don't support hardware dma coherent.
And allocate vrings use dma_alloc_coherent is a common way in kernel, as the memory interacted between two systems should use consistent memory to avoid caching effects. So here add noncoherent platform support for vop driver. Also add some related dma changes to make sure noncoherent platform works well.
Sherry Sun (4):
misc: vop: change the way of allocating vring and device page
misc: vop: do not allocate and reassign the used ring
misc: vop: simply return the saved dma address instead of virt_to_phys
misc: vop: mapping kernel memory to user space as noncached
drivers/misc/mic/bus/vop_bus.h | 2 + drivers/misc/mic/host/mic_boot.c | 8 +++ drivers/misc/mic/host/mic_main.c | 15 ++---- drivers/misc/mic/vop/vop_debugfs.c | 2 - drivers/misc/mic/vop/vop_main.c | 48 +++-------------- drivers/misc/mic/vop/vop_vringh.c | 84 +++++++----------------------- include/uapi/linux/mic_common.h | 5 +- 7 files changed, 42 insertions(+), 122 deletions(-)
Signature · 1 line
2.17.1
PATCH v3 1/4 misc: vop: change the way of allocating vring and device page
Allocate vrings use dma_alloc_coherent is a common way in kernel. As the memory interacted between two systems should use consistent memory to avoid caching effects, same as device page memory.
The orginal way use __get_free_pages and dma_map_single to allocate and map vring, but not use dma_sync_single_for_cpu/device api to sync the changes of vring between EP and RC, which will cause memory synchronization problem for those devices which don't support hardware dma coherent.
--- drivers/misc/mic/host/mic_main.c | 15 +++-------- drivers/misc/mic/vop/vop_vringh.c | 43 +++++++++---------------------- 2 files changed, 16 insertions(+), 42 deletions(-)
drivers/misc/mic/host/mic_main.c +4 -11
@@ -10,6 +10,7 @@#include <linux/module.h> #include <linux/pci.h> #include <linux/poll.h> +#include <linux/dma-mapping.h> #include <linux/mic_common.h> #include "../common/mic_dev.h"
@@ -48,18 +49,11 @@ static struct ida g_mic_ida;/* Initialize the device page */ static int mic_dp_init (struct mic_device *mdev) { - mdev->dp =kzalloc (MIC_DP_SIZE ,GFP_KERNEL ); + mdev->dp =dma_alloc_coherent (&mdev->pdev->dev,MIC_DP_SIZE , + &mdev->dp_dma_addr,GFP_KERNEL );if (!mdev->dp)return -ENOMEM ; - mdev->dp_dma_addr =mic_map_single (mdev, - mdev->dp,MIC_DP_SIZE ); -if (mic_map_error (mdev->dp_dma_addr)) { -kfree (mdev->dp); -dev_err (&mdev->pdev->dev,"%s %d err %d\n" , - __func__, __LINE__, -ENOMEM ); -return -ENOMEM ; - } mdev->ops->write_spad (mdev,MIC_DPLO_SPAD , mdev->dp_dma_addr); mdev->ops->write_spad (mdev,MIC_DPHI_SPAD , mdev->dp_dma_addr >>32 );return 0 ;
@@ -68,8 +62,7 @@ static int mic_dp_init(struct mic_device *mdev)/* Uninitialize the device page */ static void mic_dp_uninit (struct mic_device *mdev) { -mic_unmap_single (mdev, mdev->dp_dma_addr,MIC_DP_SIZE ); -kfree (mdev->dp); +dma_free_coherent (&mdev->pdev->dev,MIC_DP_SIZE , mdev->dp, mdev->dp_dma_addr); }/**
drivers/misc/mic/vop/vop_vringh.c +12 -31
87 lines not on this page · raw message
Signature · 1 line
2.17.1
kernel test robot
build botRe: [PATCH V3 1/4] misc: vop: change the way of allocating vring and device page
Thank you for the patch! Perhaps something to improve:
Hi Sherry,
Thank you for the patch! Perhaps something to improve:
[auto build test WARNING on char-misc/char-misc-testing] [also build test WARNING on soc/for-next linus/master v5.9 next-20201022] [If your patch is applied to the wrong git tree, kindly drop us a note. And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]
url: https://github.com/0day-ci/linux/commits/Sherry-Sun/Change-vring-space-from-nomal-memory-to-dma-coherent-memory/20201022-131008 base: https://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc.git f3277cbfba763cd2826396521b9296de67cf1bbc config: i386-randconfig-s002-20201022 (attached as .config) compiler: gcc-9 (Debian 9.3.0-15) 9.3.0 reproduce: # apt-get install sparse # sparse version: v0.6.3-dirty # https://github.com/0day-ci/linux/commit/6851e84ec2f16ab12b04b2a5bf61b05465d450e6 git remote add linux-review https://github.com/0day-ci/linux git fetch --no-tags linux-review Sherry-Sun/Change-vring-space-from-nomal-memory-to-dma-coherent-memory/20201022-131008 git checkout 6851e84ec2f16ab12b04b2a5bf61b05465d450e6 # save the attached .config to linux build tree make W=1 C=1 CF='-fdiagnostic-prefix -D__CHECK_ENDIAN__' ARCH=i386
If you fix the issue, kindly add following tag as appropriate
"sparse warnings: (new ones prefixed by >>)"
>> drivers/misc/mic/vop/vop_vringh.c:1052:29: sparse: sparse: incorrect type in assignment (different base types) @@ expected unsigned long @@ got restricted __le64 [usertype] address @@ >> drivers/misc/mic/vop/vop_vringh.c:1052:29: sparse: expected unsigned long >> drivers/misc/mic/vop/vop_vringh.c:1052:29: sparse: got restricted __le64 [usertype] address
vim +1052 drivers/misc/mic/vop/vop_vringh.c
1024
1025 static inline int
1026 vop_query_offset(struct vop_vdev *vdev, unsigned long offset,
1027 unsigned long *size, unsigned long *pa)
1028 {
1029 struct vop_device *vpdev = vdev->vpdev;
1030 struct mic_vqconfig *vqconfig = mic_vq_config(vdev->dd);
1031 unsigned long start = MIC_DP_SIZE;
1032 int i;
1033
1034 /*
1035 * MMAP interface is as follows:
1036 * offset region
1037 * 0x0 virtio device_page
1038 * 0x1000 first vring
1039 * 0x1000 + size of 1st vring second vring
1040 * ....
1041 */
1042 if (!offset) {
1043 *pa = virt_to_phys(vpdev->hw_ops->get_dp(vpdev));
1044 *size = MIC_DP_SIZE;
1045 return 0;
1046 }
1047
1048 for (i = 0; i < vdev->dd->num_vq; i++) {
1049 struct vop_vringh *vvr = &vdev->vvr[i];
1050
1051 if (offset == start) {
> 1052 *pa = vqconfig[i].address;
1053 *size = vvr->vring.len;
1054 return 0;
1055 }
1056 start += vvr->vring.len;
1057 }
1058 return -1;
1059 }
1060
---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all(a)lists.01.org
Re: [PATCH V3 1/4] misc: vop: change the way of allocating vring and device page
In reply to Sherry SunQuoted 20 lines
static int
mic_dp_init(struct mic_device *mdev)
{drivers/misc/mic/host/mic_main.c +2
-9- mdev->dp =kzalloc (MIC_DP_SIZE ,GFP_KERNEL ); + mdev->dp =dma_alloc_coherent (&mdev->pdev->dev,MIC_DP_SIZE , + &mdev->dp_dma_addr,GFP_KERNEL );if (!mdev->dp)return -ENOMEM ; - mdev->dp_dma_addr =mic_map_single (mdev, - mdev->dp,MIC_DP_SIZE ); -if (mic_map_error (mdev->dp_dma_addr)) { -kfree (mdev->dp); -dev_err (&mdev->pdev->dev,"%s %d err %d\n" , - __func__, __LINE__, -ENOMEM ); -return -ENOMEM ; - } mdev->ops->write_spad (mdev,MIC_DPLO_SPAD , mdev->dp_dma_addr); mdev->ops->write_spad (mdev,MIC_DPHI_SPAD , mdev->dp_dma_addr >>32 );return 0 ;
drivers/misc/mic/host/mic_main.c static int mic_dp_init(struct mic_device *mdev) · line 66
Quoted 7 lines
drivers/misc/mic/host/mic_main.c +1
-2@@ -68,8 +62,7 @@ static int mic_dp_init(struct mic_device *mdev)/* Uninitialize the device page */ static void mic_dp_uninit (struct mic_device *mdev) { -mic_unmap_single (mdev, mdev->dp_dma_addr,MIC_DP_SIZE ); -kfree (mdev->dp); +dma_free_coherent (&mdev->pdev->dev,MIC_DP_SIZE , mdev->dp, mdev->dp_dma_addr); }
This adds an over 80 char line. Also please just kill mic_dp_init and mic_dp_uninit and inline those into the callers.
drivers/misc/mic/vop/vop_vringh.c static int vop_virtio_add_device(struct vop_vdev *vdev, · line 333
Quoted 2 lines
drivers/misc/mic/vop/vop_vringh.c +2
-0+ vvr->buf =dma_alloc_coherent (vop_dev (vdev),VOP_INT_DMA_BUF_SIZE , + &vvr->buf_da,GFP_KERNEL );
Another overly long line.
drivers/misc/mic/vop/vop_vringh.c vop_query_offset(struct vop_vdev *vdev, unsigned long offset, · line 1052
Quoted 6 lines
drivers/misc/mic/vop/vop_vringh.c +1
-1@@ -1068,7 +1049,7 @@ vop_query_offset(struct vop_vdev *vdev, unsigned long offset,struct vop_vringh *vvr = &vdev->vvr[i];if (offset == start) { - *pa =virt_to_phys (vvr->vring.va); + *pa = vqconfig[i].address;
vqconfig[i].address is a __le64, so this needs an endian swap.
But more importantly the caller of vop_query_offset, vop_mmap, uses remap_pfn_range and pa. You cannot mix remap_pfn_range with DMA coherent allocations, it can only be mmaped using dma_mmap_coherent.
RE: [PATCH V3 1/4] misc: vop: change the way of allocating vring and device page
Hi Christoph,
Quoted 24 lines
Quoted 24 lines
static int
mic_dp_init(struct mic_device *mdev){drivers/misc/mic/host/mic_main.c +2
-9- mdev->dp =kzalloc (MIC_DP_SIZE ,GFP_KERNEL ); + mdev->dp =dma_alloc_coherent (&mdev->pdev->dev,MIC_DP_SIZE , + &mdev->dp_dma_addr,GFP_KERNEL );if (!mdev->dp)return -ENOMEM ; - mdev->dp_dma_addr =mic_map_single (mdev, - mdev->dp,MIC_DP_SIZE ); -if (mic_map_error (mdev->dp_dma_addr)) { -kfree (mdev->dp); -dev_err (&mdev->pdev->dev,"%s %d err %d\n" , - __func__, __LINE__, -ENOMEM ); -return -ENOMEM ; - } mdev->ops->write_spad(mdev, MIC_DPLO_SPAD, mdev-
dp_dma_addr);
mdev->ops->write_spad(mdev,MIC_DPHI_SPAD, mdev-dp_dma_addr>> 32);(quoted diff) +0
-0return 0;@@ -68,8 +62,7 @@ static int mic_dp_init(struct mic_device *mdev) /* Uninitialize the device page */ static void mic_dp_uninit(struct mic_device *mdev) {
Quoted 8 lines
drivers/misc/mic/host/mic_main.c static int mic_dp_init(struct mic_device *mdev) · line 66
Quoted 5 lines
drivers/misc/mic/host/mic_main.c +2
-2-mic_unmap_single (mdev, mdev->dp_dma_addr,MIC_DP_SIZE ); -kfree (mdev->dp); +dma_free_coherent (&mdev->pdev->dev,MIC_DP_SIZE , mdev->dp, +mdev->dp_dma_addr); }This adds an over 80 char line. Also please just kill
mic_dp_initandmic_dp_uninitand inline those into the callers.
Okay, will fix them.
Quoted 17 lines
Review static int mic_dp_init(struct mic_device *mdev) · line 67
Quoted 2 lines
(quoted diff) +1
-0+ vvr->buf = dma_alloc_coherent(vop_dev(vdev),
VOP_INT_DMA_BUF_SIZE,drivers/misc/mic/vop/vop_vringh.c static int vop_virtio_add_device(struct vop_vdev *vdev, · line 333
Quoted 1 line
drivers/misc/mic/vop/vop_vringh.c +1
-0+ &vvr->buf_da, GFP_KERNEL);Another overly long line.
Review vop_query_offset(struct vop_vdev *vdev,
Quoted 1 line
(quoted diff) +0
-0@@ -1068,7 +1049,7 @@ vop_query_offset(struct vop_vdev *vdev,unsigned long offset,
drivers/misc/mic/vop/vop_vringh.c vop_query_offset(struct vop_vdev *vdev, unsigned long offset, · line 1052
Quoted 5 lines
drivers/misc/mic/vop/vop_vringh.c +1
-1struct vop_vringh *vvr = &vdev->vvr[i];if (offset == start) { - *pa =virt_to_phys (vvr->vring.va); + *pa = vqconfig[i].address;vqconfig[i].address is a
__le64, so this needs an endian swap.
Quoted 3 lines
But more importantly the caller of
vop_query_offset,vop_mmap, usesremap_pfn_rangeand pa. You cannot mixremap_pfn_rangewith DMA coherent allocations, it can only be mmaped usingdma_mmap_coherent.
Will change to use dma_mmap_coherent in V4.
Best regards
Sherry
PATCH v3 2/4 misc: vop: do not allocate and reassign the used ring
In reply to Sherry SunWe don't need to allocate and reassign the used ring here and remove the used_address_updated flag.Since RC has allocated the entire vring, including the used ring. Simply add the corresponding offset can get the used ring address.
If following the orginal way to reassign the used ring, will encounter a problem. When host finished with descriptor, it will update the used ring with putused_kern api, if reassign used ring at EP side, used ring will be io device memory for RC, use memcpy in putused_kern will cause kernel panic.
--- drivers/misc/mic/vop/vop_debugfs.c | 2 -- drivers/misc/mic/vop/vop_main.c | 48 ++++-------------------------- drivers/misc/mic/vop/vop_vringh.c | 31 ------------------- include/uapi/linux/mic_common.h | 5 ++-- 4 files changed, 8 insertions(+), 78 deletions(-)
drivers/misc/mic/vop/vop_debugfs.c +0 -2
@@ -79,8 +79,6 @@ static int vop_dp_show(struct seq_file *s, void *pos)seq_printf (s,"Vdev reset %d\n" , dc->vdev_reset);seq_printf (s,"Guest Ack %d " , dc->guest_ack);seq_printf (s,"Host ack %d\n" , dc->host_ack); -seq_printf (s,"Used address updated %d " , - dc->used_address_updated);seq_printf (s,"Vdev 0x%llx\n" , dc->vdev);seq_printf (s,"c2h doorbell %d " , dc->c2h_vdev_db);seq_printf (s,"h2c doorbell %d\n" , dc->h2c_vdev_db);
drivers/misc/mic/vop/vop_main.c +6 -42
84 lines not on this page · raw message
drivers/misc/mic/vop/vop_vringh.c +0 -31
@@ -53,33 +53,6 @@ static void _vop_notify(struct vringh *vrh) vpdev->hw_ops->send_intr (vpdev, db); } -static void vop_virtio_init_post (struct vop_vdev *vdev) -{ -struct mic_vqconfig *vqconfig =mic_vq_config (vdev->dd); -struct vop_device *vpdev = vdev->vpdev; -int i, used_size; - -for (i =0 ; i < vdev->dd->num_vq; i++) { - used_size =PAGE_ALIGN (sizeof (u16) *3 + -sizeof (struct vring_used_elem ) * -le16_to_cpu (vqconfig->num)); -if (!le64_to_cpu (vqconfig[i].used_address)) { -dev_warn (vop_dev (vdev),"used_address zero??\n" ); -continue ; - } - vdev->vvr[i].vrh.vring.used = - (void __force *)vpdev->hw_ops->remap ( - vpdev, -le64_to_cpu (vqconfig[i].used_address), - used_size); - } - - vdev->dc->used_address_updated =0 ; - -dev_info (vop_dev (vdev),"%s: device type %d LINKUP\n" , - __func__, vdev->virtio_id); -} -static inline void vop_virtio_device_reset (struct vop_vdev *vdev) {int i;
@@ -130,9 +103,6 @@ static void vop_bh_handler(struct work_struct *work) struct vop_vdev *vdev = container_of(work, struct vop_vdev, virtio_bh_work); -if (vdev->dc->used_address_updated) -vop_virtio_init_post (vdev); -if (vdev->dc->vdev_reset)vop_virtio_device_reset (vdev);
@@ -250,7 +220,6 @@ static void vop_init_device_ctrl(struct vop_vdev *vdev, dc->guest_ack =0 ; dc->vdev_reset =0 ; dc->host_ack =0 ; - dc->used_address_updated =0 ; dc->c2h_vdev_db =-1 ; dc->h2c_vdev_db =-1 ; vdev->dc = dc;
include/uapi/linux/mic_common.h +2 -3
@@ -56,8 +56,7 @@ struct mic_device_desc {* @vdev_reset: Set to 1 by guest to indicate virtio device has been reset. * @guest_ack: Set to 1 by guest to ack a command. * @host_ack: Set to 1 by host to ack a command. -* @used_address_updated: Set to 1 by guest when the used address should be -* updated. +* @must_be_zero: Reserved because this bit is no longer needed. * @c2h_vdev_db: The doorbell number to be used by guest. Set by host. * @h2c_vdev_db: The doorbell number to be used by host. Set by guest. */
@@ -67,7 +66,7 @@ struct mic_device_ctrl {__u8 vdev_reset;__u8 guest_ack;__u8 host_ack; -__u8 used_address_updated; +__u8 must_be_zero;__s8 c2h_vdev_db;__s8 h2c_vdev_db; } __attribute__ ((aligned(8)));
Signature · 1 line
2.17.1
kernel test robot
build botRe: [PATCH V3 2/4] misc: vop: do not allocate and reassign the used ring
Thank you for the patch! Perhaps something to improve:
Hi Sherry,
Thank you for the patch! Perhaps something to improve:
[auto build test WARNING on char-misc/char-misc-testing] [also build test WARNING on soc/for-next linus/master v5.9 next-20201022] [If your patch is applied to the wrong git tree, kindly drop us a note. And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]
url: https://github.com/0day-ci/linux/commits/Sherry-Sun/Change-vring-space-from-nomal-memory-to-dma-coherent-memory/20201022-131008 base: https://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc.git f3277cbfba763cd2826396521b9296de67cf1bbc config: i386-randconfig-s002-20201022 (attached as .config) compiler: gcc-9 (Debian 9.3.0-15) 9.3.0 reproduce: # apt-get install sparse # sparse version: v0.6.3-dirty # https://github.com/0day-ci/linux/commit/6ae9d6d36b63c7bb8170f4c0409470d8e7101880 git remote add linux-review https://github.com/0day-ci/linux git fetch --no-tags linux-review Sherry-Sun/Change-vring-space-from-nomal-memory-to-dma-coherent-memory/20201022-131008 git checkout 6ae9d6d36b63c7bb8170f4c0409470d8e7101880 # save the attached .config to linux build tree make W=1 C=1 CF='-fdiagnostic-prefix -D__CHECK_ENDIAN__' ARCH=i386
If you fix the issue, kindly add following tag as appropriate
"sparse warnings: (new ones prefixed by >>)"
>> drivers/misc/mic/vop/vop_main.c:339:14: sparse: sparse: incorrect type in assignment (different address spaces) @@ expected void *used @@ got void [noderef] __iomem * @@ >> drivers/misc/mic/vop/vop_main.c:339:14: sparse: expected void *used >> drivers/misc/mic/vop/vop_main.c:339:14: sparse: got void [noderef] __iomem * >> drivers/misc/mic/vop/vop_main.c:357:35: sparse: sparse: restricted __le64 degrades to integer
vim +339 drivers/misc/mic/vop/vop_main.c
289
290 /*
291 * This routine will assign vring's allocated in host/io memory. Code in
292 * virtio_ring.c however continues to access this io memory as if it were local
293 * memory without io accessors.
294 */
295 static struct virtqueue *vop_find_vq(struct virtio_device *dev,
296 unsigned index,
297 void (*callback)(struct virtqueue *vq),
298 const char *name, bool ctx)
299 {
300 struct _vop_vdev *vdev = to_vopvdev(dev);
301 struct vop_device *vpdev = vdev->vpdev;
302 struct mic_vqconfig __iomem *vqconfig;
303 struct mic_vqconfig config;
304 struct virtqueue *vq;
305 void __iomem *va;
306 struct _mic_vring_info __iomem *info;
307 void *used;
308 int vr_size, _vr_size, err, magic;
309 u8 type = ioread8(&vdev->desc->type);
310
311 if (index >= ioread8(&vdev->desc->num_vq))
312 return ERR_PTR(-ENOENT);
313
314 if (!name)
315 return ERR_PTR(-ENOENT);
316
317 /* First assign the vring's allocated in host memory */
318 vqconfig = _vop_vq_config(vdev->desc) + index;
319 memcpy_fromio(&config, vqconfig, sizeof(config));
320 _vr_size = round_up(vring_size(le16_to_cpu(config.num), MIC_VIRTIO_RING_ALIGN), 4);
321 vr_size = PAGE_ALIGN(_vr_size + sizeof(struct _mic_vring_info));
322 va = vpdev->hw_ops->remap(vpdev, le64_to_cpu(config.address), vr_size);
323 if (!va)
324 return ERR_PTR(-ENOMEM);
325 vdev->vr[index] = va;
326 memset_io(va, 0x0, _vr_size);
327
328 info = va + _vr_size;
329 magic = ioread32(&info->magic);
330
331 if (WARN(magic != MIC_MAGIC + type + index, "magic mismatch")) {
332 err = -EIO;
333 goto unmap;
334 }
335
336 vdev->used_size[index] = PAGE_ALIGN(sizeof(__u16) * 3 +
337 sizeof(struct vring_used_elem) *
338 le16_to_cpu(config.num));
> 339 used = va + PAGE_ALIGN(sizeof(struct vring_desc) * le16_to_cpu(config.num) +
340 sizeof(__u16) * (3 + le16_to_cpu(config.num)));
341 vdev->used_virt[index] = used;
342 if (!used) {
343 err = -ENOMEM;
344 dev_err(_vop_dev(vdev), "%s %d err %d\n",
345 __func__, __LINE__, err);
346 goto unmap;
347 }
348
349 vq = vop_new_virtqueue(index, le16_to_cpu(config.num), dev, ctx,
350 (void __force *)va, vop_notify, callback,
351 name, used);
352 if (!vq) {
353 err = -ENOMEM;
354 goto unmap;
355 }
356
> 357 vdev->used[index] = config.address + PAGE_ALIGN(sizeof(struct vring_desc) * le16_to_cpu(config.num) + 358 sizeof(__u16) * (3 + le16_to_cpu(config.num)));
359 writeq(vdev->used[index], &vqconfig->used_address);
360
361 vq->priv = vdev;
362 return vq;
363 unmap:
364 vpdev->hw_ops->unmap(vpdev, vdev->vr[index]);
365 return ERR_PTR(err);
366 }
367
---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all(a)lists.01.org
Re: [PATCH V3 2/4] misc: vop: do not allocate and reassign the used ring
In reply to Sherry SunQuoted 4 lines · Sherry Sun
We don't need to allocate and reassign the used ring here and remove the
used_address_updatedflag.Since RC has allocated the entire vring, including the used ring. Simply add the corresponding offset can get the used ring address.
Someone needs to verify this vs the existing intel implementations.
drivers/misc/mic/vop/vop_main.c static struct virtqueue *vop_find_vq(struct virtio_device *dev, · line 339
Quoted 3 lines
drivers/misc/mic/vop/vop_main.c +1
-2- used = (void *)__get_free_pages (GFP_KERNEL | __GFP_ZERO, -get_order (vdev->used_size[index])); + used = va + PAGE_ALIGN(sizeof(struct vring_desc) * le16_to_cpu(config.num) +
This adds an over 80 char line.
drivers/misc/mic/vop/vop_main.c static struct virtqueue *vop_find_vq(struct virtio_device *dev, · line 357
Quoted 1 line
drivers/misc/mic/vop/vop_main.c +1
-0+ vdev->used[index] = config.address + PAGE_ALIGN(sizeof(struct vring_desc) * le16_to_cpu(config.num) +
Again.
RE: [PATCH V3 2/4] misc: vop: do not allocate and reassign the used ring
Hi Greg & Christoph,
Quoted 10 lines
Subject: Re: [PATCH V3 2/4] misc: vop: do not allocate and reassign the used ring
Quoted 4 lines · Sherry Sun
We don't need to allocate and reassign the used ring here and remove the
used_address_updatedflag.Since RC has allocated the entire vring, including the used ring. Simply add the corresponding offset can get the used ring address.Someone needs to verify this vs the existing intel implementations.
Hi Greg, @[email protected], sorry I don't have the Intel MIC devices so cannot test it, do you know anyone who can help test this patch changes on the Intel MIC platform? Thanks.
Quoted 7 lines
drivers/misc/mic/vop/vop_main.c static struct virtqueue *vop_find_vq(struct virtio_device *dev,
Quoted 5 lines
drivers/misc/mic/vop/vop_main.c +2
-2- used = (void *)__get_free_pages (GFP_KERNEL | __GFP_ZERO, -get_order (vdev->used_size[index])); + used = va + PAGE_ALIGN(sizeof(struct vring_desc) * +le16_to_cpu(config.num) +This adds an over 80 char line.
Hi Christoph, will fix it in V4, thanks.
Quoted 5 lines
Review
Quoted 3 lines
(quoted diff) +2
-0+ vdev->used[index] = config.address + PAGE_ALIGN(sizeof(struct +vring_desc) * le16_to_cpu(config.num) +Again.
Best regards
Sherry
[email protected]
maintainerRe: [PATCH V3 2/4] misc: vop: do not allocate and reassign the used ring
Quoted 6 lines · Sherry Sun
Hi Greg & Christoph,
Quoted 4 lines
Subject: Re: [PATCH V3 2/4] misc: vop: do not allocate and reassign the used ring
On Thu, Oct 22, 2020 at 01:06:36PM +0800, Sherry Sun wrote:
Quoted 8 lines
Quoted 6 lines
Quoted 4 lines
We don't need to allocate and reassign the used ring here and remove the
used_address_updatedflag.Since RC has allocated the entire vring, including the used ring. Simply add the corresponding offset can get the used ring address.Someone needs to verify this vs the existing intel implementations.
Hi Greg, @[email protected], sorry I don't have the Intel MIC devices so cannot test it, do you know anyone who can help test this patch changes on the Intel MIC platform? Thanks.
Why not ask the authors/maintainers of that code?
greg k-h
RE: [PATCH V3 2/4] misc: vop: do not allocate and reassign the used ring
Hi Sudeep & Ashutosh & Vincent,
Can you help test the patch about removing the codes of reassign used ring, and comment on the impact for Intel MIC platform? Thanks for any help.
Best regards
Sherry
Quoted 24 lines
Subject: Re: [PATCH V3 2/4] misc: vop: do not allocate and reassign the used ring
Quoted 6 lines · Sherry Sun
Hi Greg & Christoph,
Quoted 4 lines
Subject: Re: [PATCH V3 2/4] misc: vop: do not allocate and reassign the used ring
On Thu, Oct 22, 2020 at 01:06:36PM +0800, Sherry Sun wrote:
Quoted 8 lines
Quoted 6 lines
Quoted 4 lines
We don't need to allocate and reassign the used ring here and remove the
used_address_updatedflag.Since RC has allocated the entire vring, including the used ring. Simply add the corresponding offset can get the used ring address.Someone needs to verify this vs the existing intel implementations.
Hi Greg, @[email protected], sorry I don't have the Intel MIC
devices so cannot test it, do you know anyone who can help test this patch changes on the Intel MIC platform? Thanks.
Why not ask the authors/maintainers of that code?
greg k-h
Re: [PATCH V3 2/4] misc: vop: do not allocate and reassign the used ring
Quoted 3 lines · Sherry Sun
Can you help test the patch about removing the codes of reassign used ring, and comment on the impact for Intel MIC platform? Thanks for any help.
I don't have access to MIC hardware myself, either.
But this patch is quite certainly going to break it since guests using a kernel without the patch will not work with hosts with a kernel with the patch.
RE: [PATCH V3 2/4] misc: vop: do not allocate and reassign the used ring
Hi Vincent,
Quoted 4 lines
Subject: Re: [PATCH V3 2/4] misc: vop: do not allocate and reassign the used ring
On Tue, Oct 27, 2020 at 08:05:43AM +0100, Sherry Sun wrote:
Quoted 8 lines
Quoted 3 lines
Can you help test the patch about removing the codes of reassign used ring, and comment on the impact for Intel MIC platform? Thanks for any help.
I don't have access to MIC hardware myself, either.
But this patch is quite certainly going to break it since guests using a kernel without the patch will not work with hosts with a kernel with the patch.
Thanks for your reply.
This patch can be used by both guests and hosts.
I have tested it on imx8qm platform(both guest and host are ARM64 architecture), and it works well.
So I guess Intel MIC won't meet big problems when both guest and host apply this patch. But it is best if it can be tested.
Best regards
Sherry
Re: [PATCH V3 2/4] misc: vop: do not allocate and reassign the used ring
Quoted 10 lines · Sherry Sun
Quoted 10 lines
Subject: Re: [PATCH V3 2/4] misc: vop: do not allocate and reassign the used ring
Quoted 3 lines · Sherry Sun
Can you help test the patch about removing the codes of reassign used ring, and comment on the impact for Intel MIC platform? Thanks for any help.
I don't have access to MIC hardware myself, either.
Quoted 8 lines
Quoted 2 lines
But this patch is quite certainly going to break it since guests using a kernel without the patch will not work with hosts with a kernel with the patch.
Thanks for your reply.
This patch can be used by both guests and hosts.
I have tested it on imx8qm platform(both guest and host are ARM64 architecture), and it works well. So I guess Intel MIC won't meet big problems when both guest and host
apply this patch. But it is best if it can be tested.
The guest and host are different systems and it should be possible to upgrade the host kernel without being forced to upgrade the guest kernel, and vice-versa. This patch breaks that.
If MIC hardware has no users and the drivers are being deleted then of course backward compatibility doesn't matter.
A message that is not in the archive (sent off-list or not yet received): [email protected]
RE: [PATCH V3 2/4] misc: vop: do not allocate and reassign the used ring
Hi Sudeep,
Quoted 32 lines
Subject: Re: [PATCH V3 2/4] misc: vop: do not allocate and reassign the used ring
Quoted 16 lines · Sherry Sun
Hi Vincent,
Quoted 8 lines
Subject: Re: [PATCH V3 2/4] misc: vop: do not allocate and reassign the used ring
Quoted 3 lines · Sherry Sun
Can you help test the patch about removing the codes of reassign used ring, and comment on the impact for Intel MIC platform? Thanks for any help.
Quoted 5 lines
I don't have access to MIC hardware myself, either.
But this patch is quite certainly going to break it since guests using a kernel without the patch will not work with hosts with a kernel with the patch.
Quoted 5 lines
Thanks for your reply.
This patch can be used by both guests and hosts.
I have tested it on imx8qm platform(both guest and host are ARM64 architecture), and it works well.
So I guess Intel MIC won't meet big problems when both guest andHi Greg/Sherry,
Both Ashutosh and I have moved on to other projects. The MIC devices have been discontinued. I have just sent across a patch to remove the MIC drivers from the kernel tree.
Quoted 5 lines
We are very glad to see that Sherry is able to reuse some of the VOP logic and it is working well. It is best if the MIC drivers are removed so Sherry can add the specific VOP logic required for imx8qm subsequently without having to worry about other driver dependencies.
Hoping this results in a cleaner imx8qm driver moving forward.
I'm ok with your patch.
Since you have deprecated the MIC related code, may I ask do you have a better solution instead of vop/scif?
Best regards
Sherry
Quoted 3 lines
Thanks,
Sudeep Dutt
A message that is not in the archive (sent off-list or not yet received): CAK8P3a1JRx32VfFcwFpK0i6F5MQMCK-yCKw8=d_R08Y3iQ7wLQ@mail.gmail.com
Re: [PATCH V3 2/4] misc: vop: do not allocate and reassign the used ring
(resending from the kernel.org address after getting bounces again)
Quoted 12 lines · Sherry Sun
Quoted 11 lines
Subject: Re: [PATCH V3 2/4] misc: vop: do not allocate and reassign the used
Both Ashutosh and I have moved on to other projects. The MIC devices have been discontinued. I have just sent across a patch to remove the MIC drivers from the kernel tree.
We are very glad to see that Sherry is able to reuse some of the VOP logic and it is working well. It is best if the MIC drivers are removed so Sherry can add the specific VOP logic required for imx8qm subsequently without having to worry about other driver dependencies.
Hoping this results in a cleaner imx8qm driver moving forward.
Quoted 3 lines
I'm ok with your patch.
Since you have deprecated the MIC related code, may I ask do you have a better solution instead of vop/scif?
I think we should try to do something on top of the PCIe endpoint subsystem to make it work across arbitrary combinations of host and device implementations,
and provide a superset of what the MIC driver, (out-of-tree) Bluefield endpoint driver, and the NTB subsystem as well as a couple of others used to do, each of them tunneling block/network/serial/... over a PCIe link of some sort, usually with virtio.
At the moment, there is only one driver for the endpoint framework in the kernel, in drivers/pci/endpoint/functions/pci-epf-test.c, but I think this can serve as a starting point.
The PCI endpoint subsystem already uses configfs for configuring the available devices, and this seems like a good fit for making it work in general. However, there are a number of use cases that have somewhat conflicting requirements, so the first step would be to figure out what everyone actually needs for virtio communication.
These are some of the main differences that I have noticed in the past:
- The simple case would be to use one PCIe endpoint device for each virtio device, but I think this needs to be multiplexed so that hardware that only supports a single PCIe endpoint can still have multiple virtio devices tunneled through it.
- While sometimes the configuration is hardcoded in the driver, ideally the type of virtio device(s) that is tunneled over the PCIe link should be configurable. The configuration of the endpoint device itself is done on the machine running on the endpoint side, but for the virtio devices, this might be either on the host or the endpoint. Not sure if one of the two ways is common enough, or we have to allow both.
- When the link is configured, you still need one side to provide a virtio device host implementation, while the other side would run the normal virtio device driver. Again, these could be done either way, and it is independent of which side has configured the link, and we might want to only allow one of the two options, or do both, or tie it to who configures it (e.g. the side that creates the device must be the virtio device host, while the other side just sees the device pop up and uses a virtio driver).
Arnd
RE: [PATCH V3 2/4] misc: vop: do not allocate and reassign the used ring
Hi Arnd,
Quoted 30 lines
Subject: Re: [PATCH V3 2/4] misc: vop: do not allocate and reassign the used ring
(resending from the kernel.org address after getting bounces again)
Quoted 13 lines · Sherry Sun
Quoted 7 lines
Subject: Re: [PATCH V3 2/4] misc: vop: do not allocate and reassign the used
Both Ashutosh and I have moved on to other projects. The MIC devices have been discontinued. I have just sent across a patch to remove the MIC drivers from the kernel tree.
Quoted 5 lines
We are very glad to see that Sherry is able to reuse some of the VOP logic and it is working well. It is best if the MIC drivers are removed so Sherry can add the specific VOP logic required for imx8qm subsequently without having to worry about other driver dependencies. Hoping this results in a cleaner imx8qm driver moving forward.
Quoted 3 lines
I'm ok with your patch.
Since you have deprecated the MIC related code, may I ask do you have a better solution instead of vop/scif?I think we should try to do something on top of the PCIe endpoint subsystem to make it work across arbitrary combinations of host and device implementations, and provide a superset of what the MIC driver, (out-of- tree) Bluefield endpoint driver, and the NTB subsystem as well as a couple of others used to do, each of them tunneling
block/network/serial/... over a PCIe link of some sort, usually with virtio.
Quoted 4 lines
At the moment, there is only one driver for the endpoint framework in the kernel, in
drivers/pci/endpoint/functions/pci-epf-test.c, but I think this can serve as a starting point.
Thanks for your detailed reply.
Yes, the PCIe endpoint subsystem is the base code, actually we have implemented a set of pci endpoint code similar to pci-epf-test.c, combine with vop (Virtio Over PCIe).
But now the vop code is going to be removed, we planned to change to NTB framework, I saw Kishon has done some jobs based on NTB and PCIe endpoint subsystem, will get a deep look. Maybe it is a good solution.
Best regards
Sherry
Quoted 32 lines
The PCI endpoint subsystem already uses configfs for configuring the available devices, and this seems like a good fit for making it work in general. However, there are a number of use cases that have somewhat conflicting requirements, so the first step would be to figure out what everyone actually needs for virtio communication.
These are some of the main differences that I have noticed in the past:
- The simple case would be to use one PCIe endpoint device for each virtio device, but I think this needs to be multiplexed so that hardware that only supports a single PCIe endpoint can still have multiple virtio devices tunneled through it.
- While sometimes the configuration is hardcoded in the driver, ideally the type of virtio device(s) that is tunneled over the PCIe link should be configurable. The configuration of the endpoint device itself is done on the machine running on the endpoint side, but for the virtio devices, this might be either on the host or the endpoint. Not sure if one of the two ways is common enough, or we have to allow both.
- When the link is configured, you still need one side to provide a virtio device host implementation, while the other side would run the normal virtio device driver. Again, these could be done either way, and it is independent of which side has configured the link, and we might want to only allow one of the two options, or do both, or tie it to who configures it (e.g. the side that creates the device must be the virtio device host, while the other side just sees the device pop up and uses a virtio driver).
Arnd
Re: [PATCH V3 2/4] misc: vop: do not allocate and reassign the used ring
Quoted 9 lines · Sherry Sun
Quoted 1 line
Subject: Re: [PATCH V3 2/4] misc: vop: do not allocate and reassign the used ring
Thanks for your detailed reply.
Yes, the PCIe endpoint subsystem is the base code, actually we have implemented a set of pci endpoint code similar topci-epf-test.c, combine with vop (Virtio Over PCIe).But now the vop code is going to be removed, we planned to change to NTB framework, I saw Kishon has done some jobs based on NTB and PCIe endpoint subsystem, will get a deep look. Maybe it is a good solution.
Ok, great. Regarding the VoP code, I think nothing stops you from reusing anything you find useful in there and build on top of it, but we should consider it as a new submission then, which means that you are free to change the interfaces without worrying about backwards compatibility.
On the flip side, this also means we need to carefully review the interface to make sure we can cover the requirements of as many users as possible without adding too much complexity for cases that we do not care about.
Arnd
Re: [PATCH V3 2/4] misc: vop: do not allocate and reassign the used ring
In reply to Arnd BergmannQuoted 7 lines · Arnd Bergmann
I think we should try to do something on top of the PCIe endpoint subsystem to make it work across arbitrary combinations of host and device implementations,
and provide a superset of what the MIC driver, (out-of-tree) Bluefield endpoint driver, and the NTB subsystem as well as a couple of others used to do, each of them tunnelingblock/network/serial/... over a PCIe link of some sort, usually with virtio.
VOP is not PCIe-specific (as demonstrated by the vop-loopback patches I posted a while ago [1]), and it would be a shame for a replacement to be tied to the PCIe endpoint subsystem. There are many SOCs out there which have multiple Linux-capable processors without cache-coherency between them. VOP is (or should I say was since I guess it's being deleted) the closest we have in mainline to easily get generic virtio (and not just rpmsg) running between these kind of Linux instances. If a new replacement framework were to be PCIe-exclusive then we'd have to invent one more framework for non-PCIe links to do pretty much the same thing.
[email protected]
maintainerRe: [PATCH V3 2/4] misc: vop: do not allocate and reassign the used ring
Quoted 20 lines · Vincent Whitchurch
Quoted 7 lines · Arnd Bergmann
I think we should try to do something on top of the PCIe endpoint subsystem to make it work across arbitrary combinations of host and device implementations,
and provide a superset of what the MIC driver, (out-of-tree) Bluefield endpoint driver, and the NTB subsystem as well as a couple of others used to do, each of them tunnelingblock/network/serial/... over a PCIe link of some sort, usually with virtio.VOP is not PCIe-specific (as demonstrated by the vop-loopback patches I posted a while ago [1]), and it would be a shame for a replacement to be tied to the PCIe endpoint subsystem. There are many SOCs out there which have multiple Linux-capable processors without cache-coherency between them. VOP is (or should I say was since I guess it's being deleted) the closest we have in mainline to easily get generic virtio (and not just rpmsg) running between these kind of Linux instances. If a new replacement framework were to be PCIe-exclusive then we'd have to invent one more framework for non-PCIe links to do pretty much the same thing.
Quoted 1 line
If this works well, please restore the existing code and move it to a new directory and we can take it from there, right?
thanks,
greg k-h
Re: [PATCH V3 2/4] misc: vop: do not allocate and reassign the used ring
In reply to Vincent WhitchurchQuoted 21 lines · Vincent Whitchurch
Quoted 7 lines · Arnd Bergmann
I think we should try to do something on top of the PCIe endpoint subsystem to make it work across arbitrary combinations of host and device implementations,
and provide a superset of what the MIC driver, (out-of-tree) Bluefield endpoint driver, and the NTB subsystem as well as a couple of others used to do, each of them tunnelingblock/network/serial/... over a PCIe link of some sort, usually with virtio.VOP is not PCIe-specific (as demonstrated by the vop-loopback patches I posted a while ago [1]), and it would be a shame for a replacement to be tied to the PCIe endpoint subsystem. There are many SOCs out there which have multiple Linux-capable processors without cache-coherency between them. VOP is (or should I say was since I guess it's being deleted) the closest we have in mainline to easily get generic virtio (and not just rpmsg) running between these kind of Linux instances. If a new replacement framework were to be PCIe-exclusive then we'd have to invent one more framework for non-PCIe links to do pretty much the same thing.
Quoted 1 line
Right, sorry I forgot about that. I think this means we should keep having an abstraction between VOP (under whichever name) and the lower levels, and be aware that it might run on any number of these:
- PCIe endpoint, with the endpoint controlling the virtio configuration
- PCIe endpoint, with the host (the side that has the pci_driver) controlling the virtio configuration
- NTB connections
- your loopback mode
- Virtio tunnels between VM guests (see https://www.linaro.org/projects/#STR)
- Intel MIC (to be removed, but it would be wrong to make assumptions that cannot be made on that type of hardware)
- ...
The existing VOP codebase does look like a reasonable start, though I think we need to discuss whether the ioctl interface should be replaced with a configfs interface, and what other changes would be needed to make it support the generalized hardware case.
Arnd
Re: [PATCH V3 2/4] misc: vop: do not allocate and reassign the used ring
Quoted 30 lines · Arnd Bergmann
Quoted 21 lines · Vincent Whitchurch
Quoted 7 lines · Arnd Bergmann
I think we should try to do something on top of the PCIe endpoint subsystem to make it work across arbitrary combinations of host and device implementations,
and provide a superset of what the MIC driver, (out-of-tree) Bluefield endpoint driver, and the NTB subsystem as well as a couple of others used to do, each of them tunnelingblock/network/serial/... over a PCIe link of some sort, usually with virtio.VOP is not PCIe-specific (as demonstrated by the vop-loopback patches I posted a while ago [1]), and it would be a shame for a replacement to be tied to the PCIe endpoint subsystem. There are many SOCs out there which have multiple Linux-capable processors without cache-coherency between them. VOP is (or should I say was since I guess it's being deleted) the closest we have in mainline to easily get generic virtio (and not just rpmsg) running between these kind of Linux instances. If a new replacement framework were to be PCIe-exclusive then we'd have to invent one more framework for non-PCIe links to do pretty much the same thing.
Quoted 1 line
Right, sorry I forgot about that. I think this means we should keep having an abstraction between VOP (under whichever name) and the lower levels, and be aware that it might run on any number of these:
Quoted 8 lines
- PCIe endpoint, with the endpoint controlling the virtio configuration
- PCIe endpoint, with the host (the side that has thepci_driver) controlling the virtio configuration
- NTB connections
- your loopback mode
- Virtio tunnels between VM guests (see https://www.linaro.org/projects/#STR)
- Intel MIC (to be removed, but it would be wrong to make assumptions that cannot be made on that type of hardware)
A virtio interface being one between host and guest is inherently asymmetric. The whole innovation of the VOP design was to treat Linux on a PCIe device as a guest, there was even talk at some point of the "guest" being managed via libvirt. So here host and guest retain their specific role/personality. The host "inserts" devices which appear in the guest e.g. So I am not sure how this asymmetry plays in the scenarios mentioned above.
Re: [PATCH V3 2/4] misc: vop: do not allocate and reassign the used ring
Quoted 11 lines · Ashutosh
Quoted 9 lines · Arnd Bergmann
- PCIe endpoint, with the endpoint controlling the virtio configuration
- PCIe endpoint, with the host (the side that has thepci_driver) controlling the virtio configuration
- NTB connections
- your loopback mode
- Virtio tunnels between VM guests (see https://www.linaro.org/projects/#STR)
- Intel MIC (to be removed, but it would be wrong to make assumptions that cannot be made on that type of hardware)
Quoted 7 lines
A virtio interface being one between host and guest is inherently asymmetric. The whole innovation of the VOP design was to treat Linux on a PCIe device as a guest, there was even talk at some point of the "guest" being managed via libvirt. So here host and guest retain their specific role/personality. The host "inserts" devices which appear in the guest e.g. So I am not sure how this asymmetry plays in the scenarios mentioned above.
This is the reason I listed the PCIe endpoint mode twice. I expect that some use cases require the same setup as MIC, with Linux on some kind of PCIe add-on card using devices that are implemented on the host. In other cases we may need the opposite: you may have a PCIe add-on card that provides arbitrary services to the host in the same way that most PCIe endpoint devices work. An example might be a smart NIC running a standalone Linux, but implementing virtio-net to avoid the need for custom drivers on the host.
In the endpoint framework, it is always the endpoint that decides what PCI devices it wants to implement, but in case of VOP that device could be either the side that configures and implements the virtio devices, or the side that probes and uses them.
Arnd
Re: [PATCH V3 2/4] misc: vop: do not allocate and reassign the used ring
In reply to Vincent WhitchurchQuoted 21 lines · Vincent Whitchurch
Quoted 7 lines · Arnd Bergmann
I think we should try to do something on top of the PCIe endpoint subsystem to make it work across arbitrary combinations of host and device implementations,
and provide a superset of what the MIC driver, (out-of-tree) Bluefield endpoint driver, and the NTB subsystem as well as a couple of others used to do, each of them tunnelingblock/network/serial/... over a PCIe link of some sort, usually with virtio.VOP is not PCIe-specific (as demonstrated by the vop-loopback patches I posted a while ago [1]), and it would be a shame for a replacement to be tied to the PCIe endpoint subsystem. There are many SOCs out there which have multiple Linux-capable processors without cache-coherency between them. VOP is (or should I say was since I guess it's being deleted) the closest we have in mainline to easily get generic virtio (and not just rpmsg) running between these kind of Linux instances. If a new replacement framework were to be PCIe-exclusive then we'd have to invent one more framework for non-PCIe links to do pretty much the same thing.
Quoted 1 line
It may be possible to use VOP in other instances but it was optimized for PCIe. The rings were specifically placed to avoid doing reads across PCIe and only do writes across PCIe.
PATCH v3 3/4 misc: vop: simply return the saved dma address instead of virt_to_phys
In reply to Sherry SunThe device page and vring should use consistent memory which are allocated by dma_alloc_coherent api, when user space wants to get its physical address, virt_to_phys cannot be used, should simply return the saved device page dma address by get_dp_dma callback and the vring dma address saved in mic_vqconfig.
--- drivers/misc/mic/bus/vop_bus.h | 2 ++ drivers/misc/mic/host/mic_boot.c | 8 ++++++++ drivers/misc/mic/vop/vop_vringh.c | 8 +++++++- 3 files changed, 17 insertions(+), 1 deletion(-)
drivers/misc/mic/bus/vop_bus.h +2 -0
@@ -75,6 +75,7 @@ struct vop_driver {* node to add/remove/configure virtio devices. * @get_dp: Get access to the virtio device page used by the self * node to add/remove/configure virtio devices. +* @get_dp_dma: Get dma address of the virtio device page. * @send_intr: Send an interrupt to the peer node on a specified doorbell. * @remap: Map a buffer with the specified DMA address and length. * @unmap: Unmap a buffer previously mapped.
@@ -92,6 +93,7 @@ struct vop_hw_ops {void (*ack_interrupt)(struct vop_device *vpdev,int num);void __iomem * (*get_remote_dp)(struct vop_device *vpdev);void * (*get_dp)(struct vop_device *vpdev); +dma_addr_t (*get_dp_dma)(struct vop_device *vpdev);void (*send_intr)(struct vop_device *vpdev,int db);void __iomem * (*remap)(struct vop_device *vpdev,dma_addr_t pa,size_t len);
drivers/misc/mic/host/mic_boot.c +8 -0
@@ -89,6 +89,13 @@ static void *__mic_get_dp(struct vop_device *vpdev)return mdev->dp; } +static dma_addr_t __mic_get_dp_dma (struct vop_device *vpdev) +{ +struct mic_device *mdev =vpdev_to_mdev (&vpdev->dev); + +return mdev->dp_dma_addr; +} +static void __iomem *__mic_get_remote_dp (struct vop_device *vpdev) {return NULL ;
@@ -120,6 +127,7 @@ static struct vop_hw_ops vop_hw_ops = { .ack_interrupt = __mic_ack_interrupt, .next_db = __mic_next_db, .get_dp = __mic_get_dp, + .get_dp_dma = __mic_get_dp_dma, .get_remote_dp = __mic_get_remote_dp, .send_intr = __mic_send_intr, .remap = __mic_ioremap,
drivers/misc/mic/vop/vop_vringh.c +7 -1
@@ -1009,7 +1009,13 @@ vop_query_offset(struct vop_vdev *vdev, unsigned long offset,* .... */ if (!offset) { - *pa =virt_to_phys (vpdev->hw_ops->get_dp (vpdev)); +if (vpdev->hw_ops->get_dp_dma) + *pa = vpdev->hw_ops->get_dp_dma (vpdev); +else { +dev_err (vop_dev (vdev),"can't get device page physical address\n" ); +return -EINVAL ; + } + *size =MIC_DP_SIZE ;return 0 ; }
Signature · 1 line
2.17.1
PATCH v3 4/4 misc: vop: mapping kernel memory to user space as noncached
In reply to Sherry SunMapping kernel space memory to user space as noncached, since user space need check the updates of avail_idx and device page flags timely.
--- drivers/misc/mic/vop/vop_vringh.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/misc/mic/vop/vop_vringh.c +1 -1
@@ -1058,7 +1058,7 @@ static int vop_mmap(struct file *f, struct vm_area_struct *vma) } err =remap_pfn_range (vma, vma->vm_start + offset, pa >>PAGE_SHIFT , size, - vma->vm_page_prot); +pgprot_noncached (vma->vm_page_prot));if (err) goto ret; size_rem -= size;
Signature · 1 line
2.17.1
Re: [PATCH V3 4/4] misc: vop: mapping kernel memory to user space as noncached
Quoted 14 lines · Sherry Sun
Mapping kernel space memory to user space as noncached, since user space need check the updates of
avail_idxand device page flags timely.Signed-off-by Joakim Zhang <[email protected]>Signed-off-by Sherry Sun <[email protected]>--- drivers/misc/mic/vop/vop_vringh.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)drivers/misc/mic/vop/vop_vringh.c +0
-0
drivers/misc/mic/vop/vop_vringh.c static int vop_mmap(struct file *f, struct vm_area_struct *vma) · line 1061
Quoted 5 lines
drivers/misc/mic/vop/vop_vringh.c +1
-1@@ -1058,7 +1058,7 @@ static int vop_mmap(struct file *f, struct vm_area_struct *vma) } err =remap_pfn_range (vma, vma->vm_start + offset, pa >>PAGE_SHIFT , size, - vma->vm_page_prot); +pgprot_noncached (vma->vm_page_prot));
Again, memory allocated using dma_alloc_coherent can only be mapped using dma_mmap_coherent, which will use the right attributes for the mapping, which often are cached.