Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1317672 > unrolled thread

[PATCH] ntb: perf test: fix address space confusion

Back to article view | Back to linux.kernel

Contents

  [PATCH] ntb: perf test: fix address space confusion Arnd Bergmann <arnd@arndb.de> - 2016-01-26 10:40 +0100
    Re: [PATCH] ntb: perf test: fix address space confusion "Jiang, Dave" <dave.jiang@intel.com> - 2016-01-26 19:30 +0100

#1317672 — [PATCH] ntb: perf test: fix address space confusion

The ntb driver assigns between pointers an __iomem tokens, and
also casts them to 64-bit integers, which results in compiler
warnings on 32-bit systems:

drivers/ntb/test/ntb_perf.c: In function 'perf_copy':
drivers/ntb/test/ntb_perf.c:213:10: error: cast from pointer to integer of different size [-Werror=pointer-to-int-cast]
  vbase = (u64)(u64 *)mw->vbase;
          ^
drivers/ntb/test/ntb_perf.c:214:14: error: cast from pointer to integer of different size [-Werror=pointer-to-int-cast]
  dst_vaddr = (u64)(u64 *)dst;
              ^

This adds __iomem annotations where needed and changes the temporary
variables to iomem pointers to avoid casting them to u64. I did not
see the problem in linux-next earlier, but it show showed up in
4.5-rc1.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Fixes: 8a7b6a778a85 ("ntb: ntb perf tool")
---
 drivers/ntb/test/ntb_perf.c | 21 +++++++++++----------
 1 file changed, 11 insertions(+), 10 deletions(-)

diff --git a/drivers/ntb/test/ntb_perf.c b/drivers/ntb/test/ntb_perf.c
index c8a37ba4b4f9..6bdc1e7b7503 100644
--- a/drivers/ntb/test/ntb_perf.c
+++ b/drivers/ntb/test/ntb_perf.c
@@ -178,7 +178,7 @@ static void perf_copy_callback(void *data)
 	atomic_dec(&pctx->dma_sync);
 }
 
-static ssize_t perf_copy(struct pthr_ctx *pctx, char *dst,
+static ssize_t perf_copy(struct pthr_ctx *pctx, char __iomem *dst,
 			 char *src, size_t size)
 {
 	struct perf_ctx *perf = pctx->perf;
@@ -189,7 +189,8 @@ static ssize_t perf_copy(struct pthr_ctx *pctx, char *dst,
 	dma_cookie_t cookie;
 	size_t src_off, dst_off;
 	struct perf_mw *mw = &perf->mw;
-	u64 vbase, dst_vaddr;
+	void __iomem *vbase;
+	void __iomem *dst_vaddr;
 	dma_addr_t dst_phys;
 	int retries = 0;
 
@@ -204,14 +205,14 @@ static ssize_t perf_copy(struct pthr_ctx *pctx, char *dst,
 	}
 
 	device = chan->device;
-	src_off = (size_t)src & ~PAGE_MASK;
-	dst_off = (size_t)dst & ~PAGE_MASK;
+	src_off = (uintptr_t)src & ~PAGE_MASK;
+	dst_off = (uintptr_t __force)dst & ~PAGE_MASK;
 
 	if (!is_dma_copy_aligned(device, src_off, dst_off, size))
 		return -ENODEV;
 
-	vbase = (u64)(u64 *)mw->vbase;
-	dst_vaddr = (u64)(u64 *)dst;
+	vbase = mw->vbase;
+	dst_vaddr = dst;
 	dst_phys = mw->phys_addr + (dst_vaddr - vbase);
 
 	unmap = dmaengine_get_unmap_data(device->dev, 1, GFP_NOWAIT);
@@ -261,13 +262,13 @@ err_get_unmap:
 	return 0;
 }
 
-static int perf_move_data(struct pthr_ctx *pctx, char *dst, char *src,
+static int perf_move_data(struct pthr_ctx *pctx, char __iomem *dst, char *src,
 			  u64 buf_size, u64 win_size, u64 total)
 {
 	int chunks, total_chunks, i;
 	int copied_chunks = 0;
 	u64 copied = 0, result;
-	char *tmp = dst;
+	char __iomem *tmp = dst;
 	u64 perf, diff_us;
 	ktime_t kstart, kstop, kdiff;
 
@@ -324,7 +325,7 @@ static int ntb_perf_thread(void *data)
 	struct perf_ctx *perf = pctx->perf;
 	struct pci_dev *pdev = perf->ntb->pdev;
 	struct perf_mw *mw = &perf->mw;
-	char *dst;
+	char __iomem *dst;
 	u64 win_size, buf_size, total;
 	void *src;
 	int rc, node, i;
@@ -364,7 +365,7 @@ static int ntb_perf_thread(void *data)
 	if (buf_size > MAX_TEST_SIZE)
 		buf_size = MAX_TEST_SIZE;
 
-	dst = (char *)mw->vbase;
+	dst = (char __iomem *)mw->vbase;
 
 	atomic_inc(&perf->tsync);
 	while (atomic_read(&perf->tsync) != perf->perf_threads)
-- 
2.7.0

[toc] | [next] | [standalone]

#1318222

On Tue, 2016-01-26 at 10:31 +0100, Arnd Bergmann wrote:
> The ntb driver assigns between pointers an __iomem tokens, and
> also casts them to 64-bit integers, which results in compiler
> warnings on 32-bit systems:
> 
> drivers/ntb/test/ntb_perf.c: In function 'perf_copy':
> drivers/ntb/test/ntb_perf.c:213:10: error: cast from pointer to
> integer of different size [-Werror=pointer-to-int-cast]
>   vbase = (u64)(u64 *)mw->vbase;
>           ^
> drivers/ntb/test/ntb_perf.c:214:14: error: cast from pointer to
> integer of different size [-Werror=pointer-to-int-cast]
>   dst_vaddr = (u64)(u64 *)dst;
>               ^
> 
> This adds __iomem annotations where needed and changes the temporary
> variables to iomem pointers to avoid casting them to u64. I did not
> see the problem in linux-next earlier, but it show showed up in
> 4.5-rc1.
> 
> Signed-off-by: Arnd Bergmann <arnd@arndb.de>

Acked-by: Dave Jiang <dave.jiang@intel.com>

> Fixes: 8a7b6a778a85 ("ntb: ntb perf tool")
> ---
>  drivers/ntb/test/ntb_perf.c | 21 +++++++++++----------
>  1 file changed, 11 insertions(+), 10 deletions(-)
> 
> diff --git a/drivers/ntb/test/ntb_perf.c
> b/drivers/ntb/test/ntb_perf.c
> index c8a37ba4b4f9..6bdc1e7b7503 100644
> --- a/drivers/ntb/test/ntb_perf.c
> +++ b/drivers/ntb/test/ntb_perf.c
> @@ -178,7 +178,7 @@ static void perf_copy_callback(void *data)
>  	atomic_dec(&pctx->dma_sync);
>  }
>  
> -static ssize_t perf_copy(struct pthr_ctx *pctx, char *dst,
> +static ssize_t perf_copy(struct pthr_ctx *pctx, char __iomem *dst,
>  			 char *src, size_t size)
>  {
>  	struct perf_ctx *perf = pctx->perf;
> @@ -189,7 +189,8 @@ static ssize_t perf_copy(struct pthr_ctx *pctx,
> char *dst,
>  	dma_cookie_t cookie;
>  	size_t src_off, dst_off;
>  	struct perf_mw *mw = &perf->mw;
> -	u64 vbase, dst_vaddr;
> +	void __iomem *vbase;
> +	void __iomem *dst_vaddr;
>  	dma_addr_t dst_phys;
>  	int retries = 0;
>  
> @@ -204,14 +205,14 @@ static ssize_t perf_copy(struct pthr_ctx *pctx,
> char *dst,
>  	}
>  
>  	device = chan->device;
> -	src_off = (size_t)src & ~PAGE_MASK;
> -	dst_off = (size_t)dst & ~PAGE_MASK;
> +	src_off = (uintptr_t)src & ~PAGE_MASK;
> +	dst_off = (uintptr_t __force)dst & ~PAGE_MASK;
>  
>  	if (!is_dma_copy_aligned(device, src_off, dst_off, size))
>  		return -ENODEV;
>  
> -	vbase = (u64)(u64 *)mw->vbase;
> -	dst_vaddr = (u64)(u64 *)dst;
> +	vbase = mw->vbase;
> +	dst_vaddr = dst;
>  	dst_phys = mw->phys_addr + (dst_vaddr - vbase);
>  
>  	unmap = dmaengine_get_unmap_data(device->dev, 1,
> GFP_NOWAIT);
> @@ -261,13 +262,13 @@ err_get_unmap:
>  	return 0;
>  }
>  
> -static int perf_move_data(struct pthr_ctx *pctx, char *dst, char
> *src,
> +static int perf_move_data(struct pthr_ctx *pctx, char __iomem *dst,
> char *src,
>  			  u64 buf_size, u64 win_size, u64 total)
>  {
>  	int chunks, total_chunks, i;
>  	int copied_chunks = 0;
>  	u64 copied = 0, result;
> -	char *tmp = dst;
> +	char __iomem *tmp = dst;
>  	u64 perf, diff_us;
>  	ktime_t kstart, kstop, kdiff;
>  
> @@ -324,7 +325,7 @@ static int ntb_perf_thread(void *data)
>  	struct perf_ctx *perf = pctx->perf;
>  	struct pci_dev *pdev = perf->ntb->pdev;
>  	struct perf_mw *mw = &perf->mw;
> -	char *dst;
> +	char __iomem *dst;
>  	u64 win_size, buf_size, total;
>  	void *src;
>  	int rc, node, i;
> @@ -364,7 +365,7 @@ static int ntb_perf_thread(void *data)
>  	if (buf_size > MAX_TEST_SIZE)
>  		buf_size = MAX_TEST_SIZE;
>  
> -	dst = (char *)mw->vbase;
> +	dst = (char __iomem *)mw->vbase;
>  
>  	atomic_inc(&perf->tsync);
>  	while (atomic_read(&perf->tsync) != perf->perf_threads)

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web