NTB: ntb_tool: uninitialized heap data in tool_fn_write()
Subsystems (2)
PATCH 1/1 NTB: ntb_tool: uninitialized heap data in tool_fn_write()
The call to:
ret = simple_write_to_buffer(buf, size, offp, ubuf, size);
will return success if it is able to write even one byte to "buf". The value of "*offp" controls which byte. This could result in reading uninitialized data when we do the sscanf() on the next line.
This code is not really desigined to handle partial writes where *offp is non-zero and the "buf" is preserved and re-used between writes. Just ban partial writes and replace the simple_write_to_buffer() with copy_from_user().
--- drivers/ntb/test/ntb_tool.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-)
drivers/ntb/test/ntb_tool.c +5 -3
@@ -367,14 +367,16 @@ static ssize_t tool_fn_write(struct tool_ctx *tc,u64 bits;int n; +if (*offp) +return 0 ; + buf =kmalloc (size +1 ,GFP_KERNEL );if (!buf)return -ENOMEM ; - ret =simple_write_to_buffer (buf, size, offp, ubuf, size); -if (ret <0 ) { +if (copy_from_user (buf, ubuf, size)) {kfree (buf); -return ret; +return -EFAULT ; } buf[size] =0 ;
Signature · 1 line
2.35.1
Quoted 8 lines · Dan Carpenter
The call to:
ret = simple_write_to_buffer(buf, size, offp, ubuf, size);will return success if it is able to write even one byte to "buf". The value of "*offp" controls which byte. This could result in reading uninitialized data when we do the
sscanf()on the next line.
Quoted 6 lines
This code is not really desigined to handle partial writes where *offp is non-zero and the "buf" is preserved and re-used between writes. Just ban partial writes and replace the
simple_write_to_buffer()withcopy_from_user().Fixes 578b881ba9c4 ("NTB: Add tool test client")
Looks good. Thanks.
-Sergey
Quoted 33 lines
Signed-off-by Dan Carpenter <[email protected]>--- drivers/ntb/test/ntb_tool.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-)drivers/ntb/test/ntb_tool.c +5
-3@@ -367,14 +367,16 @@ static ssize_t tool_fn_write(struct tool_ctx *tc,u64 bits;int n; +if (*offp) +return 0 ; + buf =kmalloc (size +1 ,GFP_KERNEL );if (!buf)return -ENOMEM ; - ret =simple_write_to_buffer (buf, size, offp, ubuf, size); -if (ret <0 ) { +if (copy_from_user (buf, ubuf, size)) {kfree (buf); -return ret; +return -EFAULT ; } buf[size] =0 ;Signature · 1 line
2.35.1
Jon Mason
maintainerQuoted 14 lines · Serge Semin
Quoted 13 lines · Dan Carpenter
The call to:
ret = simple_write_to_buffer(buf, size, offp, ubuf, size);will return success if it is able to write even one byte to "buf". The value of "*offp" controls which byte. This could result in reading uninitialized data when we do the
sscanf()on the next line.This code is not really desigined to handle partial writes where *offp is non-zero and the "buf" is preserved and re-used between writes. Just ban partial writes and replace the
simple_write_to_buffer()withcopy_from_user().
Quoted 6 lines
Quoted 1 line
Fixes 578b881ba9c4 ("NTB: Add tool test client")Looks good. Thanks.
Reviewed-by Serge Semin <[email protected]>-Sergey
Sorry for the extremely long delay in response. This series is in my ntb branch and will be in my pull request for v5.20 which should be going out later today.
Thanks,
Jon
Quoted 34 lines
Quoted 34 lines
Signed-off-by Dan Carpenter <[email protected]>--- drivers/ntb/test/ntb_tool.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-)drivers/ntb/test/ntb_tool.c +5
-3@@ -367,14 +367,16 @@ static ssize_t tool_fn_write(struct tool_ctx *tc,u64 bits;int n; +if (*offp) +return 0 ; + buf =kmalloc (size +1 ,GFP_KERNEL );if (!buf)return -ENOMEM ; - ret =simple_write_to_buffer (buf, size, offp, ubuf, size); -if (ret <0 ) { +if (copy_from_user (buf, ubuf, size)) {kfree (buf); -return ret; +return -EFAULT ; } buf[size] =0 ;Signature · 1 line
2.35.1