cxl: Fix cleanup of port devices on failure to probe driver.
PATCH 1/1 cxl: Fix cleanup of port devices on failure to probe driver.
The device is created, and then there is a check if a driver succesfully bound to it. In event of failing the bind (e.g. failure in cxl_port_probe()) the device is left registered. When a bus rescan later occurs, fresh devices are created leading to a multiple device representing the same underlying hardware. Bad things may follow and at very least we have far too many devices.
Fix by ensuring autoremove is registered if the device create succeeds, but doesn't depend on sucessful binding to a driver.
Bug was observed as side effect of incorrect ownership in [PATCH v9 6/9] cxl/port: Read CDAT table
but will result from any failure to in cxl_port_probe().
--- drivers/cxl/mem.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)
drivers/cxl/mem.c +6 -1
@@ -29,6 +29,7 @@ static int create_endpoint(struct cxl_memdev *cxlmd, {struct cxl_dev_state *cxlds = cxlmd->cxlds;struct cxl_port *endpoint; +int rc; endpoint =devm_cxl_add_port (&parent_port->dev, &cxlmd->dev, cxlds->component_reg_phys, parent_port);
@@ -37,13 +38,17 @@ static int create_endpoint(struct cxl_memdev *cxlmd,dev_dbg (&cxlmd->dev,"add: %s\n" ,dev_name (&endpoint->dev)); + rc =cxl_endpoint_autoremove (cxlmd, endpoint); +if (rc) +return rc; +if (!endpoint->dev.driver) {dev_err (&cxlmd->dev,"%s failed probe\n" ,dev_name (&endpoint->dev));return -ENXIO ; } -return cxl_endpoint_autoremove (cxlmd, endpoint); +return 0 ; }static void enable_suspend (void *data)
Signature · 1 line
2.32.0
Quoted 35 lines · Jonathan Cameron
The device is created, and then there is a check if a driver succesfully bound to it. In event of failing the bind (e.g. failure in
cxl_port_probe()) the device is left registered. When a bus rescan later occurs, fresh devices are created leading to a multiple device representing the same underlying hardware. Bad things may follow and at very least we have far too many devices.Fix by ensuring autoremove is registered if the device create succeeds, but doesn't depend on sucessful binding to a driver.
Bug was observed as side effect of incorrect ownership in [PATCH v9 6/9] cxl/port: Read CDAT table
but will result from any failure to incxl_port_probe().Fixes 8dd2bc0f8e02 ("cxl/mem: Add the cxl_mem driver")Signed-off-by Jonathan Cameron <[email protected]>--- drivers/cxl/mem.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)drivers/cxl/mem.c +1
-0@@ -29,6 +29,7 @@ static int create_endpoint(struct cxl_memdev *cxlmd, {struct cxl_dev_state *cxlds = cxlmd->cxlds;struct cxl_port *endpoint; +int rc; endpoint =devm_cxl_add_port (&parent_port->dev, &cxlmd->dev, cxlds->component_reg_phys, parent_port);@@ -37,13 +38,17 @@ static int create_endpoint(struct cxl_memdev *cxlmd,dev_dbg (&cxlmd->dev,"add: %s\n" ,dev_name (&endpoint->dev));
drivers/cxl/mem.c static int create_endpoint(struct cxl_memdev *cxlmd, · line 49
Quoted 10 lines
drivers/cxl/mem.c +4
-0+ rc =cxl_endpoint_autoremove (cxlmd, endpoint); +if (rc) +return rc; +if (!endpoint->dev.driver) {dev_err (&cxlmd->dev,"%s failed probe\n" ,dev_name (&endpoint->dev));return -ENXIO ; }
I wonder if this code is really required? I think that if the suspend code in cxl_mem_probe was reworked a bit I think this could be removed.
But for now:
Quoted 9 lines
drivers/cxl/mem.c +1
-1-return cxl_endpoint_autoremove (cxlmd, endpoint); +return 0 ; }static void enable_suspend (void *data)Signature · 1 line
2.32.0