kernel/linux.git/drivers/mailbox/mailbox-test.c, branch v7.1

mailbox: mailbox-test: make data_ready a per-instance variable

2026-04-18T18:10:14+00:00

While not the default case, multiple tests can be run simultaneously. Then, data_ready being a global variable will be overwritten and the per-instance lock will not help. Turn the global variable into a per-instance one to avoid this problem. Fixes: e339c80af95e ("mailbox: mailbox-test: don't rely on rx_buffer content to signal data ready") Signed-off-by: Wolfram Sang Signed-off-by: Jassi Brar

mailbox: mailbox-test: initialize struct earlier

2026-04-18T18:10:14+00:00

The waitqueue must be initialized before the debugfs files are created because from that time, requests from userspace can already be made. Similarily, drvdata and spinlock needs to be initialized before we request the channel, otherwise dangling irqs might run into problems like a NULL pointer exception. Fixes: 8ea4484d0c2b ("mailbox: Add generic mechanism for testing Mailbox Controllers") Signed-off-by: Wolfram Sang Signed-off-by: Jassi Brar

mailbox: mailbox-test: don't free the reused channel

2026-04-18T18:10:14+00:00

The RX channel can be aliased to the TX channel if it has a different MMIO. This special case needs to be handled when freeing the channels otherwise a double-free occurs. Fixes: 8ea4484d0c2b ("mailbox: Add generic mechanism for testing Mailbox Controllers") Signed-off-by: Wolfram Sang Signed-off-by: Jassi Brar

mailbox: mailbox-test: handle channel errors consistently

2026-04-18T18:10:14+00:00

mbox_test_request_channel() returns either an ERR_PTR or NULL. The callers, however, mostly checked for non-NULL which allows for bogus code paths when an ERR_PTR is treated like a valid channel. A later commit tried to fix it in one place but missed the other ones. Because the ERR_PTR is only used for -ENOMEM once and is converted to -EPROBE_DEFER anyhow, convert the callee to only return NULL which simplifies handling a lot and makes it less error prone. Fixes: 8ea4484d0c2b ("mailbox: Add generic mechanism for testing Mailbox Controllers") Fixes: 9b63a810c6f9 ("mailbox: mailbox-test: Fix an error check in mbox_test_probe()") Signed-off-by: Wolfram Sang Reviewed-by: Geert Uytterhoeven Signed-off-by: Jassi Brar

mailbox: mailbox-test: free channels on probe error

2026-04-13T01:39:02+00:00

On probe error, free the previously obtained channels. This not only prevents a leak, but also UAF scenarios because the client structure will be removed nonetheless because it was allocated with devm. Link: https://sashiko.dev/#/patchset/20260327151217.5327-2-wsa%2Brenesas%40sang-engineering.com Fixes: 8ea4484d0c2b ("mailbox: Add generic mechanism for testing Mailbox Controllers") Signed-off-by: Wolfram Sang Signed-off-by: Jassi Brar

mailbox: test: really ignore optional memory resources

2026-03-29T15:56:38+00:00

Memory resources are optional but if the resource is empty devm_platform_get_and_ioremap_resource() prints an error nonetheless. Refactor the code to check the resources locally first and process them only if they are present. The -EBUSY error message of ioremap_resource() is still kept because it is correct. The comment which explains that a plain ioremap() is tried as a workaround is turned into a info message. So, a user will be informed about it, too. Signed-off-by: Wolfram Sang Reviewed-by: Geert Uytterhoeven Signed-off-by: Jassi Brar

mailbox: mailbox-test: Fix debugfs_create_dir error checking

2025-11-28T15:31:52+00:00

The debugfs_create_dir() function returns ERR_PTR() on error, not NULL. The current null-check fails to catch errors. Use IS_ERR() to correctly check for errors. Fixes: 8ea4484d0c2b ("mailbox: Add generic mechanism for testing Mailbox Controllers") Signed-off-by: Haotian Zhang Signed-off-by: Jassi Brar

mailbox: Switch back to struct platform_driver::remove()

2024-11-24T19:11:01+00:00

After commit 0edb555a65d1 ("platform: Make platform_driver::remove() return void") .remove() is (again) the right callback to implement for platform drivers. Convert all platform drivers below drivers/mailbox to use .remove(), with the eventual goal to drop struct platform_driver::remove_new(). As .remove() and .remove_new() have the same prototypes, conversion is done by just changing the structure member name in the driver initializer. Make a few indentions consistent while touching these struct initializers. Signed-off-by: Uwe Kleine-König Reviewed-by: AngeloGioacchino Del Regno Acked-by: Chen-Yu Tsai Signed-off-by: Jassi Brar

mailbox: mailbox-test: Convert to platform remove callback returning void

2024-01-14T05:08:51+00:00

The .remove() callback for a platform driver returns an int which makes many driver authors wrongly assume it's possible to do error handling by returning an error code. However the value returned is ignored (apart from emitting a warning) and this typically results in resource leaks. To improve here there is a quest to make the remove callback return void. In the first step of this quest all drivers are converted to .remove_new(), which already returns void. Eventually after all drivers are converted, .remove_new() will be renamed to .remove(). Trivially convert this driver from always returning zero in the remove callback to the void returning variant. Signed-off-by: Uwe Kleine-König Signed-off-by: Jassi Brar

mailbox: mailbox-test: Fix an error check in mbox_test_probe()

2023-09-05T15:10:42+00:00

mbox_test_request_channel() function returns NULL or error value embedded in the pointer (PTR_ERR). Evaluate the return value using IS_ERR_OR_NULL. Signed-off-by: Minjie Du Signed-off-by: Jassi Brar