[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-ID: <20240213173429.43748-1-mario.limonciello@amd.com>
Date: Tue, 13 Feb 2024 11:34:28 -0600
From: Mario Limonciello <mario.limonciello@....com>
To: Herbert Xu <herbert@...dor.apana.org.au>
CC: Mario Limonciello <mario.limonciello@....com>, Tom Lendacky
<thomas.lendacky@....com>, John Allen <john.allen@....com>, "open list:AMD
CRYPTOGRAPHIC COPROCESSOR (CCP) DRIVER - DB..."
<linux-crypto@...r.kernel.org>, open list <linux-kernel@...r.kernel.org>,
"Tim Van Patten" <timvp@...gle.com>
Subject: [PATCH v2 1/2] crypto: ccp: Avoid discarding errors in psp_send_platform_access_msg()
Errors can potentially occur in the "processing" of PSP commands or
commands can be processed successfully but still return an error code in
the header.
This second case was being discarded because PSP communication worked but
the command returned an error code in the payload header.
Capture both cases and return them to the caller as -EIO for the caller
to investigate. The caller can detect the latter by looking at
`req->header->status`.
Reported-and-tested-by: Tim Van Patten <timvp@...gle.com>
Fixes: 7ccc4f4e2e50 ("crypto: ccp - Add support for an interface for platform features")
Signed-off-by: Mario Limonciello <mario.limonciello@....com>
---
v1->v2:
* Update tag for Tim
* Reword commit message
---
drivers/crypto/ccp/platform-access.c | 11 +++++++++--
1 file changed, 9 insertions(+), 2 deletions(-)
diff --git a/drivers/crypto/ccp/platform-access.c b/drivers/crypto/ccp/platform-access.c
index 94367bc49e35..1b8ed3389733 100644
--- a/drivers/crypto/ccp/platform-access.c
+++ b/drivers/crypto/ccp/platform-access.c
@@ -118,9 +118,16 @@ int psp_send_platform_access_msg(enum psp_platform_access_msg msg,
goto unlock;
}
- /* Store the status in request header for caller to investigate */
+ /*
+ * Read status from PSP. If status is non-zero, it indicates an error
+ * occurred during "processing" of the command.
+ * If status is zero, it indicates the command was "processed"
+ * successfully, but the result of the command is in the payload.
+ * Return both cases to the caller as -EIO to investigate.
+ */
cmd_reg = ioread32(cmd);
- req->header.status = FIELD_GET(PSP_CMDRESP_STS, cmd_reg);
+ if (FIELD_GET(PSP_CMDRESP_STS, cmd_reg))
+ req->header.status = FIELD_GET(PSP_CMDRESP_STS, cmd_reg);
if (req->header.status) {
ret = -EIO;
goto unlock;
--
2.34.1
Powered by blists - more mailing lists