[PATCH 6.1 23/64] hwmon: (coretemp) Fix out-of-bounds memory access

6.1-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Zhang Rui <rui.zhang@intel.com>

[ Upstream commit 4e440abc894585a34c2904a32cd54af1742311b3 ]

Fix a bug that pdata->cpu_map[] is set before out-of-bounds check.
The problem might be triggered on systems with more than 128 cores per
package.

Fixes: 7108b80a542b ("hwmon/coretemp: Handle large core ID value")
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Cc: <redacted>
Link: https://lore.kernel.org/r/20240202092144.71180-2-rui.zhang@intel.com (local)
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Stable-dep-of: fdaf0c8629d4 ("hwmon: (coretemp) Fix bogus core_id to attr name mapping")
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/hwmon/coretemp.c | 8 ++------
 1 file changed, 2 insertions(+), 6 deletions(-)

diff --git a/drivers/hwmon/coretemp.c b/drivers/hwmon/coretemp.c
index 09aab5859fa7..a80ebf433f7c 100644
--- a/drivers/hwmon/coretemp.c
+++ b/drivers/hwmon/coretemp.c

@@ -467,18 +467,14 @@ static int create_core_data(struct platform_device *pdev, unsigned int cpu,
 	if (pkg_flag) {
 		attr_no = PKG_SYSFS_ATTR_NO;
 	} else {
-		index = ida_alloc(&pdata->ida, GFP_KERNEL);
+		index = ida_alloc_max(&pdata->ida, NUM_REAL_CORES - 1, GFP_KERNEL);
 		if (index < 0)
 			return index;
+
 		pdata->cpu_map[index] = topology_core_id(cpu);
 		attr_no = index + BASE_SYSFS_ATTR_NO;
 	}
 
-	if (attr_no > MAX_CORE_DATA - 1) {
-		err = -ERANGE;
-		goto ida_free;
-	}
-
 	tdata = init_temp_data(cpu, pkg_flag);
 	if (!tdata) {
 		err = -ENOMEM;

-- 
2.43.0