Skip to contentSkip to conversation

tpm: fix event_size output in tpm1_binary_bios_measurements_show

PATCH Superseded Thorsten Blum 2 participants 2 messages May 21 – May 22

Subsystems (3)

v1 of 2 Superseded Latest is v2 →

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 tpm: fix event_size output in tpm1_binary_bios_measurements_show

Commit 186d124f07da ("tpm_eventlog.c: fix binary_bios_measurements") split the output to write the endian-converted event header first and then the variable-length event data.

However, the split was at sizeof(struct tcpa_event) - 1, even though event_data was a zero-length array, and later a flexible array member, both of which already excluded the event data.

Therefore, the current code writes the first three bytes of event_size from the endian-converted header and then the last byte from the raw header, which can emit a corrupted event_size on PPC64, where do_endian_conversion() maps to be32_to_cpu().

Use seq_write() to write the full endian-converted header, followed by the variable-length event->event_data.

Drop the obvious comment while at it.

Fixes 186d124f07da ("tpm_eventlog.c: fix binary_bios_measurements")
---
 drivers/char/tpm/eventlog/tpm1.c | 16 ++--------------
 1 file changed, 2 insertions(+), 14 deletions(-)
drivers/char/tpm/eventlog/tpm1.c +2 -14
index e7913b2853d5..291720e89d91 100644
@@ -224,29 +224,17 @@ static int tpm1_binary_bios_measurements_show(struct seq_file *m, void *v)
 {
 	struct tcpa_event *event = v;
 	struct tcpa_event temp_event;
-	char *temp_ptr;
-	int i;
 
 	memcpy(&temp_event, event, sizeof(struct tcpa_event));
 
-	/* convert raw integers for endianness */
 	temp_event.pcr_index = do_endian_conversion(event->pcr_index);
 	temp_event.event_type = do_endian_conversion(event->event_type);
 	temp_event.event_size = do_endian_conversion(event->event_size);
 
-	temp_ptr = (char *) &temp_event;
-
-	for (i = 0; i < (sizeof(struct tcpa_event) - 1) ; i++)
-		seq_putc(m, temp_ptr[i]);
-
-	temp_ptr = (char *) v;
-
-	for (i = (sizeof(struct tcpa_event) - 1);
-	     i < (sizeof(struct tcpa_event) + temp_event.event_size); i++)
-		seq_putc(m, temp_ptr[i]);
+	seq_write(m, &temp_event, sizeof(temp_event));
+	seq_write(m, event->event_data, temp_event.event_size);
 
 	return 0;
-
 }
 
 static int tpm1_ascii_bios_measurements_show(struct seq_file *m, void *v)
maintainer
#1 lore raw
Quoted 16 lines · Thorsten Blum

Commit 186d124f07da ("tpm_eventlog.c: fix binary_bios_measurements") split the output to write the endian-converted event header first and then the variable-length event data.

However, the split was at sizeof(struct tcpa_event) - 1, even though event_data was a zero-length array, and later a flexible array member, both of which already excluded the event data.

Therefore, the current code writes the first three bytes of event_size from the endian-converted header and then the last byte from the raw header, which can emit a corrupted event_size on PPC64, where do_endian_conversion() maps to be32_to_cpu().

Use seq_write() to write the full endian-converted header, followed by the variable-length event->event_data.

Quoted 8 lines

Drop the obvious comment while at it.

Fixes 186d124f07da ("tpm_eventlog.c: fix binary_bios_measurements")
---
 drivers/char/tpm/eventlog/tpm1.c | 16 ++--------------
 1 file changed, 2 insertions(+), 14 deletions(-)

Got it, I think you're probably right.

Quoted 6 lines
drivers/char/tpm/eventlog/tpm1.c +0 -0
index e7913b2853d5..291720e89d91 100644

drivers/char/tpm/eventlog/tpm1.c static int tpm1_binary_bios_measurements_show(struct seq_file *m, void *v) · line 232

Quoted 9 lines
drivers/char/tpm/eventlog/tpm1.c +0 -3
@@ -224,29 +224,17 @@ static int tpm1_binary_bios_measurements_show(struct seq_file *m, void *v)
 {
 	struct tcpa_event *event = v;
 	struct tcpa_event temp_event;
-	char *temp_ptr;
-	int i;
 
 	memcpy(&temp_event, event, sizeof(struct tcpa_event));
 
-	/* convert raw integers for endianness */

spurious change

drivers/char/tpm/eventlog/tpm1.c static int tpm1_binary_bios_measurements_show(struct seq_file *m, void *v) · line 240

Quoted 8 lines
drivers/char/tpm/eventlog/tpm1.c +0 -4
 	temp_event.pcr_index = do_endian_conversion(event->pcr_index);
 	temp_event.event_type = do_endian_conversion(event->event_type);
 	temp_event.event_size = do_endian_conversion(event->event_size);
 
-	temp_ptr = (char *) &temp_event;
-
-	for (i = 0; i < (sizeof(struct tcpa_event) - 1) ; i++)
-		seq_putc(m, temp_ptr[i]);

Why changing condition does not fix the bug? This could be just +-1 line change.

Quoted 4 lines
drivers/char/tpm/eventlog/tpm1.c +0 -4
-
-	temp_ptr = (char *) v;
-
-	for (i = (sizeof(struct tcpa_event) - 1);

drivers/char/tpm/eventlog/tpm1.c static int tpm1_binary_bios_measurements_show(struct seq_file *m, void *v) · line 240

Quoted 10 lines
drivers/char/tpm/eventlog/tpm1.c +2 -3
-	     i < (sizeof(struct tcpa_event) + temp_event.event_size); i++)
-		seq_putc(m, temp_ptr[i]);
+	seq_write(m, &temp_event, sizeof(temp_event));
+	seq_write(m, event->event_data, temp_event.event_size);
 
 	return 0;
-
 }
 
 static int tpm1_ascii_bios_measurements_show(struct seq_file *m, void *v)

BR, Jarkko