-Wattribute-warning in drivers/scsi/libfc/fc_elsct.c

[nathanchance]

With both ARCH=arm64 allmodconfig and ARCH=x86_64 allmodconfig on linux-next, I see:

In file included from drivers/scsi/libfc/fc_elsct.c:17:
In file included from ./include/scsi/libfc.h:11:
In file included from ./include/linux/timer.h:6:
In file included from ./include/linux/ktime.h:24:
In file included from ./include/linux/time.h:60:
In file included from ./include/linux/time32.h:13:
In file included from ./include/linux/timex.h:65:
In file included from ./arch/arm64/include/asm/timex.h:8:
In file included from ./arch/arm64/include/asm/arch_timer.h:18:
In file included from ./include/linux/smp.h:13:
In file included from ./include/linux/cpumask.h:12:
In file included from ./include/linux/bitmap.h:11:
In file included from ./include/linux/string.h:253:
./include/linux/fortify-string.h:328:4: error: call to __write_overflow_field declared with 'warning' attribute: detected write beyond size of field (1st parameter); maybe use struct_group()? [-Werror,-Wattribute-warning]
                        __write_overflow_field(p_size_field, size);
                        ^
./include/linux/fortify-string.h:328:4: error: call to __write_overflow_field declared with 'warning' attribute: detected write beyond size of field (1st parameter); maybe use struct_group()? [-Werror,-Wattribute-warning]
2 errors generated.

This comes from the new fortify checks.

I noticed that @GustavoARSilva sent a patch to convert a couple of structures to use a flexible member array in drivers/scsi, which actually resolves this.

cc @kees

[GustavoARSilva]

I wonder if this patch still resolves those issues:

diff --git a/drivers/scsi/libfc/fc_encode.h b/drivers/scsi/libfc/fc_encode.h
index 74ae7fd15d8d..7dcac3b6baa7 100644
--- a/drivers/scsi/libfc/fc_encode.h
+++ b/drivers/scsi/libfc/fc_encode.h
@@ -246,7 +246,7 @@ static inline int fc_ct_ms_fill(struct fc_lport *lport,
                                   &entry->type);
                put_unaligned_be16(len, &entry->len);
                put_unaligned_be64(lport->wwnn,
-                                  (__be64 *)&entry->value[0]);
+                                  (__be64 *)&entry->value);

                /* Manufacturer */
                entry = (struct fc_fdmi_attr_entry *)((char *)entry->value +
diff --git a/include/scsi/fc/fc_ms.h b/include/scsi/fc/fc_ms.h
index 00191695233a..56a5d2b5a624 100644
--- a/include/scsi/fc/fc_ms.h
+++ b/include/scsi/fc/fc_ms.h
@@ -158,7 +158,7 @@ struct fc_fdmi_port_name {
 struct fc_fdmi_attr_entry {
        __be16          type;
        __be16          len;
-       __u8            value[1];
+       __u8            value[];
 } __attribute__((__packed__));

 /*
@@ -166,7 +166,7 @@ struct fc_fdmi_attr_entry {
  */
 struct fs_fdmi_attrs {
        __be32                          numattrs;
-       struct fc_fdmi_attr_entry       attr[1];
+       struct fc_fdmi_attr_entry       attr[];
 } __attribute__((__packed__));

 /*

[nathanchance]

@GustavoARSilva it does, thanks!

[GustavoARSilva]

Great. :) I just sent out a proper patch for this: https://lore.kernel.org/linux-hardening/20220214223903.GA859464@embeddedor/

[GustavoARSilva]

Patch applied to 5.18/scsi-staging: https://lore.kernel.org/linux-hardening/yq1lexv91gl.fsf@ca-mkp.ca.oracle.com/T/#m62f3046ac7770bf25e646155b71bdaddb5ee8274