-Wuninitialized in drivers/misc/sgi-xp/xpc_partition.c #466

nathanchance · 2019-04-30T20:26:57Z

After turning on -Wuninitialized in scripts/Makefile.extrawarn or KCFLAGS, on x86 allyesconfig

drivers/misc/sgi-xp/xpc_partition.c:73:14: warning: variable 'buf' is uninitialized when used within its own initialization [-Wuninitialized]
        void *buf = buf;
              ~~~   ^~~
1 warning generated.

The text was updated successfully, but these errors were encountered:

nathanchance · 2019-05-01T16:13:28Z

I have no idea how this variable should be properly initialized, this entire function seems odd.

Link:

static unsigned long
xpc_get_rsvd_page_pa(int nasid)
{
	enum xp_retval ret;
	u64 cookie = 0;
	unsigned long rp_pa = nasid;	/* seed with nasid */
	size_t len = 0;
	size_t buf_len = 0;
	void *buf = buf;
	void *buf_base = NULL;
	enum xp_retval (*get_partition_rsvd_page_pa)
		(void *, u64 *, unsigned long *, size_t *) =
		xpc_arch_ops.get_partition_rsvd_page_pa;

	while (1) {

		/* !!! rp_pa will need to be _gpa on UV.
		 * ??? So do we save it into the architecture specific parts
		 * ??? of the xpc_partition structure? Do we rename this
		 * ??? function or have two versions? Rename rp_pa for UV to
		 * ??? rp_gpa?
		 */
		ret = get_partition_rsvd_page_pa(buf, &cookie, &rp_pa, &len);

		dev_dbg(xpc_part, "SAL returned with ret=%d, cookie=0x%016lx, "
			"address=0x%016lx, len=0x%016lx\n", ret,
			(unsigned long)cookie, rp_pa, len);

		if (ret != xpNeedMoreInfo)
			break;

		/* !!! L1_CACHE_ALIGN() is only a sn2-bte_copy requirement */
		if (is_shub())
			len = L1_CACHE_ALIGN(len);

		if (len > buf_len) {
			kfree(buf_base);
			buf_len = L1_CACHE_ALIGN(len);
			buf = xpc_kmalloc_cacheline_aligned(buf_len, GFP_KERNEL,
							    &buf_base);
			if (buf_base == NULL) {
				dev_err(xpc_part, "unable to kmalloc "
					"len=0x%016lx\n", buf_len);
				ret = xpNoMemory;
				break;
			}
		}

		ret = xp_remote_memcpy(xp_pa(buf), rp_pa, len);
		if (ret != xpSuccess) {
			dev_dbg(xpc_part, "xp_remote_memcpy failed %d\n", ret);
			break;
		}
	}

	kfree(buf_base);

	if (ret != xpSuccess)
		rp_pa = 0;

	dev_dbg(xpc_part, "reserved page at phys address 0x%016lx\n", rp_pa);
	return rp_pa;
}

nickdesaulniers · 2019-05-02T18:31:40Z

If there are other call sites to get_partition_rsvd_page_pa, I would look at them. (oh, look, there's not).

Looks like get_partition_rsvd_page_pa is an indirect call to either xpc_get_partition_rsvd_page_pa_uv or xpc_get_partition_rsvd_page_pa_sn2. Both of those convert their first arg to a u64 then pass it along.

Honestly, I'd email the maintainers pointing this warning out, and ask what exactly is going on here.

nathanchance · 2019-05-03T03:34:09Z

Inquiry sent: https://lore.kernel.org/lkml/20190503033340.GA7980@archlinux-i9/

nathanchance · 2019-05-23T16:44:00Z

Patch sent with @stephenhines's suggestion: https://lore.kernel.org/lkml/20190523161532.122421-1-natechancellor@gmail.com/

nathanchance · 2019-05-25T02:52:46Z

Patch accepted: https://git.kernel.org/gregkh/char-misc/c/b0576f9ecb5c51e9932531d23c447b2739261841

nathanchance · 2019-07-11T23:13:30Z

Merged into mainline: https://git.kernel.org/torvalds/c/b0576f9ecb5c51e9932531d23c447b2739261841

[ Upstream commit b0576f9ecb5c51e9932531d23c447b2739261841 ] Clang warns: drivers/misc/sgi-xp/xpc_partition.c:73:14: warning: variable 'buf' is uninitialized when used within its own initialization [-Wuninitialized] void *buf = buf; ~~~ ^~~ 1 warning generated. Arnd's explanation during review: /* * Returns the physical address of the partition's reserved page through * an iterative number of calls. * * On first call, 'cookie' and 'len' should be set to 0, and 'addr' * set to the nasid of the partition whose reserved page's address is * being sought. * On subsequent calls, pass the values, that were passed back on the * previous call. * * While the return status equals SALRET_MORE_PASSES, keep calling * this function after first copying 'len' bytes starting at 'addr' * into 'buf'. Once the return status equals SALRET_OK, 'addr' will * be the physical address of the partition's reserved page. If the * return status equals neither of these, an error as occurred. */ static inline s64 sn_partition_reserved_page_pa(u64 buf, u64 *cookie, u64 *addr, u64 *len) so *len is set to zero on the first call and tells the bios how many bytes are accessible at 'buf', and it does get updated by the BIOS to tell us how many bytes it needs, and then we allocate that and try again. Fixes: 2792902 ("[IA64-SGI] cleanup the way XPC locates the reserved page") Link: ClangBuiltLinux/linux#466 Suggested-by: Stephen Hines <srhines@google.com> Reviewed-by: Arnd Bergmann <arnd@arndb.de> Reviewed-by: Nick Desaulniers <ndesaulniers@google.com> Signed-off-by: Nathan Chancellor <natechancellor@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Sasha Levin <sashal@kernel.org>

[ Upstream commit b0576f9 ] Clang warns: drivers/misc/sgi-xp/xpc_partition.c:73:14: warning: variable 'buf' is uninitialized when used within its own initialization [-Wuninitialized] void *buf = buf; ~~~ ^~~ 1 warning generated. Arnd's explanation during review: /* * Returns the physical address of the partition's reserved page through * an iterative number of calls. * * On first call, 'cookie' and 'len' should be set to 0, and 'addr' * set to the nasid of the partition whose reserved page's address is * being sought. * On subsequent calls, pass the values, that were passed back on the * previous call. * * While the return status equals SALRET_MORE_PASSES, keep calling * this function after first copying 'len' bytes starting at 'addr' * into 'buf'. Once the return status equals SALRET_OK, 'addr' will * be the physical address of the partition's reserved page. If the * return status equals neither of these, an error as occurred. */ static inline s64 sn_partition_reserved_page_pa(u64 buf, u64 *cookie, u64 *addr, u64 *len) so *len is set to zero on the first call and tells the bios how many bytes are accessible at 'buf', and it does get updated by the BIOS to tell us how many bytes it needs, and then we allocate that and try again. Fixes: 2792902 ("[IA64-SGI] cleanup the way XPC locates the reserved page") Link: ClangBuiltLinux#466 Suggested-by: Stephen Hines <srhines@google.com> Reviewed-by: Arnd Bergmann <arnd@arndb.de> Reviewed-by: Nick Desaulniers <ndesaulniers@google.com> Signed-off-by: Nathan Chancellor <natechancellor@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Sasha Levin <sashal@kernel.org>

[ Upstream commit b0576f9 ] Clang warns: drivers/misc/sgi-xp/xpc_partition.c:73:14: warning: variable 'buf' is uninitialized when used within its own initialization [-Wuninitialized] void *buf = buf; ~~~ ^~~ 1 warning generated. Arnd's explanation during review: /* * Returns the physical address of the partition's reserved page through * an iterative number of calls. * * On first call, 'cookie' and 'len' should be set to 0, and 'addr' * set to the nasid of the partition whose reserved page's address is * being sought. * On subsequent calls, pass the values, that were passed back on the * previous call. * * While the return status equals SALRET_MORE_PASSES, keep calling * this function after first copying 'len' bytes starting at 'addr' * into 'buf'. Once the return status equals SALRET_OK, 'addr' will * be the physical address of the partition's reserved page. If the * return status equals neither of these, an error as occurred. */ static inline s64 sn_partition_reserved_page_pa(u64 buf, u64 *cookie, u64 *addr, u64 *len) so *len is set to zero on the first call and tells the bios how many bytes are accessible at 'buf', and it does get updated by the BIOS to tell us how many bytes it needs, and then we allocate that and try again. Fixes: 2792902 ("[IA64-SGI] cleanup the way XPC locates the reserved page") Link: ClangBuiltLinux/linux#466 Suggested-by: Stephen Hines <srhines@google.com> Reviewed-by: Arnd Bergmann <arnd@arndb.de> Reviewed-by: Nick Desaulniers <ndesaulniers@google.com> Signed-off-by: Nathan Chancellor <natechancellor@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Sasha Levin <sashal@kernel.org>

[ Upstream commit b0576f9ecb5c51e9932531d23c447b2739261841 ] Clang warns: drivers/misc/sgi-xp/xpc_partition.c:73:14: warning: variable 'buf' is uninitialized when used within its own initialization [-Wuninitialized] void *buf = buf; ~~~ ^~~ 1 warning generated. Arnd's explanation during review: /* * Returns the physical address of the partition's reserved page through * an iterative number of calls. * * On first call, 'cookie' and 'len' should be set to 0, and 'addr' * set to the nasid of the partition whose reserved page's address is * being sought. * On subsequent calls, pass the values, that were passed back on the * previous call. * * While the return status equals SALRET_MORE_PASSES, keep calling * this function after first copying 'len' bytes starting at 'addr' * into 'buf'. Once the return status equals SALRET_OK, 'addr' will * be the physical address of the partition's reserved page. If the * return status equals neither of these, an error as occurred. */ static inline s64 sn_partition_reserved_page_pa(u64 buf, u64 *cookie, u64 *addr, u64 *len) so *len is set to zero on the first call and tells the bios how many bytes are accessible at 'buf', and it does get updated by the BIOS to tell us how many bytes it needs, and then we allocate that and try again. Fixes: 279290294662 ("[IA64-SGI] cleanup the way XPC locates the reserved page") Link: ClangBuiltLinux/linux#466 Suggested-by: Stephen Hines <srhines@google.com> Reviewed-by: Arnd Bergmann <arnd@arndb.de> Reviewed-by: Nick Desaulniers <ndesaulniers@google.com> Signed-off-by: Nathan Chancellor <natechancellor@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Sasha Levin <sashal@kernel.org>