blob: f403bbec9b1b10faf7e970829ecff946e76cd3b0 (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
|
From: =?UTF-8?q?Rafa=C5=82=20Mi=C5=82ecki?= <rafal@milecki.pl>
Date: Thu, 13 Jul 2023 17:30:59 +0200
Subject: [PATCH] nvmem: core: fix support for fixed cells NVMEM layout
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Returning -EPROBE_DEFER for "fixed-layout" makes nvmem_register() always
fail (that layout is supported internally with no external module). That
makes callers (e.g. mtd_nvmem_add()) fail as well and prevents booting
on devices with "fixed-layout" in DT.
Add a quick workaround for it.
Fixes: 6468a6f45148 ("nvmem: core: handle the absence of expected layouts")
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
---
--- a/drivers/nvmem/core.c
+++ b/drivers/nvmem/core.c
@@ -798,6 +798,19 @@ static struct nvmem_layout *nvmem_layout
return NULL;
/*
+ * We should return -EPROBE_DEFER only when layout driver is expected to
+ * become available later. Otherwise NVMEM will never probe successfully
+ * for unsupported layouts. There is no known solution for that right
+ * now.
+ *
+ * This problem also affects "fixed-layout". It's supported in NVMEM
+ * core code so there never will be layout for it. We shouldn't return
+ * -EPROBE_DEFER in such case. Add a quick workaround for that.
+ */
+ if (of_device_is_compatible(layout_np, "fixed-layout"))
+ return NULL;
+
+ /*
* In case the nvmem device was built-in while the layout was built as a
* module, we shall manually request the layout driver loading otherwise
* we'll never have any match.
|