1 #
2 # This file and its contents are supplied under the terms of the
3 # Common Development and Distribution License ("CDDL"), version 1.0.
4 # You may only use this file in accordance with the terms of version
5 # 1.0 of the CDDL.
6 #
7 # A full copy of the text of the CDDL should have accompanied this
8 # source. A copy of the CDDL is also available via the Internet at
9 # http://www.illumos.org/license/CDDL.
10 #
11
12 #
13 # Copyright (c) 2013 Joyent, Inc. All rights reserved.
14 #
15
16 Historically e1000g and igb were maintained by two different teams at Sun and
17 thus while they used identical common code from Intel, they each only ever used
18 portions of it and were not kept in sync with one another. To help make
19 maintenance and the adding of new devices easier in illumos, we have gone
20 through and made it so that igb and e1000g rely on the same set of common code;
21 however, this code is not in its own module, each has its own copy of the code
22 compiled into it for various reasons which will be laid out below.
23
24 As part of the interface with the common code, the driver is required to define
25 an e1000_osdep.h. Currently each version of the driver defines its *own* version
26 of this header file in their own driver specific directory. However, the code
27 that implements this is different in each directory, specifically e1000g_osdep.c
28 and igb_osdep.c. It's important that they have different names and not be called
29 the same thing due to how the uts makefiles work.
30
31 Deviations from the common FreeBSD code:
32
33 We have a few differences from the common version of the FreeBSD code that exist
34 so that we can both gather firmware information and that have workarounds for
35 older chipsets. While, we would like to get that to be synced up and common, it
36 is not currently.
37
38 Energy Efficient Ethernet (EEE) is not enabled by default. This technology was
39 introduced with the I350 family of parts in the igb driver. However, there have
40 been issues seen with it in the wild and thus we opt to disable it by default
41 until tests have proven that there are no longer problems.
42
43 To help make that easier, we've documented here what these extra definitions
44 are. DO NOT just blindly copy over new common code. There is more work that
45 needs to be done in terms of changed interfaces and expectations for the
46 drivers.
47
48 # e1000_defines.h
49
50 In e1000_defines.h we add the following three definitions which are not
51 currently present. These definitions allow us to attach firware revisions and
52 other information to the devinfo tree.
53
54 #define NVM_VERSION 0x0005
55 #define NVM_OEM_OFFSET_0 6
56 #define NVM_OEM_OFFSET_1 7
57
58 # Workarounds for the 82541 and 82547
59
60 There are various workarounds in place for the 82541 and 82547 due to errata
61 that exist for these devices. This has traditionally been a part of the common
62 code. Until this can get merged into the common code completely, we've spearted
63 out the changes that are the actual C functions into
64 uts/common/io/e1000g/e1000g_workarounds.c. However, this alone is not
65 sufficient. You must make sure that in e1000_hw.h that the struc
66 e1000_dev_spec_82541 actually looks like the following:
67
68 struct e1000_dev_spec_82541 {
69 enum e1000_dsp_config dsp_config;
70 enum e1000_ffe_config ffe_config;
71 u32 tx_fifo_head;
72 u32 tx_fifo_start;
73 u32 tx_fifo_size;
74 u16 spd_default;
75 bool phy_init_script;
76 bool ttl_workaround;
77 };
78
79 # EEE
80
81 By default we disable all support for EEE. To cause this to happen you must
82 make the following change in e1000_82575.c's init_mac_params.
83
84 From:
85 394 /* Enable EEE default settings for EEE supported devices */
86 395 if (mac->type >= e1000_i350)
87 396 dev_spec->eee_disable = FALSE;
88 To:
89 394 /* Enable EEE default settings for EEE supported devices */
90 395 if (mac->type >= e1000_i350)
91 396 dev_spec->eee_disable = TRUE;
92
93 Future work:
94
95 The next step here is to take the osdep portions and merge them. That would
96 allow us to build one common misc module e1000api that both igb and e1000g
97 depend on rather than building separate copies of the common code into each
98 driver. Another potential option which may prove to have less value is to take
99 all of the gld and ddi logic and have one driver export that leaving e1000g and
100 igb as small stubs which depend on that. Note however, that the latter is not
101 how our upstream is currently structuring their igb and em (FreeBSD's e1000g)
102 drivers.