diff options
author | Paul Mackerras <paulus@samba.org> | 2007-05-07 23:37:51 -0400 |
---|---|---|
committer | Paul Mackerras <paulus@samba.org> | 2007-05-07 23:37:51 -0400 |
commit | 02bbc0f09c90cefdb2837605c96a66c5ce4ba2e1 (patch) | |
tree | 04ef573cd4de095c500c9fc3477f4278c0b36300 /Documentation/ia64/aliasing.txt | |
parent | 7487a2245b8841c77ba9db406cf99a483b9334e9 (diff) | |
parent | 5b94f675f57e4ff16c8fda09088d7480a84dcd91 (diff) |
Merge branch 'linux-2.6'
Diffstat (limited to 'Documentation/ia64/aliasing.txt')
-rw-r--r-- | Documentation/ia64/aliasing.txt | 71 |
1 files changed, 37 insertions, 34 deletions
diff --git a/Documentation/ia64/aliasing.txt b/Documentation/ia64/aliasing.txt index 38f9a52d1820..9a431a7d0f5d 100644 --- a/Documentation/ia64/aliasing.txt +++ b/Documentation/ia64/aliasing.txt | |||
@@ -112,16 +112,6 @@ POTENTIAL ATTRIBUTE ALIASING CASES | |||
112 | 112 | ||
113 | The /dev/mem mmap constraints apply. | 113 | The /dev/mem mmap constraints apply. |
114 | 114 | ||
115 | However, since this is for mapping legacy MMIO space, WB access | ||
116 | does not make sense. This matters on machines without legacy | ||
117 | VGA support: these machines may have WB memory for the entire | ||
118 | first megabyte (or even the entire first granule). | ||
119 | |||
120 | On these machines, we could mmap legacy_mem as WB, which would | ||
121 | be safe in terms of attribute aliasing, but X has no way of | ||
122 | knowing that it is accessing regular memory, not a frame buffer, | ||
123 | so the kernel should fail the mmap rather than doing it with WB. | ||
124 | |||
125 | read/write of /dev/mem | 115 | read/write of /dev/mem |
126 | 116 | ||
127 | This uses copy_from_user(), which implicitly uses a kernel | 117 | This uses copy_from_user(), which implicitly uses a kernel |
@@ -138,14 +128,20 @@ POTENTIAL ATTRIBUTE ALIASING CASES | |||
138 | 128 | ||
139 | ioremap() | 129 | ioremap() |
140 | 130 | ||
141 | This returns a kernel identity mapping for use inside the | 131 | This returns a mapping for use inside the kernel. |
142 | kernel. | ||
143 | 132 | ||
144 | If the region is in kern_memmap, we should use the attribute | 133 | If the region is in kern_memmap, we should use the attribute |
145 | specified there. Otherwise, if the EFI memory map reports that | 134 | specified there. |
146 | the entire granule supports WB, we should use that (granules | 135 | |
147 | that are partially reserved or occupied by firmware do not appear | 136 | If the EFI memory map reports that the entire granule supports |
148 | in kern_memmap). Otherwise, we should use a UC mapping. | 137 | WB, we should use that (granules that are partially reserved |
138 | or occupied by firmware do not appear in kern_memmap). | ||
139 | |||
140 | If the granule contains non-WB memory, but we can cover the | ||
141 | region safely with kernel page table mappings, we can use | ||
142 | ioremap_page_range() as most other architectures do. | ||
143 | |||
144 | Failing all of the above, we have to fall back to a UC mapping. | ||
149 | 145 | ||
150 | PAST PROBLEM CASES | 146 | PAST PROBLEM CASES |
151 | 147 | ||
@@ -158,7 +154,7 @@ PAST PROBLEM CASES | |||
158 | succeed. It may create either WB or UC user mappings, depending | 154 | succeed. It may create either WB or UC user mappings, depending |
159 | on whether the region is in kern_memmap or the EFI memory map. | 155 | on whether the region is in kern_memmap or the EFI memory map. |
160 | 156 | ||
161 | mmap of 0x0-0xA0000 /dev/mem by "hwinfo" on HP sx1000 with VGA enabled | 157 | mmap of 0x0-0x9FFFF /dev/mem by "hwinfo" on HP sx1000 with VGA enabled |
162 | 158 | ||
163 | See https://bugzilla.novell.com/show_bug.cgi?id=140858. | 159 | See https://bugzilla.novell.com/show_bug.cgi?id=140858. |
164 | 160 | ||
@@ -171,28 +167,25 @@ PAST PROBLEM CASES | |||
171 | so it is safe to use WB mappings. | 167 | so it is safe to use WB mappings. |
172 | 168 | ||
173 | The kernel VGA driver may ioremap the VGA frame buffer at 0xA0000, | 169 | The kernel VGA driver may ioremap the VGA frame buffer at 0xA0000, |
174 | which will use a granule-sized UC mapping covering 0-0xFFFFF. This | 170 | which uses a granule-sized UC mapping. This granule will cover some |
175 | granule covers some WB-only memory, but since UC is non-speculative, | 171 | WB-only memory, but since UC is non-speculative, the processor will |
176 | the processor will never generate an uncacheable reference to the | 172 | never generate an uncacheable reference to the WB-only areas unless |
177 | WB-only areas unless the driver explicitly touches them. | 173 | the driver explicitly touches them. |
178 | 174 | ||
179 | mmap of 0x0-0xFFFFF legacy_mem by "X" | 175 | mmap of 0x0-0xFFFFF legacy_mem by "X" |
180 | 176 | ||
181 | If the EFI memory map reports this entire range as WB, there | 177 | If the EFI memory map reports that the entire range supports the |
182 | is no VGA MMIO hole, and the mmap should fail or be done with | 178 | same attributes, we can allow the mmap (and we will prefer WB if |
183 | a WB mapping. | 179 | supported, as is the case with HP sx[12]000 machines with VGA |
180 | disabled). | ||
184 | 181 | ||
185 | There's no easy way for X to determine whether the 0xA0000-0xBFFFF | 182 | If EFI reports the range as partly WB and partly UC (as on sx[12]000 |
186 | region is a frame buffer or just memory, so I think it's best to | 183 | machines with VGA enabled), we must fail the mmap because there's no |
187 | just fail this mmap request rather than using a WB mapping. As | 184 | safe attribute to use. |
188 | far as I know, there's no need to map legacy_mem with WB | ||
189 | mappings. | ||
190 | 185 | ||
191 | Otherwise, a UC mapping of the entire region is probably safe. | 186 | If EFI reports some of the range but not all (as on Intel firmware |
192 | The VGA hole means the region will not be in kern_memmap. The | 187 | that doesn't report the VGA frame buffer at all), we should fail the |
193 | HP sx1000 chipset doesn't support UC access to the memory surrounding | 188 | mmap and force the user to map just the specific region of interest. |
194 | the VGA hole, but X doesn't need that area anyway and should not | ||
195 | reference it. | ||
196 | 189 | ||
197 | mmap of 0xA0000-0xBFFFF legacy_mem by "X" on HP sx1000 with VGA disabled | 190 | mmap of 0xA0000-0xBFFFF legacy_mem by "X" on HP sx1000 with VGA disabled |
198 | 191 | ||
@@ -202,6 +195,16 @@ PAST PROBLEM CASES | |||
202 | This is a special case of the previous case, and the mmap should | 195 | This is a special case of the previous case, and the mmap should |
203 | fail for the same reason as above. | 196 | fail for the same reason as above. |
204 | 197 | ||
198 | read of /sys/devices/.../rom | ||
199 | |||
200 | For VGA devices, this may cause an ioremap() of 0xC0000. This | ||
201 | used to be done with a UC mapping, because the VGA frame buffer | ||
202 | at 0xA0000 prevents use of a WB granule. The UC mapping causes | ||
203 | an MCA on HP sx[12]000 chipsets. | ||
204 | |||
205 | We should use WB page table mappings to avoid covering the VGA | ||
206 | frame buffer. | ||
207 | |||
205 | NOTES | 208 | NOTES |
206 | 209 | ||
207 | [1] SDM rev 2.2, vol 2, sec 4.4.1. | 210 | [1] SDM rev 2.2, vol 2, sec 4.4.1. |