Searched hist:"5 a122a6e1b620e79c5c6807e83f1cd75fe872fb6" (Results 1 – 7 of 7) sorted by relevance
/openbmc/smbios-mdr/src/ |
H A D | pcieslot.cpp | diff 5a122a6e1b620e79c5c6807e83f1cd75fe872fb6 Wed May 03 23:25:03 CDT 2023 Brandon Kim <brandonkim@google.com> mdrv2: Attempt to update D-Bus objects in place
When systemInfoUpdate() is called multiple times, there's a potential race condition of a user polling for the objects while the array of D-Bus objects are erased and being repopulated. This results in incomplete set of memory or CPU counts, which can lead to unforseen consequences.
Tested: Verified that a corner case that was hitting this case consistently (when SMBIOS was transferred using ipmi-blob repeatedly) goes away with this implementation.
Signed-off-by: Brandon Kim <brandonkim@google.com> Change-Id: I312aa91cd11b1dd06502d04272889922108d39a2
|
H A D | cpu.cpp | diff 5a122a6e1b620e79c5c6807e83f1cd75fe872fb6 Wed May 03 23:25:03 CDT 2023 Brandon Kim <brandonkim@google.com> mdrv2: Attempt to update D-Bus objects in place
When systemInfoUpdate() is called multiple times, there's a potential race condition of a user polling for the objects while the array of D-Bus objects are erased and being repopulated. This results in incomplete set of memory or CPU counts, which can lead to unforseen consequences.
Tested: Verified that a corner case that was hitting this case consistently (when SMBIOS was transferred using ipmi-blob repeatedly) goes away with this implementation.
Signed-off-by: Brandon Kim <brandonkim@google.com> Change-Id: I312aa91cd11b1dd06502d04272889922108d39a2
|
H A D | dimm.cpp | diff 5a122a6e1b620e79c5c6807e83f1cd75fe872fb6 Wed May 03 23:25:03 CDT 2023 Brandon Kim <brandonkim@google.com> mdrv2: Attempt to update D-Bus objects in place
When systemInfoUpdate() is called multiple times, there's a potential race condition of a user polling for the objects while the array of D-Bus objects are erased and being repopulated. This results in incomplete set of memory or CPU counts, which can lead to unforseen consequences.
Tested: Verified that a corner case that was hitting this case consistently (when SMBIOS was transferred using ipmi-blob repeatedly) goes away with this implementation.
Signed-off-by: Brandon Kim <brandonkim@google.com> Change-Id: I312aa91cd11b1dd06502d04272889922108d39a2
|
H A D | mdrv2.cpp | diff 5a122a6e1b620e79c5c6807e83f1cd75fe872fb6 Wed May 03 23:25:03 CDT 2023 Brandon Kim <brandonkim@google.com> mdrv2: Attempt to update D-Bus objects in place
When systemInfoUpdate() is called multiple times, there's a potential race condition of a user polling for the objects while the array of D-Bus objects are erased and being repopulated. This results in incomplete set of memory or CPU counts, which can lead to unforseen consequences.
Tested: Verified that a corner case that was hitting this case consistently (when SMBIOS was transferred using ipmi-blob repeatedly) goes away with this implementation.
Signed-off-by: Brandon Kim <brandonkim@google.com> Change-Id: I312aa91cd11b1dd06502d04272889922108d39a2
|
/openbmc/smbios-mdr/include/ |
H A D | pcieslot.hpp | diff 5a122a6e1b620e79c5c6807e83f1cd75fe872fb6 Wed May 03 23:25:03 CDT 2023 Brandon Kim <brandonkim@google.com> mdrv2: Attempt to update D-Bus objects in place
When systemInfoUpdate() is called multiple times, there's a potential race condition of a user polling for the objects while the array of D-Bus objects are erased and being repopulated. This results in incomplete set of memory or CPU counts, which can lead to unforseen consequences.
Tested: Verified that a corner case that was hitting this case consistently (when SMBIOS was transferred using ipmi-blob repeatedly) goes away with this implementation.
Signed-off-by: Brandon Kim <brandonkim@google.com> Change-Id: I312aa91cd11b1dd06502d04272889922108d39a2
|
H A D | cpu.hpp | diff 5a122a6e1b620e79c5c6807e83f1cd75fe872fb6 Wed May 03 23:25:03 CDT 2023 Brandon Kim <brandonkim@google.com> mdrv2: Attempt to update D-Bus objects in place
When systemInfoUpdate() is called multiple times, there's a potential race condition of a user polling for the objects while the array of D-Bus objects are erased and being repopulated. This results in incomplete set of memory or CPU counts, which can lead to unforseen consequences.
Tested: Verified that a corner case that was hitting this case consistently (when SMBIOS was transferred using ipmi-blob repeatedly) goes away with this implementation.
Signed-off-by: Brandon Kim <brandonkim@google.com> Change-Id: I312aa91cd11b1dd06502d04272889922108d39a2
|
H A D | dimm.hpp | diff 5a122a6e1b620e79c5c6807e83f1cd75fe872fb6 Wed May 03 23:25:03 CDT 2023 Brandon Kim <brandonkim@google.com> mdrv2: Attempt to update D-Bus objects in place
When systemInfoUpdate() is called multiple times, there's a potential race condition of a user polling for the objects while the array of D-Bus objects are erased and being repopulated. This results in incomplete set of memory or CPU counts, which can lead to unforseen consequences.
Tested: Verified that a corner case that was hitting this case consistently (when SMBIOS was transferred using ipmi-blob repeatedly) goes away with this implementation.
Signed-off-by: Brandon Kim <brandonkim@google.com> Change-Id: I312aa91cd11b1dd06502d04272889922108d39a2
|