Add Support for Phy rx eom log #2056
Conversation
igaw
force-pushed
the
phy-rx-eom-log
branch
from
795d25a
to
3fd1667
Compare
|
rebased on current master to get latest libnvme changes. |
The printing backend rework is still WIP. The goal is to have all commands producing output for all formats (at least stdout and json). Currently there are a lot of missing hooks. So it would be great to have also the json output part added. I am unsure if we want have a binary output as well. I suspect for some commands it would make sense. |
I don't mind it that it is in this PR. |
bpaupore-wdc
force-pushed
the
phy-rx-eom-log
branch
from
e44b407
to
9ef05a0
Compare
Noticed this when adding similar behavior for phy_rx_eom_log's odp, it doesn't make much sense to right-shift the byte by 8. Signed-off-by: Brandon Paupore <[email protected]>
bpaupore-wdc
force-pushed
the
phy-rx-eom-log
branch
from
9ef05a0
to
0e92568
Compare
This attempts to preserve the functionality of realloc while also ensuring the final memory is aligned. Signed-off-by: Brandon Paupore <[email protected]>
bpaupore-wdc
force-pushed
the
phy-rx-eom-log
branch
from
0e92568
to
d860837
Compare
This implements support for TP4119a, adding a new command nvme phy-rx-eom-log for issuing Get Log Page with the new Log Identifier and allowing for configuration/validation of the Log Specific Parameters. Signed-off-by: Brandon Paupore <[email protected]>
bpaupore-wdc
force-pushed
the
phy-rx-eom-log
branch
from
d860837
to
10fde77
Compare
|
Thanks a lot! |
This primarily introduces a new command
phy-rx-eom-logfor getting that log page and allowing the user to specify the lsi/lsp values for it. These changes depend on libnvme commite923016.One question I have is how to handle failing allocations in
nvme-print-json.c. Currently these changes just silently fail to include the printable eye data in the json root, but if there's some way this error should be reported I can make that change.The second commit addresses a very minor bug obscuring non-zero reserved fields for fdpa. I could split this out into its own PR if that would be preferred, otherwise it seemed simple enough to include here as a separate commit.