pussuw opened a new pull request, #16031:
URL: https://github.com/apache/nuttx/pull/16031
## Summary
This replaces the big kernel lock (enter/leave_critical_section()) with
local spinlocks from all MPFS drivers that still use it. The reason is
obviously to alleviate big kernel lock c
pussuw commented on code in PR #16031:
URL: https://github.com/apache/nuttx/pull/16031#discussion_r202052
##
arch/risc-v/src/mpfs/mpfs_dsn.c:
##
@@ -91,9 +98,7 @@ int mpfs_read_dsn(uint8_t *dsn, size_t len)
while ((getreg32(SERVICES_SR) & SCBCTRL_SERVICESSR_BUSY) && --r
xiaoxiang781216 commented on code in PR #16031:
URL: https://github.com/apache/nuttx/pull/16031#discussion_r2007997947
##
arch/risc-v/src/mpfs/mpfs_dsn.c:
##
@@ -91,9 +98,7 @@ int mpfs_read_dsn(uint8_t *dsn, size_t len)
while ((getreg32(SERVICES_SR) & SCBCTRL_SERVICESSR_BUS
pussuw commented on PR #16031:
URL: https://github.com/apache/nuttx/pull/16031#issuecomment-2743447683
> @pussuw nice work! Is there some way to test and display how much it
improved the NuttX performance?
I cannot share logs but in our use case total CPU load drops by ~3-5% in SMP
m
pussuw commented on code in PR #16031:
URL: https://github.com/apache/nuttx/pull/16031#discussion_r2008035543
##
arch/risc-v/src/mpfs/mpfs_dsn.c:
##
@@ -91,9 +98,7 @@ int mpfs_read_dsn(uint8_t *dsn, size_t len)
while ((getreg32(SERVICES_SR) & SCBCTRL_SERVICESSR_BUSY) && --r
nuttxpr commented on PR #16031:
URL: https://github.com/apache/nuttx/pull/16031#issuecomment-2743116471
[**\[Experimental Bot, please feedback
here\]**](https://github.com/search?q=repo%3Aapache%2Fnuttx+13552&type=issues)
The PR mostly meets the NuttX requirements but is missi
xiaoxiang781216 merged PR #16031:
URL: https://github.com/apache/nuttx/pull/16031
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscr...@nu
xiaoxiang781216 commented on code in PR #16031:
URL: https://github.com/apache/nuttx/pull/16031#discussion_r2013932328
##
arch/risc-v/src/mpfs/mpfs_dsn.c:
##
@@ -91,9 +98,7 @@ int mpfs_read_dsn(uint8_t *dsn, size_t len)
while ((getreg32(SERVICES_SR) & SCBCTRL_SERVICESSR_BUS
xiaoxiang781216 commented on code in PR #16031:
URL: https://github.com/apache/nuttx/pull/16031#discussion_r2008619207
##
arch/risc-v/src/mpfs/mpfs_coremmc.c:
##
@@ -525,7 +526,7 @@ static void mpfs_configxfrints(struct mpfs_dev_s *priv,
uint32_t xfrmask,
{
irqstate_t flags
acassis commented on PR #16031:
URL: https://github.com/apache/nuttx/pull/16031#issuecomment-2743389771
@pussuw nice work! Is there some way to test and display how much it
improved the NuttX performance?
--
This is an automated message from the Apache Git Service.
To respond to the messa
pussuw commented on code in PR #16031:
URL: https://github.com/apache/nuttx/pull/16031#discussion_r2008078388
##
arch/risc-v/src/mpfs/mpfs_dsn.c:
##
@@ -91,9 +98,7 @@ int mpfs_read_dsn(uint8_t *dsn, size_t len)
while ((getreg32(SERVICES_SR) & SCBCTRL_SERVICESSR_BUSY) && --r
pussuw commented on code in PR #16031:
URL: https://github.com/apache/nuttx/pull/16031#discussion_r2008033805
##
arch/risc-v/src/mpfs/mpfs_coremmc.c:
##
@@ -525,7 +526,7 @@ static void mpfs_configxfrints(struct mpfs_dev_s *priv,
uint32_t xfrmask,
{
irqstate_t flags;
- fl
12 matches
Mail list logo
