- increase pmc cpuid field from 8 to 12 bits
- add cpuid version string to initialize entry in the log
so that filter can identify which counter index an
event name maps to
- GC unused config flags
- make fixed counter assignment more robust as well as the
changes needed to be properly identified for filter
This adds the -U options to pmcstat which will attribute in-kernel samples
back to the user stack that invoked the system call. It is not the default,
because when looking at kernel profiles it is generally more desirable to
merge all instances of a given system call together.
Although heavily revised, this change is directly derived from D7350 by
Jonathan T. Looney.
Obtained from: jtl
Sponsored by: Juniper Networks, Limelight Networks
- avoid the need to call a function to get size of known array. I'll
likely re-arrange some of the indirect in a later to avoid the magic
constants.
- use correct type
- add const
- replace caddr_t with void*. This corrects an alignment warning.
- remove duplicated include from immediately prior commit
Under base clang we're now down to:
- 3 warning in top.c, 1 warning in mahcine.c, 4 warning in display.c,
- 1 warning in utils.c
Tested with base clang, gcc7, gcc9, base gcc (mips)
The only syscalls in this class are close, closefrom, munmap, and revoke.
Submitted by: aniketp
MFC after: 2 weeks
Sponsored by: Google, Inc. (GSoC 2018)
Differential Revision: https://reviews.freebsd.org/D15650
- Add const where helpful
- add missing 'static' for file-local functions
- use nitems where possible
- convert manual abort() to assert
- use strndup instead of homegrown version
Tested with clang, gcc7, and gcc9
With -lpl, code surrounded by parentheses in continuation lines is lined up
even if it would extend past the right margin.
With -nlpl (the default), such a line that would extend past the right
margin is moved left to keep it within the margin, if that does not require
placing it to the left of the prevailing indentation level.
These switches have no effect if -nlp is selected.
Submitted by: Tom Lane
With -lp, if a line has an opening paren which is not closed on that line,
then continuation lines will be lined up to start at the character position
just after the opening paren.
Submitted by: Tom Lane
gcc8 warns that "verf" was set but not used. This was because the code
that uses it is disabled via a "#if 0".
This patch adds a "#if 0" to the variable's declaration and assignment
to get rid of the warning.
This way the code could be re-enabled without difficulty.
Requested by: mmacy
MFC after: 2 weeks
Rewrite the macros so that they take a parameter. Consumers use it to signal
how much room in the buffer they need; this lets them do that once when
required space is known instead of doing the check once every loop step.
Also take the parameter value into consideration when resizing the buffer;
the requested space may be larger than the constant 400 bytes that the
previous version used - now it's the sum of those two values.
On the consumer side, don't copy strings byte by byte - use memcpy().
Deduplicate code that copied base 2, base 8 and base 16 literals.
Don't advance the e_token pointer once the token has been copied into
s_token. This allows easy calculation of the token's length.
The troff output in indent was invented at Sun and the online documentation
for some post-SunOS operating system includes this:
The usual way to get a troffed listing is with the command
indent -troff program.c | troff -mindent
The indent manual page in FreeBSD 1.0 already lacks that information and
troff -mindent complains about not being able to find the macro file.
It seems that the file did exist on SunOS and was supposed to be imported
into 4.3BSD together with the feature, but that has never happened.
Removal of troff output support simplifies a lot of indent's code.
vgrind(1) seems to be a promising replacement.
It was a shorthand for checking if ps.procname is a non-empty string; the
same can be done with ps.procname[0] which avoids the need for updating
is_procname after every call to lexi().
The trick is to copy everything from the start of the line into the buffer
that stores newlines and comments until indent finds a brace or an else.
pr_comment() will use that information to calculate the original indentation
of the boxed comment.
This requires storing two pieces of information: the real start of the
buffer (sc_buf) and the start of the comment (save_com).
lexi() reads the input stream and categorizes the next token. indent will
sometimes buffer up a sequence of tokens in order rearrange them. That is
needed for properly cuddling else or placing braces correctly according to
the chosen style (KNF vs Allman) when comments are around. The loop that
buffers tokens up uses lexi() to decide if it's time to stop buffering. Then
the temporary buffer is used to feed lexi() the same tokens again, this time
for normal processing.
The problem is that lexi() apart from recognizing the token, can change
a lot of information about the current state, for example ps.last_nl,
ps.keyword, buf_ptr. It also abandons leading whitespace, which is needed
mainly for comment-related considerations. So the call to lexi() while
tokens are buffered up and categorized can change the state before they're
read again for normal processing which may easily result in changing
interpretation of the current state and lead to incorrect output.
To work around the problems:
1) copy the whitespace into the save_com buffer so that it will be read
again when processed
2) trick lexi() into modifying a temporary copy of the parser state instead
of the original.
"while (...)" and "else" or "{"
* Don't flush newlines - there can be multiple of them and they can happen
before a token that isn't else or {. Instead, always store them in save_com.
* Don't dump the buffer's contents on newline assuming that there is only
one comment before else or {.
* Avoid producing surplus newlines, especially before else when -ce is on.
* When -bl is on, don't treat { as a comment (was implemented by falling
through "case lbrace:" to "case comment:").
This commit fixes the above, but exposes another bug and thus breaks several
other tests. Another commit will make them pass again.
In fixing issues with uid > INT_MAX, I broke the uid without username
case. The latter is more important so return the old state.
Discussed with: allanjude
One of the downsides of using numeric WARNS is that if we only have a
single type of issue we get no protection from other changes. For
example, we got no warning for missing variable declaration, due to
the issues with "const".
For this utility, explicitly list out the warnings which are failing.
They should still be fixed, so only reduce them to warning instead of
error.
Tested with: clang base (amd64, i386), gcc6, gcc7, gcc9, gcc base (mips)
Even though there do appear to be more artificial frames, with 12, stack
traces no longer list at all. Revert until a better, more stable value can
be determined.
- use bool instead of int [0]
- use calloc correctly [0]
(this also caught an incorrect sizeof argument) [1]
- use size_t over int [2]
- correct style
Reported by: pfg [0], scan-build [1], gcc [2]
Plenty of allocation sites pass M_ZERO and sizes which are small and known
at compilation time. Handling them internally in malloc loses this information
and results in avoidable calls to memset.
Instead, let the compiler take the advantage of it whenever possible.
Discussed with: jeff
- Prefer calloc over malloc. This is more predicable and we're not in a
performance sensitive context. [1]
- Remove bogus comment (obsolete from prior commit). [2]
- Remove void casts and type casts of NULL
- Remove redundant declaration of 'quit'
- Add additional const
Reported by: kib [1], vangyzen [2]
Summary: Included VSX registers in powerpc core dumps (both kernel and gcore)
Submitted by: Luis Pires
Differential Revision: https://reviews.freebsd.org/D15512
Currently all the primitives are waiting for a rewrite, tidy them up in the
meantime.
Vast majority of cases pass sizes which are multiple of 8. Which means the
following rep stosb/movb has nothing to do. Turns out testing first if there
is anything to do is a big win across the board (cpus with and without ERMS,
Intel and AMD) while not pessimizing the case where there is work to do.
Sample results for zeroing 64 bytes (ops/second):
Ryzen Threadripper 1950X 91433212 -> 147265741
Intel(R) Xeon(R) CPU X5675 @ 3.07GHz 90714044 -> 121992888
bzero and bcopy are on their way out and were not modified. Nothing in the
tree uses them.
