On Wed, Aug 20, 2025 at 08:02:14AM +1000, Richard Henderson wrote:
> On 8/20/25 06:27, Daniel P. Berrangé wrote:
> > void qemu_log(const char *fmt, ...)
> > {
> > FILE *f;
> > - g_autofree const char *timestr = NULL;
> > -
> > /*
> > - * Prepare the timestamp*outside* the logging
> > - * lock so it better reflects when the message
> > - * was emitted if we are delayed acquiring the
> > - * mutex
> > + * Prepare the context*outside* the logging
> > + * lock so any timestamp better reflects when
> > + * the message was emitted if we are delayed
> > + * acquiring the mutex
> > */
> > - if (message_with_timestamp) {
> > - g_autoptr(GDateTime) dt = g_date_time_new_now_utc();
> > - timestr = g_date_time_format_iso8601(dt);
> > - }
>
> Something that really ought to be handled better is that we currently use
> qemu_log to print pieces of a larger message. E.g. looping over the general
> registers to print the whole cpu state.
>
> (1) We don't want timestamps in the middle, and
> (2) we don't want unnecessary memory allocation in the middle.
Interestingly I found that for CPU_LOG_TB_CPU we don't use the qemu_log()
function at all. TCG calls qemu_log_try_lock() to acquire the raw FILE *
handle and writes directly to it, so we miss all prefix info entirely.
If other parts of QEMU that do big multi-line log dumps do the same,
then we're ok.
With regards,
Daniel
--
|: https://berrange.com -o- https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org -o- https://fstop138.berrange.com :|
|: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|
